Proto Sociology VOL. 13, 1999 - REASONING AND ARGUMENTATION

Proto Sociology An International Journal of Interdisciplinary Research VOL. 13, 1999 REASONING AND ARGUMENTATION Dieter ...

Author: Dieter Mans | Gerhard Preyer (eds.)

36 downloads 761 Views 820KB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Proto Sociology An International Journal of Interdisciplinary Research VOL. 13, 1999 REASONING AND ARGUMENTATION Dieter Mans, Gerhard Preyer (eds.) Content

Introduction: On Contemporary Developments in the Theory of Argumentation Gerhard Preyer, Dieter Mans

3

Ralph H. Johnson Reasoning, Argumentation and The Network Problem

14

Leo Groarke The Fox and the Hedgehog: On Logic, Argument, and Argumentation Theory

29

J. Anthony Blair Presumptive Reasoning/Argument: An Overlooked Class

46

Robert C. Pinto Argument Schemes and the Evaluation of Presumptive Reasoning: some Reflections on Blair’s Account

61

Douglas Walton The New Dialectic: A Method of Evaluating an Argument Used for Some Purpose in a Given Case

70

Manfred Kienpointner Comments on Douglas Walton’s Paper: The New Dialectic: A Method of Evaluating an Argument Used for Some Purpose in a Given Case

92

Christopher W. Tindale The Authority of Testimony

96

John Woods Peirce’s Abductive Enthusiasms

117

Henry W. Johnstone, Jr. “‘Any,’ ‘Every,’ and the Philosophical Argumentum ad Hominem”

126

ON CONTEMPORARY PHILOSOPHY Hans Lenk Interdisziplinarität und Interpretation

133

Ellery Eells Causal Decision Theory

159

Authors

181

Imprint

182

On ProtoSociology

183

Volumes

184

Bookpublications

190

Subscription

204

Digital Publications – Special Offer

205

Introduction: On Contemporary Developments in the Theory of Argumentation

3

INTRODUCTION: On Contemporary Developments in the Theory of Argumentation Stephan Toulmin’s book – The Uses of Argument (1958) – was a bold attack against the subsumption of theories of argumentation under formal logic. Toulmin’s book received very critical reviews but even his harshest critics cannot deny that many contemporary developments in the field are consistent with his basic line of reasoning. Since the 70's there are broad researches on the theory of argumentation called “informal logic” or “critical thinking”. Many flowers have grown – some say too many – and our biotope shows a great diversity. Some points can be made to show the relevance of this biotope. Here is one of them: The methods of scientific research is observation, description, and theorizing. The medium of understanding scientific and other cultural knowledge are texts with their arguments and explanations. Written language is a medium which takes structural effects in the semantics of social structure.1 Content analysis is a tool for understanding texts and thereby society and it is therefore a basic research method in the social sciences. The analysis of arguments2 is a special form of content analysis and given the frequency of arguments we should expect a lot of work in that area done by social scientists. That is not the case. Protosociology intends to popularize the theory of argumentation as a challenge for theoretical research and a possibility for new applications. We publish some papers by leading researchers in the field to outline the topology of a new and powerful interdisciplinary theory. We have compiled these results and hope that this compilation will spur the interest in the field. If so, we intend to publish an issue which shows the possibility of applications for the social sciences. In contemporary research we recognize elaborated results, expansions of discovery, and new developments of conceptualization of the initial subject. Some of the leading questions in particular in our collection are:

1 See in particular N. Luhmann, Die Gesellschaft der Gesellschaft I (2 Vol.), Frankfurt am Main 1997, pp. 249-290. 2 We prefer the terms “analysis of arguments” and “theories of argumentation” over “informal logic” or “critical thinking” simply because the word “argumentation” points to the type of work we are doing.

4

Dieter Mans, Gerhard Preyer

How shall we distinguish between argumentation and reasoning? What is the meaning of crucial terms like inference and implication? What type of inferences do we apply in everyday argumentation? One of the significant problems in contemporary research is the analysis of presumptive arguments because such arguments are a feature of common argumentation. How fruitful is the conception of New Dialectic to the study of fallacies? What is an argument scheme and how can it be used to criticize an argument?1 When we write down a coherent sequence of sentences (may be a scientific paper, a report on a sport event, a recipe for cooking, or a letter to a friend) some of these sentences may be identified as an argument. Suppose some days later you read your own paper and you say to yourself: “I tried to find an argument, but I did not succeed.” But even if you were not successful in finding the intended argument, your paper contains the trace of a mental activity. Now if you try to describe, what was going on, you might say, “I thought about x” or “I reasoned about x”. Most – if not all! – researchers assent to the following statement: You cannot find an argument without reasoning, but you can reason without finding an argument. It follows, that reasoning and argument cannot be the same. Ralph H. Johnson’s contribution gives a highly readable exposition which clarifies fundamental terms in the theory of argumentation. His leading question is: How shall we distinguish between argumentation and reasoning? What is relationship between reasoning and rationality, intelligence, knowledge and thinking? He reviews some of the positions that other theorists have taken on the relationship between argumentation and reasoning. Specifically, he addresses the views expressed by Finochiarro (1984), Govier (1987), and Walton (1990) and discusses their shortcomings. His answer is that reasoning can best be understood as the generic concept of which argumentation is a species. Besides that, his paper indicates that there is productive confluence of cognitive science, philosophy and psychology in the theory of argumentation. For the novice the theory of argumentation may seem more like a network of different disciplines.2 That is not the use Johnson makes of the word “net(work)” but it is a fine metaphor to characterize the multi-disciplinary approach of our discipline. He uses the word to analyze the relationship between the family of terms like reasoning, knowledge, rationality, intelligence

1 See on the topics, the new developments and problems of a theory of argumentation in particular the contributions of R. H. Johnson, D. Walton, and J.A. Blair, in this vol. 2 But that view would disregard the genuine contributions of the theory of argumentation.


5

etc. Maybe that these philosophical family resemblances are one reason, why the theory of argumentation is not too popular in the social sciences. Suppose you are well acquainted with the family resemblances in the argument/reasoning community and you have identified an argument. That must be an easy task – at least under normal conditions – because taking part in a discussion and replying to arguments is a standard social activity. Probably your next step would be to identify certain parts of the argument. You might reason: That must be easy too, because if I can identify an argument, I can identify the parts of an argument. But it is not. Arguments like many other abstract entities are very difficult to break up into smaller pieces, compare the difficulties we have in linguistics with sentences and their parts. Thereby the subject of theorizing shows us a holism like in the semantics of natural language (D. Davidson). It is irritating – not only for the novice – that a reconstructed argument – the real structure of the argument – can be very different from the real argument in a text. What are the premises and what exactly links the premises to the conclusion? The cement of the argument1 which glues together premises and conclusion(s) is at the center of an intensive debate – see also Gilbert Harman (1986), he suggests it is a mistake to “conflate” reasoned change of belief with argument and proof, Don S. Levi (1995) rejects the assumption that arguments are sets of premises and conclusions, Michael Gilbert (1997) argues for a radically non-standard, non-linear, and non-logical conception of “coalescent argument” which is needed to explain everyday argumentation. In his contribution, Leo Groarke presents a wonderful parable about a hedgehog and a fox. The fox uses 36 tricks to escape his enemies, the hedgehog has only one, but this one and only trick works perfectly. The fox symbolizes of course the type of argumentation theory, which uses many type of glues while the hedgehog has only logic to tie premises and conclusion together. To understand the hedgehog in his “one glue fits all” theory, lets look at an example: (P1) John is a mountain climber. (P2) All mountain climbers are well trained. Therefore (C) John is well trained.

1 This is a variant of J.L. Mackie’s, Cement of the Universe, Oxford, 1974, which refers to causality.

6


This is a perfect argument in the following sense. Since the conclusion is a logical consequence of the premises, you cannot find a better glue between premises and conclusion. Yet suppose you try to find a better connection: (P1) John is a mountain climber. (P2) All mountain climbers are well trained. (P3) If (P1) and (P2), then (C) John is well trained. Therefore (C) John is well trained. Someone who tries to find an implicit premise would have misunderstood the nature of logical inference. There is nothing hidden and there is nothing to make explicit. The same is true for the next trial, the change of a (deductive) logical conclusion in an inductive one. Such an argument would be absurd: (P1) John is a mountain climber. (P2) All mountain climbers are well trained. (P3) In all known cases: If (P1) and (P2), we had (C) John is well trained. Therefore (C) John is well trained. The moral is: A deductive logical argument is complete since the conclusion follows from the premises. Any non deductive argument can be turned into a deductive logical argument, by adding an implicit premise. It is this asymmetry which seem to give the hedgehog his superiority over the fox. Groarke’s point is not, that there are no other types of argument. You cannot change an inductive argument into a deductive one by any trick, if you consider the force of the argument. But is the distinctive property of deductive arguments that defines a challenge: Are there arguments which cannot be reduced to deductive arguments in the sense Groarke describes? Groarke illustrates his discussion of persuasion with visual images which are normally taken as paradigm examples of non-logical persuasion. And he argues that the discussion of argumentation, reasoning and persuasion in a variety of disciplines much too quickly dismisses the classical approach that he proposes. The argumentation fauna is inhabited by much more foxes than hedgehogs. Probably Groarke is the only hedgehog in our collection. J. Anthony Blair discusses the questions that a complete theory of inference/argument schemes should answer. His aim is a metatheory of inference/argument schemes. But what makes non deductive arguments so attractive? Why does the 36 – 1 tricks attract so much attention?


7

Blair gives an answer by looking at an very important class of arguments. It is no exaggeration to say, that this class of arguments is by far the largest. That an argument as presented in a text does not meet the standards of deductivism is of course no problem for Groarke. What Blair intends to show is, that you cannot turn those arguments into deductive ones without changing their very nature.1 Here is a realistic example: (P1) In Kosovo people are murdered by paramilitary forces. Therefore (C) The use of military force for the protection of the people is legitimate. This argument seems to be a good one, many would agree. But you cannot deduce (C) from (P1). Arguments of this type are called presumptive.2 Blair gives a lot of examples for this type of argument. You can of course add a premise, which makes the argument deductive and turns a presumptive in a deductive argument: (P1) In Kosovo people are murdered by paramilitary forces. (P2a)If in Kosovo people are murdered by paramilitary forces then the use of military force for the protection of the people is legitimate. Therefore (C) The use of military force for the protection of the people is legitimate. Now the premises entail the conclusion but (P2) is too special. It is not the added premise which gives the argument its force. If you ask someone “why do you believe (C) given (P1), (P2a)” will not be the answer. Let’s try a more general solution: (P1) In Kosovo people are murdered by paramilitary forces. (P2b) If people are murdered by paramilitary forces – in Kosovo or else-where – then the use of military force for the protection of the people is legitimate. Therefore (C) The use of military force for the protection of the people is legitimate.

1 But Groarke might still insist, that his framework is the best way to analyze arguments. 2 This is a rather rough classification. Presumptive arguments are those where you jump from the premises to the conclusion(s). The exact nature of presumptive arguments is the focus of an intense debate.

8


(P2b) is a much better candidate for the answer you will get, when you ask someone why he believes (C) given (P1). But (P2b) is too general; it is easy to find counterexamples. The problem is to find a missing premise which belongs to the argument – i.e. the added premise must be an arguers premise – and not only a missing premise for the sake of deductive argument. You will find presumptive arguments in virtually any text. The popularity of this argumenttype poses a serious challenge to the theory of argumentation: What is the glue between premises and conclusion in a presumptive argument? We can phrase the question in another way: How can we differentiate between a good presumptive argument – where the premises support the conclusion – and a bad one – where this is not the case? This leads us to Robert C. Pinto’s comment to Blair’s analysis of argument schemes referring also to Walton’s account. Pinto evaluates Blair’s concept of presumptive reasoning. He agrees with Blair’s that we do not evaluate all given inferences with our traditional logical procedures, but he makes an important difference concerning the use of argument schemes. Argument schemes in presumptive reasoning in contrast to those used in deductive reasoning have no normative force. To conform to an argument scheme is a necessary condition for the correctness of a presumptive argument, it shifts the burden of proof, but it is not a sufficient condition for its correctness. Douglas Walton’s paper raises many fundamental questions of argumentation theory and that is no surprise, since Walton has published extensively on nearly any aspect of the subject. He outlines a new dialectical framework – called New Dialectic – for the evaluation of arguments. Having its origins in the old dialectic of the ancient world, the new dialectic is centrally concerned with presumptive reasoning of a kind that is not well represented by either the forms of deductive or inductive reasoning. To represent the forms of argument of this third type of reasoning, argumentation schemes are used that model the structures of common types of arguments like argument from sign, argument from analogy and argument from commitment. We will concentrate our discussion on this type of argument and include some of the results published in his book on presumptive reasoning.1 For a social scientist, the inference from correlation to causality defines an important class of presumptive arguments. Look at the following example: In a certain 1 D. Walton, Argumentation Schemes for Presumptive Reasoning; Mahawh, N.J., Erlbaum 1996. A book in German language on the subject is by M. Kienpointner, Alltagslogik: Struktur und Funktion von Argumentationsmustern, Stuttgart 1992.


9

population 50% of all woman and 20% of all men get low payment. Using the terminology of social science, we can say, that there is a correlation between sex and payment. You want to infer that sex is the cause of the difference in payment, but a social scientist knows that this inference is defeasible. We can test the inference by asking critical questions.1 Walton specifies 7 critical questions: Q1: Is there a positive correlation between A and B? Q2: Are there a significant number of instances of the positive correlation between A and B? Q3: Is there good evidence that the causal relationship goes from A to B, and not just from B to A? Q4: Can it be ruled out that the correlation between A and B is accounted for by some third factor – a common cause – that causes both A and B? Q5: If there are intervening variables, can it be shown, that the causal relationship between a and B is indirect – mediated through other causes –? Q6: If the correlation fails to hold outside a certain range of causes, then can the limits of this range be clearly indicated? Q7: Can it be shown that the increase or change in B is not solely due to the way B is defined ... or classified? Two of the seven answers are easy: Q1 “yes” and Q7 “yes”. The answer to Q2 is “yes” too, but with a qualification. If we compare the results of different studies, we will find a considerable variance between different correlations. In few cases, we may even find, that woman are better paid which implies a change in the sign of correlation between sex and payment. This variance poses a serious problem which might be cured by an answer to Q4. But any attempt to find a third factor – or many third factors – will not lead to a stable correlation. We simply do not have theories in the social sciences, which can precisely give us the factors to control. But we have some knowledge how to influence the correlation between sex and income – and do that all the time, at least we try to change it. But since a causal relation cannot be changed by any means, we probably do not have a causal relation between sex and income. Why should the non existence of a causal relationship between sex and income be of any methodological importance in the theory of argumentation? The problem is that over and above the criteria given by Walton – or any other

1 These critical questions belong to an argumentation scheme. Walton discusses 25 of these schemes in his book in the framework of his dialectical approach to argumentation.

10


criteria – you need some substantial field dependent knowledge1 to judge the correctness of this type of presumptive argument. If this is correct, we will never – in principal – be able, to have an “Alltagslogik” as a kind of generalized logic to judge the correctness of an argument. Argumentation schemes are of great value for a practical criticism of arguments. But one fundamental difference between logic and Alltagslogik remains. Logic works in any context, Alltagslogik cannot be used, without considering the context of an argument.2 We can differentiate between two hypotheses: (H1) To judge the correctness of a presumptive argument you can use a set of (sufficient) rules, which define a reasoning scheme. We do not know at present if there are 25 or 250 schemes, but research in the theory of argumentation will find these schemes.3 (H2) Schemes for presumptive reasoning are of great pragmatic value for judging arguments, but you can never give a set of sufficient rules. You will – always? / in a certain percentage of cases? – need some domain specific knowledge which cannot be captured in rules. The difference between H1 and H2 may be residual importance from a practical point of view. But if one looks at the possible shapes of a theory of argumentation, the difference is large and reflects another difference: Between those who think, that human rationality consists of a set of rules – or algorithms, if you like – and those who deny that. The slow development of Artificial Intelligence seems to support the skeptical position.4 Manfred Kienpointner responses to Walton’s conception to outline an alternative against the classical scientific philosophy and contemporary postmodernism5, and evaluate his New Dialectic. He emphasizes the fruitful frame 1 This term has been coined by S. Toulmin (1958). 2 On the foundation of contexts in the theory of interpretation see G. Preyer, “Interpretation and Rationality: Steps from Radical Interpretation to the Externalism of Triangulation, Protosociology Vol. 11, 1998: Cognitive Semantics II - Externalism in Debate. 3 C. Lumer has defended an algorithmic/logical theory of argumentation in Praktische Argumentationsrheorie, Brauschweig 1990. On Lumer’s account D. Mans, “Argumentation im Kontext, Exkurs: Zu C. Lumers “Praktischer Argumentationstheorie”, in: G. Preyer, M. Ulkan, A. Ulfig, Intention - Bedeutung - Kommunikation. Kognitive und handlungstheoretische Grundlagen der Sprachtheorie, Frankfurt am Main 1997. 4 One might reply, that argument schemes can never give sufficicient conditions. They can only indicate, when to shift the burdon of proof. But I doubt that you will get very far with that substitution. We might still ask “how many rules do you need in a certain context to shift the burdon?”. You can’t escape contextuality. 5 On modernism and postmodernism see also G. Preyer “Moderne and Postmoderne im Kontext von Globalisierung”, http://www.rz.uni-frankfurt.de/protosociology


11

of Walton’s research but articulates also some reservations toward his criticism of ancient dialectic, and the conceptualizing of presumptive and abductive arguments. Cartesian epistemology has rejected the “argument from authority” in principle, and the refusal of this argument takes effect in modern thinking. From this point of view the judgement of other people can not be the foundation of knowledge. Yet central parts of our knowledge is not grounded in personal experience but relies on expert knowledge. One of Walton’s 25 argumentation schemes1 is the argument from expert opinion or appeal to authority. Chris Tindale takes up Walton’s argumentation scheme but his contribution discusses the appeal to authority from a different and surprising perspective. Suppose you are told by a medical expert, that taking a certain sort of pills will cure your disease without serious side effects. You – as a layperson – can believe the expert or you can begin to investigate his claims. In most cases – since your no pharmacologist - the second route is not open for you. You cannot think through the matter that obviates any appeal to authority. But frequent appeals to authority reduce your autonomy. This cognitive threat to our autonomy by our dependence on authority raises some questions: What is the relation between my cognitive autonomy and the cognitive autonomy of other people? How is our empirical knowledge formed by the fact, that we are social beings? And what are the necessary cognitive capabilities to create and maintain our autonomy? Tindale reasons convincingly, that testimony is a key to answer these questions. My knowledge about Patagonia for example, can go back to two sources: Personal experience, and testimony by others.2 To criticize their testimony, I have to develop a certain cognitive competence. Since testimony seems to be a more fundamental source of knowledge than expert opinion, this cognitive competence is the kind of competence we need to criticize expert opinion. Tindale names basic skills in scientific methodology, reasoning and ethics. His contribution is a further example for the strong connections the theory of argumentation has with other disciplines especially with philosophy. A trained philosopher might not be surprised, that the basic cognitive mechanism of acquiring knowledge are of great importance for the theory of argumentation. But all others will benefit as well from his intriguing and thorough discussion of testimony.

1 D. Walton, op.cit. 1996. 2 To simplify the matter we disregard, pictures, films etc., since it is not easy to separate these sources from testimony.

12


John Woods’ contribution is about abduction or abductive inferences. Abduction is a sibling of induction and deduction and Peirce is its philosophical father. If a bag contains red marbles and you take out one marble, you may infer that the marble is red. This is deduction. If you do not know the color of the marbles in the bag and take out one marble and it is red, you may infer that all marbles in the bag are red. This is induction. But if a bag of red marbles is standing at some place and a red marble lies in the vicinity of the bag, you may infer, that the marble is from the bag. And that is abduction. What you infer from the premises – “A bag of red marbles is standing at some place.”, “A red marble lies in the vicinity of the bag.” – is the conclusion – “The marble is from the bag.” The hypothesis “the marble is from the bag” could serve as part of an explanation for the fact, that a red marble lies on the floor. As you will have noticed, abductive reasoning is a special kind of presumptive reasoning.1 Abductive reasoning is part of our everyday life – if we are looking for explanations - and it is very essential part of any science. Finding explaining hypotheses is one of the most challenging creative task and that may be the reason why stories like those about Newton’s apple are so popular. Abduction is an inference to a potential explanation. Those sociologists who are engaged in hermeneutic text interpretation know the relevance of abduction and its problems too. We will illustrate the point with one of Peirce’s examples. You find a stone which looks like a petrified fish far in the interior of the country. You can explain the finding by any of the following hypothesis: a b c

The sea once washed over this land. John Doe has left the petrified fish at that place to fool me. The petrified fish was transported by some geological process to the location.

How many potential explaining hypothesis exist? There are good reason that one will always find an additional potential explanation beyond a given set of potential explanations. Thus there is a potentially infinite set of potential explanations. How do you rule out a potential explanation as incorrect – or probably incorrect –? Just take same deductive consequences with an empirical content and test them. In real life situations very few from a small set of hypothesis will survive – in many cases there is only one hypothesis to test. That very abbreviated 1 If you take your first steps in the theory of argumentation, you are advised to differentiate carefully between explanations and arguments. Abduction ties both together, which indicates that you have left the initial stage.


13

description points to three ingredients in any theory of abduction. The first ingredient is a hypothesis generator, the second ingredient is a generator of testable deductive consequences and the third ingredient is theory of refutation. Even the best understood ingredient – the second one – is very difficult to formalize, because only a small subset of deductive consequences is “empirically interesting”. If you still ask yourself: “what makes abduction so terribly difficult, after all, even the layperson does it all the time?” you should carefully read the first sentence of Woods’ paper. Any research on the computational aspects of abduction requires a high degree of formalization. So Woods’ contribution touches an ambitious aspect of argumentation theory, namely the goal to clarify crucial concepts with formalized methods. The explication of creativity within a computational framework would have far reaching consequences over and above the theory of argumentation. Henry W. Johnstone’s contribution reminds us, that reasoning about reasoning has a long tradition. Probably men begun to use deductive and non deductive arguments not too long after they begun to use language.1 But the logic of Aristotle has been for a long time the only tool for the analysis of arguments. Therefore we are pleased to include a paper with a strong reference to one of the most influential of all philosophers. Nonetheless our edition may have convinced some readers that modern theory of argumentation is more than just a footnote to Aristotle.

Dieter Mans, Gerhard Preyer J. W. Goethe-University, Frankfurt am Main, Germany

1 We are not aware if the search for “the oldest published argument” is a topic in the community. Hints are welcome.

14

Ralph H. Johnson

RALPH H. JOHNSON Reasoning, Argumentation and The Network Problem I. Introduction The question is: What is reasoning? Why is this question important generally and to the cognitive sciences specifically? One strong incentive to ask this question has come from the increased attention to the importance of teaching reasoning, which naturally leads to the question of just what reasoning is. (For discussion of this entry point into the question, see my paper “Reasoning, Critical Thinking and The Network Problem” (1996).) This question is also of great interest to the cognitive sciences, because one important tributary that flows into cognitive science is research on AI which historically has been interested in developing computer models to simulate reasoning. But what is it that they are simulating? It is one thing to develop programs that model deduction, quite another thing to write programs which will model argumentation. It is clear that how we answer this question hinges on careful delineation of the term reasoning. However, here we encounter the problem that there is no uniform understanding of this term. As we shall see, many have adopted the view that reasoning can more or less be equated with inferring; others see reasoning as a change in view. I will argue here for a third view. This same kind of divergence attends the term argument. To illustrate, the following specimen is classified by some logicians as an argument: (S1) If the moon is not made of green cheese, then arithmetic is complete. The moon is not made of green cheese. So arithmetic is complete. So also: (S2) The sky is blue. Grass is green. Therefore, it is not the case that tigers are carnivorous.1 Other logicians, myself included, demur. Probably some would include as arguments the following: 1 See Lambert and Ulrich (1980) and Govier (1987:182).

Reasoning, Argumentation and The Network Problem

15

(S3) 99% of ravens are black. That bird is a raven. Probably, that bird is black. (S4) Tweety is a bird Probably, Tweety can fly. Now I although I do not challenge the view that (S1) through (S4) are specimens of reasoning (though the credentials of (S1) and (S2) are somewhat suspect), my sense is that it is mistaken and unhelpful to classify them as arguments. The issues I want to address in this paper is the nature of reasoning, and the relationship between reasoning and argument. If we are going to get a clearer read on these issues, we are not going to be able to do without some awareness of the multiple meanings associated with these words. For, on the one hand, there are legitimate senses of both reasoning and argument in which they are virtually synonymous. Thus, one speaks of a line of reasoning and more or less means an argument. On the other hand, there are uses of both terms according to which they have quite different meanings. Such divergent usage means that any position on this issue should be taken as stipulating how these terms can best be used and should provide appropriate rationale for such stipulations. But there is more, for these terms not only have a close association with one another; they have close associations with other terms, so close that it does not seem feasible to give any significant account of one of them in isolation from the remainder. I refer to this situation as The Network Problem. Let me now turn to it. II. The Network Problem Once we begin to focus our attention on reasoning, we open up another a host of questions such as: What is the connection between reasoning and rationality? What is the relationship between reasoning and intelligence? reasoning and knowledge? reasoning and thinking? These terms seem to form a family or network in the sense that achieving clarity about any one of them appears to be dependent, to some degree, on seeing it in relationship to the others.1

1 The situation here is not unlike the way that Grice and Strawson (1976) construe the situation of terms like “analytic” and “synthetic” which they see as part of a larger family (or network) which includes “a priori,” “contingent” etc.

16

Ralph H. Johnson

reasoning knowledge intelligence

rationality thinking

argument THE NETWORK PROBLEM-1.1 Much work has been done on the theory of knowledge, the theory of rationality, the theory of intelligence, but very little work has been done on the theory of reasoning, though there are indications of its emergence beginning with Finocchiaro (1984). The term argument itself is closely connected to a family of terms: argument inference

implication

reasoning NETWORK PROBLEM-2 In my view, inquiry into both forms of The Network Problem is the task of the theory of reasoning. I shall not attempt to solve any of the three forms of that problem in this paper. I will however be setting forth my position regarding third form of the problem. Hence this paper is intended as a contribution to the theory of reasoning. In the next section, I review the positions of some prominent logicians on the nature of reasoning.2

III. The Nature of Reasoning: Some Positions A. Introduction The key issue for the theory of reasoning is to get clear on the nature of reasoning itself. In this section I examine the views of Finocchiaro, Walton and Harman indicate why I find them inadequate. That will pave the way for my attempt in Section IV to provide a more adequate account.

1 See my (1996) where I identify two forms of The Network Problem, the first of which has to do with critical thinking and problem solving. It is the second form of that problem that we are concerned with here, but to avoid confusion, I shall refer to it here as The Network Problem-1, to distinguish it from yet another form of the problem, to be introduced shortly. 2 I shall sometimes phrase the problem in the material mode – as the nature of reasoning, and sometimes in the formal mode – the meaning of reasoning (“reasoning”). (For the distinction between the formal and material modes, see Carnap (1951).)


17

B. Finocchiaro on reasoning In “Informal Logic and the Theory of Reasoning” (1984), Finocchiaro called for the development of the theory of reasoning and in the course of that provides his own conception of reasoning. The focus of Finocchiaro’s inquiry is reasoning rather than argumentation. This is a deliberate move, made so as to include “besides the study of arguments, such activities as problem-solving, decision-making, persuasion and explaining which cannot be equated with argumentation but which may involve reasoning in an essential way” (3). Here Finocchiaro endorses the view that reasoning and argument are different, and seems to be advocating for a broad conception of reasoning – one which makes evident the connection between the theory of reasoning and The Network Problem-2 above. Finocchiaro now adds a series of clarifying comments about his definition of the theory of reasoning, after which he anticipates and responds to four objections, each of which challenges the philosophical legitimacy of the theory of reasoning. It is the first objection – the one he calls “the most fundamental”– that is crucial for our purposes. It is that no such subject matter (as reasoning) really exists. This objection, says Finocchiaro, is based on the view that reasoning is “an epiphenomenal illusion deriving from using a general label to refer to a number of disparate activities” (4). From the vantage point of such a conception, it would seem to follow that “a theory of reasoning per se, as distinct from theorizing about particular instances or types or fields of reasoning, makes no more sense that a theory of success in general” (4).1 Finocchiaro’s rejects this view of reasoning and proposes, as a countersuggestion, that “the essential feature of all reasoning is the interrelating of individual thoughts in such a way that some follow from others and that the normal linguistic expression of such interrelated thinking involves the use of particles like ‘because,’ ‘therefore’” (4). Finocchiaro refers to this as a “minimal conception,” by which I think he means that he is aiming just to get at the basics. Even so, my chief reservation about this way of conceiving reasoning is that makes inference to be the core of reasoning. For “the interrelation of thoughts so that some follow from others” could serve equally well as a definition of inference, and the particles he mentions typically attend inferences (and perhaps arguments). I refer to this view as inferentialism and discuss it later. That then is the first problem with his minimal conception. I see two other problems. First, what happens when the reasoner thinks that one thought follows from another but it does not follow, as happens 1 Finocchiaro is here rejecting an approach to conceptualizing reasoning that approximates the one I shall later present.

18

Ralph H. Johnson

when the reasoner produces what is called a non-sequitur? Is such a specimen to be categorized as reasoning? I think it should be, but it would not qualify under Finocchiaro’s definition because the one thought does not follow from the other, though the reasoner thinks it does. As stated, then, the definition is too narrow because it would rule out what seem like clear cases of (bad) reasoning. Finocchiaro can modify his concept so that it read s”the interrelation of thoughts some of which are believed to follow from others.” Still he faces the question of what it is for one thought to follow from another? That is the second problem. If one starts down this road, one is immediately going to think of deduction, where one proposition is deduced from another. But this would be much too limited as a definition of inference, never mind reasoning. Thus, he would have to broaden the notion of inference to allow for a “weaker” kind of “following from.”1 Even if he makes this adjustment, he remains within the grips of the idea that reasoning is essentially inference. Not only are we then saddled with the problem of defining inference, which is task that must be embraced at some point, but the conception of reasoning has been narrowed unduly. I conclude that Finocchiaro’s minimal conception won’t satisfy our demands and turn to another account, that of Walton. C. Walton on reasoning and argument (1990) In “What is Reasoning? What is an Argument?” (1990), Walton raises a number of important questions pertinent to our inquiry. He writes: It is problematic to see how reasoning is related to argument. Are reasoning and argument essentially the same thing? Or is one a proper subpart of the other? Or can you have reasoning that is not in argument? It seems hard to know where to begin replying to these questions. (400)

Walton begins by citing Govier’s way of demarcating between argument and reasoning: An argument is a publicly expressed tool of persuasion. Typically it takes thinking to construct an argument. Reasoning is distinguished from arguing along these lines: reasoning is what you may do before you can argue, and your argument expresses some of your best reasoning. But much reasoning is done before and outside the context of argument. (401)

I have some difficulties with this view. First, Govier seems to equate thinking and reasoning, whereas I will later argue that it is worthwhile to distinguish

1 In private conversation, Finocchiaro has said he would now change his definition and use the notion of “depend upon” rather than “follow from” in which case my objections here will lose their force.


19

them. Second, in Govier’s view, reasoning generally takes place before and outside of the context of argument, which suggests that argument is not for her a kind of reasoning. This raises the question of exactly what Govier means by argument. If argument is not itself reasoning, then what is it, and how does it stand related to argument? Once again we are dealing with the second form of The Network Problem, and we see how much hinges on how we construe the fundamental terms here – argument and reasoning. Walton begins his own account by citing the definition given by Angeles in his Dictionary of Philosophy (1981), according to whom there are three senses of reasoning, the first of which is “the process of inferring conclusions from statements.” Walton says: “The first definition is positive and does seem to capture the basic idea behind reasoning” (401). (The second and third senses do not enter into discussion and so I omit any discussion of them here.) Thus Walton also seems to endorse an inferential approach to reasoning. He adds a way of distinguishing the two: inferring leads from premises to conclusion, but we may reason not only to propositions but also to actions. Walton likewise agrees with Angeles that reasoning can be used for a variety of purposes. (Note that argument and reasoning are here conceived as non-equivalent terms.) Following further reflection on Angeles’s definition, Walton arrives at the view that “Reasoning is the making or granting of assumptions called premises (starting points) and the process of moving toward conclusions (end points) from these assumptions by means of warrants” (403). Walton perhaps thinks that this statement lines up with his earlier statement, but I detect some nonnegligible differences. Specifically, while this second version remains inferentialist, it has begun to move the definition of reasoning in the direction of argument. Indeed the terminology Walton employs – premise, warrant, conclusion – is the language of argument, especially reminiscent of Toulmin’s approach (1958, 1979) to the structure of argument (or reasoning). Walton offers us two somewhat different definitions of reasoning, one of which seems clearly inferentialist (i.e., construes reasoning as inferring), while the second leads in the direction of construing reasoning as argument. Walton’s account thereby reflects the tension and the ambiguity I am confronting in this paper. What is needed is an account of reasoning that will at least initially set it apart from inferring and arguing. But before I turn to that matter, I want to pause to say something about the position I have labeled inferentialism.

20

Ralph H. Johnson

D. About Inferentialism I have claimed that both Finocchiaro and Walton adopts an inferentialist view of reasoning. It is time to say more about just what that means. By “inferentialism,” I mean the view that reasoning is essentially inference (or else reducible to it). This the bias in favor of inference (particularly deductive inference) as the paradigm is evident in much of the empirical-psychological research. A classic study, Human Reasoning (1978) by Revlin and Mayer, contains no explicit definition of reasoning, but appears to operate on the assumption that reasoning is essentially inference. In the Introduction we read: The present volume contains a series of manuscripts concerning current research methods, results, and theories on human reasoning. While this collection does not represent an exhaustive survey of research on human inference ... (7)

Note how readily and naturally Revlin and Mayer move from discussing reasoning to inference, as though they were pretty much the same thing. The papers in Human Reasoning almost all take syllogistic reasoning to be the most representative form of reasoning. So although the title of the book refers to reasoning, all of the articles deal with inference and the vast majority with deductive inference. This is also true Wason’s landmark study The Psychology of Reasoning (1972) which with no less justice might have been titled The Psychology of Deductive Inference. Not just inference, then, but deductive inference has enjoyed pride of place when it comes to theorizing about reasoning. There are many reasons for this focus. Aristotle’s syllogistic laid the foundations for logic – for what we might call the systematic normative study of reasoning. The term “syllogismos” in Greek could equally be rendered by the term “reasoning.” Even though Aristotle’s own work on reasoning far outstrips his work on syllogism in Prior and Posterior Analytics, still many retain the belief that deductive inference is either synonymous with reasoning or else certainly among the most central forms. In psychological research on reasoning, the tendency to focus on inference is perhaps explained both by historical influence,1 as well as by the fact that clear cut answers exist to the question what follows from what, so that empirical studies can be undertaken – something less easily accomplished with argumentation and explanation. Further, there is no denying that inference is prominent in many of our cognitive endeavors. The last 20 years have witnessed tremendous advances in cognitive psychology and the study of reasoning. There has been a marked movement to broaden the scope of psychological research to include other forms of inference. 1 Piaget believed that “reasoning is nothing more than the propositional calculus itself” (12).


21

Here one thinks of Human Inference (1980) by Nisbett and Ross and also Judgement under Uncertainty (1982) by Kahneman, Slovic and Tversky. Yet the fixation on inference abides. In his book, Mental Models (1984), Johnson-Laird sets forth “seven goals for evaluating any theory of reasoning.” A careful read will show that in his formulation of these goals, Johnson-Laird is thinking not so much of reasoning as inference.1 The thought here is that we can benefit from a wider and different view. One author who has offered a different approach is Harman. I turn next to a somewhat brief consideration of his views. E. Harman on Reasoning In Change in View (1986), Harman develops the view that reasoning can be characterized by the idea of “change in view” (belief, intent). However, I believe that this way of characterizing reasoning is both too broad and too narrow. It is too broad because unless one specifies that the change in view must be the result of reasoning (in which case the definition becomes circular), there are many instances where people have changed their view but have not reasoned. For instance, over time, someone’s view may change with any explicit attempt at modification, simply by the press of circumstances. Someone might say: “I used to think that .... but now I don’t.” And when asked “What changed your mind?” he or she might reply: “I’m not sure.” In short sometimes our views change because we have reasoned and thought about them; and sometimes they change without direction attention or focus. Harman’s conception is too narrow because although reasoning often does produce a change in view, it can just as often produce a reinforcement of the view. That is, I may review the arguments offered for a position and decide that those arguments do indeed support that position. I have engaged in reasoning but my view has not changed. Harman’s approach, though a welcome break from the tradition of inferentialism, seems open to serious criticism. 1 The criteria given by Johnson-Laird are as follows (65-66): 1. A descriptively adequate theory must account for the evaluation of conclusions, the relative difficulty of different inferences, and the systematic errors and biases that occur in drawing spontaneous conclusions. 2. The theory should explain he differences in inferential ability from one individual to another. 3. The theory should be extensible in a natural way to related varieties of inference rather than apply solely to a narrow class of deductions. 4. The theory should explain how children acquire the ability to make valid inferences. 5. The theory must allow that people are capable of making valid inferences, that is, they are potentially rational. 6. The theory should shed some light on why formal logic was invented and how it was developed. 7. The theory should ideally have application to the teaching of reasoning skills.

22

Ralph H. Johnson

Without wishing to claim completeness for this inventory of how philosophers and logicians have construed reasoning, we have seen that there are problems with the inferentialist accounts offered by Finocchiaro and Walton, and with Harman’s alternative. In the next section, I present my own account. IV. Reasoning and Argument A. Reasoning The position I advocate attempts to adhere to the sense of the word common in everyday discourse, where reasoning would likely mean something like “figuring it out.” Taken this way, reasoning would be almost the same as thinking. However, I believe there is something to be gained by distinguishing the two.1 I propose that we take thinking to be the having and/or processing of mental representations. Thus, daydreaming would be a case of thinking but not reasoning, because it lacks the direction associated with reasoning. Similarly, remembering a joke would be thinking, but not reasoning. Thinking thus would be the genus of which reasoning is a species. To understand reasoning and set it apart from other modes of thinking, I would begin with the nominal conception already stated: reasoning is the seeking and/or the giving of reasons. One who seeks reasons is reasoning, as is one who gives reasons. However, we humans do not simply seek and have and give reasons. We seek reasons for a reason, and we give reasons for a reason. Thus, what distinguishes thinking (the having of representations – like daydreaming) from reasoning (e.g. investigating the role of daydreaming in our emotional lives) is that in the latter case the representations are controlled by purpose. Reasoning then is thinking directed by purpose.2 In line with this admittedly broad specification, it seems to me that the following specimens all qualify as instances of reasoning. 1. When I find a puddle of water in my basement and seek an explanation for it, I am looking for a reason. I am reasoning. When I then give that explanation to my wife (“it’s just water from the pipes because they are sweating”), I am giving her my reasons. I am reasoning; so explaining is one type of reasoning. 2. When I predict the outcome of the Super Bowl game by figuring out which team has the best chance of winning, I am finding reasons. I am reasoning; so predicting is another type of reasoning.

1 Max Black (1945) also makes this distinction, though not in quite the same way. 2 I am reminded here of L. Susan Stebbing’s phrase – “thinking to some purpose.”


23

3. When I make a controversial assertion and back it up with reasons, I am giving reasons; I am reasoning; so asserting is another type of reasoning. 4. When I develop an argument to support my view that Canada should not commit forces to an invasion of Iraq, by considering possible objections and replying to them, I am presenting reasons to support my position. I am reasoning; so arguing is another type of reasoning. 5. When I offer a definition or clarification of the term “pornography” (against the background of the current debate), I am finding reasons to think of “pornography” in a specific way. I am reasoning; so defining and clarifying are types of reasoning. 6. When I infer that we must do our shopping on Friday because it has to be either Friday or Saturday but Saturday is out, I am reasoning; so deducing is another type of reasoning. And so on. All of these are instances of purposive thought in which the finding and giving of reasons plays a crucial role. All of these are instances of reasoning. Now either all of these operations are essentially inferential in character; or, there is some operation – reasoning – that is individuated in each of them (this seems to be Angeles view); or else there is some third possibility, which is what I hold to be the case. Even if one wishes to claim that inferring occupies pride of place in one’s account of reasoning; i.e., that defining, persuading, predicting, arguing are essentially inferential in character, that will have to be shown by some form of argumentation. I am doubtful that such an argument can be made successfully, but the possibility must be admitted. The second alternative – to hold that reasoning is something over and above these instances in which it is embodied – also needs to be shown. Anyone minded to give such an account owes us an explanation of what reasoning is such that it is different from all of these, and yet instantiated in each of them. I here defend the third alternative, which is close to what Finocchiaro termed the “epiphenomenal” position. Even though all reasoning may be described as thinking under the governance of purpose, there is no one specific mental activity that can properly be identified with reasoning. Semantically, then, “reasoning” is a generic term. By that I mean that there is no one activity which just is reasoning, full-stop. Rather reasoning is the “name” for a host of different activities.1 If this recommendation makes sense, then we can see why it is a mistake to identify reasoning with inferring, or indeed with any of its 1 This is not unlike Wittgenstein’s position on language which, in Zettel, he talks about “language” as “a name for a collection” (#322).

24

Ralph H. Johnson

species. That would be a mistake comparable to identifying fruits with apples. It may be that apples are the most common fruit and the one that immediately comes to mind whenever we speak of fruits. But it remains a mistake to identify the genus with an of its species. Just as there is nothing which is fruit, over and above its various species, so too there is nothing that is reasoning, over and above its various instantiations. From my account it follows that while all inferring is reasoning, the reverse is not true. There are forms of reasoning which it does not make sense to identify as inferring. Pre-eminent among them is argument. I cannot here present a complete account of argument, but I can at least attempt to show how I would characterize so as to distinguish it from inference and reasoning.1 B. Argument It is clear that argument, or better arguing, is one type of reasoning. My immediate concern now is to show that argument is not the same as inference. This needs to be shown because in the 20th century there is a strong tendency to merge the two. As evidence of this let me cite the definition provided by Copi in his Introduction to Logic – one of the oldest and best selling logic textbooks in North America. An argument, in the logician’s sense, is any group of propositions of which one is claimed to follow from the others which are regarded as providing evidence for the truth of that one. (7, 2nd ed., 1961)

(In later editions, “evidence” is replaced by “support or grounds” (1986, 7th ed., p. 6).) A few lines later, Copi adds: “An argument is not a mere collection of propositions, but has a structure” p. 7). Here we see the same essential structure and relationship as has been used to characterize inference, and so the impression is created that argument and inference mean pretty much the same thing. To further complicate matters, the notion of “follows from” is ambiguous as between implication and inference, and so we are right smack in the middle of the third form of The Network Problem. So the three terms – argument, inference and implication – have become enmeshed. I cannot here undertake a full attempt to disentangle but let me at least indicate the line of reasoning that I would pursue. To begin, I cite an example of what I take to be clear cut instance of each Implication: Boston is a city and Boston is in the United States. 1 For a better treatment of both argument and the Network Problem, see my Manifest Rationality (forthcoming from Lawrence Erlbaum).


25

Therefore Boston is in the United States. The first statement implies (entails) the second, and this is so quite apart, from whether any human reasoner ever reasoned in just this way. (S1) then is an instance of an implication. Inference: From the fact that the dog did not bark, it follows that the intruder was known to the dog. (Based on Sherlock Holmes) Here the second thought (proposition) is in some sense derived from the first (whether it implies it or not). The reasoner has inferred the second from the first. Argument: The President should be impeached; for he lied to the grand jury; he lied in a deposition in the Paula Jones case; and he has obstructed justice. These are certainly impeachable offenses. The objection that he should not be impeached because the American people don’t support impeachment makes two unwise assumptions: first, that the polls are accurate; second, that representatives are obliged to vote as their constituents say. This is an argument. The reasoner is attempting to persuade someone else of the truth of a particular claim and the doing of this requires evidence or reasons. Clearly the reasoner thinks that the reasons cited are good reasons for the claim and he wants others to think so also. This may be true even when (a) the reasoner has not himself made any inferences, and (b) the statements do not take the form of implicatures. Notice as well that the reasoner has produced not only a line of reasoning that leads to the conclusion but has as well attempted to defuse a possible objection. Let me now return to the examples with which I began. (S1) If the moon is not made of green cheese, then arithmetic is complete. The moon is not made of green cheese. So arithmetic is complete. No one who is serious about arguing that arithmetic is complete would produce such a specimen of reasoning to make his point or to attempt to persuade others. I suspect that no one has ever drawn such an inference. In fact, such specimens as (S1) and do not originate in the real life practice of argumentation but rather are artificial specimens constructed to illustrate a point about the propositional calculus – here that the validity of an implication does not establish the truth of the conclusion. (It also illustrates the relationship that

26

Ralph H. Johnson

some have called material implication.1) (S3) and (S4) might well be taken as examples of inference. However, they do not seem to be example of arguments, because their conclusions are not the sorts of controversial matters which spark arguments. What about (S2)? It is a strange specimen, no doubt. It does not appear to be an implication, nor an inference, nor yet an argument (though it was been labeled “argument” by the authors who produced it (Lambert & Ulrich, p. 24). Is it then reasoning? I don’t know what to say about (S2). On the one hand, I want to say that it is just atrocious reasoning, which means, of course, that it is reasoning. On the other hand, (S2) does not much resemble any actual reasoning, how any human reasoner would think. It seems like just a series of statements randomly linked by “therefore,” as if someone generated the following sequence of words “windows I furiously spring door at” and asked if these constitute a sentence. Whatever we say about its status, (S2) is a queer specimen. Though we want our theory of reasoning to deliver a verdict on its status, it is not the sort of example we wish our theory to illuminate. It may help if we look at how (S2) came into being. It was invented by Lambert and Ulrich to make a series of points about the difficulties of providing a formal account of informal fallacy. I would say, then, that while it is not itself an example of reasoning (since it does not satisfy the definition produced earlier), it was produced in order to make a point about one particular kind of reasoning. Perhaps then it is best relegated to the category of “thinking” rather than “reasoning.” To summarize this section: By the term argument then, I understand an intellectual product, essentially public and social in character, which seeks to persuade rationally. By inference I understand a movement (of the mind) from one item (usually a thought represented in a proposition) to another, and where the former serves as the basis for and leads to the latter. An inference may come to be the conclusion of an argument, but it need not be. By implication I understand a relationship between two proposition, such that one of them follows from the other(s). An implication may well be nested in an argument but it need not be. An inference can give rise to an argument but is itself a quite different type of reasoning. If I am right, argument cannot be reduced to a concatenation of implicatures or inferences.

1 I here bypass the whole debate about whether material implication is in fact a kind of implication. For discussion of this, see Anderson and Belnap (1975).


27

V. Conclusion In this paper, I have attempted to set forth a position on the nature of reasoning that differs from the dominant view in the 20th century – inferentialism. I have also attempted to provide an account of how reasoning and argument are related. That attempt has involved me in an effort to distinguish argument from inference and implication. For too long the intellectual culture of the 20th century (influenced as it was by the Positivist Ideal) has tended to merge the identity of these three. I hope to have given some basis for distinguishing argument from both implication and inference. All three are important; but they have different conditions and confer different benefits and play importantly different and indeed complementary in the intellectual life of both the individual and the culture. Argument, inference and implication are all types of reasoning. Their full and proper definition and differentiation is the subject of what I call the theory of reasoning. This paper has been an effort to make a contribution to that inquiry. I hope that my attempt to answer these questions, limited and controversial though it is, will suggest to other just how important this inquiry is. References Anderson, A.R. and Nuel D. Belnap Jr. 1975. Entailment: The Logic of Relevance. Princeton: Princeton University Press. Black, Max. 1945. Critical Thinking: An Introduction to Logic and Scientific Method. New York: Prentice-Hall. (2nd ed, 1952). Carnap, Rudolf. 1956. “Empiricism, Semantics and Ontology”. In: Meaning and Necessity. Chicago: University of Chocago Press. Copi, Irving. 1961. An Introduction to Logic. New York: Macmillan. (7th ed., 1986.) Finocchiaro, Maurice. 1984. “Informal Logic and the Theory of Reasoning”. Informal Logic 6, 38. Govier, Trudy. 1987. Problems in Argument Analysis and Evaluation. Dordrecht: Foris. Grice, H. P. and P. F. Strawson. (1956). “In Defense of a Dogma”. In: Rosenberg, J.F. and Travis, C. (Eds.), Readings in the Philosophy of Language, pp. 81-94, Englewood Cliffs, NJ: PrenticeHall. Harman, Gilbert. 1986. Change in View. Cambridge: The MIT Press. Inhelder, Barbel. 1958. The Growth of Logical Thinking from Childhood to Adolescence. New York: Basic Books. Johnson, Ralph H. 1996. “Reasoning, Critical Thinking and The Network Problem.” Chapter 14 of The Rise of Informal Logic. Johnson, Ralph H. 1996. The Rise of Informal Logic. Newport News, VA: Vale Press. Johnson, Ralph H. Manifest Rationality: “A Pragmatic Theory of Argument”. Mahwah, NJ: Lawrence Erlbaum (forthcoming). Johnson-Laird, P. N. 1983. Mental Models: Towards a Cognitive Science of Language, Inference and Consciousness, Cambridge,: Harvard University Press.

28

Ralph H. Johnson

Kahneman, Daniel, Paul Slovic and Amos Tversky. 1982. Judgment Under Uncertainty: Heuristics and Biases, Cambridge, Cambridge University Press. Lambert, Karl and William Ulrich. 1980. The Nature of Argument. New York: Macmillan. Nisbett, Richard and Lee Ross, Human Inference: Strategies and Shortcomings of Social Judgment. 1980. Englewood Cliffs, NJ: Prentice-Hall. Revlin, Russell and Richard E. Mayer, Eds. 1978. Human Reasoning, Washington, D.C.: V. H. Winston, 1978, pp. 4-5. Stebbing, L. Susan. 1939. Thinking to Some Purpose. Harmondsworth (Middlesex): Penguin Books. Toulmin, Stephen E. 1958. The Uses of Argument. Cambridge: Cambridge University Press. Toulmin, Stephen E., Richard Reike and Allan Janik. 1984. An Introduction to Reasoning (2nd ed.). New York: Macmillan. (1st ed, 1979). Walton, Douglas. 1990. “What is Reasoning? What is an Argument?” The Journal of Philosophy 87, 399-419. Wason, Peter C. 1972. The Psychology of Reasoning: Structure and Content, Cambridge: The Harvard University Press. Wittgenstein, Ludwig. 1967. Zettel. Trans. G.E.M.Anscombe.Oxford: Basil Blackwell.

The Fox and the Hedgehog: On Logic, Argument, and Argumentation Theory

29

LEO GROARKE The Fox and the Hedgehog: On Logic, Argument, and Argumentation Theory Introduction According to a traditional Aesop’s fable, the fox once asked the hedgehog how many tricks he knew for outwitting the hunters and their dogs. “Only one” the hedgehog answered. “I know how to play dead.” The fox was aghast at this response and proceeded to boast of the thirty-six tricks he knew. Just as he was finishing, a pack of hounds came over the hill. Immediately, the hedgehog fell to the ground and played dead. The hounds came yelping up to him, poked him with their noses, and ran after the fox. The hunters ignored the hedgehog as they went rushing by. Thirty-six tricks later, an exhausted fox finally managed to escape the hounds. As it was walking back to its lair, it met the hedgehog. “How,” it said, “did you ever manage to elude the hounds and hunters with just one trick?” “I have just one trick,” the hedgehog replied, “but it’s a good one.” It may seem a long jump from this fable to contemporary argumentation theory, but we can make the jump if we properly understand the latter’s aims and its origins. The goal of argumentation theory might be described as a comprehensive understanding of argument and persuasion. This encompasses an understanding of argument in scientific and (less so) mathematical contexts but argumentation theorists have been primarily concerned with the study of argumentation in non-formal contexts – in advertising, political commentary, legal argument, public discourse, advocacy and other forms of everyday discourse. It is tempting to describe such argumentation as “natural language” argument, but this characterization is too narrow, for argumentation theorists also study instances of argument and persuasion which employ non-verbal forms of communication like visual images, physical movement, and other forms of “body language” (see, e.g., Birdsell & Groarke 1996 and Gilbert 1997). Within the last fifteen years, the field has expanded – one might say exploded – to become an interdisciplinary field of inquiry which is influenced by philosophy, formal and informal logic, traditional and contemporary rhetoric, communications studies, linguistics and cognitive psychology.

30

Leo Groarke

In many ways, argumentation theory is a reaction against the concept of argument which characterizes traditional formal logic – a concept which takes as its paradigm argument and proof as they occur in mathematical and scientific contexts. In reacting against this narrow focus, argumentation theorists have maintained that logic’s account of argument is too narrow to explain the broad range of arguments that they study. Toulmin (1958), Mortenson & Anderson (1970), Perelman (1982), Johnson (1996) and innumerable other commentators have, therefore, roundly criticized the account of argument which characterizes traditional logic and formal analyses of argument more broadly. Though they concede the usefulness of this account in a narrow range of contexts, they argue that the attempt to understand argumentation more broadly requires a fundamentally different account of argument. In the present paper I want to question argumentation theory’s attempt to formulate conceptions of argument and persuasion which are fundamentally distinct from the account that is assumed by traditional formal logic. A full defence of the latter is beyond the scope of the present paper, but I will argue that there is no clear evidence which shows that the notion of argument embedded in traditional logic cannot account for the complexities of ordinary argument. For though there are ways in which this notion must be broadened and relaxed if we are to make sense of ordinary discourse, this can be accomplished without countenancing the radical alternatives which have been proposed. In criticizing these alternative accounts of argument I will, in particular, criticize a tendency to promote a view of argument and persuasion which distinguishes a great many categories of argument which require different kinds of analysis. Instead of adopting the diverse notions of argument analysis this implies, I argue that we will do better to retain the one which is implicit in traditional logic. Like the hedgehog in the fable, this approach leaves us with only one trick, but it is a trick which is a good one, and one which promises a more effective analysis of ordinary argument than the many tricks which tend to characterize argumentation theory in its present state of development. Argument Traditionally Conceived We might describe the “traditional” concept of argument simply. It defines an argument as a set of reasons (premises) in support of a conclusion. Sound arguments are arguments with true premises which are deductively valid. A deductively valid argument is in argument in which it is impossible for the premises to be true and the conclusion false. It is in this sense that the conclusion follows from the premises.


31

Though they are often misconstrued in argumentation texts (see Groarke 1999), these are familiar notions. In the context of argumentation theory, the important point is that the notion of argument which logic has assumed is easily detached from formal logic, and is not necessarily tied to the attempt to provide a formal account of argument. Many commentators assume the latter, but this is an elementary mistake. Arguments which are sound and/or valid need not be arguments which are proved so in some standard formal logic. In ordinary contexts, it is often redundant to back such arguments with formal proofs. In other contexts, proofs may be difficult or impossible because there is no formal system which adequately captures the features of ordinary inference an argument depends on. The attempt to create formal systems which better capture such complexities is one of the (many) driving forces which motivates the development of new formal logics. In the present paper, my concern is the concept of argument traditionally assumed by formal logic rather than the attempt to formalize it. In detaching this notion from formal logic and applying it to the kinds of contexts studied by argumentation theory, it must be granted that ordinary argument and persuasion are not governed by the strict standards that characterize argument and proof in formal contexts like science, logic and mathematics. Even in cases of good reasoning, arguments in ordinary discourse usually rely on premises which are acceptable rather than true or certain. This is an inevitable consequence of the fact that the claims (and values) which are the basis of public discussion, political debate, advertising, conflict resolution, etc. are characterized by contentiousness rather than consensus. At least from a practical point of view, a definitive treatment of relevant facts and values is frequently impossible. It follows that argumentation in ordinary discourse must often depend on premises which are acceptable, probable, plausible or likely rather than true or certain. As rhetoricians emphasize, this may also mean that premises may be acceptable or not depending on the audience one addresses. Or as pragmadialectics puts it, attempts to resolve differences of opinion often must be conducted in a way that reflects the beliefs and the attitudes of the specific parties to the dispute. The role that premise acceptability must play in argumentation contexts suggests that soundness sets too crude a standard when we judge ordinary arguments. In the frequent contexts in which we cannot clearly demarcate true and false premises, we are in no position to judge soundness or unsoundness. While this means that we must look for some other paradigm of good argument, it does not follow that the basic idea behind soundness is mistaken, but only that the assessment of ordinary argumentation requires an analogue which understands a good argument as a valid argument with acceptable rather

32

Leo Groarke

than true or certain premises. This is an important change to the traditional account of good argument – most significantly because it allows the same argument to be good in one context and not another – but there is an obvious sense in which it modifies soundness rather than rejects it. Another way in which ordinary argument and persuasion departs from the ideals which motivate traditional formal logic is in their reliance on implicit premises and assumptions. Outside of mathematics, logic, science, and legal and scholarly endeavors, attempts to fully enunciate the structure and content of an argument are rare. Instead, arguers rely on a host of implicit understandings that play a central role in the argumentative exchange. Arguments are forwarded without explicit premise and conclusion indicators. Enthymematic constructions leave crucial premises unstated and rhetorical questions are used to obliquely state conclusions. Methodological understandings are assumed rather than elaborated, loaded terms are used to make implicit judgments, stylistic goals (not a concern for content or explicitness) determine the form of many claims, and sarcasm, insinuation, allusion and irony abound. In some cases, arguments may be entirely implicit because they are not forwarded verbally, but visually or through other non-verbal forms of communication. In many ways, pragma-dialectics sums up these aspects of argumentation when it points out that argumentative speech acts are often implicit and indirect (Van Eemeren & Grootendorst 1992, 44-72). One might easily contrast the implicitness in ordinary argumentation with the extreme explicitness which tends to characterize the arguments traditionally emphasized by formal logic. But it is a mistake to think that this implicitness requires a rejection of the conception of argument which logic traditionally assumes. It implies only that one must analyze ordinary argument and persuasion in a way that recognizes their frequently implicit nature and renders it explicit. The goals that this implies are already assumed in virtually all variants of argumentation theory, which maintain that argument analysis must incorporate some attempt to recognize implicit arguments, premises and assumptions. Even non-verbal arguments can be accommodated in this way. It is in this regard worth noting that traditional logic recognizes that propositions may be expressed in a variety of ways. This recognition is easily expanded to include non-verbal forms of communication. This expansion requires only that we recognize that a proposition may be expressed in ways that do not use language in the ordinary sense.1 To take a pertinent example, the proposition that someone is lying about some crime that they committed can be expressed with 1 I am indebted to Lorne Falkenstein for this point.


33

a variety of sentences (“He’s lying.” “Don’t believe him.” “That’s a tall tale if I ever heard one.” etc.) but also in non-verbal ways which may include facial expressions (the raising of one’s eyebrows accompanied by a look of exasperation) or a caricature which portrays the person accused of lying as a Pinnochio look alike with an absurdly extended nose. So long as we recognize that nonverbal claims of this sort have verbal counterparts we can treat them as propositions in a very ordinary way, and can understand arguments which contain them as arguments in the traditional sense. It follows that the implicit character of much ordinary argumentation does not show that it is impossible to formulate an account of ordinary argument and persuasion which is based on the account of argument formal logic traditionally assumes. If this account is to be challenged by argumentation theory, it must do more than point out the role that implicitness and acceptability play in ordinary reasoning. To see that the challenges which it has proposed are problematic or at least open to debate, we need to consider some of the competing conceptions of argument which argumentation theorists have suggested and the questions which they raise. Argument in Argumentation Theory Though many argumentation theorists have criticized the notion of argument which is embedded in traditional formal logic, most have conceded that it is a powerful tool in the analysis of some kinds of argument. Argumentation theorists who argue that it is too narrow to account for ordinary argumentation therefore grant that the traditional model is useful in the context of logic and mathematics, but maintain that it cannot be usefully extended to other contexts. As a result, argumentation theory frequently endorses an account of argument which distinguishes between argumentation which does and does not conform to argument as it is traditionally understood, and proposes a new model of argument in the latter contexts. When Johnson (1996) criticizes the notion of argument which has been assumed by formal deductive logic (“FDL”), he therefore writes “that FDL had in mind one important subset of arguments, but the realm of argumentation was much broader” (90). Judged from this perspective, the development of argumentation theory has been the development of a typology of argument which distinguishes different kinds of argumentation which are assessed by different kinds of standards. A small subset of the arguments which are studied by argumentation theorists are understood in terms of logic’s traditional model of argument. Other kinds of argument are understood and assessed in ways that employ other models of good reasoning. Deductive arguments are, for example, understood in the

34

Leo Groarke

traditional way, while inductive arguments are understood and assessed in other ways. There is much that might be said about this particular distinction (see below), but it this is only the beginning of the story, for most argumentation theorists have argued that other distinctions are needed if we are to adequately classify the different kinds of argument and persuasion that characterize ordinary discourse. This trend is already implicit in Govier (1987), who criticizes the inductive/deductive distinction on the grounds that it fails to recognize fundamental distinctions between different kinds of non-deductive arguments. As she puts it, “The great divide between deductive and inductive arguments is spurious and theoretically dangerous, because it makes it too easy to ignore the many nondeductive arguments which are not classically inductive.” (1987, 53) The different kinds of argument which have been distinguished by theorists pursuing this line of reasoning are too numerous to be catalogued here, but some examples can be illustrative. In her criticism of the deductive/inductive distinction, Govier distinguishes statistical generalizations, consistency arguments by analogy, “conductive” arguments, and non-conclusive philosophical arguments (1987, 50-51). Authors like Perelman (1982) distinguish between arguments which are dialectical and those which are not, maintaining that the traditional account of argument cannot account for the former. Johnson (1996) distinguishes between inference, which is properly studied by formal logic, and “an argument in the complete sense,” which “can only develop against the background of heterogeneity of point of view and of other arguments” (1996, 94). And Walton distinguishes between different kinds of dialogues which are characterized by different sets of assumptions and procedures (see, e.g., Wlaton & Krabbe 1995; Walton 1998). In its most radical form, argumentation theory proposes kinds of argument which are notable precisely because they constitute a radical rejection of the model which logic has traditionally assumed. Levi (1995) has, for example, argued that it is a mistake to think that arguments in ordinary discourse must contain premises and conclusions. Gilbert (1997) maintains that “the logical mode” is viable in restricted contexts, but argues that we must recognize three alternate modes if we are to understand the full range of ordinary argumentation and persuasion. The “emotional” mode legitimates appeals to emotions, the “visceral” mode is “primarily physical and can range from a touch to classical nonverbal communication, i.e., body language, to force” (84), and the “kisceral” mode relies on “the intuitive, the imaginative, the religious, the spiritual, and the mystical” (86). In attempting to understand and assess argumentation in terms of these three “alternate” modes, Gilbert argues that we must transcend the linear approach to argument which logic has traditionally assumed.


35

Another important instance of argumentation theory’s willingness to countenance forms of argumentation which are radically distinct from argument as it is traditionally conceived is reflected in a distinction Gilbert has labeled the “convince/ persuade dichotomy.” As he puts it: ...to convince is to use reason, dialectic and logic, while to persuade is to rely on emotion, prejudice, and language. This distinction has moral as well as logical implications insofar as “convincing” has been considered to be a superior method. “Persuading” appeals to the “baser” components of the human psyche, namely, the emotions, while ‘convincing’ speaks to the “higher” aspects, namely, reason. On this view, one who is persuaded may be so for reasons that have little to do with the value of the arguments or the truth of the premises put forward. (Gilbert 1997: 4)

Gilbert criticizes the preference for convincing implied in this account, but still assumes a clear distinction between convincing and persuading which holds that persuasion must be understood and assessed in a different way than arguments in the traditional sense. Johnson and Blair (1994) employ a variant of the convince/persuade dichotomy when they explain the dynamics of argumentation within advertising. According to their account, advertising “have the appearance of argumentation” but this in reality “a facade,” for “most advertising works not at the rational level but at a deeper level.” It follows that “advertising has a ‘logic’ of its own” which must be distinguished from “the logic of real arguments” (221) on the basis of a proposed distinction between “rational” and “psychological” persuasion. As Johnson and Blair put it: “[M]any advertisements have the facade of arguments. They look like premises leading to a conclusion and like exercises in rational persuasion. In fact, we are persuaded by the school that holds that advertising is best viewed as psychological persuasion – an attempt to use psychological strategies to implant the name of a product in our unconscious minds. Hence criticism of advertising as a form of argumentation is misconceived. Learning how to decode ads and making ourselves aware of the strategies that advertisers use is more useful than looking for fallacies in the arguments.” (225). One might easily expand this account of argumentation theory by introducing other kinds of argument which are said to contradict the traditional account of argument. In the present context, it is enough to say that it is increasingly characterized by such distinctions, and the different methods of analysis and assessment they imply. If we compare this approach to argumentation to the attempt to analyze all arguments in terms of some form of the traditional conception of argument, then it will be clearer how the fable of the fox and the hedgehog might be applied to argumentation theory. For there is

36

Leo Groarke

a sense in which its development has been characterized by the fox’s approach to problem solving – i.e. by the development of a great variety of tricks for analyzing ordinary argumentation (I believe that Dutch pragma-dialectics is in some ways an exception to this rule, but I cannot discuss it in detail here). In contrast, an approach which attempts to preserve and extend the traditional conception of argument is hedgehog-like insofar as it is built upon a single way of understanding arguments, which is proposed for use in every circumstance. A definitive comparison of the distinct approaches to argumentation this implies is impossible in a single article, but in the remainder of the present one I will sketch some reasons why we should be skeptical of the fox-like approach which now characterizes argumentation theory – an approach which is much more problematic than argumentation theorists have assumed. Kinds of Argument and the Deductive/Inductive Distinction Argumentation theory’s attempt to distinguish different kinds of argument is in part problematic because the kinds of argument it distinguishes are not “natural kinds” in the sense that they are not clearly distinguished within ordinary reasoning and persuasion. It is argumentation theorists – not ordinary reasoners – who distinguish between deductive, inductive and conductive arguments; between attempts at convincing and persuading; between “persuasion,” “negotiation,” “pedagogical,” “information-conveying,” “cooperative informing,” and “eristic” dialgoues; and between logical, emotional, visceral and kisceral arguments. The extent to which such distinctions are removed from ordinary contexts is itself a matter of concern. Ordinary reasoners do not distinguish the kinds of argument and persuasion which argumentation theory proposes as fundamentally different kinds of argument. Consider the most entrenched distinction within argumentation theory – the distinction between inductive and deductive arguments. It is said to show that there are some everyday arguments which cannot be understood in traditional terms. But accounts of the distinction are in many ways problematic. Textbook treatments regularly misconstrue the meaning of deductive validity and suggest that inductive arguments are arguments which are intended inductively rather than deductively (see Groarke 1999a). Even if we ignore problems that this raises about intentions (which are in an important sense unobservable) it makes little sense to think that ordinary reasoners intend their arguments deductively or inductively, for most ordinary reasoners are unaware of the distinction and are, in view of this, in no position to intend their arguments one way or the other. As Govier puts it “People who argue do (at least implicitly) distinguish conclusions from premises.... But they often do not, even


37

implicitly, make claims about what sort of connection is supposed to hold between these premises and their conclusion” (1987, 46). It follows that the inductive/deductive distinction is a theoretical distinction which is imposed on ordinary argumentation, rather than one which guides it in practice. And it is a distinction which is very difficult to apply in practice, for those arguments which are normally classified as inductive arguments can usually be construed as deductive enthymemes which include implicit premises which establish a deductive link between the premises and conclusion. The argument “I know the apples on the table are good because I have had two of them” is a paradigm example of the kinds of inductive inference that characterize ordinary reasoning. It can be described as a statistical generalization based on a sampling of two. But it can easily be construed as a deductive enthymeme which conveys the argument: Premise: I had two of the apples on the table. (Unexpressed) Premise: They were good. (Unexpressed) Premise: They are like the other apples. Conclusion: The apples on the table are good. As Gerritsen (1994) and Groarke (1992, 1999) have pointed out, it is in principle possible to treat all inductive arguments in a similar way. This does not eliminate the uncertainty which is characteristic of inductive arguments, but it locates this uncertainty in implicit premises which allow one to treat inductive arguments as attempts at deductive reasoning. The implicit premises which are in the process recognized are, as pragma-dialectics has noted, a natural extension of the treatment of enthymemes and implicit premises which is a necessary part of any account of ordinary argument. In the present context, the important point is that such a strategy shows that one can extend the traditional conception of argument to the kinds of contexts studied by argumentation theory without relinquishing its deductive nature. But even if one grants some version of the deductive/inductive distinction, it is difficult to justify the plethora of distinctions which now characterize many variants of argumentation theory. For the distinctions between different kinds of dialogue; between emotional, visceral, kisceral and logical arguments; and between other forms of argument are much less clear and well established than the distinction between deductive and inductive arguments. Though a discussion of all of these distinctions is impossible in the present context, the problems that they raise can be illustrated in terms of the distinction between convincing and persuading. It is of special significance in the present context given that persuading as it is standardly understood encompasses the kinds of

38

Leo Groarke

argumentation which are standardly assumed to be most removed from argument as it is traditionally understood. Convincing and Persuading Reconsidered One might describe the proposed distinction between convincing and persuading as a distinction between two ways of changing (or reinforcing) belief. The former is an appeal to reason, which attempts to change or strengthen belief by weighing evidence in the literal, rational, deliberate way that characterizes science, mathematics or detective novels. Persuasion is, in contrast, an appeal to emotion which relies on emotional associations, hidden meanings, humour, and other psychological ploys. One might question the tenability of the convince/persuade distinction is by questioning its assumption that emotions are in some important sense not rational. One might do so by developing an account of emotion that understands it as some kind of cognitive phenomenon (as does De Sousa 1987). But it is more important, in the context of argumentation theory, to consider whether the persuade/convince distinction clearly separates two different kinds of ordinary reasoning. We can see that there are problems with the convince/persuade distinction even if we restrict ourselves to clear examples of arguments in the traditional sense, for it is a mistake to think that the emotional dynamics that characterize persuasion are absent from such arguments when they are used in day to day discourse. The argument “Homosexuality should be illegal because the Bible condemns it and whatever it condemns should be illegal.” is a deductive argument which is a paradigm example of an argument in the traditional sense – its premises and conclusion are easily identifiable and it is deductively valid. Yet the use of such an argument in everyday contexts is inevitably affected by complex, pervasive and powerful emotional resonances (both positive and negative) about conservative religion and homosexuality. It follows that arguments in the traditional sense cannot be easily separated from emotional considerations in the way suggested by the convince/persuade distinction. One might reply that there is a sense in which the emotion which surrounds this example is irrelevant – does not provide legitimate evidence for or against the conclusion and that it should not play a role in the argument’s assessment. Two things can be said in response to this suggestion. First, that much the same could be said about paradigm instances of persuasion, in which the emotional ploys which are used typically do not provide legitimate evidence for or against the point (or the behavior) which the persuasion attempts to establish. Second, it cannot be assumed that emotion has no legitimate role to play in judging the acceptability or unacceptability of arguments in ordinary


39

reasoning. An impassioned plea for an end to genocide in Kosovo will inevitably appeal to emotion but it would be strange to say that it does so inappropriately. In the contexts which charaterize traditional formal logic – mathematics. logic, and science – emotional distance is a goal, but the same cannot be said when we deal with contentious practical issues. On the contrary, most social and political issues have an emotional element which legitimately comes into play when we judge whether the premises of an ordinary argument are acceptable or unacceptable, even when the argument is a paradigm instance of argument in the traditional sense. Much more needs to be said about the way in which this emotional element should be judged and assessed is in principle compatible with the account of argument which has already been proposed. As examples of traditional argument may legitimately involve emotion, examples of persuasion are often best understood as appeals to reason and rationality (and as convincing in this sense). Consider, to take a hypothetical example, an instance in which fear is the basis of persuasion. Imagine that you are kidnapped, taken to an apartment, and held captive. When you think the kidnapper isn’t looking, you reach for the phone intending to call the police. But your abductor hears you and spins around. He gives you a threatening look and thrusts his gun in your direction. You stop reaching for the phone. What has happened in this scenario? In some sense, the kidnapper made you change your mind about your phone call. But did he rationally convince you, or did he just persuade you? He has not said anything, so there is a clear sense in which he does not explicitly forward premises and a conclusion. Moreover, the situation is a thoroughly emotional one. The kidnapper plays on your fear of being hurt and it is this fear which directs your decision to accede to his wishes. One might conclude that this is a paradigm instance of persuasion and yet there is a very clear sense in which the kidnapper forwarded an argument in the traditional sense which has already been elaborated. For even though he has said nothing, his actions function as a way of elaborating an argument which might be summarized as follows: Premise: Stop or I’ll shoot. Premise: You don’t want me to shoot (more than you want to continue). Conclusion: You should stop what you’re doing. In principle, this argument could have been made verbally, in a variety of ways. Instead of pointing his gun, the kidnapper might have uttered the rhetorical question “Do you want me to shoot you or do you want to continue?”

40

Leo Groarke

Or he might have said “If you don’t stop, I’ll shoot. You don’t want me to shoot. So you should stop what you’re doing.” In real life, a kidnapper is unlikely to make such an argument fully explicit just because this takes time and because there is no need for him to do so. According to argumentation theory, the argument is conveyed more directly when it is verbalized, but there is a sense in which this is wrongheaded: an aggressive pointing of the kidnapper’s gun gets the point across more directly than a verbal claim. Certainly one is likely to appreciate the force of his argument more strongly when a gun is thrust in one’s direction. Whatever needs to be said in this regard, the important point is that the kidnappers thoroughly emotional attempt at persuasion can easily be understood as an attempt to forward an argument which conforms to the traditional conception of an argument. Like our previous examples, this suggests that there is no easy way to separate convincing and persuasion in ordinary discourse. Still more importantly, the decision to treat this example of persuasion as a case of convincing and thus argument has significant theoretical advantages. For it allows us to assess it in the same way we assess its verbal counterpart. Considered from this point of view, we can recognize the argument as a variant of disjunctive syllogism and the traditional ad bacculum. It follows that the example can be understood, not merely as an argument in the traditional sense, but also as a widely recognized form of argument which can be assessed accordingly. Though a detailed evaluation is beyond the scope of the present article, it is worth noting that the argument in question is arguably a good one which has the characteristics which convince Wreen (1988) and Woods (1995) that there can be good instances of ad bacculum. In their way a decision to treat this instance of persuasion as a case of argument in the traditional sense pays dividends, for it permits us to understand it and evaluate it in standard ways, using well understood criteria. In contrast, an adherence to the persuade/convince dichotomy requires that we understand it in terms of criteria which have not been clearly and convincingly established. A full discussion of convincing and persuasion lies beyond the scope of the present paper, but it can still be said that such examples suggest that the distinction is not clear in ordinary argumentation and that ordinary instances of argument and persuasion frequently contradict its attempt to distinguish appeals to reason and emotion. In view of such considerations, we may do better to recognize instances of persuasion as attempts to convey arguments in the traditional sense, albeit attempts that rely on more subtle and more implicit means. Such an account intuitively corresponds to our ordinary use of terms like “convince” and “persuade” and still allows an analysis of both argument and persuasion which recognizes their affinities to each other, and to the


41

traditional account of argument. It will take time to see how far such an approach can be taken, though it must at least be said that it has a potential which most argumentation theorists have failed to appreciate. Expanding the Realm of Argument So far, I have criticized argumentation theory on the grounds that it has been too quick to embrace conceptions of argument which radically reject the conception of argument embedded in traditional logic. Historically, its enthusiasm for such conceptions is a reaction against logic’s preoccupation with a very narrow range of argument in science, logic and mathematics. As Perelman (1982) emphasizes, argumentation in real life is a much broader enterprise that involves many other contexts. As Johnson (1996) points out, traditional attempts to teach logic in a deductive, formal way have often emphasized peculiar examples of argument which have little relevance to ordinary argument and persuasion. This much must be granted. But these historical facts do not show that the traditional conception of argument is necessarily tied to the kinds of contexts which have characterized formal logic. In assuming that it is, argumentation theorists have overlooked the possibility that the notion of argument assumed in such contexts can be expanded and extended so that it encompasses a much broader range of argumentation. We have already seen that we can move in this direction by detaching this conception of argument from formal concerns and by modifying it in ways that make room for the features of ordinary argumentation which most clearly distinguish it from formal arguments (i.e. implicitness and acceptability). The extent to which the traditional concept of argument can be extended beyond mathematical, scientific and logical contexts can be illustrated in the context of non-verbal arguments. Though some argumentation theorists have recognized that they play a prominent role in ordinary argument, the examples of argument which are most prominent in discussions in argumentation theory are still predominantly verbal. Such examples present a typical argument as a set of sentences which includes premises that support some conclusion. This is the model which has been traditionally assumed in logical and mathematical proofs, though Barwise and Etchmendy (1992) have recently argued for the possibility of visual deductions. In everyday contexts, it can similarly be said that visual elements play a largely unappreciated role in the argumentation that characterizes political debate, advertising, and so on. The kidnapping example we considered earlier already shows that nonverbal argumentation can sometimes be accounted for in a way that uses rather than rejects traditional notions of argument, for this is a case in which the

42

Leo Groarke

kidnapper presents his argument in a non-verbal way. A famous example which also illustrates this basic point is found in the traditional Life of Aesop (Daly 1961). According to its biography of the famous fabulist, he was born an ugly mute slave and received the gift of speech (logos) in return for a pious favor to a messenger of Isis. The incident which illustrates the possibility of non-verbal arguments occurs before this transpires, when Aesop is still unable to express himself verbally – a situation which forces him to argue in a nonverbal way. In the incident in question, Aesop’s fellow slaves try to take advantage of him by stealing the master’s figs, eating them, and blaming it on Aesop. Though he cannot speak, he manages to outwit them by swallowing warm water, putting his finger down his throat and forcing himself to throw up, the results proving that he has not eaten the figs (when he points to the other slaves and they are forced to do the same, the evidence literally falls from their mouths and they are punished both for stealing the figs and for trying to frame a man who cannot defend himself). How should we understand this incident? Clearly, Aesop has not forwarded a verbal argument. But he has still managed to convince his master that he is innocent by providing evidence to this effect. We might summarize the implicit argument as: Premise: If I was guilty of eating the figs, they would be in my vomit. Premise: They are not in my vomit. Conclusion: I am not guilty of eating the figs. This is a straightforward instance of modus tollens which can be represented as: Premise: g Ú v Premise: v Conclusion: g where g = I am guilty of eating the figs; v = They are in my vomit. In demonstrating his innocence with this argument, Aesop provides an example of a simple propositional argument which is conveyed non-verbally. In the process he provides an example of a physical argument which is “visceral” in Gilbert’s sense (1997, 75-86), but one which shows that it is a mistake to think that such arguments need to be understood and assessed in a way that makes it useful to differentiate them from “logical” modes of argument. Other cases of non-verbal argument can be dealt with similarly. Consider a pharmaceutical company which tries to persuade us that we should buy their brand of cough syrup by airing television advertisements in which men and women in white lab coats enunciate its superior effects. If one restricts one’s


43

attention to their verbal claims, one might easily recognize a verbal argument that maintains that we should buy their brand because it has the qualities they enumerate. But this is a small part of what is going on. Visually, we can easily discern another argument. Considered from a visual point of view, the lab coats that the actors wear are not accidental, but an attempt to present them as experts on the medical properties of cough syrup. If they have stethoscopes as well as lab coats, we are being visually told that they are doctors. We can thus recognize another argument which might be summarized as follows. Premise: These doctors recommend this brand of cough syrup. Premise: You should accept their recommendation given their status. Conclusion: You should buy this cough syrup. Though this argument is never explicitly stated, it can still be usefully treated as a case of argument in the traditional sense – more specifically, as an attempt to construct an appeal to authority which should be assessed accordingly. It goes without saying that it may, in the process of assessing an argument of this sort, be useful to say something about our psychological reactions to authority (or white lab coats or stethoscope). This is an important aspect of the argument about which many useful things might be said. The problem is the prevalent assumption that the importance of this aspect of the argument implies that we cannot treat it as an argument in a very traditional way. Rather, it should be said that what needs to be said in this regard is entirely compatible with the attempt to see this visual message as an argument in the traditional sense. It is in keeping with this that much recent work on “visual arguments” suggests that a great many non-verbal visual images are used to forward implicit premises and conclusions, implying that the standard notion of argument can be use fully employed in the attempt to explain and assess persuasion in such contexts (see, e.g., Birdsell & Groarke 1996, Blair 1996, and Groarke 1999). Concluding Remarks In the present paper I have argued that argumentation theory has been too willing to embrace alternatives to the notion of argument which logic has traditionally assumed. Instead of adopting these alternatives, I have argued that it would do better to extend this traditional notion and understand ordinary argument and persuasion in terms of it. There is no way to conclusively establish this approach in a single article, but it can still be said that this approach to ordinary argumentation has not been given the serious consideration it deserves, and that the problems with the more usual approach have been

44

Leo Groarke

ignored rather than appreciated. If nothing else, the approach I have suggested challenges argumentation theory to more convincingly demonstrate (1) that there are ordinary which show that some forms of argumentation cannot be understood as arguments in the traditional sense; (2) that non-logical forms of argumentation are in some way irreducible to argument in the traditional sense; and (3) that the distinctions which are standardly made between different kinds of argumentation can withstand careful scrutiny. I think the examples I have already provided suggest that this is much more difficult than usually imagined, and that the standard approaches to argumentation theory tend to needlessly complicate our understanding of ordinary argument and persuasion. Until argumentation theorists show otherwise, we will do better to develop an approach to argumentation which has more affinities to the traditional account of argument which argumentation theory has rejected. It is better to be a hedgehog with one good trick than a fox with a bag of tricks that do not work as well. References Birdsell, David & Leo Groarke (1996). “Toward A Theory of Visual Argument.” Argumentation and Advocacy, Vol. 33, No. 2. Blair, J. Anthony (1996). “The Possibility and the Actuality of Visual Arguments.” Argumentation and Advocacy, Vol. 33, No. 2. Daly, Lloyd W. 1961. Aesop Without Morals. New York: Thomas Yoseloff. de Sousa, Ronald (1987). The Rationality of Emotion. Cambridge: The MIT Press. Barwise, John & John Etchmendy. (1992). “Visual Information and Valid Reasoning.” In Leslie Burkholder, ed. Philosophy and the Computer. Boulder: Westview Press. Gerritsen, S. (1994). “A Defense of Deductivism in Reconstructing Unexpressed Premises,” in van Eemeren & Grootendorst, 1994. Gilbert, Michael (1997). Coalescent Argumentation. Mahwah: Lawrence Erlbaum Associates. Govier, Trudy. (1987). Problems in Argument Analysis and Evaluation. Amsterdam: Foris Publications. Groarke, Leo (1992). “In Defense of Deductivism: Replying to Govier,” in Van Eemeren et. al., eds. (1992) —— (1996). “Logic, Art and Argument.” Informal Logic. 18, Nos. 2&3, 105-129. —— (1999). "Deductivism Within Pragma-Dialectics," Argumentation 13, 1-16. Johnson, Ralph H. (1996). The Rise of Informal Logic. Newport News: Vale Press. —— & J. Anthony Blair (1994). Logical Self-Defense. (United States Edition) New York: McGraw-Hill. Levi, Don S. “The Case of the Missing Premise,” Informal Logic Vol 17, No. 1, Winter 1995. Mortenson, G.D. & R.L. Anderson (1970). “The Limits of Logic.” Journal of the American Forensic Association, Vol. 7. Perelman, Chaim 1982. The Realm of Rhetoric. Notre Dame: University of Notre Dame Press. Toulmin, Stephen 1958. The Uses of Argument. Cambridge: Cambridge University Press. Walton, Douglas (1992). The Place of Emotion in Argument. University Park: The Pennsylvania State University Press.


45

—— (1998). Argument Structure: A Pragmatic Theory. Toronto: University of Toronto Press. —— and Erick C. Krabbe (1995). Commitment in Dialogue. Albany: State University of New York Press. Woods, John (1995). “Appeal to Force.” In Hans V. Hansen and Robert C. Pinto, Fallacies: Classical and Contemporary Readings. University Park: The Pennsylvania State University Press. Wreen, Michael J. (1988). “May the force be with you,” Argumentation 2. Van Eemeren, Frans H. & Rob Grootendorst 1992. Argumentation, Communication, and Fallacies: A Pragma-Dialectical Perspective. London: Lawrence Erlbaum Associates. ——, eds. 1994. Studies in Pragma-Dialectics. Amsterdam: International Centre for the Study of Argumentation. ——, J.A. Blair, and C.A. Willard, eds. (1992). Argumentation Illuminated. Amsterdam: International Society for the Study of Argumentation (ISSA).

46

J. Anthony Blair

J. ANTHONY BLAIR Presumptive Reasoning/Argument: An Overlooked Class1 (a) Mother to child: Jay, you mustn’t say that your grandmother’s trifle looks like the dog’s throw-up. That hurts her feelings, and it’s wrong to hurt someone’s feelings. Jay: But Mum, it’s true! Mother: Never mind. (b) Matthias: Professor Pinto, I would like you to reconsider my grade, please. You gave Wichert’s term paper an A-. I read it, and I can’t see how my paper is any worse, but your teaching Assistant gave mine a C-. Professor Pinto: Certainly. If you’re right either your grade should go up or Wichert’s should come down. (c) Sally (calling from another room): Scott, what time is it? Scott: Nine-thirty. Sally: Thanks. I’d better get going: I promised Frans I’d meet him at ten. (d) Town mayor to petitioning merchants: Ladies and gentlemen, we can’t possible declare the first week in June the town’s official “Ice-Cream Week.” If we did that, then the vendors of every product sold in town would want their own official “Week,” and we’d have to give it to them. But there are more than fifty-two products sold in town, so we’d have to either turn some down, or have double or even triple official “Weeks.” And there’d be competition for weeks deemed “better” than others. The result would be a disastrous legal nightmare or an equally disastrous confusing tangle of product promotions, or both. I’m sorry, but it’s out of the question. (e) Coach Scotty Bowman (addressing the Detroit Red Wings hockey team, April 1999): Men, we all want to win the Stanley Cup for a third year in a row. You know as well as I do what that means. We’ve got to fore-check, we’ve got to back-check, we’ve got to move the puck up-ice fast when we get a turn-over, we’ve got to get somebody in front of the net when we get 1 Many of the ideas in this paper owe their origins to what I took away from conversations with my colleague, R.C. Pinto, who may wish me to take complete responsibility for them. My debt to Douglas Walton’s work is, I hope, clearly indicated by the numerous citations of it. I also thank Professor Dieter Mans, whose comments on an earlier version of the paper have led to corrections of fact and improvements of detail and exposition.

Presumptive Reasoning/Argument: An Overlooked Class

47

the puck in their end, we’ve got to clog between the blue lines when they’re trying to bring the puck up ice, we’ve got to use the body aggressively, we’ve got to protect the goalie, we can’t take retaliation penalties, we’ve got to work, work, work! You know all this. You’ve done it before. Let’s do it again! (f) Father:Son, I don’t want to catch you smoking again. Smoking is bad for you. Son:But Dad, you smoke. (g) Consumer’s Digest magazine: The Zero refridgerator rates “excellent” for quick-freezing, temperature stability, and storage space; the shelf-configuration and flexibility score “very good,” as does the service record; its noise level and ease of cleaning the air-intake grill are only “fair”; and price and resale value rate “poor.” On balance, and in comparison to what else is on the market, Consumer’s Digest rates the Zero a fair-to-good buy in a refridgerator, about in the middle of the pack. 1. Introduction The aim of this paper is to draw attention to an important, distinctive, relatively neglected class of reasoning or argument. Theorists in Artificial Intelligence and Expert Systems who have wanted their modelling or simulations of human reasoning to be empirically informed (and not based on a prioristic assumptions about its nature) have been giving the large class of reasoning and argument discussed here their study for some time. (See the discussions of nonmonotonic reasoning reported in Gabbay, Hogger and Robinson, 1994, for example.) Philosophers and others, too, at least to the extent that they are interested in accounting for human reasoning and argument, should give it their close attention. The paper begins with some terminological clarifications, because words like ‘reasoning’ and ‘argument’ are slippery, yet are often, even in scholarly discourse, used in ill-define ways. The concept of presumptive reasoning/argument is then introduced via a discussion of one way of assessing reasoning or arguments. There follows a list of some of the types of presumptive reasoning/argument and a sketch of the main generic properties of the class. The paper ends with a rejoinder to a challenge to the claim that presumptive reasoning/argument is a distinct class of reasoning and argument.

48

J. Anthony Blair

2. Reasoning and argument, and their norms Reasoning, whatever else it may entail, is uncontroversially at least the mental act of inferring. One plausible analysis of inferring is: changing or reinforcing a belief or other attitude in response to exposure to a body of information (see Harman 1986). How good the reasoning is depends on two things: whether the reasoner takes the appropriate epistemic stance toward the information, and whether the relation of the body of information to the propositional attitude inferred from it satisfies the appropriate norms. Thus, first, reasoning is not good if the reasoner drew an inference from information he took to have a given value (for example, to be true) when he should have known it did not have that value (for example, it was clearly false or was certainly dubious). Garbage in, garbage out. And, second, reasoning is not good if the reasoner took the information to support his conclusion to a certain degree when in fact it provided a different degree of support. He might have mistaken strong support for weak support, moderate support for strong support, and so on. (We also say that reasoning is bad if the agent failed to draw any inference at all from information he was aware of when in fact that information supported some attitude of importance to him. Example: “When you noticed that you passed that hotel a third time, did it not occur to you that the taxi driver was taking you for a ride?” This point suggests that the norms of reasoning have a pragmatic element.) Some might prefer to restrict reasoning, or its assessment, to the second of the above factors. On that conceptualization, if the person’s conclusion is well supported by his premises, assuming them to be true, then his reasoning may be said to be good, even if it should turn out that his premises were false and he should have known that. On the other hand, to accept a premise that is patently implausible seems also to be, or to imply, a failure of reasoning, for the person who accepts a wildly implausible claim and then reasons from it fails to recognize the inconsistency between that premise of his reasoning and other, well-established beliefs that he holds. For example, the person who begins making large financial commitments on the basis of having received in the mail an announcement from Publisher’s Clearinghouse that he has, out-ofthe-blue, “won” $10 million, has not “thought it through,” as we might say. This consideration argues for the broader conception of reasoning assessment, which in turn implies a broader conception of reasoning than just the act of inferring from information received. The word ‘argument’ is a veritable Hydra: besides its seven meanings described in the Oxford English Dictionary (1971), the more it is analyzed, the


49

more senses are distinguished (or stipulated). In one sense, common in philosophy, the information from which a reasoner inferred a conclusion, together with the conclusion expressed as having been inferred from that information, is termed an “argument,” or that person’s “argument.” We say of such a person that he reasoned, not that he argued; but we, or philosophers, at least, nonetheless speak of the “argument” of, or implicit in, or corresponding to, his reasoning, or indeed, of “his argument”: “Corresponding to every possible inference is an argument, . . .” (Copi and Cohen 1990: 6). The merits of such arguments are (unsurprisingly) determined in exactly the same way that the merits of the product of the reasoning represented by such arguments are determined, noted above, for the two labels denote the same thing. That first sense of ‘argument’ is frequently confused with a somewhat different one, namely, the proposition/s communicated by one person (the arguer, the person arguing or making the argument) to another or others (the audience or the interlocutor) as being true or belief-worthy and as supporting some further proposition, such that, in the attempt to persuade or convince the audience to accept that further proposition, the arguer invites the audience to infer the latter from the former (Pinto 1995: 276). In addition, sets of propositions that might represent reasoning, although no one has reasoned in that way, or that might be used to try to convince an audience, although they have never been so used, are also in many quarters termed “arguments.” These might more accurately be called possible arguments. Notice that someone can use an argument to try to convince an audience (‘argument’ in the second sense) although it does not represent his own reasoning (‘argument’ in the first sense) – as, for example, when an arguer presses an audience with the implications of its previously-expressed belief commitments, although he does not share those commitments. The merits of the products of arguments of the second sort are much more difficult to determine than those of the first sort, because arguments of this second sort are much more complicated than the former. They are not just sets of propositions, but sets of propositions put to particular uses in a complex social practice. As van Eemeren and Grootendorst (1983) have pointed out, such arguments are multi-dimensional: functional, social, expressed in language and dialectical (or interactional). As a result, their evaluation must be multidimensional as well. These various aspects of any such argument are in addition to the argument’s “logical” (Walton 1996b) or “illative” (Blair 1995) core (the grounds adduced in support of claims). Thus such an argument can go wrong by failing to persuade, by failing to communicate, by violating speech-

50

J. Anthony Blair

act conventions, or by failing to respond to the interlocutor’s challenges as well as to by appealing to false premises or by inviting a strength of commitment to a conclusion inappropriate to the force of the evidence. For whatever reasons, philosophers interested in the norms that govern reasoning or argument have traditionally focused on the illative core, in abstraction from the social practice in which it is embedded in arguments of the second sort. There can be no objection to that predilection, provided that all the factors germane to the assessment of the argument’s illative core can be identified in abstraction from the surrounding context of argumentation, and as long as the verdict on the illative core is not presumed to apply equally to the entire argument (in the second sense). Classically, philosophers have conceived the normative principles governing the two aspects of the illative core – the premises’ adequacy in themselves, and the adequacy of the premises’ support for the conclusion – to belong to two distinct fields: epistemology and logic, respectively. On the basis of that traditional conception (with no commitment to its correctness), the subject matter of this paper can be identified as belonging to logic, for the proposal of the paper has to do with the criteria for the adequacy of the support that stipulated premises provide for an argument’s conclusion in the illative core of argumentation. Consider one type of illative core support structure, the type in which the grounds adduced in support of the claim entail it. In such an argument or reasoning, the conjunction of sentences expressing those grounds and the negation of that claim is an inconsistent set. Entailment in this sense is almost always uncontroversially a virtue of arguments and reasoning. When an illative core structure has the property that the premises entail the conclusion, it is, with rare exceptions, a good argument or good reasoning in that respect – that is, with respect to the support the premises, taken as given, provide for the conclusion. (The qualification is needed for the rare case in which the premises are synonymous with, or presuppose, the conclusion, that is, for the case of circular or question-begging reasoning or arguments.) Is entailment a necessary as well as a (normally) sufficient condition of good argument or reasoning support? Are there other ways an argument’s or reasoning’s support can be good besides by entailing the conclusion (using ‘entailment’ as stipulated above)? Many philosophers think that entailment is not the only virtue of argument or reasoning support. The following is an argument for that position. (1) The general criterion for any type of premise-conclusion support is that it be rational, on the basis of accepting its premises, to accept the conclusion of any argument or reasoning with that type of support. (2) There are various types of support to be found in arguments or reasoning such that, given the


51

premises and their relation to their conclusion, it is rational to accept that conclusion on the basis of those premises, although the premises do not entail the conclusion. Hence, (3) premise-conclusion entailment is not a necessary condition of good premise-conclusion support in reasoning or argument. Since the conclusion of this argument is entailed by its premises, the argument does not beg the question by assuming to be acceptable the possibility of the rational, non-entailment premise-support that it is used to argue for. Examples that support premise (2): “All observed chimpanzees have had a vegetarian diet, so chimpanzees do not eat meat.”; “60% of a large representative sample of voters have indicated their intention to vote Tory in the next election, so the Tories will win the next election.”; “Cats tend to dislike water, so probably this cat will dislike water.”; and so on. These examples suggest that various types of reasoning/argument with inductive support are rational. But there are other types of support as well that are rational yet non-deductive and to these we now turn. 3. Presumptive reasoning/argument So far, the discussion has reviewed old ground. Consider next the following classes of reasoning or argument exemplified by the invented, but realistic, examples provided at the beginning of the paper. (The list of types is meant to be illustrative, not exhaustive.) (a) Inferences or arguments from general moral principles, plus descriptions of particular circumstances, to judgements about what ought to be done in particular cases. When Jay’s mother appeals to the principle that one ought not to hurt another’s feelings, she need not take that principle to be exceptionless. She merely insists that it applies in this case. In fact she takes it in this case to override the principle, invoked by Jay, that one should tell the truth, which thus itself cannot be a moral principle that has no exceptions. Presumably, both principles can be overridden. To this date moral theorists have agreed to no meta-rules that strictly settle conflicts of moral principles. (Sir David Ross’s classic discussion of “duties” remains illuminating in this respect (1930, Chapter 2).) The result is that such inferences as the one Jay’s mother makes and urges on him are not entailments. Nor, obviously, are they inductive in any scientific sense. (b) Case-by-case analogical reasoning (of the following form): “A token of action/judgement-type X was correct in case A; case B appears to be similar to

52

J. Anthony Blair

case A in the salient respects; so a token of action/judgement-type X is correct in case B.” In Matthias’s argument to persuade Dr. Pinto to reconsider the grade his essay received, Matthias is claiming that his essay was analogous Wichert’s essay in relevant respects (presumably, with respect to those of its properties that satisfied the criteria for an A-range grade), and that therefore his grade should be similar to Wichert’s. Dr. Pinto must judge not only whether the two essays are similar in the respects noted by Matthias, but also whether they are similar in all respects relevant to the grade. For example, were Matthias’ arguments as thorough, was his exposition as lucid, was his paper’s organization as perspicuous as those of Wichert? Such assessments are, as we say, “judgement calls.” Furthermore, even if the papers were similar, unless they were identical to a degree that suggests plagiarism, Dr. Pinto must judge whether the degree of similarity justifies a similar evaluation. So far, we have no calculus for the determination of such judgements. (c) Appeals to testimony: reasoning of the form, “According to S, P is true; so P is probably true.” – and the many variants of it. Some argue that basing a belief on another’s testimony in simple cases, such as when you ask them the time of day or for simple directions (“Which way to downtown, please?”), entails no inference, but rather is epistemically akin to basing a belief on what one sees (for example, believing there is a tree ahead on the ski-run because one sees a tree ahead on the ski-run). If they are right, then the example of Sally’s accepting that it is 9:30 on the basis of Scott’s sayso is not a good example of a presumptive inference because it is not an example of an inference. Without taking the time to dispute that position here, it can be said that there are other kinds of case in which beliefs based on appeals to testimony are uncontroversially inferential. A major class consists of appeals to (expert) authority – reasoning that a proposition is (presumptively) true on the grounds that one or more other people who have a privileged epistemic access in such cases have declared it to be true. Thus, when several independent oncologists consulted by a cancer patient produce the same diagnosis and recommend the same treatment, it is reasonable for the patient to conclude that he has the condition they name and would be well-advised to take the treatment they prescribe. Yet, famously, medical diagnoses can be wrong and recommended treatments can turn out to have been ill-advised. So the patient’s inference is no entailment. And, again, it is not an induction in any standard sense.


53

(d) All “balance of considerations” or “pro and con” reasoning or arguments. This is reasoning or argumentation in which there are both reasons that support the conclusion and reasons that support the contradictory of the conclusion. Typically, there is more than one line of support on each side. The reasoner must weigh the force of each line of support, sum the weight of considerations on each side, and decide which side’s reasons or arguments are, on balance, more compelling. In the example given, the town mayor lists only the disadvantages of the proposal that the ice-cream sellers have made. In arguments like this, which are designed to persuade an audience, it is not uncommon to find reference to the reasons on one side only. The mayor can safely assume that the ice-cream vendors know the reasons they advanced in favor of declaring the first week of June “Ice-Cream Week.” Still, the mayor’s reasoning tacitly includes the judgement that the “con” arguments he mentions outweigh the “pro” arguments the vendors had advanced or others that might exist. While heuristics would doubtless be helpful, there obviously can be no mechanical decisionprocedure for such reasoning. (e) All means-ends reasoning or arguments: for or against an action or policy on the basis of the value of the predicted consequences of performing or implementing it. There are many variants of such reasoning. The Detroit Red Wings’ coach, Scotty Bowman, listed what seemed to him to be the salient means to the end of his team’s winning hockey games in the National Hockey League playoffs in the spring of 1999. Given the premise that the team wanted the end (to win the Stanley Cup), if Bowman is right that these are means to that end, then he has given the team a good argument for executing those means. But there is no entailment, for a couple of reasons. For one thing, the members of the team may want that end, but want other ends more (for instance, to avoid career-threatening injuries). For another thing, they might agree that the means Bowen cites are good ones, yet also believe that other, conflicting, means are in some circumstances superior. [Historical note: the Detroit Red Wings were eliminated in the second round of the 1999 Stanley Cup playoffs by the Colorado Avalanche, 4 games to 2 in the best of 7 games series.] (f) All reasoning or arguments for or against someone’s credibility on the grounds of his/her past conduct or character. The son’s reply to his father’s admonition not to smoke is a classic example of a circumstantial ad hominem argument. Walton (1998) discusses this example, and argues persuasively that such reasoning can be non-fallacious, even though the son’s retort identifying the inconsistency between his father’s

54

J. Anthony Blair

advice and his behaviour does not succeed in discrediting his father’s position. In general, we usually have nothing else to rely on than (the record of) a person’s past conduct when we assess their credibility or their suitability for a job or a role. This reasoning is not statistical (the candidate conducted herself well in S1, S2, S3, . . . Sn, so she will conduct herself will in Sn+1), for we allow for balance-of-considerations reasoning in coming to a conclusion about the net indication of a record that is a mix of favourable and unfavourable performances, and we select certain past circumstances to be more salient as predictors of the present case than others. (g) The reasoning involved in evaluations of any kind. This class is huge. The field of evaluation is a professional and academic discipline (see Scriven 1991). Product evaluation of the sort exemplified by the Consumer’s Digest magazine assessment of the Zero refridgerator, is but one large subclass. There are also, among others, personnel evaluation, program evaluation, policy evaluation, technology assessment, and environmental impact assessment. Beyond its “professional” exercise, “evaluation is what distinguishes food from garbage, lies from truth, and science from superstition. In short, it is the sine qua non of intelligent thought and action . . .” (ibid.: 139-140). A high percentage of the inferences involved in evaluative reasoning are presumptive. The identification of the relevant criteria of evaluation, the determination of the appropriate standards to use in applying those criteria to the particular case, the grading of the entity being assessed using those criteria and standards, all involve complex prima facie inferences. Heuristics are useful, but this is another type of reasoning in which mechanical decision-procedures seem to have limited application, if any. In vast numbers of the arguments in such reasoning the premises do not entail the conclusion, or so I will argue. In just this sense, these are not types of “deductive” reasoning or argument. Neither is any of these a type of “inductive” reasoning or argument in any but the vacuous sense that they are non-deductive. None exhibits the standard kinds of inductive reasoning or argument typically found in science. (a) – (g) are all types of what Walton (1996a) has called “presumptive” reasoning or arguments. Presumptive reasoning has the following properties. (1) The reasoning-types have identifiable patterns. These are termed reasoning (or argument) “schemata” or “schemes” (singular: “schema” or “scheme,” respectively) (see Schellens 1987, van Eemeren and Kruiger 1987, Kienpointner 1987, 1992a, 1992b, Walton 1996a). It is useful to distinguish between schemes that are patterns of good reasoning (here termed “normative” schemes) and


55

schemes that simply describe how people tend to reason or have reasoned in the past (called here “descriptive” schemes). These normative schemes can be expressed with greater or less generality, and may be classified in various ways. The scheme of relying on a spouse’s sayso for the time of day is a particular type of the more general scheme of relying on a person’s sayso for information. And clearly, relying on a physician’s judgement for a medical diagnosis belongs to the same genus of schemes as the first two, yet is a particular type of the more general scheme of relying on expert judgement for an assessment or classification of a phenomenon – which in turn is related to, but different from, relying on expert judgement for a recommendation of a treatment or a solution to a problem. To mention another example, involving a different type of reasoning/argument scheme, Walton (1998: 261) identifies altogether 18 species, subspecies, sub-subspecies, and sub-sub-subspecies of the generic ad hominem argument, for instance, the circumstantial ad hominem (broad sense), the direct ethotic ad hominem (including the variants: from veracity, from prudence, from perception, from cognitive skills, from morals), the bias ad hominem, poisoning the well, guilt by association, two wrongs, tu quoque, and so on. (2) (a) Each normative scheme describes a pattern of reasoning or argument the premises of which, if true, provide support for the conclusion, but only if other things are equal. This point may be expressed in a variety of other ways. (b) A normative scheme is a default form of good reasoning, and any such scheme is subject to override in a variety of circumstances. (c) Reasoning or arguing in the pattern of a normative schemes is prima facie good reasoning, but whether it is good reasoning all things considered depends on the circumstances in that case. (d) The grounds that instantiate the premise-types of a given normative scheme create a presumption in favor of the proposition instantiating the conclusion of that scheme. Whether that presumption is borne out depends on the particulars of the case. (e) In an dialogue in which arguments are traded by the interlocutors, the party with grounds instantiating a normative scheme thereby shifts the burden of proof to the other party. This variety of idioms attests to the ubiquity of this kind of reasoning. An example will illustrate the point concretely. Having promised to do something is a reason to do it, but (pace Kant) it is a decisive reason only in the absence of countervailing considerations. Perhaps in this case keeping the promise conflicts with a stronger obligation, or would cause enormous harm and produce little benefit. In either of these, or similar, circumstances, the promise is overridden. However, in their absence, and that of other similar disqualifiers, the promise is a decisive reason to perform the promised action.

56

J. Anthony Blair

Similar examples could be given for the other types of reasoning or argument listed. (3) For any given reasoning/argument scheme, the types of circumstances under which other things are not equal (in which the default does not apply / the reasoning is not good all things considered / the presumption is overridden) cannot be exhaustively specified, because (a) the variations on each scheme are innumerable, and (b) the circumstances of the instantiations of each scheme are too varied for the exempting conditions of any one instantiation to be generalizable to all of them. (4) However, in spite of their capacity for infinite variation, human motivation and conduct tend to follow a limited number of well-defined paths. Hence, it is possible to outline in a general way for any scheme the principal classes of exceptions, and so to provide useful general, if not universal, guidelines for the employment of that scheme. In the literature on schemes some theorists (Schellens 1987, van Eemeren and Kruiger 1987, Walton 1996a) have formulated what they call “critical questions” for each scheme, the role of which is to remind its user of the types of circumstances that typically derail reasoning of the pattern represented by that scheme. The critical questions function as a check-list to help determine whether any of the standard types of excepting conditions that should cancel the default represented by the scheme are presented in that particular instances of its employment. An example will illustrate. The following are Walton’s (1996a: 77-79) scheme and critical questions for “argument from analogy”: Scheme: Generally, case C1 is similar to case C2. A is true (false) in case C1. Therefore, A is true (false) in case C2. Critical Questions: 1. Are C1 and C2 similar, in the respect cited? 2. Is A true (false) in C1? 3. Are there differences between C1 and C2 that would tend to undermine the force of the similarity cited? 4. Is there some other case C3 that is also similar to C1, but in which A is false (true)? The thesis of this paper is that presumptive reasoning/argument represents a sui generis class of reasoning/argument. As noted, computer scientists have long been aware of this type of reasoning, and a few philosophers have also (see Govier 1987, 1999, who has drawn attention to Wellman 1971 and Wisdom). But it has not, at least in philosophy, received much attention as a distinctive class with a wide variety of sub-types that is on a par with, and deserving as careful attention as, “deductive” and “inductive” reasoning or


57

argument. Scriven (1987) was the first philosopher I am aware of who discussed it as a set of types of reasoning or argument exhibiting a distinctive logic (or set of logics), which he called probative logic, but his highly suggestive paper is rarely cited in this connection. Argument schemes lurk in the background of discussions of informal fallacies, as Walton (1996a) noted. He deserves credit for having made explicit the role of presumptive reasoning/argument in informal fallacies. However, although he explicitly recognizes presumptive reasoning/arguments as a distinctive type, Walton limits his discussion of it to its application in illuminating informal fallacies. Walton (1996a, Ch. 2) identifies the rationality of this type of reasoning with the rationality of practical reasoning – reasoning about what to do. That seems true for means-ends reasoning, reasoning from moral principles and case-by-case reasoning applied to action decision or judgement. It does not, however, neatly fit the rationality of case-by-case reasoning applied to classifications, to testimony assessment, to balance-of-considerations reasoning, to credibility judgements, or to evaluations. A case could be made that the latter sorts of reasoning either always, or for the most part, bear eventually on reasoning about what to do. But that risks stretching “practical reasoning” beyond any useful denotation. Better, at least at the outset, to remain openminded about the possibility that each type within this broad class obtains its rationality in a distinctive way. 4. The challenge of deductivism Whether these variants of presumptive reasoning/arguments represent a distinct class presupposes that their premise-conclusion relation cannot adequately be modeled as deductive or by some well-established inductive model. To be sure, deductive reconstruction is always possible for any reasoning or argument. That is, with the addition of “missing”/ “tacit”/ “unexpressed” premises, the reasoning or argument can be interpreted so that the premises entail the conclusion. But are reasoning and arguments of these types correctly understood “as attempts at deductive arguments” or reasoning (see Groarke 1999: 1)? Are they, for instance, best interpreted as follows? 1. P. 2. If P, then C, other things being equal. 3. In this case, other things are equal. j 4. In this case, C. The trouble with this interpretation is that 3 is too strong. It is not possible to be justified in asserting categorically that other things are equal. Thus, a re-

58

J. Anthony Blair

sponsible reasoner or arguer would not assert 3, but rather something like 3’: “In this case, other things appear to be equal.” But on that formulation, 1-3’ do not entail 4. Perhaps there is another interpretation of the reasoning that does permit responsible deductive reconstruction: 5. P. 6. If P, then, if other things are probably equal, probably C. 7. In this case, probably other things are equal. j 8. In this case, probably C. The trouble with this interpretation is that premise 6 is too tentative. In presumptive reasoning or argument the contention is that the default definitely applies, in the absence of reasons or evidence to the contrary. By replacing 2 with 6, deductive reconstruction is obtained only by distorting the original reasoning or argument. Groarke (1999: 5) correctly points out that deductive reconstruction does not imply that the conclusion is certain, but only that the premises necessarily imply the conclusion. Now, this claim is standardly taken to mean that the set of the premises and the denial of the conclusion is inconsistent – that it is a contradiction to affirm the premises and deny the conclusion; that there is no possible world in which the premises are true and the conclusion false. Yet in the case of presumptive reasoning schemes, it is impossible be sure that other things are equal. A world in which the evidence fitting a scheme all points to the conclusion, and a thorough consideration of the circumstances uncovers no excepting condition, yet some excepting condition has been overlooked, is always possible, however unlikely. (In this respect, presumptive reasoning/arguments are akin to inductive reasoning/arguments.) In addition, it is no defect in such reasoning or arguments that their premises do not entail their conclusions. These conclusions are always, in principle, subject to reconsideration in the light of new information which reveals a fuller understanding of the circumstances. It would appear, then, that deductive reconstruction without distortion is not possible, and that it is not plausible to suppose those who reason or argue presumptively are trying to reason or argue in a way that can be modeled as a deduction. Assuming that inductive reasoning and argument are also obviously different, it then follows that presumptive schemes represent a distinct sort of reasoning or argument.


59

5. Conclusion This paper has been programmatic. The goal has been join my voice to others’ (Govier’s 1987, 1999, Scriven’s 1987, Walton’s 1996a) in drawing attention to what appears to be a distinctive class of arguments or reasoning that cannot be reduced to the standard categories of “deductive” or “inductive.” As the examples listed illustrate, presumptive reasoning and arguments represent a significant subset of all human reasoning and argument. They deserve thorough and systematic study. References Blair, J. Anthony. 1995. Premise adequacy. In F.H. van Eemeren, R. Grootendorst, J.A. Blair and C.A. Willard, (eds.), Proceedings of the Third ISSA Conference on Argumentation.Vol. III, Reconstruction and Application, 191-202. Amsterdam: SicSat. Copi, Irving M. and Cohen, Carl. 1990. Introduction to Logic. 8th edition. New York: Macmillan Publishing Company. Eemeren, Frans H. and Grootendorst, Rob. 1983. Speech Acts in Argumentative Discussions. Dordrecht: Foris. van Eemeren, Frans H. and Kruiger, Tjark. 1987. Identifying argumentation schemes. In Frans H. Eemeren, Rob Grootendorst, J. Anthony Blair, and Charles A.Willard, (eds.), Proceedings of the Conference on Argumentation 1986, Vol. 2, Argumentation: Perspectives and Approaches, 70-81. Dordrecht-Holland/Providence-U.S.A.: Foris Publications. Gabbay, Dov M., Hogger, C. J. and Robinson, J.A. (Eds.). 1994. The Handbook of Logic in Artificial Intelligence and Logic Programming, Vol. 3: Nonmonotonic Reasoning and Uncertain Reasoning. Oxford: The Clarendon Press. Govier, Trudy. 1987. Problems in Argument Analysis and Evaluation. DordrechtHolland/Providence-U.S.A.: Foris. ——. 1999. The Philosophy of Argument. Newport News, VA: Vale Press. Groarke, Leo. 1999. Deductivism within pragma-dialectics. Argumentation, 13.1: 1-26. Harman, Gilbert. 1986. Change in View, Principles of Reasoning. Cambridge, MA: MIT Press (Bradford). Kienpointner, Manfred 1987. Towards a typology of argumentative schemes. In Frans H. Eemeren, Rob Grootendorst, J. Anthony Blair, and Charles A.Willard, (eds.), Proceedings of the Conference on Argumentation 1986, Vol. 1 Argumentation: Across the Lines of Discipline, 275287. Dordrecht-Holland/Providence-U.S.A.: Foris Publications. ——. 1992a. How to classify arguments. In F.H. van Eemeren, R. Grootendorst, J.A. Blair and C.A. Willard, (Eds.), Argumentation Illuminated, 178-188. Amsterdam: SicSat. ——. 1992b. Alltagslogik, Struktur und Funktion von Argumentationsmustern, Stuttgart-Bad Cannstatt: frommann-holzboog. Oxford English Dictionary. 1971. Compact Edition. Oxford: Oxford University Press. Pinto, Robert C. 1995. The relation of argument to inference. In F.H. van Eemeren, R. Grootendorst, J.A. Blair and C.A. Willard, (eds.), Proceedings of the Third ISSA Conference on Argumentation.Vol. I, Perspectives and Approaches, 271-286. Amsterdam: SicSat. Ross, W.D. 1930. The Right and the Good. Oxford: The Clarendon Press. Schellens, Peter Jan. 1987. Types of argument and the critical reader. In Frans H. Eemeren, Rob Grootendorst, J. Anthony Blair, and Charles A.Willard, (eds.), Proceedings of the Conference on

60

J. Anthony Blair

Argumentation 1986, Vol. 3, Argumentation: Analyses and Practices, 34-41. DordrechtHolland/Providence-U.S.A.: Foris Publications. Scriven, Michael. 1987. Probative logic: review and preview. In Frans H. van Eemeren, Rob Grootendorst, J. Anthony Blair, and Charles A.Willard, (eds.), Proceedings of the Conference on Argumentation 1986, Vol. 1 Argumentation: Across the Lines of Discipline 7-32. Dordrecht: Foris. ——, 1991. Evaluation Thesaurus. Fourth edition. Newberry Park, CA: Sage Publications. Walton, Douglas. 1996a. Argument Schemes for Presumptive Reasoning. Mahwah, NJ: Lawrence Erlbaum Associates. ——. 1996b. Argument Structure, A Pragmatic Theory. Toronto: University of Toronto Press. ——. 1998. Ad Hominem Arguments. Tuscaloosa and London: The University of Alabama Press. Wellman, Carl. 1971. Challenge and Response: Justification in Ethics. Carbondale & Edwardsville: Southern Illinois University Press. Wisdom, John. 1957. Explanation and proof. Lectures presented at the University of Virginia. Unpublished manuscript. Cited in Govier, 1987, 1999.

Argument schemes and the evaluation of presumptive reasoning

61

ROBERT C. PINTO Argument Schemes and the Evaluation of Presumptive Reasoning: some Reflections on Blair’s Account In “Presumptive Reasoning/Argument: An Overlooked Class,” Professor Blair makes a largely convincing case that “presumptive reasoning/argument represents a sui generis class of reasoning/argument” (Blair 1999, p.56). The strength of Blair’s paper lies in the fact that a.) he demonstrates through the use of examples that there is a large class plausible arguments/inferences that don’t fit the usual models of good deductive or of good inductive reasoning b.) he sets out a list of four properties of presumptive reasoning – properties whose joint presence might be taken to be criterial for the class in question. I agree with Blair that there is a large number of arguments/inferences that cannot be assessed in terms of the techniques of formal deductive logic or of the probabilistic standards usually applied to so-called inductive inferences. I agree, moreover, that recent work on argument schemes and the critical questions associated with them promises to yield useful tools for dealing with such reasoning. However, I have reservations about two aspects of the story that Blair tells, and in what follows I will attempt to lay out the grounds for those reservations. 1) In what sense does presumptive reasoning/argument constitute a sui generis class? Blair says that presumptive reasoning/argument “has not, at least in philosophy, received much attention as a distinctive class with a wide variety of subtypes that is on a par with, and deserving as careful attention as, ‘deductive’ and ‘inductive’ reasoning or argument” (see Blair, p. 57). Clearly, he wants to contrast presumptive reasoning with both “deductive” and “inductive” reaso-

62

Robert C. Pinto

ning – though (wisely, in my view1) he puts the words ‘deductive’ and ‘inductive’ in scare quotes. My reservation – albeit a minor one – is over the attempt to contrast presumptive reasonings with “inductive reasonings” (or perhaps better: to contrast “valid” presumptive reasonings with reasonings that are inductively strong). That contrast is problematic, for at least three reasons. First, the concepts of ‘inductive argument’ and ‘inductive strength’ are not sharply defined. Blair himself acknowledges that presumptive reasonings are inductive “in the vacuous sense that they are non-deductive”. But, he says, “[n]one exhibits the standard kinds of inductive reasoning found in science” (see Blair, p. 54). Yet there is, as far as I know, no precise criterion for determining which of the reasonings that occur in science are inductive and which are not.2 It might be tempting to identify inductive inference with those inferences whose strength can be measured in straightforward ways by the calculus of probability (and in particular with the help of inferential or projective statistics).3 But that I think yields too narrow a concept of inductive inference – since (absent the success of the Bayesian program) it would exclude the confirmation of theoretical hypotheses, which these days is a paradigm of (so-called) inductive reasoning. Second, there are kinds of reasoning which arguably belong to the class of presumptive reasonings, but which are also standardly classed as inductive arguments. Thus both Blair and Walton count argument from analogy as presumptive, while Copi treats it as a species of inductive argument (see Copi and Cohen 1990, p. 357), and even Trudy Govier recognizes two species of argument from analogy, one of which she calls inductive analogy (see Govier 1989, p. 141). Moreover, Walton includes Argument from Evidence to Hypo1 In the literature of logic it is common to find people trying to classify arguments or inferences as deductive or inductive, and maintaining that the appropriate standards for judging an argument or inference depend on the category to which it belongs. With Brian Skyrms, I hold that to be a mistake. Rather, we can ask about any argument, ‘Is it deductively valid?’ And if the answer to that question is ‘No,’ we should then proceed to ask, ‘Is it inductively strong or is it inductively weak?’ See Skyrms 1966. The usual attempts to classify arguments as deductive or inductive inevitably bog down in questions of whether the argument’s author “claims” or intends it to be taken as deductively valid or inductively strong – and in many cases there is simply no answer to that question (e.g., where the arguer lacks the logical sophistication and conceptual apparatus to think of arguments in terms of deductive validity and inductive strength). 2 Authors like Marcello Pera have argued, persuasively I think, that scientific discourse contains a broad range of arguments that he calls rhetorical, at least some of which would fall into Blair’s or Walton’s listings of presumptive arguments. See Pera 1994, esp. chapters 3 and 4. If that is so, then we can’t define inductive (as opposed to presumptive) arguments/inferences as those that occur in science. 3 Indeed, Balir’s examples of inductive support of are of this sort (Blair, p. 51).


63

thesis among his list of argumentation schemes for presumptive reasoning (see Walton 1995, pp. 67-71 – though on p. 46 he expresses some hesitation about its inclusion in the list), while Blair’s suggestion (Blair, p. 54) that inductive inferences are those found in science would seem to lead to the view that Argument from Evidence to Hypothesis is inductive. Finally, one of the principal features of presumptive reasonings is the fact that they are defeasible or that they are instances of nonmonotonic1 reasoning. That is to say, additional information, consistent with the premisses of the original argument or inference, can undermine or override the support that those premisses give to their conclusion. (Thus ‘Tweety is a bird’ supports the conclusion that Tweety can fly; but the additional information that Tweety has a broken wing cancels or overrides the inference from ‘Tweety is a bird’ to the conclusion that Tweety can fly.) But defeasibility has long been recognized to be a feature of “inductive” or probabilistic inferences as well (see, for example, Chisholm 1977, pp. 71-72). And indeed, it is necessarily a feature of any argument or inference that is not deductively valid.2 In virtue of this, most of the things that Blair says about presumptive reasonings on pp. 54-56 seem also to apply to inferences that are inductively strong (at least if patterns of “inductive inference” can be counted as argument schemes). That leaves us with the question of whether and how the defeasibility of presumptive inferences differs from the defeasibility of inductive and/or probabilistic inferences. These considerations do not, I think, undermine anything central to what Blair is saying about presumptive reasoning. And I am inclined to think presumptive reasonings should be contrasted with the inferences usually deemed inductive. But these considerations indicate the need to elaborate the precise relation between presumptive reasoning and inductive inference and to clarify the force of the claim that presumptive reasoning is sui generis. 2) Are there normative argument schemes? In characterizing the relationship of argument schemes to presumptive reasonings, Blair distinguishes between descriptive schemes and normative schemes. The latter in contradistinction to the former are described as “patterns of good reasoning” (Blair, p. 54, emphasis mine). Moreover, he characterizes the nor1 On nonmonoticity, see the discussion in Walton 1995, pp. 21-23, and van Bentham’s explanation in van Bentham 1996, esp. pp. 29-30. 2 Let [p, q, r] be a set of premisses that support but do not entail a conclusion c. From the fact that they don’t entail c, it follows that not-c is consistent with that set of premisses. But then the enhanced set of premisses [p, q, r, not-c] will fail to support c, since it will in fact entail not-c.

64

Robert C. Pinto

mativity of such schemes in the following ways (which he appears to treat as roughly equivalent): C.1 Each scheme “describes a pattern of reasoning or argument the premisses of which, if true, provide support for the conclusion, but only if other things are equal.” C.2 Each scheme “is a default form of good reasoning,” though “any such scheme is subject to override in a variety of circumstances.” C.3 Each instance of a normative scheme “is prima facie good reasoning, but whether it is good reasoning all things considered depends on the circumstances of the case.” C.4 “The grounds that instantiate the premise-types of a given normative scheme create a presumption if favor of the proposition instantiating the conclusion….” C.5 In a dialogue, “the party with grounds instantiating a normative scheme thereby shifts the burden of proof to the other party.” Implicit in these claims, I think, is Walton’s description of argumentation schemes as “a formal pragmatic structure of arguments that is the counterpart to logical forms of inference in semantics” (Walton 1995, p. x). In discussing arguments of a deductive sort, Blair himself has said: When the illative core structure [of an argument or inference] has the property that the premisses entail the conclusion, it is, with rare exceptions, a good argument or good reasoning in that respect – that is, with respect to the support that the premisses, taken as given, provide for the conclusion. (Blair, p. 50).

And of course in standard cases, we determine that premisses of an argument entail its conclusion by determining that the argument instantiates one of the logical forms of inference.1 Analogously, Blair seems to be suggesting in C.1C.5 that we can determine whether the premisses of an argument create a presumption in favor of its conclusion by determining whether we are dealing with an instance of a normative argument scheme. Now I want to question whether this way of viewing the function of argument schemes in the appraisal of arguments and inferences is correct and/or fruitful. That is to say, I want to question whether the argument schemes that Blair and Walton recognize really do have the normative force as1 Interestingly enough, as Blair himself concedes (Blair p. 39), the fact that premisses entail a conclusion does not guarantee that they support it – circular or question-begging arguments often entail but do not genuinely support them. (There are other exceptions as well; see Pinto 1994, pp. 121-122.). Indeed, one can say of instances of the logical forms of inference what Blair says about instances of normative argument schemes in C.3 – namely, that each such instance “is prima facie good reasoning, but whether it is good reasoning all things considered depends on the circumstances of the case.”


65

cribed to them. Since I am questioning whether these argument schemes are in fact normative, I shall henceforth refer to the schemes in question simply as recognized argument schemes, in order to leave open the question whether they are indeed normative. Let me begin by distinguishing between the two following two claims: P.1 Any good presumptive argument/inference creates a presumption in favor of its conclusion (and shifts to burden of proof to anyone who would dispute that conclusion). P.2 Any instance of a recognized argument scheme should be presumed to be a good presumptive argument/inference (though that presumption can be overridden in special circumstances). I take (P.1) to be uncontroversial in this context – indeed, I think that it captures part of the essence of what it is to be a good presumptive argument/inference. The second of these claims (P.2) is, I think, pretty much equivalent to Blair’s C.2 – and, I suspect, to C.1 and C.3 as well. And it is (P.2) that I want to question. To help motivate my reservations here, let me use an example of a fairly simple piece of presumptive reasoning, and then canvas several of the kinds of criticisms that would lead us to reject either the reasoning or its conclusion. Let me take as an example the Argument from Sign, whose scheme is represented in Walton 1995 (p. 49) as follows: A is true in this situation. B is generally indicated as true when its sign, A, is true, in this kind of situation. Therefore, B is true in this situation. Consider now two examples that fit this scheme (the first is from Walton, p. 49): Case 1: Bob is covered with red spots. Therefore, Bob has the measles. Case 2: When I purchased that ring from a reputable jewelry store, it came in a box that was labeled “24 karat gold” Therefore, the ring is made of 24 karat gold. Notice that each of arguments requires an unexpressed generalization to the effect that one thing is a reliable sign of another: that being covered with red spots is a reliable sign of having the measles, that the label on a box in a reputable store is a reliable indicator of the contents of the box.

66

Robert C. Pinto

Consider now some of the potential criticisms that might be made of these arguments, or some of the potential defects that might be found in them: 1. The first expressed premiss might be defective or might be challenged. For example, Bob’s face and torso might have seemed to be covered with red spots because Bob was viewed through a stained window pane. 2. The unexpressed generalization might be challenged, for a number of different sorts of reasons a) The validity or reliability of the generalization might be challenged. For example, are red spots a reliable sign of measles, and/or is it reasonable to assume that they are? As a matter of fact, measles is just one of many medical conditions that could be responsible for red spots, so that this generalization might well be challenged. b) Is the strength of the correlation between the sign and the signified great enough to warrant a presumption? For example, the presence of red spots gives some support to the hypothesis that a subject has measles; but in the estimation of many the support would be too weak to create a presumption that measles is present, or to shift the burden of proof to anyone who would resist that conclusion. c) Does the generalization, even if reliable, provide the appropriate sort of evidence for the context at hand. In the context of informal dealings with friends and family members, the store label probably does provide an adequate ground for presuming the metal to be 24 karat gold. But for insurance purposes it would surely be considered insufficient; for insurance purposes something like written assurance from a qualified expert, or the results of chemical tests, would be required. 3. Though the initial premiss is OK (the box was indeed labeled “24 karat gold”), and though the unexpressed generalization is sufficiently strong and of a sort appropriate to the context at hand, there are additional facts that undermine the inference. Suppose, for example, that an earthquake occurring near the jewelry store in question had emptied most containers of their contents, and that a stock clerk with little knowledge of precious metals was responsible for placing items of jewelry back in their containers. Knowledge that such a thing had occurred would undermine the inference in Case 2 – without, it should be noted, providing evidence that the conclusion of that inference is false. 4. Though the inference does not suffer from any of the kinds of defects catalogued so far, there is additional evidence that overrides the inference in question, by supporting the negation of its conclusion. For example, when the ring in Case 2 is brought to a highly reputable jeweler for


67

repair, he reports after careful examination that it is made of gold that’s only 14 karats. Now I want to call attention to what I consider to be an important difference between the defects catalogued in (1) and (2), and the defects catalogued in (3) and (4). The defects catalogued in (1) and (2) render an inference one that should not have been made in the first place. As I see it, a respondent who objects to a proposed presumptive inference on grounds that fall under (1) or (2) is claiming in effect that proponent has failed to shift the burden of proof. Let me call the allegation that a proposed inference suffers from such defects a Type I criticism of that inference. It is worth noting, moreover, that the kinds of objections envisaged in cases (1), (2a) and (2b) are each of a sort that can be raised against arguments or inferences normally be classed as “deductive” or “inductive.” They are in no way specific to presumptive reasonings. However, as I see it, in cases (3) and (4), the initial presumptive inference is not defective. The inference goes through, as it were – in the sense that it in a dialogue it shifts the burden of proof to the respondent or in a monolectical situation it gives a reasoner a genuine ground for holding the conclusion. In the absence of additional factual information about the case under scrutiny, the conclusion retains its status as a presumption. And when we say that arguments and inferences which avoid the defects catalogued in (1) and (2) shift the burden of proof, what we mean amounts simply to this: it is incumbent upon a respondent who would dispute the conclusion of such an argument to bring forth precisely additional factual information that will either undermine the inference, as happens in (3), or override the inference, as happens in (4). Let me call the allegation that a presumptive inference can be undermined or overridden in these ways a Type II criticism of the inference. Now, if you examine the examples at the beginning of Blair’s paper and his gloss on those examples in section 3, I think you will find in every case that he is calling attention to inferences and arguments that avoid the defects catalogued in (1) and (2), but have conclusions that are open to revision on the basis of additional information that undermines or overrides those inferences, on the pattern of (3) and/or (4) – he is calling attention to the possibility of Type II criticisms of such inferences and arguments. Let me return now to the claim I want to dispute, namely (P.2) Any instance of a recognized argument scheme should be presumed to be a good presumptive argument/inference (though that presumption can be overridden in special circumstances). I would maintain, rather, that when confronted with an instance of Argument from Sign, we ought not to assume it is a good presumptive argument until we

68

Robert C. Pinto

have satisfied ourselves that it is not subject to a Type I criticism – anymore than we should assume that a deductively valid argument is a good argument until we have satisfied ourselves that its premisses are reasonable to believe, that it does not beg the question, etc.1 Indeed, consider the two critical questions that Walton (following Hastings) associates with Argument from Sign (Walton 1995, p. 48): 5. What is the strength of the correlation of the sign with the event signified? 6. Are there other events that would more reliably account for the sign? I would maintain that until we’ve answered the first of those two questions, we should not presume that an Argument from Sign even gets off the ground. Question 1 does not ask after information that would “override” the presumption that the argument in question is a good one; rather it asks for information without which we can’t form a reasonable opinion about whether the argument is a good one. Question 1 belongs, I think, to Type I criticism. By contrast, question 2 could well be interpreted as a request for additional information that would undermine or override any presumption afforded to the conclusion by the sign cited in the premisses of the argument. Question 2 is an attempt to instigate Type II criticism. Yet if argumentation schemes aren’t normative in the way Blair claims them to be – if they are not the counterparts of the logical rules of inference in semantics – what is their role in the evaluation of arguments and inferences? As I’m inclined to see the matter, the point of identifying argument schemes lies in the critical questions associated with them. It isn’t the schemes that do the evaluative work, it’s we who do the evaluative work by judiciously challenging premisses, identifying the risks of basing conclusions (if only presumptive conclusions) on those premises, and enriching our grasp of potentially relevant fact that might alter the significance of the facts brought to light in the premisses. Evaluating an argument or inference is a matter of probing its strengths and weaknesses, and we do well if the initial impetus for our probing comes from the critical questions that have, in the past, proved fruitful in the probe of this, that or the other sort of argument. Associating arguments with schemes is an effective way of associating them with particular sets of critical questions – questions that will initiate the probes that may issue in evaluative verdicts. 1 Nor should we presume that an argument which resembles inductively strong arguments in a general way is itself inductively strong, until we have satisfied ourselves both that its premisses are reasonable, and that the degree of support it lends to its conclusion is such as to warrant acceptance of its conclusion as probable.


69

I don’t claim to have shown that Blair’s account of the force and function of argumentation schemes is wrong, or that mine is right. I do hope, however, that I’ve brought to light an alternative to Blair’s way of conceiving the matter. And I hope, as well, to have made it clear that at this stage more work needs to be done on these issues. References Benthem, Johan van (1996). “Logic and argumentation”. In van Benthm, et al. 1996, pp. 27-42. Benthem, Johan van, et al. (1996). Logic and Argumentation. North Holland, Amsterdam: Royal Netherlands Academy of Arts and Sciences. Blair, J. Anthony (1999). “Presumptive Reasoning/Argument: An Overlooked Class”. Protosociology 13 (1999). Chisholm, Roderick (1977). Theory of Knowledge, second edition. Englewood Cliffs, N. J.: Prentice-Hall, Inc. Copi, Irving and Carl Cohen (1990). Introduction to Logic, Eight Edition. New York: Macmillan Publishing Company. Govier, Trudy (1989). Analogies and Missing Premisses. Informal Logic XI,3, pp. 141-152. Johnson, Ralph and J. Anthony Blair (1994). New Essays in Informal Logic. Windsor, Ontario: Informal Logic. Pera, Marcello (1994).The Discourses of Science, trans. Clarissa Botsford. Chicago and London: University of Chicago Press. Pinto, Robert (1994). Logic, “Epistemology and Argument Appraisal”. In Johnson and Blair 1994, pp. 116-124. Skyrms, Bryan (1966). Choice and Chance; an introduction to inductive logic. Belmont, Calif.: Dickenson Pub. Co. Walton, Douglas (1995). Argument Schemes for Presumptive Reasoning. Madison, Wisconsin: Lawrence Erlbaum Associates.

70

Douglas Walton

DOUGLAS WALTON The New Dialectic: A Method of Evaluating an Argument Used for Some Purpose in a Given Case The purpose of this article is to explain to the reader how to evaluate an argument critically with respect to how that argument was supposedly used for some purpose in a goal-directed type of conversational exchange. Of course, only so much can be explained in a short article. Nevertheless, by introducing the reader to the recent literature on argumentation, and to the main methods that are being developed in that literature, some insight into how to use the new techniques can be given. The subject is controversial, as well. Some deny that there can be any binding standards for judging when an argument is reasonable or not, or whether one argument is better than another, as used in a given case, in natural language discourse. Others feel that the only objective methods that can or should be used to support such judgments are those of deductive and inductive calculi of the kind that have for so long been central in the field of logic. Because of the controversial nature of the subject, many fundamental logical and philosophical questions are raised in this article. What is the difference between argumentation and reasoning? What are the forms of presumptive inference commonly used in everyday argumentation? How are such inferences chained together to make up sequences of reasoning? How is such reasoning used for different purposes in different kinds of arguments where two parties are involved in a dispute? What are the relationships between argument and explanation? How can presumptive arguments be seen as instances of inference to the best explanation that provide a tentative basis for accepting a conclusion? How does arguing fix an arguer’s commitment to specific propositions that can then be attributed to her, as representing her position? How do new argument moves in a disputation change that commitment? How should change of commitment be organized in different contexts of dialogue when two parties reason together for some collaborative purpose? In particular, what are the rules for retraction of commitments? Can these different normative models of dialogue be formalized, with precisely defined components and sets of rules? These are the new kinds of questions that are being asked. In this article, an exposition of developments in argumentation theory is presented that gives the reader a revealing glimpse into how these questions are

The New Dialectic : A Method of Evaluating an Argument

71

being answered. The framework in which the questions are answered is called the “new dialectic”. The new dialectic is mainly concerned with the most common kinds of everyday arguments, and is based on presumptive reasoning rather than deductive or inductive logic. Presumptive reasoning takes the from of an inference in which the conclusion is a guess or presumption, accepted on a tentative basis, and subject to retraction as a commitment, should new information come in. The new dialectic shares many common features with the old dialectic of Plato and Aristotle, but is also different from it in other features. In the new dialectic, argumentation is analyzed and evaluated as used for some purpose in a type of dialogue underlying a conversational exchange. Each type of dialogue has its own standards of plausibility and rationality against which to measure the successful use of an argument. Thus the new dialectic has a relativistic aspect that makes it different from the classical positivistic philosophy. But it also has a structure with logical standards of evaluation of argument use, which makes it different from postmodern anti-rationalism. 1. Old and New Ways of Thinking about Thinking The positivistic philosophy that has been the orthodox way of thinking in the universities since the Enlightenment, and that has become even more dominant in the twentieth century took science, especially the hard sciences of mathematics, physics and chemistry, as the models of correct reasoning. The kinds of reasoning used outside science, like the kind of thinking used in everyday deliberations, or the kind for reasoning used in law and ethics, were simply dismissed as “subjective”. Scientific reasoning, seen as consisting of deductive logic and inductive logic of the kind represented by the probability calculus, was taken to represent all of logic. This positivistic philosophy has failed to yield a theory of reasoning and argumentation that was useful for cognitive science, to explain how human or robotic agents can reason in collaboratively carrying out practical tasks on the basis of communicating shared assumptions, or in criticizing the views of another agent. This positivistic view point saw deductive logic, of the kind one would find in the reasoning in Euclidean geometry, as the model of correct reasoning. The positivistic viewpoint saw reasoned thinking along the lines the receiving of knowledge and the revising of beliefs. Even the notion of an agent or thinker came to be abstracted out of the equation. An argument was seen as simply a “designated” set of propositions. Beyond propositional and quantificational logic, further conceptualization came to be based on the highly abstract notion of a possible world. Reasoning was thought to be about various kinds of “accessibility relations” between pairs of possible worlds.

72

Douglas Walton

In the positivistic viewpoint, concern with thinking centered on a highly abstract subject called “epistemology”. In epistemology, knowledge and belief existed, and the central problem was how to connect the two concepts. The subject matter seemed to consist mainly of endless controversies about whether knowledge could be defined as justified true belief. Both these lines of advance turned out to be dead ends. Spectacularly so. Traditional logic and analytical philosophy turned out to have not enough to say that was useful to tell those in the field of artificial intelligence how rational thinking should work. The aging priests of analytical philosophy still talk enthusiastically about possible worlds and justified true beliefs, at some centers of learning. But judging from the little respect accorded logic, the humanities and philosophy in recent years, not many are listening any more. To fill the gap, postmodern theories came along to advocate throwing aside any pretense to rationality. But this viewpoint did not turn out to be a success either. Then along came argumentation theory. A rational argument was now described in terms of acceptance (commitment) instead of belief or knowledge. Departing from the impersonal framework of deductive logic, argumentation theory saw an argument as a dialogue exchange between two parties who are reasoning together. No longer exclusively concerned with deductive and inductive forms of argument, argumentation theory considered many forms of presumptive inference based on intelligent guessing that leads to a tentatively acceptable conclusion on one side of a dialogue. This new viewpoint is not acceptable to traditional analytical philosophers who are so used to looking at the world through positivistic lenses. But computing, and artificial intelligence in particular, has taken to argumentation like a duck to water, finding this new theory very useful for all kinds of purposes. Philosophers will also accept these views at some point in the future, once they realize that scientists have accepted it. 2. The New Dialectic This article presents an introduction to and outline of a new dialectic designed to be used to normatively evaluate any arguments used in a given case. Originating in the old dialectic of the ancient philosophers, the new dialectic is centrally concerned with the most common kind of arguments used in everyday conversations, which is based on presumptive reasoning rather than deductive or inductive logic. Presumptive reasoning is always based on default by an argument from ignorance that tilts a burden of proof one way or another on an unsettled issue. Presumptive reasoning can be used in a closed world


73

situation or an open world situation. In what Reiter (1980, p. 69) calls the closed world assumption, all the positive information in a data base is listed, and therefore negative information is represented by default. For example, in case 1 below (Reiter, 1980, p. 69), the design of an artificial intelligence question-answering system is considered. Case 1 Consider a data base representing an airline flight schedule and the query “Does Air Canada flight 113 connect Vancouver with New York?” A deductive question-answering system will typically treat the data base together with some general knowledge about the flight domains as a set of premises from which it will attempt to prove CONNECT(AC113, Van,NY). If this proof succeeds, then the system responds “yes”. But, as Reiter indicates, the interesting fact is that if the system does not succeed, it will typically respond “no”. In other words (Reiter, 1980, p. 69), “Failure to find a proof has sanctioned an inference.” Such an inference by default has often traditionally been called a lack-of-knowledge inference or an argumentum ad ignorantiam. The same kind of inference drawn by Reiter’s artificial intelligence system might be drawn by a human reasoner who is scanning the flight monitor at the airport. When such a person sees the listing of flights on the airport monitor, he presumes that all the flights are listed, and infers by default that any flight not listed is not offered. Of course, he can always check by asking at the desk. But the inference by default is most likely a pretty good guess, because the person is plausibly justified in assuming that the closed world assumption is met in this case. If the closed world assumption is in place, then the inference is more than just a hypothesis or assumption. Once the data base is closed off, the negative inference by default from the data base is such that we can say that the conclusion is known to be true (assuming that all the data in the data base are known to be true). But most argumentation in everyday conversational exchanges is based on the open world assumption, where we are uncertain whether the data base is complete, or think it is incomplete, and we have to make a guess about conclusion to infer. It is precisely in such a guesswork situation that presumptive reasoning becomes most useful. Hence the real practical importance in argumentation of the form of reasoning often called the argument from ignorance, outlined in section five below. Presumptive reasoning works by making a guess, in the form of drawing a conclusion and accepting it on a tentative basis, subject to possible retraction as a commitment, should new argumentation alter the case. A presumptive

74

Douglas Walton

inference gives an arguer a reasons for accepting a conclusion, even though that conclusion may later have be withdrawn if critical questions are asked in the dialogue. Nevertheless, although such forms of inference are neither deductively valid nor inductively strong, they do have a certain standing or bindingness in a dialogue. The form of inference does have a logical structure. It tells you that if you accept the premises, and the form of the argument is structurally correct, then unless you can ask the right critical questions, you must accept the conclusion. Such forms of inference called argumentation schemes represent the logical structures of these kinds of arguments. In the new dialectic, reasoning is defined as a chaining together of inferences. Reasoning can be used for various purposes. It is used in explanations as well as in arguments. An argument is a use of reasoning to fulfill a goal of a dialogue, of one of six basic types. All argument is to try to settle some issue that has two sides. The present view is called dialectical because every argument is seen as a case of two parties reasoning together for some purpose. What is primarily important in the dialectical system of evaluating arguments is not (at least centrally) knowledge or belief, but something called commitment. This term refers to the acceptance of a proposition by a participant in a dialogue. Commitments do not have to be logically consistent with each other. But if a proponent’s set of commitments are apparently inconsistent in a case, a respondent can challenge that inconsistency, and call for some resolution or explanation of the apparent inconsistency. The new dialectic is amenable to formalization, as shown by Hamblin (1970), and Walton and Krabbe (1995). But the formal structure required is quite different from that of the traditional deductive propositional and quantifier logics. The formalization is a game-like structure in which, there are two participants, a proponent and a respondent. Each takes turn, making moves generally asking question, replying to question, and putting forward arguments. The type of dialogue, as a whole, has a goal, and each participant has an individual goal (or role). The rules define what kinds of moves are considered legitimate for the purpose of contributing collaboratively to the goal of the dialogue. In some types of dialogues, the individual goals of the participants are opposed to each other. Other types of dialogue are not adversarial in this same sense, and the participants are supposed to cooperate with each other and help each other to work towards the goal together. 3. The Old Dialectic The ancient art of dialectic was a philosophical activity in which two persons took part. The questioner first poses a problem, the respondent chooses a


75

position, and then the questioner draws inferences based on the respondent’s answers (Kapp, 1942, p. 12). Evidently the questioner’s aim was to draw out conclusions that raise doubts about the respondent’s position, perhaps even conclusions that appear to be inconsistent with the respondent’s position. But the exact purpose of dialectic as an organized and goal-directed activity is obscure to the modern reader of ancient philosophical texts. It is an art that fell into obscurity after the fall of the ancient world. Plato called dialectic “the art concerning discussions” (Robinson, 1953, p. 69), but tended to shift his meaning of the term to describe “the ideal method, whatever that might be.” (Robinson, 1953, p. 70). Zeno of Elea was supposed by many in the ancient world to be the inventor of dialectic (Kneale and Kneale, 1962, p. 7), but it is not known exactly what Zeno had in mind. According to the Kneales (p. 7), Plato in the Parmenides refers to Zeno’s claim to have written a book in which he draws out some absurd philosophical consequences of another person’s philosophical view. This reference may have been the basis of Aristotle’s later remark, quoted by Diogenes Laertius and Sextus Empiricus, that Zeno was the inventor of dialectic. For a clearly articulated explanation or analytical theory of what dialectical argument is supposed to be, Aristotle’s account is probably the most useful source. Aristotle, in On Sophistical Refutations (165b3-165b4), defined dialectical arguments as “those which, starting from generally accepted opinions (endoxa), reason to establish a contradiction.” According to the Topics (100b22), generally accepted opinions are “those which commend themselves to all or to the majority or to the wise - that is, to all of the wise or to the majority or to the most famous and distinguished of them. “ For Aristotle then, dialectic was the use of reasoning to draw logical consequences from premises that are generally accepted opinions. What kind of activity was this? It seems to be an art of the gadfly. The best example we can seem to come up from the ancient world with is the critical use of argumentation by Socrates, as portrayed in the Platonic dialogues. Socrates questioned the opinions of those who thought themselves to be wise, and were assumed to be wise by others or the majority. He also probed and questioned conventionally accepted views. He often drew contradictions and logical problems from these views, using logical reasoning. His method was of asking a sequence of questions, where each question is based on the previous answer given by a speech partner. The ancients attached quite a lot of importance to dialectic as an art. In addition to its use to teach skills of arguing, and for arguing in casual conversations, Aristotle even saw dialectic as being useful for questioning and discussing the axioms or first principles (archai)of the sciences (Topics, 101b4). This idea is simply not acceptable to the modern way of thinking since the

76

Douglas Walton

so-called Enlightenment. Pascal and Descartes argued that the model of good reasoning should be that of Euclidean geometry, where theorems are rigorously deduced from self-evident axioms. This paradigm of scientific reasoning as the all-powerful method of reasoning represents the modern way of thinking, where it is assumed that, time and time again, science has proved “common sense” to be wrong. The idea of some kind of reasoning outside science, being brought to bear on science and used to question or criticize the assumptions of science, where such critical argumentation commands rational assent, is alien to our modern ways of thinking. The ancient art of dialectic had no place in the modern way of thinking. When ancient logic was rediscovered in the middle ages, it was Aristotle’s theory of the syllogism (deductive reasoning) that came to dominate as the paradigm of logical argumentation. Aristotle’s fallacies retained a toe-hold in the logic textbooks, but was never again taken seriously as a central part of logic. When the Stoic logic of propositions was formalized around the beginning of the twentieth century, deductive formal logic eclipsed all other branches of the subject. Dialectic was a lost art. 4. Types of Dialogue The new dialectic is built on the pragmatic foundation introduced by Grice, in his famous paper on the logic of conversation (1975). According to Grice, an argument should be seen as a contribution to a conversation between two parties, and should be evaluated as a good (useful) argument or not, on the basis of whether it made a collaborative contribution to the moving forward of the conversation towards its goal. This new approach was tied in much more closely to how we use and judge arguments in everyday conversational exchanges. The argument was now to be evaluated with respect to how it was used for different purposes in different types of conversational exchanges. But what are these different types of exchanges, and what are their goals? By specifying the precise rules and goals of the different types of conversational exchanges, The New Dialectic (1998) offered a new method for evaluating arguments that could be applied to the informal fallacies (sophistical refutations) that held such a place of importance in the early applied logic of Aristotle. By going back to the Aristotelian roots of logic as an applied, practical discipline, the new dialectic brought out and formulated, in modern terms adequate for state of the art argumentation theory, many of the leading ideas expressed in the ancient works on dialectical argument that heretofore appeared obscure, and were for so long treated as peripheral in logic. For the first time it becomes possible to apply objective logical standards of evaluation to

77


arguments in everyday conversational exchanges on controversial topics where real conflicts of opinions exist. The new dialectic offers a framework of rationality for judging an argument as correct or incorrect insofar as it has been used adequately or not in a given case to contribute to goals of dialogue appropriate for the case. An argument is judged to have been used in a fallacious way in a dialogue insofar as it has been used in such a way as to impede the goals of the dialogue. A dialogue, to use our generic term, or a conversation to use Grice’s term, is defined as a goal-directed conventional framework in which two speech partners reason together in an orderly way, according to the rules of politeness, or normal expectations of cooperative argument appropriate for the type of exchange they are engaged in. Each type of dialogue has distinctive goals, turn-taking moves, and methods of argumentation used by the participants to work towards these goals together. Six basic types of dialogue are described in the new dialectic - persuasion dialogue, the inquiry, negotiation dialogue, information-seeking dialogue, deliberation, and eristic dialogue. TYPES OF DIALOGUE Type of Dialogue

Initial Situation

Participant’s Goal

Goal of Dialogue

Persuasion

Conflict of Opinions

Persuade Other Party

Resolve or Clarify Issue

Inquiry

Need to Have Proof

Find and Verify Evidence

Prove (Disprove) Hypothesis

Negotiation

Conflict of Interests

Get What You Most Want

Reasonable Settlement that Both Can Live With

Information-Seeking

Need Information

Acquire or Give Information

Exchange Information

Deliberation

Dilemma or Practical Choice

Co-ordinate Goals and Actions

Decide Best Available Course of Action

Eristic

Personal Conflict

Verbally Hit Out at Opponent

Reveal Deeper Basis of Conflict

Table 1

78

Douglas Walton

Each of these types of dialogue is put forward in the new dialectic as a normative model which specifies broadly how a given argument should be used, in one of these contexts, in order to be correct, or to be defensible against the criticism that it is incorrect, erroneous or fallacious. The most central type of dialogue, for the typical purposes of applied logic as it is taught in classrooms today, in courses of critical thinking, is the persuasion dialogue. In this type of dialogue, the proponent has a particular thesis to be proved, and the respondent has the job of casting doubt on that thesis by raising questions about it. In some instances however, the dialogue can be symmetrical. Both participants have a thesis to be proved, and each has the aim of persuading the other to accept his or her thesis. In a persuasion dialogue, each party takes the initial concessions of the other as premises, and then by a series of steps, tries to use these premises in arguments designed to persuade the other party, by means of using rational argumentation, to give up his original thesis. One very common problem is that during the sequence of argumentation, the dialogue may shift from a persuasion dialogue to another type of dialogue, say to a negotiation or a quarrel. Such dialectical shifts can be very confusing, and are associated with many of the major informal fallacies. Another problem associated with the job of evaluating many common arguments is to apply the dialectical method to cases of mixed discourse, like sales pitches, political debates, and legal arguments in trials. Such cases are characterized by the mixing of two or more types of dialogue. They also frequently involve dialectical shifts from one type of dialogue to another, during the same sequence of argument. 5. Presumptive Reasoning Recent concerns with the evaluation of argumentation in informal logic and speech communication have more and more begun to center around nondemonstrative arguments that lead to tentative (defeasible) conclusions, based on a balance of considerations. Such arguments do not appear to have structures of the kind traditionally identified with deductive and inductive reasoning. However, they are extremely common, and are often called “plausible” or “presumptive,” meaning that they are only tentatively or provisionally acceptable, even when they are correct. These arguments shift a weight of evidence to one side of a balance, thus supporting a conclusion that was previously in doubt. But such a weight can, as the argument continues, be shifted back to the other side. Presumptive reasoning is based on pragmatic implicatures drawn out by a hearer on the basis of what a speaker’s remarks can normally be taken to imply


79

in the context of an orderly, cooperative dialogue. Hence presumptive reasoning is more subject to contextual interpretation, and also more subject to defeat (and error) than logical reasoning of the more familiar deductive and inductive sorts. Perhaps for these reasons, presumptive reasoning has generally been ignored in logic, and excluded from serious consideration as inherently “subjective,” in the past. However (Walton,1996) shows that the inference structures of presumptive reasoning are well worth investigating, and do help us to critically evaluate argumentation of the kind that powerfully influences people in everyday speech, on all kinds of controversial issues where presumptive conclusions are drawn. What kind of support is given to a conclusion on the basis of presumptive reasoning? The kind of support given is different from that given by a deductively valid argument or an inductively strong argument. For logicians long accustomed to working with deductive and inductive standards of argument support, the move to a third standard is not easy to make, especially when the new standard typically gives only a weaker kind of support that is tentative in nature, and subject to withdrawal in many instances. One way to introduce the new idea is through the idea of a generalization. In deductive logic, the universal quantifier is used to stand for a kind of generalization, ‘For all x, if x has property F then x has property G’, in which a single counter-example defeats the generalization. This type of generalization could be called absolute, in the sense that it is equivalent to ‘There are (absolutely) no x such that x has F, but does not have G.’ In contrast, inductive generalizations, of the form ‘Most, many, or a certain percentage (expressed numerically as a fraction between zero and one) of things that have property F also have property G’. This kind of generalization is not absolute, because it allows for a certain number of counter-examples (but not too many). The kind of generalization characteristic of presumptive reasoning is based on a type of generalization of the form, ‘Normally, but subject to exceptional cases, if something has property F, it may also be expected to have property G.’ This kind of conditional is subject to defeat in unusual or unexpected situations that are not normal, or what one would normally expect. Our confidence in it is tentative, because, as we find more out about a situation, it can come to be known that it differs from the normal type of situation. For example, we normally expect that if something is a bird, it flies. But in a particular case, we may find out that Tweety, a bird, is a penguin, or has a broken wing. This new information will defeat the inference based on the normal presumption that Tweety, since he is bird, is an individual that flies. Many (statisticians, in particular) feel that presumptive reasoning can be shown to be a species of inductive reasoning, perhaps so-called “subjective

80

Douglas Walton

probability”. This claim appears dubious, because encountering something that is not normal (and is an exception) in a particular case is often a surprise, and does not appear to be based on statistical regularities. However, it should not be entirely rules out, perhaps, that some sort of statistical model of inference may be found that fits presumptive reasoning. So far, however, no numerical formula for evaluating presumptive reasoning appears to have been found, or at least any criterion that fits all kinds of cases. Presumptive reasoning is highly familiar in computer science, where it is frequently identified with abductive inference, or what is often called “inference to the best explanation”. But abductive inference does not appear to be the same thing as presumptive reasoning, even though the two kinds of reasoning appear to be closely related. The terminology on these questions is not settled yet, and there are many different theories about how these two kinds of reasoning are related. 6. Abductive Inference What is called abductive reasoning in computer science, or “inference to the best explanation” in philosophy, is a distinctive kind of inference that goes from given data to a hypothesis that best explains the data. An example from ordinary conversation is given by Josephson and Josephson (1994, p. 6) : Case 2 Joe : Why are you pulling into this filling station? Tidmarsh : Because the gas tank is nearly empty. Joe : What makes you think so? Tidmarsh : Because the gas gauge indicates nearly empty. Also, I have no reason to think that the gauge is broken, and it has been a long time since I filled the tank. Classified as an argumentation scheme, we would say that the argument in this case is an instance of argument from sign. The gas gauge indicating “nearly empty”, or being low, is a sign that the tank is nearly empty. Giving such a sign or indication is what this instrument is designed for. But we can also see the reasoning as an inference to the best explanation. Tidmarsh considers two possible explanations for the indication on the gas gauge. One is that the gas is low. The other is that the gauge is broken. But, as he says, there is no reason to think that the gauge is broken. So the best explanation, from what is known in the case, is that in fact the gas in the tank is nearly empty. Abductive reasoning is common in science (Josephson and Josephson, 1992, p. 7). Some would even argue that the typical type of reasoning to a


81

scientific hypothesis from given data is abductive in nature. According to Harman (1965), when a scientist infers the existence of atoms, she is reasoning from the best explanation of the given scientific data. Peirce (1965, p. 375) classified all inference as falling into three classifications - deduction, induction, and what he called “hypothesis”, which corresponds to abductive inference. Peirce, in a work called ‘The Proper Treatment of a Hypothesis’ (Eisele, 1985, pp. 890-904) described abduction as a kind of guessing, characteristic of scientific reasoning at the discovery stage, which can save experimental work by narrowing down the possible hypotheses to be tested to the most plausible candidates. Peirce clearly identified abductive reasoning as a distinctive type of inference that is important in science, and described it as a kind of plausible reasoning or “guessing” (Eisele, 1985, p. 898). He also frequently wrote about abduction as a kind of “explaining” process in ‘The Proper Treatment of a Hypothesis’ (Eisele, 1985, p. 899). Peirce ( 1965, p. 375) defined hypothesis as occurring in the following kind of instance : “where we find some very curious circumstance, which would be explained by the supposition that it was a case of a certain general rule, and thereupon adopt that supposition.” He gives the following two examples (p. 375). Case 3 I once landed at a seaport in a Turkish province ; and, as I was walking up to the house which I was to visit, I met a man upon horseback, surrounded by four horsemen holding a canopy over his head. As the governor of the province was the only personage I could think of who would be so greatly honored, I inferred that this was he. This was an hypothesis. Case 4 Fossils are found ; say, remains like those of fishes, but far in the interior of the country. To explain the phenomenon, we suppose the sea once washed over this land. This is another hypothesis. Case 4 is clearly an example of a scientific kind of hypothesis, while case 3 is the kind of inference to the best explanation that is so common in everyday reasoning. These two cases illustrate very well how presumptive reasoning is typically based on a kind of inference to the best explanation that Peirce called “hypothesis” or abductive inference. One thing that is very interesting about abductive reasoning is that it combines the two functions of argument and explanation. An abductive inference may be used as an argument to support a conclusion, but the basis of that support utilizes an explanation, or a series of explanations. We normally think of argument and explanation as two different speech acts, or uses of

82

Douglas Walton

discourse. The purpose of putting forward an argument to a hearer is normally to prove some proposition that is in doubt to the hearer. The purpose of offering an explanation of to a hearer is to take a proposition that both the speaker and hearer presume is true, and to make it understandable to the hearer. These two speech acts are inherently different, but in abduction they are combined. The explanation function is part of what supports or makes possible the carrying out of the argument function. The distinction is often explained in computer science as one of the direction of the reasoning. Normally, the reasoning in an argument moves forward. That is, the inference goes from the premises to the conclusion. But in abductive reasoning, there is also a backwards movement of inference. The conclusion is taken as a given data, and then a search back is made to try to determine the best explanation for this data. The general form an abductive inference can be represented as follows compare (Josephson and Josephson, 1994, p. 14). Form Abduct: D is a collection of data. Hypothesis H explains D. No other hypothesis explains D as well as H. Therefore H is plausibly true (acceptable). Another kind of scientific reasoning that has been recognized as being based on abductive inference is medical diagnosis, a species of argumentation from sign that reasons from the given data to the best explanation. For example, in the case of a diagnosis of measles, the physician might reason as follows. Measles: If the patient has red spots, then the patient has measles. The patient shows red spots. Therefore, the patient (plausibly) has measles. The conclusion drawn in the Measles Case is tentative, and based on the assumption that there is no better explanation of the red spots. The converse of the major premise, ‘If the patient shows red spots, the patient has the measles’ is not true, since showing red spots is only one sign of having the measles, and it s not a conclusive sign, by any means. It is just one sign that can be used abductively as evidence for measles, in the absence of any better explanation of the red spots. At any rate, it is this sort of case analysis that is the basis of the often-expressed theory that abductive reasoning takes the form of argument called affirming the consequent. This analysis has its tricky aspects, however,


83

and the reader might be referred to (Walton, 1996b, pp. 264-281) for a fuller discussion of the forms of abductive inference and argumentation from sign. Josephson and Josephson (1994, p. 266) report that use of a seven-step scale of plausibility values worked very well in modeling plausible reasoning in medical diagnostic systems. Such use of numerical values might suggest that plausible reasoning could be formalized using the probability calculus. But they report (p. 268) that interpreting plausibility as probability just didn’t work out very well. After going into the various technical possibilities in such a modeling, they conclude (p. 269) that there is no “significant computational payoff” in it. They conclude (p. 270) that there is “a need to go beyond probability”, and look in some other direction for a way to model plausible reasoning. They conclude (p. 272) that is unlikely that plausibility, of the kind characteristic of abductive reasoning, can ever be quantified, in the way that probability is quantified in the probability calculus. To make “smart machines” that can reason plausibilistically, they conclude, we need to go in a different direction. How is one to judge, by some clearly defined standard, whether a particular presumptive inference is structurally correct or not, in a given instance? This has become an extremely important question in recent times, and could even rightly be called the central question of argumentation theory. The problem is that while we in the field of logic are highly familiar with deductively valid forms of argument, and somewhat familiar with inductively strong forms of argument, we appear to lack forms of argument corresponding to cases of presumptive reasoning. However, there is a literature on what are called argumentation schemes. 7. Argumentation Schemes Argumentation schemes are the forms of argument (structures of inference) that enable one to identify and evaluate common types of argumentation in everyday discourse. In (Walton, 1996), twenty-five argumentation schemes for presumptive reasoning are identified. Matching each argumentation scheme, a set of critical questions is given. The two things together, the argumentation scheme and the matching critical questions, are used to evaluate a given argument in a particular case, in relation to a context of dialogue in which the argument occurred. An argument used in a given case is evaluated by judging the weight of evidence on both sides at the given point in the case where the argument was used. If all the premises are supported by some weight of evidence, then that weight of acceptability is shifted towards the conclusion, subject to rebuttal by the asking of appropriate critical questions.

84

Douglas Walton

One premise of an argumentation scheme typically takes the form of a presumptive generalization, of the kind described above, to the effect that if x has property F, then normally x will also have property G. For example, the argumentation scheme for argument from sign is the following (Walton, 1996, p. 49) : A is true in this situation. B is generally indicated as true when its sign, A, is true, in this kind of situation. Therefore, B is true in this situation. The second premise is a presumptive generalization which says that if A is true, then generally, but subject to exceptions, B is also true. But such a generalization is defeasible. It, taken with the other premise of the scheme, shifts a weight of acceptance to the conclusion. But counter-argumentation in a case may subsequently overturn acceptance of the argument by withdrawing that weight, or even introducing new evidence that places a weight against it. The list of presumptive argumentation schemes given in (Walton, 1996) offers a useful, modern, accessible, systematic and comprehensive account that the reader can use as an aid in interpreting, analyzing and evaluating natural language argumentation in everyday conversations. Perelman and Olbrechts-Tyteca (1958) identified many distinctive kinds of arguments used to convince a respondent on a provisional basis. Arthur Hastings’ Ph.D. thesis (1963) made an even more systematic taxonomy by listing some of these schemes, along with useful examples of them. Recently Kienpointner (1992) has produced an even more comprehensive outline of many argumentation schemes, stressing deductive and inductive forms. Among the presumptive argumentation schemes presented and analyzed in (Walton, 1996) are such familiar types of argumentation as argument from sign, argument from example, argument from commitment, argument from position to know, argument from expert opinion, argument from analogy, argument from precedent, argument from gradualism, and the slippery slope argument. Helpful examples of each type of argumentation are given and discussed. In other recent writings on argumentation, like van Eemeren and Grootendorst (1992),there is a good deal of stress laid on how important argumentation schemes are in any attempt to evaluate common arguments in everyday reasoning as correct or fallacious, acceptable or questionable. The exact nature of the relationship between argument from sign and abductive inference is an interesting question. The measles inference above is clearly an instance of argument from sign, and it is also an instance of abductive inference, or inference to the best explanation. Many instances of argument


85

from sign can also be very well analyzed as cases of inference to the best explanation. Consider the example of argument from sign cited in (Walton,1996, p. 47). Case 5 Here are some bear tracks in the snow. Therefore, a bear passed this way. In this case, the premise is based on the observing of a particular shape and appearance of imprints in the snow identified as a bear tracks. The best explanation of the existence of such tracks would be (in the right context) that a bear passed this way. But not all cases of argument from sign appear to fit the inference to the best explanation format this well. For example, dark clouds could be a sign of rain, but can we say that the rain is the best explanation of the dark clouds? Not without some twisting and stretching, which leads one to suspect that not all cases of argument from sign fit the abductive model. Nonetheless, it is clear that abductive inference and argument from sign are very closely related. 8. Arguments from Ignorance A useful and encouraging aspect of the new dialectic is that it shows how the presumptive argumentation schemes are the essential underlying structure needed for the analysis of the traditional informal fallacies. Govier (1988, p.34) has rightly stressed that the fallacies approach to argument evaluation is incomplete, precisely because it needs to be based on a prior understanding of the various types of good argument involved. She notes that the traditional fallacies are most often based on good arguments that are not “propositionally valid”, but nonetheless represent “ways of arguing well”. Not knowing exactly what these “ways” are has been the biggest obstacle to the analysis of the fallacies. This new dialectic represents a breakthrough by showing exactly what these ways of arguing well amount to. Three important informal fallacies that have been analyzed in depth by the dialectical method elsewhere can be used to illustrate this point. These three are: argument from ignorance (argumentum ad ignorantiam), hasty generalization (secundum quid), and argumentation from consequences (argumentum ad consequentiam). The argument from ignorance has the following simple form: it has not been shown that propositions A is true, therefore it may be presumed that A is false. To see how common this type of inference is, consider the following dialogue.

86

Douglas Walton

Case 6 Bob : Is Leona Helmsley still in jail? She’s probably out by now. Helen : Maybe she’s still in there, because we’d probably hear about it if she got out. How the argument from ignorance is used in this dialogue can be better understood by placing it in the following sequence of reasoning : (1) we would probably hear about it if Helmsley got out (because the story would be widely reported in the media), but (2) we haven’t heard about it, therefore (3) she’s probably not out, i.e. (4) she’s still in there. Neither Bob nor Helen has any definite evidence, one way or the other, yet Helen’s presumptive inference that “Maybe she’s still in there.” seems to justify the drawing of a reasonable conclusion by default. It is more of a conjectural than a solid conclusion, but it would seem to be an exaggeration to call her argument fallacious. It is an argument from ignorance that can be evaluated within the context of dialogue in which it was used in the above case, as a nonfallacious use of presumptive reasoning. In the standard treatment of the logic textbooks, this type of argument has traditionally classified as a fallacy. But in this case, and in many other cases studied in (Walton, 1996a), the argument is evaluated dialectically as a presumptively reasonable argument that could function as tie-breaker in a balance-of-considerations case, thereby being used to draw a justifiable conclusion in a dialogue. The argument from ignorance still continues to be condemned as a fallacy by many commentators. Gaskins (1992) portrays it as a powerful and all-pervasive argumentation strategy used especially influentially in modern discourse to base conclusions on suspicion about all forms of authority. According to Gaskins, the skillful modern advocate uses the following form of the argument from ignorance: “I win my argument unless you can prove me wrong.” According to Gaskins, this fallacious form of argument has come to dominate not only legalistic argumentation, but also scientific inquiry, and modern moral disputes on public policy. However, the argument from ignorance has had powerful defenders. Socrates, in the Apology, was allegedly told by the Oracle at Delphi that he was the wisest man of all, because he was the only one who admitted his ignorance. Here a subtle form of ignorance, knowing what you do not know – was used as a premise to support the conclusion that awareness of limitations could be a kind of wisdom. Recently, Witte, Kerwin and Witte (1991) have championed this Socratic attitude of trying to teach medical students an awareness of the limitations of medical knowledge, instead of the more usual me-


87

thod of instruction, which they see as dogmatic memorizing of facts, as a better method of medical education. They have set out a Curriculum of Medical Ignorance that advocates use of the argument from ignorance as a model of medical reasoning. Another common use of the argument from ignorance is in computer science. It is a very familiar kind of reasoning in this area to search through a knowledge base, find that a particular proposition sought for is not there, and then presumptively conclude that this proposition is false. This is called default reasoning in computer science. A familiar example (Reiter, 1987) similar to case 1would be the kind of case where you look at an airport monitor listing all the stops on a flight between Vancouver and Amsterdam, and you see that Winnipeg is not listed as one of the stops. When you infer that the plane does not stop at Winnipeg, you are using an argumentum ad ignorantiam. But this argument could be reasonable, assuming the convention that all stops are listed on this monitor (what Reiter calls the closed world assumption). The dialectical examination of the argument from ignorance, indicated by the cases above, suggest that in many cases, it is a reasonable argument that has legitimate and common uses in scientific and medical reasoning, and in the kind of knowledge-based reasoning common in computer science. It can also be shown to commonly used, and quite reasonable in many cases, in legal reasoning, where, for example, it is reflected in the basic principle of criminal law that a person should be presumed to be not guilty, in the absence of proof of guilt. 9. Applying the New Dialectic to Cases In the new dialectic, each case is unique, and a given argument needs to be judged on the basis of the text of discourse available, representing the information in that case. To evaluate the argument, we have to ask whether the closed world or the open world assumption is appropriate. And we have to ask what type of dialogue the participants were supposedly engaged in, as far as the information given in the case indicates. If these facts are not determined by the information given in the case, then the best we can do is to evaluate the case hypothetically, based on assumptions that may hold or not, from what we know about the case. Any assessment of this kind is contextual. We have to look at the case as a whole, and then evaluate the argument in light of how it was used in that case, as far as we can determine the relevant details of the case. A dialectical assessment of a particular argument as used in a case appears to be quite different from the usual use of deductive logic to assess whether a given argument in natural language is valid or invalid. But it may not be as

88

Douglas Walton

different as is widely assumed. Even so, formal treatment would appear to be more limited in the dialectical assessment, because in many cases of everyday conversational exchanges, there is little or no explicit agreement between the participants on exactly what type of dialogue they are supposed to be taking part in. Political argumentation, for example, is typically mixed, being partly persuasion dialogue, but also partly negotiation and eristic dialogue (as well as involving information-seeking dialogue and deliberation, in many cases). Even so, each type of dialogue does have a formal structure, and once the argument is modeled in a given structure, formal techniques, of the kind currently in use in AI, can be brought to bear on it. Nonexplicit premises can be articulated, and so forth. Formalization is both possible and helpful, but since the data in a case can be massive, how the formal structure is to be applied to the known data in a case requires considerable preparatory work in interpreting what the text of discourse should be taken to mean, in a given case.. So here is the problem. It is not so much a problem of any difficulties of formalization, as it is a problem of determining the body of data one takes to be the case. The new dialectic has many uses. But among the foremost of these uses is that of evaluating an argument found in a given text of discourse in a specific case. Logic has long dealt with the evaluation of such arguments, but the assumption has always been that the argument is just a designated set of propositions -a set of premises and a conclusion - and that everything else about it is trivial or unimportant from a viewpoint of its logical evaluation. In the new dialectic, what is now important is not only the set of propositions, but the context of dialogue in which these propositions have (presumably) used for some purpose. Now each case needs to be looked at with respect to the argument is supposedly being used - is it being used to persuade, to negotiate, or to deliberate, for example? The same argument could be seen as quite reasonable if it was supposed to a negotiation tactic, whereas it could be rightly judged to be fallacious if supposedly used as a contribution to a critical discussion on some specific issue. For example, the same threat that is relevant in a negotiation dialogue could be irrelevant if used as an argument if it is supposed to be part of a critical discussion. The first task in evaluating any given argument in a text of discourse is to identify the sequence of reasoning using an argument diagram to pinpoint each proposition, and to identify the inferences dawn from such propositions to the conclusions that were derived. This task requires the filling in of nonexplicit premises and conclusions. Hence the context of dialogue is vitally important even at this stage, because judging how to charitably fill in such missing links should be guided by dialectical factors, like the arguer’s commitments, as known in the case. Once agreement is reached on what the premises and


89

conclusions of the given argument are supposed to be, as judged from the textual and contextual evidence of the case, the next step is to examine each inference in the chain of reasoning. Where weaknesses are found, the appropriate critical questions need to be asked. Then finally, looking over the whole sequence of reasoning exhibited in the diagram, the question of relevance needs to be raised. Where is the argument leading? Is it leading towards the ultimate conclusion that is supposed to be proved in the type of dialogue exchange of which it is supposedly a part? These are the questions relating to dialectical relevance. 10. Uses of the New Dialectic The new dialectic offers a practical method of identifying, analyzing and evaluating authentic cases of everyday argumentation that does not require an abstraction form the realities of discourse in a natural language setting. A new option is offered. No longer do arguments have to be judged solely in relation to standards of deductive and inductive reasoning. In the past, the dominance of this more restrictive approach has led to a distorted view of many everyday, presumptive arguments, often leading to the conclusion that such arguments are somehow inherently defective, or even fallacious. In the new dialectic, the different contexts of use of such arguments are taken into account - included are such factors as the type of dialogue, the stage of a discussion, the commitments of the discussants, and other factors that are specific to a case of argumentation in which two speech partners are attempting to reason together for some collaborative purpose. The targeting of the new dialectic to factors of how an argument was used in a specific case gives a more practical way of evaluating everyday argumentation. Judged by such practical standards, an argument can be evaluated as weak in certain respects, and open to appropriate critical questions, without being so badly off that it should be condemned as fallacious, implying an error or defect that is beyond repair. As well as providing new tools for the analysis of arguments by teachers of courses on critical thinking and informal logic, the new dialectic has other important fields of application. It is clearly applicable to many common kinds of arguments, and problems of argumentation, in fields like artificial intelligence, experts systems, legal and medical reasoning, and use of evidence in academic research (not excluding scientific argumentation). The new dialectic is a framework for reasoning that strikes a healthy balance between descriptive empirical research on argumentation and normative or abstract logical methods of setting standards for good arguments. Such a balance, although lacking in the past, is healthy because neither the empirical

90

Douglas Walton

or the normative approach, by itself, can provide a method of argument evaluation that is both objective in standards and that fits the realities of real cases of argumentation in a way that is practically useful. Real arguments in conversational exchanges are mixtures of different types of dialogue, and have different standards of plausibility and rationality. Each type of dialogue has its own distinctive goals, its own procedural rules, and its own standards of burden of proof. An argument that could be appropriate and reasonable might be highly inappropriate, or even fallacious, in another type of dialogue. Cases also frequently involve mixtures of two or more types of dialogue, and shifts from one type of dialogue to another. In the new dialectic, judging how an argument was used in a given case is a contextual matter. Much depends on what type of dialogue the participants were supposedly engaging in, when the argument in question was put forward by one of the parties in the discussion. The evidence on which to judge a case, therefore, must be sought in the context of use of the argument. As shown in section nine above, sometimes there is plenty of such evidence available in a given case, but in other cases, the best that can be done is to make a conditional evaluation of the case, based on what evidence is given in that case. Such conditional evaluations, despite their hypothetical and incomplete nature, can, in many cases, be extremely helpful in diagnosing the logical strengths and weaknesses of an argument.

Acknowledgements I would like to thank Manfred Kienpointer for comments that enabled me to avoid several errors, and helped me to improve the exposition of many points. This paper was supported by a research grant from the Social Schiences and Humanities Research Council of Canada. I would also like to thank the Canada-U.S. Fulbright Program and the Department of Communication Studies of Northwestern University for funding that helped to support the work in this paper. References Aristotle, On Sophistical Refutations, Loeb Classical Library, Cambridge, Mass., Harvard University Press, 1928. Aristotle, Topics, Loeb Classical Library, Cambridge, Mass., Harvard University Press, 1939. Carolyn Eisele, Historical Perspectives on Peirce’s Logic of Science, vol. 1, Berlin, Mouton, 1985. Richard H. Gaskins, Burdens of Proof in Modern Discourse, New Haven, Yale University Press, 1992. Trudy Govier, ‘Ways of Teaching Reasoning Directly,’ Critical Thinking : Proceedings of the First British Conference on Informal Logic and Critical Thinking, ed. Alec Fisher, East Anglia, University of East Anglia, 1988, 30-38. J. Paul Grice, ‘Logic and Conversation’, in The Logic of Grammar, ed. Donald Davidson and Gilbert Harman, Enrice, Carlifornia, 1975, 64-75. Gilbert Harman, ‘The Inference to the Best Explanation’, Philosophical Review, 74, 1965, 88-95. Arthur C. Hastings, A Reformulation of the Modes of Reasoning in Argumentation, Evanston, Illinois, Ph.D. Dissertation, 1963.


91

John R. Josephson and Susan G. Josephson, Abductive Inference : Computation, Philosophy, Technology, New York, Cambridge University Press, 1994. Ernst Kapp, Greek Foundations of Traditional Logic, New York, Columbia University Press, 1942. Manfred Kienpointner, Alltagslogik : Struktur und Funktion von Argumentationsmustern, Stuttgart, Fromman-Holzboog, 1992. William and Martha Kneale, The Development of Logic, Oxford, Clarendon Press, 1962. Charles S. Peirce, Collected Papers of Charles Sanders Peirce, vol 2, ‘Elements of Logic’, ed. Charles Hartshorne and Paul Weiss, Cambridge, Mass., Harvard University Press, 1965. Chaim Perelman and Lucie Olbrechts-Tyteca, The New Rhetoric : A Treatise on Argumentation, trans. J. Wilkinson and P. Weaver, 2nd ed., Notre Dame, University of Notre Dame Press, 1971 (First published, as La Nouvelle Rhetorique, in 1958). Raymond Reiter, ‘A Logic for Default Reasoning’, Artificial Intelligence, 13, 1980, 81-132. Raymond Reiter, ‘Nonmonotonic Reasoning,’ Annual Review of Computer Science, 2, 1987, 147-186. Richard Robinson, Plato’s Earlier Dialectic, 2nd ed., Oxford, Clarendon Press, 1953. Frans H. van Eemeren and Rob Grootendorst, Argumentation, Communication and Fallacies, Hillsdale, N.J., Lawrence Erlbaum Associates, 1992. Douglas Walton, Argumentation Schemes for Presumptive Reasoning, Mahwah, N. J., Erlbaum, 1996. Douglas Walton, Arguments from Ignorance, University Park. Pa., Penn State Press, 1996a. Douglas Walton, Argument Structure : A Pragmatic Theory, Toronto, University of Toronto Press, 1996b Douglas Walton, The New Dialectic : Conversational Contexts of Argument, Toronto, University of Toronto Press, 1998. Douglas N. Walton and Erik C. W. Krabbe, Commitment in Dialogue : Basic Concepts of Interpersonal Reasoning, Albany, State University of New York Press, 1995. Charles L. Witte, Ann Kerwin, and Marlys H. Witte, ‘On the Importance of Ignorance in Medical Practice and Education,’ Interdisciplinary Science Reviews 16, 1991, 295-2.

92

Manfred Kienpointner

MANFRED KIENPOINTNER Comments on Douglas Walton’s Paper: The New Dialectic: A Method of Evaluating an Argument Used for Some Purpose in a Given Case In his article, Walton outlines an alternative to both classical positivistic philosophy and postmodern anti-rationalism (section 1). Starting from insights of ancient dialectic and modern argumentation theory, he develops a framework called ‘New Dialectic’ (section 2). New Dialectic is mainly concerned with the most common kinds of everyday arguments, which is based on presumptive reasoning rather than deductive or inductive logic. What is presumptive reasoning? Walton’s answer is: ‘Presumptive reasoning works by making a guess, in the form of drawing a conclusion and accepting on a tentative basis, subject to possible retraction as commitment, should new argumentation alter the case’ (p. 73). Walton then outlines characteristic features of the old dialectic developed by ancient philosophers like Plato and Aristotle. He proceeds to criticize its lack of practical applicability and Aristotle’s highlighting of common sense as the basis of all reasoning, including scientific reasoning (section 3). In section 4, Walton describes a central assumption of the New Dialectic, namely, that argumentation has to be analyzed relative to the type of dialogue underlying a conversational exchange. He establishes a typology of dialogues with differing standards of plausibility and rationality (among them, the persuasion dialogue, the inquiry, the negotiation dialogue, the eristic dialogue). He also makes clear that real dialogues often are mixtures of two or more types of dialogue and that a dialogue may undergo a dialectical shift from one type to another. Next, Walton provides a closer description of presumptive reasoning. It is characterized by the fact that it only produces tentative conclusions, shifts the ‘weight of evidence to one side of a balance’, but is ‘more subject to defeat (and error)’ and ‘to contextual interpretation’ than familiar deductive and inductive types of arguments (p. 79). But he also states that ‘some sort of statistical model of inference may be found that fits presumptive reasoning’ (p. 80), which would make it possible to see presumptive reasoning as a sub-type of inductive reasoning. Walton also assumes that presumptive reasoning is closely related to abductive reasoning (section 5). In section 6, Walton deals with argumentation schemes as forms or structures of presumptive reasoning. In Walton (1996), he has established a typology

Comments on Douglas Walton’s Paper: The New Dialectic

93

of 25 argumentation schemes which can be used as an aid in interpreting, analyzing and evaluating everyday argumentation. Each argumentation scheme is accompanied with a set of critical questions which can point out potential weaknesses of inferences according to the schemes. Walton then points out that the New Dialectic can provide a fruitful approach to study traditional informal fallacies. Recent developments in fallacy theory have shown that many alleged instances of informal fallacies are defeasible arguments which show certain weaknesses, but are nevertheless acceptable given certain contextual circumstances. It it therefore not acceptable to dismiss all informal fallacies as completely unacceptable kinds of arguments. Walton gives the example of the argumentum ad ignorantiam, which has the following form: ‘it has not been shown that proposition A is true, therefore it may be presumed that A is false’ (p. 86). He demonstrates that there are a variety of cases and contexts where this type of argument can be used in a plausible way, like other kinds of presumptive reasoning. One example is default reasoning in computer science: if a particular proposition sought for in a knowledge base is not there, one can presumptively – which does not mean: deductively – conclude that it is false. Another example would be medical reasoning, where even a Curriculum of Medical Ignorance has been developed for teaching students an awareness of the limitations of medical knowledge (section 7). In the final section of his paper, Walton stresses that in the New Dialectic, ‘each case is unique and a given argument needs to be judged on the basis if the text if discourse available’ (p. 88). This does not preclude formalization, but creates the problem that in any given case a massive amount of data has to be taken into account to answer the following questions: which kind of dialogue underlies the argument? How can the sequence of reasoning be diagrammed? How can the missing premises and conclusions be filled in? Which critical questions can appropriately be asked? What is the relevance of the argument? (section 8). I now would like to give some comments concerning strenghts and weaknesses of Walton’s approach. In general, I would like to stress that Walton’s New Dialectic is a highly interesting and fruitful synthesis of traditional approaches and modern argumentation theory, including his many important works on fallacies, dialogue games, argumentation schemes and other areas of the study of argumentation (cf. e.g. Walton 1989, 1992, 1996). The New Dialectic offers a promising and realistic framework for looking at authentic instances of everyday argumentation, which does not require to idealize away the real world complexities of discourse. With the important concept of presumptive reasoning, New Dialectic is able to take a positive

94

Manfred Kienpointner

view towards the specific merits and deficiencies of everyday arguments. This means that they are no longer solely judged in relation to established kinds of reasoning like deductive or inductive resoning, which in the past often has led to the conclusion that they are always somehow deviant or even fallacious. Both for the analysis and the evaluation of everyday arguments it is indispensable to take into account the various contextual factors like the type of dialogue, the stage of the discussion, the commitments of the discussants etc. Therefore, Walton’s insistance on the individual case is to be appreciated. Moreover, the New Dialectic has many uses, as Walton rightly points out (p. 89). It provides a set of tools for the analysis and evaluation of everday arguments which can be used by teachers of informal logic and critical thinking courses. Furthermore, it has other important fields of application like artificial intelligence, expert systems, legal and medical reasoning. To me, the possibility of a broad practical application of a theoretical approach is far from being of only secondary importance. Finally, like other recent approaches in argumentation theory (e.g. the Pragmadialectics of the Amsterdam school; cf. Van Eemeren/Grootendorst 1994), the New Dialectic tries to establish a fruitful balance between descriptive and normative perspectives on argumentation. This is to be praised because neither purely empirical research nor ideal models of rational argumentation can do justice to all aspects of argumentative discourse. My more critical remarks concern only three minor points, which could be easily integrated into the research program of the New Dialectic. First of all, I think that Walton’s criticism of ancient dialectic is a bit too harsh. He writes that ‘the exact purpose of dialectic as an organized and goaldirected activity is obscure to the modern reader of ancient philosophical texts’ (p. 75). It is true that the dialogue rules listed in the eighth book of the Aristotelian Topics (155b3ff.) could not easily be transferred to everyday debates and discussions. But they were intended as part of the teaching practice at philosophical schools and not as principles for, say, political debates (cf. Moraux 1968). Moreover, when Walton criticizes that Aristotle even recommended dialectic as being useful for the discussion of the axioms and first principles of the sciences (‘This idea is simply not acceptable to the modern way of thinking, where it is assumed that, time and time again, science has proved ‘common sense’ to be wrong’, p.5), one could defend the Aristotelian position by pointing out that there is hardly another way of discussing axiomatic principles of sciences apart from using common sense arguments. Second, I would like to suggest that the relationship between presumptive and abductive arguments is clearified in some more detail. Peirce’s description

Comments on Douglas Walton’s Paper: The New Dialectic

95

and examples (1965: 372ff.) show that his abductive arguments are very close to presumptive arguments, and also follow some of the argumentation schemes (e.g. argument from sign) which are introduced by Walton (p. 81). A closer look at the similarities and differences of abductive and presumptive reasoning could be set on the agenda of future work on the New Dialectic. Third, I would like to make clear that in my book on argumentation schemes (Kienpointner 1992: 230, 245) I remain completely open to the idea that the classical deductive-inductive dichotomy might not be sufficient to deal with all types of arguments which can be empirically established. Therefore, it was not my primary intention ‘to stress deductive and inductive forms’ of arguing (cf. Walton, p. 85).

References Aristotle (1960): Posterior Analytics. Ed. and transl. by H. Tredennick/Topica. Ed. and transl. by E.S. Forster. London: Heinemann. Van Eemeren, F.H./Grootendorst, R. (eds.)(1994): Studies in Pragma-Dialectics. Amsterdam: SicSat. Kienpointner, M. (1992): Alltagslogik. Stuttgart: Frommann-Holzboog. Moraux, P. (1968): La joute dialectique d’après le huitième livre des Topiques. In: G.E.L. Owen (ed.)(1968): Aristotle on Dialectic. The Topics. Oxford: Oxford UP. 277-311. Peirce, Ch.S. (1965): Deduction, Induction and Hypothesis. In: Ch. Hartshorne & P. Weiss (eds.): Collected Papers. Vol. II: Elements of Logic. Cambridge, Mass.: Harvard Univ. Press. 372-388. Walton, D. N. (1989): Informal Logic. Cambridge: Cambridge Univ. Press. Walton, D.N. (1992): The Place of emotion in Argument. University Park: Pennsylvania State Univ. Press. Walton, D.N. (1996): Argumentation Schemes for Presumptive Reasoning. Mahwah, N.J.: Erlbaum.

96

Christopher W. Tindale

CHRISTOPHER W. TINDALE The Authority of Testimony ‘Be wary of the argument from authority.’ – Sokal and Bricmont (1998:178) Introducion One of the older and, perhaps, more ill-traveled forms of argument is what is known as the ‘appeal to authority’, the simple strategy of using the say-so of a recognized authority (person, source, or institution) as evidence in support of a claim or belief. I say ‘ill-traveled’ because for many people the ‘appeal to authority’ has been judged inherently fallacious, as implied in the warning from Sokal and Bricmont above. But there has developed in the textbook treatments a general sense that such appeals can be reasonable if they meet certain criteria. To take only one such textbook account as an example: the authority’s credentials must be stated; those credentials must be relevant to the claim/case in question; the authority must be free from bias in the case; there should be wide agreement among relevant experts; and the claim must concern an area of knowledge wherein such consensus would be possible (Groarke, Tindale, & Fisher, 1997: 275)1. Such criteria work well up to a point (usually helping students develop reasoning skills and recognize the better ways to proceed); but they sketch over or simply avoid some of the more difficult questions that can arise, for example: Should we distinguish between authorities and experts? How do we appraise competing claims of ‘authorities’ in cases where there is obviously no consensus, but consensus might be in the process of being formed? And how, most importantly, do we deal directly with authorities and evaluate their responses? The last question in particular points to an important problem involving the impact of authority on argumentation that takes us beyond the simple ‘appeal to authority’, narrowly defined. While the layperson may find herself listening to an arguer who appeals to an authority and then making a judgement about the appropriateness of that authority, other relevant discussions of authority will often involve cases where the layperson is directly confronted 1 Similar accounts can be found in a tradition stretching back to Bentham’s Handbook of Political Fallacies (1816).

The Authority of Testimony

97

with the authority in person, with no intermediary ‘appealer’. Some of the basic conditions for assessing an authority can still apply here, but the layperson also needs to be able to assess the authority’s responses, and there the gulf between them may seem impenetrable: whether we refer to authorities or experts, they are by definition more knowledgeable on the question at hand than is the layperson, who may simply be reduced to silence. An apparent paradigm case of this could be seen in the following: In a courtroom, the layperson takes the form of a jury member who must evaluate the testimony of an expert witness. But how is this to be done? What kind of background experience can assist the juror? I raise this case because it is not the kind of case that principally interests me in this paper. In this case, obviously the ability of the juror to appraise the testimony of the expert witness depends on the nature of the witness’ testimony. If he/she testifies to general matters of perception, like what can be seen in certain lighting or what can be lifted by a woman with a certain build, then the juror can rely on notions of plausibility and likelihood in judging the testimony. If the expert testifies on technical knowledge, then the juror is in a weaker position. But in fact, this ‘paradigm’ case is, in this respect, artificial in the sense that the expert witness in such a trial has a burden of proof upon her or him to make the testimony understandable to the juror. Properly conducted, the structure of the trial should facilitate that understanding. More difficult are the ‘everyday cases’ where we consume the advice of experts and authorities without those experts feeling compelled to explain themselves. Two things help here, and these are what I intend to explore throughout the paper: (i) Expert knowledge must in some way conform to objective conditions. In fact, increasingly in accounts of the argument form, we see an implicit recognition that authority is subordinate to objective evidence. But this evidence, this ‘objectivity’ has its source in general testimony as a primary way of knowing. (ii) Secondly, we can consider what kinds of questions need to be asked by the layperson of both the expert and the expert’s testimony. And, more importantly, how the gradual understanding of such questions and the ability to employ them factor in the development of a certain type of autonomous, competent knower. The Nature of the Problem The power of authority permeates all sectors of life and any appreciation of history indicates that this is no modern feature but one rooted in human development and culture. Of course, there is an important difference between authority that is derived from political or social (or institutional) standing, and

98


that which is derived from knowledge. But at a deeper level, these have often been intertwined as those ‘who knew’ were in a position to assume other types of authority. Along such lines Alberto Manguel speaks of ‘the argument of authority’ (1996:270), whereby a classical text and its author were often deemed infallible (Aristotle is an obvious example). Thus it became a common argumentative move to attribute one’s opinions to others, thus rendering them immune to criticism. Latter-day disdain for the ‘argument from authority’ may well be rooted in such early practices, even though we may recognize that the ‘argument from authority’ as a strategy for providing contributory evidence for a claim, and the ‘argument of authority’ to foreclose debate on a question are by no means synonymous. Furthermore, modern rejections of the argument form may well be based on apparently reasonable grounds. Consider two examples, one from literature and another from science. In E.M. Forster’s 1934 obituary of art critic Roger Fry, he eulogized him in the following terms: ‘What characterized him and made him so precious in twentieth-century England was that, although he was a modern, he believed in reason....[He] rejected authority, mistrusted intuition’(1936:50-51). As Forster explains this, Fry, when presented with an appeal to authority would express a preference to rather think about it. Hence, the contrast of appeals to authority and ‘reason’. There is an irony in this of course, since Forster’s plea for the mistrust of authority is based in part on the appeal to Fry’s practice. But the point is well made that accepting the views of an authority somehow involves circumventing the task of thinking through the matter oneself, and if one could do this thinking, it would be preferable. I emphasize this phrase because it captures the crux of the problem that should appear: can we (laypeople) think through the matter in away that obviates any appeal to authority? It is after all the nature of the authority as the possessor of some particular knowledge or experience which distinguishes them from the layperson in the matter at hand. If Fry can equally (or sufficiently) think through the matter for himself, then the point of the appeal is made redundant in his case. The point is illustrated further in the domain of science. By way of clarifying their invocation to be wary of the argument from authority1, Sokal and 1 This is probably the place to seriously consider one of the questions raised earlier, whether we should draw a distinction between ‘an authority’ and ‘an expert’, because here in the domain of science it may be thought to matter. Popular science writer, Carl Sagan insisted, for example, that because authorities have made mistakes in the past and will continue to do so, ‘[a]rguments from authority carry little weight’ (1995:210). Or, as he also puts it: ‘in science there are no authorities; at most, there are experts’ (Ibid.). We might legitimately wonder why experts would not succumb to the same fallibility that seems to plague authorities. But, this aside, ‘expert’ has a narrowness that in cognitive, and related argumentative, matters seems to have a clarity that is more helpful


99

Bricmont recommend following ‘the best of the natural sciences’ methodological principles, starting with this one: to evaluate the validity of a proposition on the basis of the facts and reasoning supporting it, without regard to the personal qualities or social status of its advocates or detractors’ (1998:178). This remark has both a positive and negative aspect to it. Negatively, it appears to interpret the ‘argument from authority’ in its non-cognitive, institutional sense, where the authority carries weight not necessarily because of what they know but because of their position or some feature of their character. Yet at the same time, they seem aware of the real nature of such an appeal because, positively, they advocate a personal investigation of the facts and reasoning supporting a proposition. Again, a commendable sentiment, but one which founders on two fronts. In the first instance, it does not seem to conform with actual scientific practice. Take as an example the investigation of the rock sample ALH84001, deemed to be a martian meteorite. Scientists working on the case used various forms of reasoning, from causal to analogical reasoning, in arriving at the conclusion that phenomena connected to the rock were ‘evidence for primitive life on Mars’ (McKay, et.al.,1996). But they also relied upon the ‘appeal to authority’ in its cognitive sense. While the thrust of their experimentation was on the chemical make-up of the rock, they based their research on the assumption that the rock was of martian origin and supported this assumption by citing the work of other scientists as reported in a paper by a further scientist in the journal Meteoritics. The correctness of that earlier research is crucial to the acceptability of McKay et.al.’s conclusions about the rock sample. If this case is representative of the kinds of activities that go on in scientific practice, and I submit that it is, then it illustrates that not only do scientists appeal to authorities, or experts, they have to make such appeals in order to draw information from cognate fields and avoid replicating studies that it would be impractical to repeat. The second problem relates back to the Roger Fry case. Here it is not a matter of whether laypeople should investigate for themselves rather than relying on the say-so of experts; it is a question of whether they able to do otherwise than rely on authorities. Charles Willard (1990) captures what may be considered the crux of the problem of the layperson’s confrontation with authority and what is troubling in both the Fry and Sokal/Bricmont recommendations. than the broader ‘authority’. Experts do not possess the wide authority of an Aristotle that transcends fields, but invariably have a more particular or specific knowledge of a limited field of inquiry or experience. While continuing to refer to the argument form ‘argument from authority’ I will understand it to refer to experts in this narrower sense.

100


As Willard presents the problem of authority (and twentieth century dependence upon it), it is not so much that we have difficulty assessing the claims made, but more that the predisposition to acquiesce to authorities threatens autonomy. Willard argues that public decision-making is firmly rooted in a reliance on authorities: ‘Public decision-making doesn’t use knowledge, it uses testimony – a tapestry of positions maintained by authoritative representatives of knowledge domains who presumably bridge the gap between disciplines and public decision-making’(16). But this poses a real dilemma, because while Willard believes that ‘it is presumptively rational in a consensualist world to argue from and acquiesce to authorities’ (11), this has the tendency to undermine personal autonomy. This is seen in the problems of appraisal. Willard makes the important observation that we are not concerned here with the appraisal of whether experts are correct – that is the job for experts themselves. We are concerned with deciding how expert testimony should be taken; whether it’s the direct attesting to some fact or experience, or the challenge of one expert by another. The extent of this concern emerges when he turns to the responses that have been made by argumentation theorists and critical thinkers. It has become the norm for such scholars to argue that the individual must learn to defend themselves against the ‘tyranny’ of authority by developing certain skills. Willard quotes Mickunas (1987) as typical of the kind of thing advocated by such people: While submitting to the authority of what has been attained and established by scientific enlightenment – the formal and technical disciplines – the person should reach through education, a level of understanding, independent judgment, and creativity, at which the scientific-technical values and norms are established. Only at this level the person can become competent to understand the rules, and their embodiment in concrete social life, and be in a position to pass an autonomous judgment and engage in free debate. (Mickunas, 1987: 336-337; cited in Willard: 19)

It will be recognized how well this conforms with the sentiments expressed by Fry and Sokal & Bricmont: the ‘problem’ of authority is that it subverts the power and opportunity of the individual to think/decide for her- or himself. But as Willard suggests, the real problem lies in this very kind of thinking about authority because it seems to advocate an ideal which is virtually unattainable. Willard writes: ‘If it puts a burden on individuals that no individual in the late 20th century can discharge, then it is an empty posture – a kind of arm-waving in the face of the problem of modernity’ (19). In this sense, we see in the appeal to experts a challenge to the concept of autonomy which has


101

significant implications for what we mean by ‘competence’ (for the layperson) and is an indictment of the activity of most critical thinking advocates. So we face a dilemma: expertise is indispensable, but with its rise comes an almost commensurate loss of autonomy. This is a dilemma that should be astutely felt by those working in argumentation and Informal Logic. As Willard notes: Argumentation and Informal Logic pedagogies won’t come to much unless based on a coherent stance toward authority. They are premised on rhetorics of mass enfranchizement that often do not square with their commitments to acquiescence to expertise, and – more important – they build a naive picture of the competent citizen (20).

Implied here is a considerable challenge. But it also announces an important shift of focus in discussions of the ‘appeal to authority’ argument, away from the authorities or experts and onto the layperson who must operate in their midst. Argumentation, with its employment and study of argument forms, is not a set of principles or tools which leave the user unaffected. Reasoning, specifically evaluating arguments, affects the evaluator. And, it seems, the ‘appeal to authority’ is an argument form that brings this into focus, because it allows, calls for, an entry into considerations of the competence and autonomy that should underlie effective argumentation. Expertise and Experience I want to take this pursuit of the competent layperson into an examination of Douglas Walton’s treatment of the ‘argument from authority’ (1997), where we can begin to develop the two threads of this paper. Walton takes a dialectical approach to argumentation (1995, 1998), focusing principally on the contexts in which a two-person dialogue frames the discourse. Thus he identifies different types of dialogue that can be involved in an exchange (1998), and has defined a fallacious argument as one that involves the inappropriate shift from one dialogue type to another (1995: 255-56). His treatment of the argumentum ad verecundiam, or ‘appeal to authority’, falls squarely within this perspective. He stresses the dialectical nature of discourses involving laypersons and experts, and argues that ‘[t]he dialectical factor is a subtlety that appears to be overlooked by the rather simplistic accounts of the logical form’ of the argument (1997: 153). Likewise, he discovers that at the heart of the distinction between legitimate and fallacious appeals to authority ‘is a dialectical failure... [where] the asking of appropriate critical questions is blocked by making it seem improper’(249). In Walton’s treatment, the common conditions for assessing the argument form that were noted above are filled out with a series

102


of critical questions related to each condition. These questions are the means by which the layperson can interrogate the expert, and it is behind this set that I will later want to look to consider what kind of competence must be held by a person capable of asking such questions. In all fairness, this is to lay the stress where Walton does not. He writes as if his interest is in the argument form as it sits out there between expert and consumer (or appealer and consumer). His critical questions have a certain “objectivity” to them, even if they are used only in specific contexts. But they still raise the question of whether there is a developing model of a competent individual behind Walton’s analysis. Walton’s treatment includes a discussion of the history of this argument form, and it is within such a history that we can find the tension between expertise and experience. Not surprisingly, Walton’s history accentuates the dialectical perspective as it can be uncovered from Aristotle through to Locke. Strictly speaking, the ad verecundiam begins with Locke in the Essay Concerning Human Understanding (1690), where he identifies it as backing one’s ‘tenets with such authorities [he] thinks ought thereby to carry the cause’ (cited in Walton, 1997: 53). Through a detailed analysis of Locke’s account, Walton comes to describe his version of the ‘argument from authority’ as ‘a characteristic type of exchange of argument: moves between two participants (or three, if you include the authority) in a dialogue framework’ (61). Hence, it becomes clearly a dialectical account, with the principal participants being the one making the appeal, and the one receiving it (being asked to accept it). In spite of this ‘origin’ in Locke, the strategy predates him in a number of forms, particularly as a feature of Aristotle’s collection of dialectical propositions (41). In fact, these two serve as important moments in the history because, although Walton details other relevant accounts in those of Plato, Galileo, the Port-Royal Logic and Bentham, the Lockean account, as he conceives it, that represents a return to the dialectical model that has been lost since Aristotle and anticipates what Walton himself will do with the argument form. Another important feature of this history, perhaps moreso for our purposes, is the way it acknowledges a clash between the acceptance of dogmatic appeals to authority and other criteria of evidence, principally scientific experimentation (34). Such a tension can even be seen in the early accounts of Plato and Aristotle. Plato’s treatment of experts is far more ambiguous than Walton’s discussion allows. The ‘one who knows’ is often a central concern. But while the expert is lauded in Plato’s dialogues (with the ultimate expert appearing in the form of the philosopher ruler), his early philosophy is essentially an attack on


103

certain experts.1 In fact, it is difficult to accept Walton’s claim that the Socratic dialectic indicates Plato’s belief that the layperson can critically assess the expert. Those who are critically assessed in these early dialogues are not experts. The competent layperson, if such we can imagine Socrates (and this would be a difficult model to emulate), assesses the non-expert in search of the expert, and consistently fails to succeed in the quest. An equally plausible reading of these dialogues would suggest that their intent is not to allow Socrates to assess the expert, but to provide the alleged expert with the opportunity to assess themselves and discover their failings as experts. Thus the rationale of the method is not to benefit the layperson but to guide the ‘expert’ in self-critique. On the other hand, in the Socratic figure we do have a model of epistemic competence, tied to the asking of critical questions, and it is interesting (even surprising) that Walton does not make more of this. This is a moment in the history that is left and not carried forward, in spite of its apparent relevance to Walton’s own account. In the ancient model that is chosen, that of Aristotle, we find another hint of trouble that deserves to be explored. Walton (40-41) identifies in Aristotle’s account an appeal to both the wise and the many. As he notes, these later will become the appeal to authority and the ad populum. But this should lead us to consider the ad populum as a type of appeal to authority. That is, what is the relationship between these two argument forms? Certainly, there is a relationship that interests Aristotle. In the Topics, the wise, or philosophers, are identified as the most notable of men. One can assent to their views ‘if it be not contrary to the views of most men’ (I.104a). This is an explicit condition that immediately limits the authority of the wise. It also flies in the face of Socrates’ insistence that the one who knows should be trusted over the many. But here Aristotle seems to provide a concession to the much broader experience of the majority. While there are certainly problems in such a suggestion, it proposes an empirical base for evidence that will characterize later accounts. This is also the first hint of the tension between the expert and direct observation that becomes more pronounced in later philosophers. In Walton’s history, it is Galileo who first stresses this tension: ‘History had reached a point of clear conflict between the institutional authority that prevailed at the time and the claims of empirical science’ (46). Galileo (1967:128) indicates clearly a preference for looking at how an authority, like Aristotle, arrived at a conclusion, rather than simply accepting the conclusion on the authority’s say-so, where this can be determined. In such a case, the layperson’s access to relevant immediate experiences is superior to the word of 1 This is also to question Feyerabend’s claims(1987) about the ‘cult’ of the expert in Plato’s work.

104


the authority. The authority becomes a substitute for that experience. As Walton notes, ‘...the say-so of any authority could come into conflict with scientific findings, based on observation and experiment’ (48). But whose observation is involved here? This very same point is emphasized in Arnauld’s The Art of thinking (or Port-Royal Logic) (Walton, 1997: 51) and even in Locke1: Of course, Locke was an empiricist, and his general viewpoint reflects the empiricist orientation of modern experimental science. In any conflict between a conclusion drawn from the opinion of an authority and a finding drawn from the “light arising from the nature of things themselves,” the latter kind of ad judicium argument wins. (54)

This appears to give us a tension between expertise and observation/experience. But it simply reduces to two sets of experience, one from the expert and another from another source. That source might be the layperson, or Aristotle’s majority, but in either case it sends the individual back to her or his own resources, to their own observations. In such cases, we might infer, the appeal to authority or expertise has no sway because it produces evidence that is no better than, and may be in conflict with, what others’ experience. However, an alternative tact to take is to ask whether in fact the expert might not sometimes be in a better position to interpret ‘direct evidence’. That is, someone’s status as an expert qualifies them to speak on certain types of experience with an authority that the layperson does not have. The non-knowledgeable individual, after all, can be mistaken. And in the relevant areas, when there is a conflict with the expert’s experience, the layperson is more likely to be in error than the expert. But when the expert conflicts with general observation, then we may have a problem. A related concern that derives from this history involves the way in which an appeal to an expert constitutes ‘evidence’ for a position. What is meant here by evidence? For example, Walton tells us (72) that to be reliable, an appeal must be based indirectly on objective evidence that can presumably be produced if requested. This puts “objective evidence” prior to the word of the authority. But in confirming an empirical base, this move also gives testimony a prior place, as we will see. Authority and Testimony The tension between the say-so of experts and the evidence of experience, and in particular the question of whose experience is at stake, deserves further exploration. 1 A similar point is identified by Walton in Bentham’s account (1997:57).


105

This question is involved, for example, in Hume’s appeal to ‘uniform experience’. In his ‘Essay on Miracles’1 the tension is expressed between reports of miraculous events (testimony) and uniform experience: It is no miracle that a man, seemingly in good health, should die on a sudden: because such a kind of death, though more unusual than any other, has yet been frequently observed to happen. But it is a miracle, that a dead man should come to life; because that has never been observed in any age or country. There must, therefore, be a uniform experience against every miraculous event, otherwise the event would not merit that appellation. (1964: 93)

Earlier (91), Hume had observed the lack of any a priori connection between testimony and reality. Rather, we find that they customarily conform to each other and this is why we find testimony credible. But our understanding of reality, presented as ‘uniform experience’ must itself be based upon testimony.2 Hume tests one type of testimony – reports of the miraculous – against another – uniform experience. For example, countering the testimony that Lazarus returned to life is ‘our’ uniform experience that the dead do not return to life (Bitzer, 1998: 192). But this reasoning has been shown to involve a vicious circularity: ‘the experience upon which our reliance upon testimony as a form of evidence is supposed to rest is itself reliant upon testimony which cannot be reduced in the same way’ (Coady, 1992: 81). As a type of inductive inference, the paradox of induction being its own standard of legitimacy is transferred to testimony. Hume is criticized for treating reasoning from testimony as one kind of reasoning from experience (analogy being another, for example) without differentiating its nature (Bitzer: 217). Moreover, a stronger criticism, attributed to John Leland (Bitzer: 211-12) argues that it is a mistake to reduce reasoning from testimony to reasoning from experience because the latter concludes with expectation, conjecture or prediction, while the former results in a judgment of a fact’s merits. Experience tells me what to expect; testimony tells me whether something is the case. As Bitzer (212) notes, it is unclear whether Hume did in fact conflate these two types of reasoning. But the distinction in the criticism is not a fine one. If the difference between the two is that between general inductions centered somehow on my own experiences and reports that I gather from others, it overlooks the extent to which ‘my’ general inductions are dependent on a ‘uniform experience’ – that is, essentially, reports gathered from others. Because I do not observe ‘uniform 1 Section 10 of Philosophical Essays and section 10 of An Enquiry Concerning Human Understanding. Page references are to the Enquiry in volume II of the Philosophical Works (1964), edited by T.H. Green and T.H. Grouse. 2 It is common experience, of course, that validates the laws of nature that miracles are deemed by Hume to violate.

106


experience’, I infer it from what I am told and read. Some reports are of such a heightened status that I use them to judge facts. But they are not so different in kind from the general reports which allow me to judge what to expect or whether something could have been predicted. What Hume’s account masks is what he recognizes elsewhere: testimony is our fundamental type of reasoning. My ‘experience’ is drawn from others’ accounts. Part of our failure to appreciate this stems from the way testimony is interwoven with other sources of evidence, like perception, memory and inference (Coady, 149). What one perceives is in part a function of what one expects; and what one expects is in part a function of what one has been told. It is also hindered by the modernist assumption that, epistemically, the self is somehow independent and self-reliant. Yet testimony is not only my most fundamental link to reality, essential for making sense of what I observe; it also connects me to society. If we assume the absence of innate knowledge, we need others for linguistic competence (Audi, 1997: 416). In a similar vein, Coady (85-95) develops an extended example of a Martian community to show how any ‘community’ relies upon testimony statements in order to exist. On the other hand, the assumption of the epistemically isolated self is fraught with problems. When asked how such a self would acquire knowledge, the expectation is that the individual’s perceptions would be fundamental. Coady invokes Condillac’s imaginary case of a marble statue1 gradually being brought to life to explore such a scenario. Starting with olfactory sensations, the statute slowly becomes fully attentive. Without overlooking its merits, Coady observes the way prejudices and assumptions of a certain cultural and philosophical predilection inform the thought experiment. Importantly, A statue does not belong to a species nor does it have a naturally communal life. But if we shift focus in this way it will be natural to see our starting-point as encompassing our knowledge and not exclusively my knowledge. There will be no problem of the epistemological priority of my perceptions over our perceptions, though there will be plenty of room for discrepancies. (Coady, 150)

From the egocentric perspective, I assume things that cannot be the case: that uniform experience is somehow rooted in and corroborated by my own, rather than vice versa; that we come each to know the world and then turn to communicate our understanding to others and contribute to the common fund, rather than that our understanding is predicated upon the common fund of experience. And testimony is our link to all this. This begins to suggest that the reason appeals to empirical facts have historically been preferred over expert testimony, when the two conflict, is because those empirical facts arise 1 This is drawn from Condillac’s A Treatise on the Sensations.


107

from general testimonial evidence and, hence, testimony is a more fundamental source of knowledge than expert opinion. Of course, reliance upon testimony alone would be naive. The point so far has been to indicate the extensive dependence that we should have upon testimony in argumentation, far beyond the more limited case of the ‘appeal to authority’. But something else must be required to make this testimonial evidence reliable, particularly if it is to form the foundation for our competent layperson. As we have seen, testimony, while fundamental, does not stand alone as a source of belief or knowledge. It functions in combination with perception, inference, and memory (Coady: 147; 175). In John Locke’s treatment of testimony (1959/1690: Bk.4, Ch.15) he remarks that when one hears reports of things which have been previously observed by the hearer, then we are ‘disposed by the nature of the thing itself to assent to it’. There is, as it were, a burden of proof favoring the speaker because what is testified to conforms readily to the hearer’s experience. But when what is reported falls outside the hearer’s experience, testimony may cause belief if ‘the relators are more in number, and of more credit, and have no interest to speak contrary to the truth’. However, in the famous case of the king of Siam who is told by the Dutch ambassador that water freezes in cold weather in Holland, Locke concedes that testimony so contrary to experience ‘will scarce be able to find belief’. Several things can be noted here that appear in later accounts of conditions for judging testimonial and expert evidence: the number of testifiers can be a factor; the perceived interests of the testifier are to be considered; and, importantly, the extent to which the report is contrary to the hearer’s experience will weigh in the hearer’s judgement. As Locke implies, human psychology is such that we cannot believe what falls too far outside our experience. Hume does give more power to testimony. As he takes up the case it involves an Indian prince (1964: 92). The freezing of water places nature in a situation quite unknown to the prince such that even the rules of analogy cannot help him. Still, Hume notes that it ‘requires a pretty strong testimony, to render it credible to people in a warm climate’ (92fn). This is because it is not miraculous, only extraordinary, since it is not contrary to uniform experience of the course of nature in cases where all the circumstances are the same. The inhabitants of SUMATRA have always seen water fluid in their own climate, and the freezing of their rivers ought to be deemed a prodigy: But they never saw water in MUSCOVY during the winter; and therefore they cannot reasonably be positive what would there be the consequence. (1964: 94fn)

108


Now we might observe that it is not the freezing of the water in Sumatra that is in question. But this is Hume’s point: if it were, then they would be asked to believe something miraculous. But instead they are asked to believe something about which they can at most reasonably offer scepticism – they cannot assent to it, but nor can they discount it. Thus, it is possible for ‘pretty strong testimony’ to bring about belief. Interestingly, this note from Hume (a late emendation of the text) turns the ‘uniform experience’ completely around and recognizes the value of general testimony from a slightly different perspective. Extrapolating from what is said about the Indian prince, it is not my experience which is being generalized. Rather, I am being asked to consider that my experience may not conform to what is uniform, as long as what is involved does not explicitly contradict my experience. This reading of the role of ‘uniform experience’ conforms nicely to the insights that derived from shifting from the egocentric to the communal perspective above. This critical scepticism is a necessary tool of the competent layperson. It opens one to general testimonial reports and the more specific ones from experts. Here, one’s autonomy is not threatened, but rather strengthened. It involves the development of a judgment both of character and with respect to the plausibility of events, two features not brought into the conditions for assessing the ‘appeal to authority’. In Robert Audi’s (1997) recent treatment of testimony, he suggests basic conditions very similar to these. Testimony is a source of basic beliefs (409), but not itself a generator of knowledge. For that transition to occur, other things are required: ‘[A] belief that p based on testimony thereby constitutes knowledge (i.e. counts as testimonially based knowledge) provided that the attester knows that p and the believer has no reason to doubt either that p or the attester’s credibility concerning it’ (412). These provisions effectively capture what is involved in the hearer’s appraisal of expert statements, since there is a difference between accepting something on the basis of an expert’s say-so and stating that one thereby knows what it is that has been attested to. It also indicates the complex context surrounding the ‘appeal to authority’ argument. Required here are further premises over and above those evaluating the expert as expert (does she have credentials in a relevant field? etc.) and what is said by the expert (is it agreed upon by other experts? etc.). We must ask also whether the expert is credible: should she be believed? Is this a context in which she would be expected to be serious? And we must ask whether what is said is plausible, given what one knows about the world and what others, in general non-expert reports, tell one about it.


109

The case of the prince/king and the frozen water also suggests another matter about general testimony and expert testimony. Living as he does in a warm climate, what is the consequence for him of believing or not believing? Insofar as the Humean king can be persuaded, his general beliefs about nature and his experience of it will change in a way not experienced by Locke’s unconvinced prince. But this is a very general change that may not be evident in any particular decision. After all, what can he do differently in light of this information? In other cases, and particularly those involving experts, there may be some immediate consequence for believing or not. If I choose not to believe that rock specimen ALH84001 originated on Mars, no consequence of any significance results.1 But if I choose not to believe my doctor’s advice that I alter my diet, there is likely to be a clear consequence arising from my nonbelief. So the consequences of acceptance and non-acceptance must be entered as a condition for deciding how much weight should be given to an expert’s testimony. This, a competent layperson must do. But this, of course, is to draw attention to the fact that we are talking here not about the competent layperson per se but about the tools (conditions) that person must employ. What we want finally to focus upon is the person who stands behind these conditions, the one using them. The Competent Layperson Douglas Walton’s dialectical treatment of the ‘appeal to authority’ argument results in six critical questions that should be asked in the assessment of an expert, along with a set of sub-questions for each. As the most extensive treatment of the argument form so far, it is a good place to turn in pursuit of the competent layperson. As noted earlier, Walton’s interest is in the questions themselves, set, as it were, between the evaluator and the expert.2 We need to step back and inquire into the sort of person who is implicated in these questions. That is, what kinds of competencies are assumed by the questions? I will discuss this with respect to each major question and its subset, then collect the relevant ideas and reflect on them in light of what has gone before. The first question raises the matter of expertise: ‘Expertise question: How credible is E as an expert source?’ (1997: 223). The assumption here is that the 1 Of course, the choice to not believe is not an arbitrary one; I still need good reasons for dismissing the claims of experts. In fact, just talking about choosing not to believe sounds strange, since beliefs are not things that I can adopt or reject at will. 2 Whether the expert is involved directly or is appealed to by a third party is not a concern at this point.

110


plausibility of the expert’s claim is directly related to the degree of credibility conceded to that person as an expert (rather than as a trustworthy character, for example). To this end five sub-questions are employed: 1. What is E’s name, job, or official capacity, location and employer? 2. What degrees, professional qualifications or certification by licensing agencies does E hold? 3. Can testimony of peer experts in the same field be given to support E’s competence? 4. What is E’s record of experience, or other indications of practiced skill in S? 5. What is E’s record of peer-reviewed publications or contributions to knowledge in S? (223). Obviously these are highly demanding questions and Walton is not suggesting that they must, or could, all be answered in every case. But they do cover the kind of information required to satisfy the first critical question. Our question in turn, is what is assumed here about the person who can ask such sub-questions competently? Ironically, the layperson must here be able to judge competence. To this end while they do not themselves have to be experts in the field in question, they need a more general ability to research and evaluate background. They need to know where to go to find records of publication and the corroborative testimony of peers, and a sufficient level of education to be able to measure the quality of what they read, particularly with respect to sub-question #5. What critical question #1 involves are the general qualifications of the expert. The layperson must understand those enough to accept the person as an expert. In some contexts, like a courtroom, much of this information would be provided as the expert is presented, and any problems with the credentials (sub-question #3, for example) would likely come to light through the cross-examination of lawyers. This first critical question doesn’t directly help the person to decide on the issue at stake; it merely establishes that what is involved is an appeal to expertise. But it still tells us something about the competent layperson. In a similar way to #1, critical question #2 helps to further establish the kind of argumentation that is at stake: ‘Field question: Is E an expert in the field that A is in?’ (223). Here, A is a proposition in a certain field, and the intent is to locate E and A in the same field. The sub-questions for this are as follows: 1. Is the field of expertise cited in the appeal a genuine area of knowledge, or area of technical skill that supports a claim to knowledge?


111

2. If E is an expert in a field closely related to the field cited in the appeal, how close is the relationship between the expertise in the two fields? 3. Is the issue one where expert knowledge in any field is directly relevant to deciding the issue? 4. Is the field of expertise cited an area where there are changes in techniques or rapid developments in new knowledge, and if so, is the expert up-to-date in these developments? (224). Clearly, these questions demand far more of the competent layperson. They assume an understanding of what counts as a genuine area of knowledge, which in turn implies a general comprehension of such fields. This is particularly difficult when the very question of whether a field is legitimate is so often in debate, particularly in law courts or public fora. Coady (1992: 286-88), for example, draws distinctions between sciences, pseudo-sciences, and nascent sciences, primarily in terms of their methodologies and agreed results. The last category (the nascent) allows that a field about which there may still be some disagreement may yet grow into a mature science. He cites stylometry as a plausible candidate for such a designation. Walton himself raises similar kinds of concerns (176-181). But this in turn does help us to a certain degree. What the competent layperson needs to know about fields is that they require an accepted and recognized set of methodological principles. They must be subject in principle to challenge and testability. And the basic understanding we already have from observing the world and receiving testimony about it should help us here. What is at stake in a field of knowledge is no different in kind from these general underlying experiences that we have. In critical question #3 Walton turns to what is said by the expert: ‘Opinion question: What did E assert that implies A?’ (223). The sub-questions associated with this clearly do not imply that a direct dialogue between the expert and layperson is being engaged, but allow more generally for reports about experts: 1. Was E quoted in asserting A? Was a reference to the source of the quote given, and can it be verified that E actually said A? 2. If E did not say A exactly, then what did E assert, and how was A inferred? 3. If the inference to A was based on more than one premise, could one premise have come from E and the other from a different expert? If so, is there evidence of disagreement between what the two experts (separately) assert? 4. Is what E asserted clear? If not, was the process of interpretation of what A said by the respondent who used E’s opinion justified? Are

112


other interpretations plausible? Could important qualifications be left out? (225) These sub-questions assume a basic logical competence. Sub-questions #2 and #3 in particular require an understanding of how inferences work. And in a related way, #1 requires some sense of the difference between direct and indirect communication and attribution, and sub-question #4 assumes an appreciation of language and meaning. Apropos Willard’s concerns about the ability of argumentation and Informal Logic courses to provide the skills necessary for competence in the face of experts, these are all basic skills that are fundamental to the early stages of any such course. The fourth critical question focuses on a feature that was recognized in the earlier discussion: the character of the expert. ‘Trustworthiness question: Is E personally reliable as a source?’ (223). Far from being an irrelevant consideration when it is the matter of expertise that is being considered, we’ve seen the personal credibility of the testifier integral to all testimonial evidence. This is understood in Walton’s sub-questions: 1. Is E biased? 2. Is E honest? 3. Is E conscientious? (217; 227) These questions, as Walton notes, (227), also recognize the close relationship between the ‘appeal to authority’ argument and ad hominem argumentation, here construed as the critical assessment of a person’s character as it relates to what is said.1 This assumes in the competent layperson a general moral acumen such that they can evaluate the qualities of character and in particular honesty. They need to separate legitimate biases (which all reasoners have) from those that are illegitimate and may cloud the judgment of the expert on a particular issue. This involves an appreciation of what is relevant to the case – another basic argumentation skill. The third question, according to Walton, ‘pertains to E’s scholarly habits or carefulness and professionalism as a skilled technician’ (217). Presumably, if E has sloppy work habits, then the research or results may be unreliable. A reasonable assessment of this question would involve more knowledge of E’s background and work history than most laypersons would possess. But it is conceivable that such information might come to light, and it would be relevant to the evaluation without requiring any specific skill on the part of the layperson.

1 See my (1999: Chapter 3) for a discussion of how these argument forms are related under the general umbrella of ethotic argumentation.


113

Critical question #5 is one of consistency: ‘Consistency question: Is A consistent with what other experts assert?’ (223). It is one of the more difficult challenges involved with the ‘appeal to authority’ argument form. The subquestions acknowledge this in that Walton offers two pairs. The first pair is straightforward: 1. Does A have general acceptance in S? 2. If not, can E explain why not, and give reasons why there is good evidence for A? (222) If the answer to #1 is ‘no’, then the second question needs to be asked. Here the competent layperson must be able to evaluate competing claims. But insofar as the context of such debates puts the onus on the disputants to support their position, the requirement reduces to the ability to judge ‘good reasons’. This is not to sketch over the very real nature of the problem. Consider the recent debate over ‘Cold Fusion’,1 one which is still ongoing. Here the dispute between the original claimants, Fleischmann and Pons, and their opponents was played out in a very public forum as the media recognized the revolutionary nature of the claim if it was established. But the layperson was left bemused by the researchers’ confession (one that still stands) that while they had achieved the result of cold fusion in a bottle, they could not explain the principles involved. Here, Hume’s critical scepticism would counsel waiting for further evidence before accepting the claim. The aspect of consistency is better captured in Walton’s provision of further questions when the answer to critical question #5 above is ‘no’: 1. Is it because the question is one on which you would normally expect experts in this field to disagree? 2. Is it a question where the preponderance of other highly qualified experts agree on an opinion other than the opinion of the expert source cited? (221) The very asking of question #5 assumes the competent layperson can recognize inconsistency and thus has an understanding of what consistency entails. This again is a general critical thinking skill. From this perspective the question set aids in judging the apparent inconsistency and deciding on its importance in accepting a claim. In the Cold Fusion debate, for example, it was crucial that other experts could not replicate the results of the principal researchers.

1 The controversy erupted in 1989 when chemists Martin Fleischmann and B. Stanley Pons announced at the University of Utah that they had succeeded in creating fusion in a bottle. In November of that year, a panel of scientists reported negatively on the claims. See Eugene F. Mallove’s ‘The Cold Fusion Mystery Continues’ Tech Talk (4/11/90).

114


This point takes us to the sixth and last of the critical questions for assessing experts: ‘Backup evidence question: Is A’s assertion based on evidence?’ (223). The failure to replicate an experiment is a matter of external evidence and on this ground we return to the historic tension discussed earlier. The sub-questions for this are as follows: 1. What is the internal evidence the expert used herself to arrive at this opinion as her conclusion? 2. If there is external evidence, e.g., physical evidence reported independently of the expert, can the expert deal with this adequately? (218). 3. Can it be shown that the opinion given is not one that is scientifically unverifiable? It might be objected here that if the critical layperson knew the answers to these question, he or she would not need to depend on the expert’s say-so, they would be on equal footing with the expert. Thus, these questions appear to echo the criticisms of Fry and Sokal and Bricmont. But to be useful (and our earlier discussion showed that such questions must be involved) we can interpret the questions as ascertaining generally that such evidence exists without requiring that the questioner know the evidence. In this sense what is assumed is general epistemic competence: the ability to recognize what counts as evidence and an appreciation of the principle involved in verifiability. (In fact, the point behind question #3 is really subsumed by the background required to ask the earlier question.) Again, sub-question #2 is a concession to the power of external evidence, based on general testimony, such that the expert must be able to account for it. In questions of conflict, the implication is, the external evidence will be deemed superior. Conclusion: Competence and Argumentation Behind much of the discussion about experts and their relation to us we find an assumption that specialized knowledge somehow distances such people from us, and, insofar as we might be judged experts in certain fields, it distances us from others. It is as if we were being looked down upon from above and viewed each in isolation. But, as we have seen, the extent to which this is a legitimate picture can be challenged by the lessons learned from testimony and the pervasive way it draws us together and provides the fund out of which we emerge as individuals. This unifying base is implicated not only in the historical treatments of the ‘appeal to authority’ argument, but also in the more developed contemporary accounts of argumentation theorists like Walton. Several of his critical questions assume a competence on the part of


115

the inquirer that relates back to how we function in the world, how we emerge and grow as epistemic beings. It is because of the fundamental role that testimony plays in our lives that the competent layperson is open to other people’s reports. From this derives a critical scepticism toward the idea that our own experience is central and representative. As such it cannot be assumed to conform with general experience. Rather, such conformity has to be established, especially with respect to astounding and complex reports. This predisposition to believe the accounts of others is a prerequisite for a reasonable attitude toward the testimony of experts. Our self-governance, or autonomy, is developed against this backdrop, conditioned by it, rather than threatened. But in order to evaluate the testimony of experts, we have seen that more specific traits must be possessed. Broadly conceived, these traits amount to basic skills in scientific methodology, reasoning, and ethics. The competent layperson draws on the general epistemic competence developed in our general functioning in the world in order to recognize what counts as evidence. Likewise, our general understanding of the world allows for the discrimination of fields of expertise as they admit of methodological principles that allow claims related to them to be assessed. The evaluation of an expert’s statements involves the logical competencies of inferencing, identifying relevance, identifying consistency, and a general appreciation of what count as good reasons. The assessment of character requires an understanding of basic moral categories and associated sensibilities. These general and specific traits coalesce in a fluid working definition of ‘competence’. Charles Willard has done the argumentation community an important service in pointing out its tendency to idealize its role in society and exaggerate its accomplishments. But that a problem could be identified in the postures of some theorists and educators does not mean that answers cannot be found in the same material. As Willard (1990: 21) himself implies in stressing that he has offered diagnosis rather than therapy, a way forward often involves a going back to see if we have been doing what we thought we were doing. It will not be lost on readers that the model of competence advocated here is very much rooted in the skills of argumentation and Informal Logic. As has been indicated, a course in argumentation or Informal Logic is exactly the place where many of the requisite skills are first identified, taught, and developed. In fact, the ‘appeal to authority’ argument seen in its full complexity suddenly appears as the perfect vehicle for the recognition and employment of these assorted skills. It is also the place where we can come to reflect on the full impact of testimony on our understandings of the world.

116


We do not need to learn to be experts as much as we need to learn to be non-experts, to function competently in a world of competing claims from specialized fields. The education and training we acquire that gives us specialized knowledge is all too obvious, hence it is easily recognized and promoted by governments and the media. But the education that gives us the broader kinds of competence often fails to receive the respect and recognition it deserves. It is, on reflection, very much the foundation of what has come to be called a liberal arts and science education.

References Audi, R. (1997) ‘The Place of Testimony in the Fabric of Knowledge and Justification’, American Philosophical Quarterly 34:405-422. Bentham, J. (1971) Handbook of Political Fallacies [1861]. New York: Thomas Crowell. Bitzer, L.F. (1998) ‘The “Indian Prince” in Miracle Arguments of Hume and His Predecessors and Early Critics’, Philosophy and Rhetoric 31:175-230. Coady, C.A.J. (1992) Testimony: A Philosophical Study. Oxford: Clarendon Press. Feyerabend, P. (1987) Farewell to Reason. London: Verso. Forster, E.M. (1936) ‘Roger Fry: An Obituary Note’, from Abinger Harvest, London: Edward Arnold (Publishers) Ltd.:50-53. Groarke, L., et al.(1997) Good Reasoning Matters!: A Constructive Approach to Critical Thinking. Toronto: Oxford University Press Canada. Hume, D. (1964) An Enquiry Concerning Human Understanding, in Philosophical Works. Volume II, ed. T.H. Green and T.H. Grouse. Scientia Verlag Aalen. Locke, J. (1959) An Essay Concerning Human Understanding [1690]. ed. A.C. Fraser, New York: Dover. Mallove, E.F. ‘The Cold Fusion Mystery Continues’ Tech Talk (4/11/90). Manguel, A. (1996) A History of Reading. New York: Alfred A. Knopf. McKay, D.S. et al. ‘Search for Past Life on Mars: Possible Relic Biogenic Activity in Martian Meteorite ALH84001’, Science Magazine (August 16, 1996). Mickunas, A. (1987) ‘Perelman on Justice and Political Institutions’, Practical Reasoning in Human Affairs, eds. J.C. Golden and J.J. Pilotta, Dordrecht: Reidel. Sagan, C. (1995) The Demon-Haunted World: Science as a Candle in the Dark. New York: Random House. Sokal, A. and J. Bricmont (1998) Intellectual Imposters: Postmodern Philosophers’ Abuse of Science. London: Profile Books. Tindale, Christopher W. (1999) Acts of Arguing: A Rhetorical Model of Argument. New York: State University of New York Press. Walton, D. (1995) A Pragmatic Theory of Fallacy. Tuscaloosa: The University of Alabama Press. Walton, D. (1997) Appeal to Expert Opinion: Arguments from Authority. University Park, Pennsylvania: Penn State Press. Walton, D. (1998) The New Dialectic: Conversational Contexts of Argument. Toronto: University of Toronto Press. Willard, C. A. (1990) ‘Authority’, Informal Logic 12:11-22.

Peirce’s Abductive Enthusiasms

117

JOHN WOODS Peirce’s Abductive Enthusiasms I am part of a team producing a multi-volume Handbook of Practical Reasoning: Computational and Theoretical Aspects, to be published in the fullness of time by Kluwer. One of our tasks is to get clear about the role of abduction in practical reasoning. So it is natural to begin with Peirce. In 1898 Charles Peirce gave a series of eight lectures at Harvard1. Delivered “off-campus”, these astonishingly fruitful pieces were presented James’ behest, in yet another gesture of generosity towards his beleaguered friend. Peirce’s range was striking. He had original and prescient things to say about probability, likelihood and randomness; a bout quantification (of which, with Frege, he was independent co-discoverer); about causality, space and time, and cosmology. Of particular note is Peirce’s emphasis on abduction (or retroduction) as the dominant method in science, in fact, as the only purely scientific method beyond brute observation. For Peirce, scientific reasoning stands in sharp contrast to practical reasoning or, as he also says, to reasoning about “matters of Vital Importance”.2 In this contrast he was hardly alone, although it may be mentioned in passing that it was a matter on which he disagreed fundamentally with Mill, for whom “a complete logic of the sciences would also be a complete logic of practical business and common life”. Peirce saw abduction as hypothetico-deductive reasoning under conditions of trial. Thus the by now familiar pattern: If H were true, certain consequences C1,...,Cn would follow. We see that C1,...,Cn do in fact obtain. So provisionally we suppose that H.

Supposing is not inferring, and neither then is abduction. It is rather a strategy for making refutable conjectures. Abducted conclusions are not matters for belief or for probability. In abduction “not only is there no definite probability to the conclusion, but no definite probability attaches even to the mode of inference [sic]. We can only say that ... we should at a given stage of our inquiry try a given hypothesis, and we are to hold to it provisionally as long as the

1 Recently reprinted as Peirce [1992]. 2 Peirce [1992], 110.

118

John Woods

facts will permit. There is no probability about it. It is a mere suggestion which we tentatively adopt”.1 This paper investigates Peirce’s fledgling positive theory. “When [we] say that a retroductive inference [sic] is not a matter for belief at all,. ..[an] example is that of the translations of the cuneiform inscriptions which began in mere guesses, in which their authors could have had no real confidence.”2 [We see, then, that Science] takes an entirely different attitude towards facts from that which Practice takes.3

Practice involves belief ineradicably, “for belief is the willingness to risk a great deal upon a proposition. But this belief is no concern of science ... .”4 Belief contrasts with acceptance, as can be gathered from this passage, which might well have been penned nearly a century later by Cohen:5 ...whether the word truth has two meanings or not, I certainly do think that holding for true is of two kinds; the one is that practical holding for true which alone is entitled to the name of Belief, while the other is that acceptance of a proposition which in the intention of pure science remains always provisional.6

Hence, I hold that what is properly and usually called belief, that is, the adoption of a proposition as [a possession for all time].. .has no place in science at all. We believe the proposition we are ready to act upon. Full belief is willingness to act upon the proposition in vital crises, opinion is willingness to act upon it in relatively insignificant affairs. But pure science has nothing at all to do with action. The proposition it accepts, it merely writes in the list of premises it proposes to use.7 And since 1 Ibid., 142. Peirce distinguishes (1) induction, or the establishment of frequencies in populations by sampling and (2) probable inference, or the inference from a known frequency in a randomly selected sample. He here anticipates Neymann-Pearson statistical sampling theory with its notion of likelihood. By requiring that inductions have a premiss to the effect that sampling is random, Peirce thought that all inductions turn on prior discernment of lawlike statements. That a method of sampling is random requires recognition of the equality of certain frequencies, and so is a kind of lawlike knowledge, that is, knowledge of generals. Of course, Peirce didn’t get randomness right. No one did or could until the development of recursion theory well into the present century. Putnam is good here. See his “Comments on the Lectures” in Peirce [1992], 61 and 68. See also Lecture Two, 123-142. 2 Peirce, op. cit., 176. 3 Ibid, 177. 4 Ibid. 5 Cohen [1992], 4: “...to accept that p is to have or adopt a policy of deeming, positing, or postulating that p – i.e. of including that proposition or rule among one’s premisses for deciding what to do or think in a particular context, whether or not one feels it to be true that p.” 6 Peirce, op. cit., 178. 7 Ibid., 112.


119

[n]othing is vital for science [and] nothing can be..., [t]here is.. .no proposition at all in science which answers to the conception of belief.1

Whereupon, ...the two masters, theory and practice, you cannot serve.2

When writing about logic, Peirce anticipates Carnap and Brouwer (and Kripke and Harman, too). And, of course, Quine. My proposition is that logic, in the strict sense of the term, has nothing to do with how you think ... . Logic in the narrower sense is that science which concerns itself primarily with distinguishing reasonings into good and bad reasonings, and with distinguishing probable reasonings into strong and weak reasonings. Secondarily, logic concerns itself with all that it must study in order to draw those distinctions about reasoning, and with nothing else.3

Concerning these things, “it is plain, that the question of whether any deductive argument, be it necessary or probable, is sound is simply a question of the mathematical relation between.. .one hypothesis and.. .another hypothesis.”4 Concerned as it is with the presence or absence of that mathematical relation between propositions. [i]t is true that propositions must be expressed somehow; and for this reason formal logic, in order to disentangle itself completely from linguistic, or psychical, considerations, invents an artificial language of its own, of perfectly regular formation, and declines to consider any proposition under any other form of statement than in that artificial language.5 ... As for the business of translating from ordinary speech into precise forms,...that is a matter of applied logic if you will....6

But applied logic stands to logic much as fool’s gold stands to gold, for it is “the logical part of the science of the particular language in which the ex-

1 Idem. 2 Idem. 3 Ibid., 143. 4 Ibid.., 144. 5 A Carnapian emphasis. 6 Ibid., 144-145.

120

John Woods

pressions analyzed [or translated] occur.”1 That is to say, it is a part of the linguistics of natural language. Logic, then, has nothing to do with how we think. Still less has it to do with how we think about vital affairs. Once you become inflated with [what good deducers, i.e. mathematicians, are up to,] vital importance seems to be a very low kind of importance indeed. But such ideas [i.e. the procedures of deductive thinkers] are only suitable to regulate another life than this. Here we are in this workaday world, little creatures, mere cells in a social organism itself a poor little thing enough, and we must look to see what little and definite task circumstances have set before our little strength to do. The peformance of that task will require us to draw upon all our powers, reason included. And in the doing of it we should chiefly depend not upon that department of the soul which is most superficial and fallible, – I mean our reason, – but upon that department that is deep and sure – which is instinct.2

All forms of reasoning, even deduction, require observation. What is observation? It is the enforced element in the history of our lives. It is that which we are constrained to be conscious of by an occult force residing in an object which we contemplate. The act of observation is the deliberate yielding of ourselves to that force majeure, – an early surrender at discretion, due to our forseeing that we must whatever we do be borne down by that power, at last. Now the surrender which we make in Retroduction, is a surrender to the Insistence of an Idea. The hypothesis, as The Frenchman says, c’est plus fort que moi.3

The passage is striking and at seeming odds with Peirce’s insistence that is not useful in ordinary affairs and that it has nothing to do with belief or probability. (After all, the most common fallacy of abduction is Bayesianism, that is, to “choose the most probable hypothesis”4). But belief is like observation. It presses in on us and demands our surrender. And what is so insistent about the Insistence of an Idea if it does not somehow call for and eventuate in belief?

1 Ibid., 145. Peirce, like Mill, is somewhat ambivalent about the primacy of induction. “As for ... Induction and Retroduction, I have shown that they are nothing but apogogical transformations of deduction and by that method the question of the value of any such reasoning is at once reduced to the question of the accuracy of Deduction” (Ibid.). On the other hand, the “marvelous self-correcting property of Reason, which Hegel made so much of, belongs to every sort of science, although it appears as essential, intrinsic, and inevitable only in the highest type of reasoning which is induction” (Idem., 168). Of Peirce’s transformations we have the space to say only that he regarded induction, probabilistic inference and retroduction as distinct forms of reasoning of which Aristotle’s first three syllogistic figures were limiting cases. See Lecture Two, 131-141. 2 Ibid., 121. 3 Ibid., 170. 4 Putnam [1992], 78.


121

We begin to see the irreducibly practical aspect of abduction. It shows itself in two ways, in the Insistence of an Idea and in the inductions that let loose our sentiment, configure our habit and culminate in belief. This last aspect is a scientific, an all too human acquiescence in a proposition by which our vital affairs might be guided. But the first is purely scientific, an intrinsic part of abduction itself, and a seeming contradiction of Peirce’s settled and repeated views. Let us see. Abduction is two things, (a) surrender to an Idea, to a force majeure; and (b) a method of testing its consequences. In surrendering to an idea, say the existence of quarks, or the promise of quantum electrodynamics, we do not come to a belief in quarks or to a belief that QED is true, but rather to a belief that quarks are a good thing to suppose and a good bet for testing, or that QED has a good explanatory record or at least promise of such. In this, abduction resembles practical reasoning (and observation, too). At the heart of this resemblance is the concept of a “vital affair”. Here, too, there are two conceptions to ponder. []A vital affair is made so by its subject matter, i.e., its vitality is topicintrinsic (e.g., ethics possibly). []A vital affair is made so by its methods of enquiry irrespective of its subject matter. It seems to be Peirce’s view that the vitality of an affair is entirely a matter of the manner in which we conduct its enquiry. There may be matters that especially conduce to such methods, but there seems to be nothing intrinsic to the methods that preclude them from application to any subject matter. In the case of abduction, these methods are applied to the important business of making conjectures for subsequent test. Essential to this is the conviction that a hypothesis H is a good enough conjecture to warrant our making it. That is a vital affair for scientist and laymen alike. What this shows is that belief is indispensable for abduction, but it does not show an inconsistency in Peirce’s account. For here the object of belief is not hypothesis H, but rather is some proposition of which H is the subject, viz., the proposition that H is a good conjectural bet (or, in some variations, that H is a good explanation of some set of target phenomena). Equally, a test of H is no good if the experimenter is in any doubt – to say nothing of a state of indifference – about whether these things are consequences of H or about whether this other thing implies the negation of a consequence of H.

122

John Woods

We have been suggesting that abduction has been revealed to have an irreducibly practical component. if that were true, it would matter for of a theory of abduction. And it would matter greatly if there could be such a thing as a theory of practical reasoning (an affirmative answer to which being presupposed by the title and subtitle of my Handbook). We note here for subsequent consideration the possibility that while abduction has been shown to have a practical aspect, it has not been shown to be irreducible in the requisite sense of being an aspect that any competent theory of abduction must take note of and accommodate within its principled disclosures. For the present we merely mark Peirce’s ambivalence about the possibility of a science or a theory of practical reasoning. Certainly there is for him little prospect of there being a logic of practical reasoning. Standing in the way is logic’s necessary silence on how to represent a piece of practical reasoning in the artificial language in which the logic transacts it proper business. “In everyday business”, says Peirce, “reasoning is tolerably successful; but I am inclined to think that it is done as well without the aid of a theory as with it”.1 This suggests not the impossibility of a theory so much as its unhelpfulness, and it leaves open an option of a type that he has already acknowledged in connection with deductive logic. Given that deductive logic doesn’t fix the representation relation that takes natural language arguments into logic’s artificial language, this is a task that is left for the science of the natural language in question. By these lights, there is no logic of abduction, but it might well have a scientific theory for all that. if so, the science of abduction will itself be abductive – as all genuine science is. And therewith is met an interesting problem. It is the problem of specifying what it is about our abductive practice that calls out for abductive theoretical analysis and disposition. If we agree that abduction is au fond the making of conjectures in the light of their consequences and the testing of those consequences in the light of new information, then what are the abductive questions to put about these practices? We are not here attempting to ready the reader for an assault upon the Problem of Abduction, as we might say; but we might briefly pause over the point before moving on. If we ask by what might abduction be justified, there is little chance that we will like the suggestion that its legitimacy is the conclusion of a deductive proof or the conclusion of a forceful abductive argument. The first doesn’t exist and the second commits the fallacy of petitio principii. Of course, [a]fter a while, as Science [i.e., abduction] progresses, it comes upon more solid ground. It is now entitled to reflect, this ground has held a long time without 1 Peirce, op. cit., 109.


123

showing signs of yielding. I may hope that it will continue to hold for a great while longer... For a large sample has now been drawn from the entire collection of occasions in which the theory comes into comparison with fact, and an overwhelming proportion, in fact all the cases that have presented themselves, have been found to bear out the theory. And so ... I can safely presume that so it will be with the great bulk of the cases in which I shall go upon the theory, especially as they will closely resemble those which have been well tried.1

This of course is induction, and induction has no place in science. If, on the other hand, the justification is not scientific, it is practical. This is not sufficient ground to disentitle induction, but it won’t work as a justification of abduction. The reason is that induction itself is abductively justified. For what justifies our inductive practices if not the conjecture that if nature were uniform our practices would be successful, together with the observation that our practices have been successful up to now? This leaves the justification of abduction with a circularity problem and then some. (Perhaps we could call it abduction’s double helix.) With that aside, we return to our present question. Regarding our abductive practices, is there a subject matter for a scientific theory? If there were, there would be something for that theory to explain by conjecture and refutation. But what is this explanandum? Evidently that science proceeds by looking for an explanation which it then tests. Then what is the explanation, the explanans, of this fact? That this is how science is best done, presumably. But now: how is this conjecture tested? If the conjecture were that this is how science is done, it would be no conjecture, but a matter of observation; and without conjecture, the theory could not be abductive. It matters that the conjecture at hand is that this is how science is best done, that this is the right way to do science. Now Hume hovers nearby, and chuckles. We find ourselves stretched out on the prickly barbs of the ought-is divide. What this suggests is that we are ill-served by any generalized abductive enthusiasm for abduction as such. A better – or anyhow more hopeful – option is to abandon our global ambitions in favor of a more provincial approach to abduction. What is needed are theories of the following things, one by one.2 First is required a scientific account of hypothesis formation, itself a subtheory of belief fixation. Peirce insists that no logic is fitted for such a task, and in this we may suppose him to be right. It may seem that in as much as belief fixation involves, essentially, the holding of beliefs, and various other probabilistic raptures, no science is available for such phenomena. But this is quite 1 Ibid.., 177. 2 See Lipton [1991].

124

John Woods

wrong. It confuses subject matter with method. The belief-theorist is meant to have or to hold no beliefs about belief, except, at the limit, his own beliefs about what are the fruitful conjectures to make about belief. We may liken this science to a branch of cognitive psychology, and nothing we have said here prevents the theorist from proceeding as abductively as he pleases, as far as abduction can in general go. All that is required is that the theory give an explanatory account under conditions of trial of how abducers form the hypotheses they provisionally adopt under conditions of trial. Second comes a theory of consequence. It is necessary for the theorist to have an account of that relation in virtue of which conjectured hypotheses are testable. Nothing we have said here precludes our calling deductive logic into service. Thirdly, we must have a theory of refutation. Our friend modus tollendo tollens will apply for the job, no doubt. He is not up to it. He will give hopelessly indeterminate guidance except for the simplest kinds of case. The best general theory of refutation produced so far is Aristotle’s. It is a beautifully crafted exploitation of properties of the syllogism and of conditions that bind certain kinds of rational, even if adversarial, discourse. Two-and-a-half millenia have come and gone since its launching. No doubt it could do with some tarting up; but it is unlikely that its central core will require, or tolerate, much change. So conceived of, a theory of abductive reasoning is a triple AL=. Of these three, only the theory of consequence lays any claim to consensual maturity, while the third awaits contemporary revival. This leaves the first, the theory of conjecture or of belief-fixation with respect to, as we may now say, plausible conjecture. On this the polls are open, and early results are only trickling in. Its various subcomponents show promise of fruitful union, what with recent initiatives in cognitive psychology and M. We would also conjecture a useful role for something like a logic (to speak loosely) of plausibility as a minimal constraint, and a full-bore account of analogy, perhaps along the lines of Woods and Hudak [1989] or along the rather different lines of Holyoak and Thagard [1994]. Pages back, we said that abduction was inherently practical in certain of its aspects. The abducer is always driven by a belief that “this proposition is worth entertaining”, a belief held by the abducer independently of whether he or anyone else actually believes the proposition in question. Given that the abducer’s belief is always a function of the Insistence of an Idea, abduction is always practical and, as such, ascientific. if this is right, the same will hold for


125

any theory of abduction itself which aspires to be abductive. It may seem to follow that there can be no theory of abduction, but we may think that this goes too far. The fallacy of composition looms. It is here the fallacy of supposing that if a theory has an ascientific component the theory as a whole is ascientific. That thinking so would be a fallacy in the present case is suggested by the fact that the abducer’s sole belief (hence his sole practical contribution) is his belief about what is a plausible candidate for testing. Of course, it may turn out that the candidate tests negatively, but this wouldn’t show that it wasn’t worth testing. And so we might find ourselves ready to concede that such beliefs are ascientific in the degree that they are untestable. But ascientific or not, such beliefs don’t discredit the scientific enterprise. We will be led by such beliefs to no scientific theories of a sort not already admissible under the testing provisions of abduction generally. If these let in bad theories, it won’t have been for the ascienticity of the beliefs that set them up as conjectures in the first place. So it cannot be supposed that abduction entire is ascientific just on account of the ascientificity of that aspect. Suppose, then, that we like this specification of our triple AL enough to be getting on with. Getting on with it in dominantly the business of elucidating the mechanisms of what we might call the theoretical imagination. If tellable at all, it would tell the story of how in Quine’s words theories are free for the thinking up, that they are the exercise of our conceptual sovereignty, no more and no less.

Bibliography L. Jonathan Cohen. 1992. An Essay on Belief and Acceptance, Oxford: Clarendon Press. Keith J. Holyoak and Paul Thagard. 1994. Mental Leaps, Cambridge, MA: MIT Press. Peter Lipton. 1991. Inference to the Best Explanation, London: Routledge 1991. C.S. Peirce. 1992. Reasoning and the Logic of Things. Ed. Kenneth Laine Ketner. Cambridge, MA: Harvard University Press. Hilary Putnam. 1992. “Comments on the Lectures”, in Peirce [1992], 55-102. John Woods and Brent Hudak. 1989. “By Parity of Reasoning”, Informal Logic XI, 125-140.

126

Henry W. Johnstone

HENRY W. JOHNSTONE, JR. “‘Any,’ ‘Every,’ and the Philosophical Argumentum ad Hominem” Aristotle obviously has in mind a geometrical proof a later version of which is included in Euclid’s Elements . This proof concerns an arbitrary triangle – any triangle – and shows that on the basis of postulates, axioms, and previously proven theorems, such an arbitrary triangle has an angle-sum of two right angles. Hence this property applies to every triangle. That every triangle has a certain property is thus demonstrated by reasoning about any triangle. “Every,” we see, belongs to the statement of a theorem; “Any” belongs to a step in the proof of the theorem, and provides the basis for the use of Universal Generalization. The same machinery operates in the philosophical argumentum ad hominem . If some arbitrary person any person holding the view in question can be attacked ad hominem for holding this view, then everyone holding this view can be similarly attacked, and the view itself comes into question.

-° C#/ Á%9'3J -C-J, ²- %¤ -#º -/3C-#+ ¤ %'Ë-#/ J½- ... µ -#?/ -° -/3° %'Æ-# J?/- E# ±'qO 3# ´-#º x#, -#E-# %'Ë-Ç Á%w'3J C#/. (Aris-

totle, Posterior Analytics 73b33 74a1)

Tredennick translates this passage: “An attribute only belongs to a subject universally when it can be shown to belong to any chance (%¤ -#º -/3C-#+) instance of that subject, and to belong to that subject primarily.1 . .. Thus that which can be shown in any chance (-/3C) instance to fulfill the condition of containing the sum of two right angles, or any other requirement, is the subject to which that universal attribute primarily belongs” (pp. 47-49). But if we render -/3C and the prepositional phrase %¨ -#º -/3C-#+ as “arbitrary” rather than “chance” or “by chance,” it is clear (as Barnes saw in 1994, when he used “arbitrary” instead of “chance” or “by chance” in his translation of this passage on p. 8 as well as in his commentary on it on p. 119) that Aristotle anticipates the Rule of Universal Generalization, according to which what is true of an arbitrary individual is true of every individual. For example, if the sum of the angles of an arbitrary triangle can be shown to equal two right angles, the sum of the angles of every triangle must be equal to two 1 In this article , I will not be concerned with how “primarily” (%'Ë-#/, %'Ë-Ç) functions in this passage, either here or in the rest of what I quote.

‘Any,’ ‘Every,’ and the Philosophical Argumentum ad Hominem

127

right angles. This is of course Aristotle’s example.1 (In an earlier translation in 1984, Barnes, too, had rendered -/3C and %¤ -#º -/3C-#+ as “chance.” This rendition apparently prevented him from seeing that at this point Aristotle was anticipating the Rule of Universal Generalization. I don’t think the question whether Tredennick’s identical rendition kept him from seeing Aristotle’s statement as anticipating this Rule arises. The Rule anticipated here had not yet been generally recognized at the time Tredennick’s translation appeared.)2 The adjective “chance,” used by Tredennick in his renderings of -/3C and %¤ -#º -/3C-#+ is a near-synonym of “random.” I am thus emphasizing the difference, in the present context, between “random” and “arbitrary.” One way of making this difference clear is to consider what would be involved in showing that the sum of the angles of any random triangle equals two right angles. This is a statistical problem, involving the use of a protractor or similar instrument to measure the angles of a large number of triangles and show that their sum turns out in all these cases to be that of two right angles. This is an inductive argument, while the place of the Rule of Universal Generalization is in a deductive context. There are cases in which while “random” is appropriate, “arbitrary” is not. “Random” is an adjective often applied to samples. Of a random sample of say, wheat drawn from the hold of a ship (to borrow an example used by Peirce in “The Doctrine of Necessity Examined”; see 1982, 6.40), it cannot be said that nothing matters about it except that it is a sample of wheat. What matters is the quality of the wheat in this particular sample. But if we ever had occasion to speak of “an arbitrary sample of wheat,” all that would matter the only premise from which we could reason would be that it was a sample of wheat. It is easy to understand the temptation to translate the adjectival form -/3C and the prepositional phrase using its genitive -/3C-#+ as “chance” or “random.” The adjectival form -/3C is the aorist participle of the verb -/97, one important new meaning of which is “to happen by chance.”

1 Aristotle is here talking about -'?7C - -'?7#.,, “a triangle quâ triangle.” Perhaps such an interpretation of this passage in the Posterior Analytics sheds some light on the meaning of the celebrated phrase -°² ², ‘Being quâ being,” in , e.g.., Metaphysics 1003a21. 2 Tredennick’s translation appeared in 1924. The paper by Gentzen in which the Rule is for the first time formalized. appeared in 1934. It is unlikely that Tredennick would have seen any need to translate -/3C in an unusual way.

128

Henry W. Johnstone

When Aristotle uses this participle in the sense of “arbitrary,” this may well be the first time the word had been used in this sense.1 What Aristotle must have in mind in speaking of an arbitrary figure is the part such a figure plays in a proof on the order of the one formalized later by Euclid (I.e., Bk. I, Dem. 32) to show that the angles of every triangle are equal to two right angles. This theorem is proved in terms of an arbitrary triangle showing on the basis of axioms, postulates, definitions, and previously proved theorems that the arbitrary triangle under consideration (that is, any triangle) does indeed necessarily have an angle-sum of two right angles. Euclid’s proof begins “Let there be a triangle ABC.” (`)-7 -'?7# -° ) It then proceeds to its conclusion that the sum of the angles of ABC is two right angles without making use of any property of ABC except that it is a triangle. For ABC is an arbitrary triangle. Instead of the )-7 construction, Euclid might have spoken of a -/3° -'?7#. Euclid does in fact often use -/3C to mean “arbitrary” (See, for example, Dem. 5, Line 11, where Euclid uses the expression -/3° )J¥#, “an arbitrary point.” He seems, in fact, throughout the Elements to reserve the participle -/3C for points. But there are, as we have seen, alternative ways in Greek of expressing the idea of “arbitrary”; for example, using )-7 to introduce an arbitrary entity like a triangle.) Aristotle and Euclid each expresses in his own way the proposition that if any arbitrary triangle has an angle-sum of two right angles, then every triangle must have the same property. Inferences of this sort, falling under the rubric of “Universal Generalization,” may seem to demand that we follow Aristotle and Euclid in designating some term involved in the proof as “arbitrary.” But it should be noted that the systems of logic introduced by Gentzen under a heading that has acquired the name of “Natural Deduction” are formalized in such a way that the characterization of a variable as “arbitrary” is inappropriate and in any event unnecessary, since the function of such a characterization is performed by formal restrictions on the introduction of the variable to which the universal quantifier may be applied. Some examples of such formalization of Universal Generalization are to be found in Price (p. 95), LeBlanc and Wisdom (p. 205) and Anderson and Johnstone (p. 170) as well as in Gentzen himself (p. 186). But formalizations of this sort, in my view, have no tendency to show that there never was a need for the concept of the arbitrary, or that there is not still a need for it. Euclid did not have to wait for Gentzen to systematize geometry. And at the end of this article I want to exhibit the 1 I assume, as most do, that the Posterior Analytics is earlier than the De Caelo, where Stocks translates-/3C as “arbitrary” in, for example, “[an] arbitrary fragment of earth.” (297b8)


129

function of the unformalized concept of the arbitrary in the analysis of certain kinds of philosophical arguments and conclusions. The examples I have given of the uses of the concept of the arbitrary in Aristotle and Euclid are applications of the general principle that while “any,” when it appears, does so in a step of a demonstration, “every” introduces the statement of the theorem demonstrated. What I am presenting here is a sort of ideal lexicography: in an ideal version of the English language, “every” would introduce a theorem and “any” (provided it function as a universal quantifier1) a step in its proof. An infraction of this ideal rule is perhaps a little more painful when “every” is used to express “arbitrary” than when “any” introduces a theorem. Thus when Heath’s translation of the theorem demonstrated by Euclid’s Demonstration 32 begins “In any triangle ...,” (p. 316) I, perhaps with linguistic hypersensitivity, but at any rate in conformity with Euclid’s Greek, would have expected “In every triangle ...” (%-°O -'I#/., as Euclid in fact has it.) . I do not intend the scope of these remarks to be restricted to theorems and proofs in Euclid. I don’t know what would prevent them from applying to any mathematical theorem, provided only a generalization were being proved and the proof hinged on the introduction of arbitrary instances. (But see below for reference to a theorem in which this is not the case.) (I suspect that the distinction I am trying to draw here is not unique to the English language. Isn’t it the case that while “tout” normally introduces a theorem, “n’importe quel”2 is more suitable in a step of a theorem? And that the same relation holds between “jeder” and “irgendeiner”? It should be recalled, however, that our English “any” in negative, conditional, and interrogative contexts behaves semantically in an idiosyncratic way.3) In any event, leaving such idiosyncrasies aside, I am here distinguishing “every” and “any” only within the context of proofs making uses of the Rule of Universal Generalization, which can be expressed semi-formally as “If 1a, where a is an arbitrary individual, then for every x , 1x .” I have not yet said anything about Universal Instantiation, in a Gentzenian formulation a logical

1 In this phraseology I am attempting to take account of the possibility that “any” might express an existential quantifier, as in “if any x is P, then every x is /g” – a example suggested to me by Hans Hansen. (Regarding the phrase “if any”, see Footnote 3, below. 2 This phrase literally “It doesn’t matter which” seems to expresses more vividly than “any” the arbitrariness of what it governs. 3 Consider “Not any S is P.” If there is any S, it is P,” and “Is anything S?” In none of these examples could “Every” be substituted for “Any” without radically changing the meaning. I know of no other language in which such idiosyncrasies are exhibited. But this study is not linguistic, and I don’t feel obliged to pursue the matter.

130

Henry W. Johnstone

movement in a direction opposite to that of Universal Generalization. According to this rule, from “For every x , 1x ,” we can deduce “1a .” Here the individual a can, but need not, be arbitrary. It can, for example, be either an arbitrary triangle or a specific one. But if we wish now to continue our train of reasoning by applying Universal Generalization to premises making use of what we have inferred through Universal Instantiation, we can deal only with an arbitrary individual say, an arbitrary triangle rather than a specific triangle the properties or dimensions of which, over and above its being a triangle are to be taken account of in our reasoning. ***** In Johnstone 1996a (p. 95), I deal with the following problem: In an argumentum ad hominem in philosophy, as well as many non-philosophical contexts, the attempt is made through the use of this argument to show that on his or her own principles, a person ought to reject a thesis he or she has asserted. If this argument is valid, the person attacked is under an obligation to withdraw or modify the thesis in question. But to attack a person in such a manner is not yet by any means to establish a general philosophical position. In my view, however, a valid philosophical ad hominem does establish a general position, provided in attacking the holder of the position we have taken account of nothing about the holder except that he/she takes the position in question. In other words, we treat the holder of the position as an arbitrary individual.1 And we apply Universal Generalization. In the so-called “third man” (but more appropriately called “the third greatness”) argument, an ad hominem addressed against Socrates by Parmenides in that eponymous Platonic dialogue, (in 132ff), Socrates is shown that his theory of forms leads to an infinite regress, and so is untenable. The argument is ad hominem because each step of the regress is powered by Socrates’ own assent to consequences invoked one after another by Parmenides. The form of greatness must share some further form with the great objects to which it is common. (I have written about this effect in Johnstone 1996b, and elsewhere.) This argument appeals to no feature of Socrates except his espousal of a certain theory of forms. Socrates is, then, an arbitrary holder of that theory. By Universal Generalization it follows that it is not merely Socrates’ maintenance of the theory in question that is under attack; it is the maintenance of the 1 To treat the person attacked as an arbitrary individual would completely undercut the very point of an argumentum ad hominem in the abusive category, and would not often be appropriate for one considered circumstantial .


131

theory by everyone who now maintains it or might be tempted to maintain it in the future. It is the theory itself that is untenable. A philosophical argumentum ad hominem has been used to reach, via a Universal Generalization, a conclusion independent of the homines who hold it. This conclusion has actually required two steps: from any thinker who might hold the theory in question to every such thinker; and then from such universal concurrence to the falsehood of the theory itself. Aristotle, in his argument1 more appropriately than Plato’s called “the third man” because it invokes, in addition to the illustrative men Socrates and Plato (who are incidentally treated in this example as arbitrary men, since none of their specific properties is at issue here except that of being men) the form in which they both participate. Aristotle then sets this form up beside the two men, considering that there must now be a further form common to all three. He carries the inquiry still one step beyond the conclusion reached in the Parmenides , basing on the general rejection of the theory of forms his own positive view of substance and attribute. To summarize: an argumentum ad hominem valid against any person who is arbitrary in the sense that the only property of this person of which we are taking account in arguing is his/her commitment to the point of view under attack is valid against every such person, and hence refutes the point of view in question itself. From this point we may proceed, as Aristotle did, to establish positive conclusions in philosophy conclusions having their origin in argumenta ad hominem . It might be objected that when a position is universally refutable ad hominem, the situation can be characterized without recourse to the idea of an argumentum ad hominem by simply stating that the position itself is refutable in its own terms the homo need not be called on to participate. For example, the version of the theory of Forms that gives rise to the third man simply generates an infinite regress, and must be rejected for that reason regardless whether anyone actually holds this theory and can, like Socrates confronted by Parmenides, be attacked ad hominem for holding it. But in Johnstone 1989 I argue at length that all valid philosophical arguments are ad hominem. That this consideration especially applies to infinite regress arguments I try to show in Johnstone 1996b. But it is by no means restricted to them. The underlying reason why I think that all valid philosophical arguments are ad hominem is that none of them, in my opinion, is ad rem . For there is in philosophy no generally accepted res no set of facts beyond controversy to 1 Summarized by Alexander of Aphrodisias in Commentarius in Metaphysica. 83.33-84.7 and 84.22-85.3

132

Henry W. Johnstone

which one could properly i.e., validly appeal as a basis for arguing against another view. Consider, for example, the issue between the Protagorean view that man is the measure of all things and what is presupposed by the Biblical question “What is man that thou art mindful of him?”1 To what res what evidence could anyone appeal as evidence supporting either of these views? But if the ad rem is unavailable, our only resource is the ad hominem; e. g., to apply the Protagorean dictum against Protagoras himself. (But of course, “Protagoras” can be taken as designating any arbitrary adherent to this dictum; the argument establishes the falsity of the dictum itself.)2

References Alexander of Aphrodisias. Commentarius in Metaphysica. In Barnes 1984a, Anderson, John M., and Henry W. Johnstone, Jr. 1962. Natural Deduction, The Logical Basis of Axiom Systems. Belmont, CA: Wadsworth. Barnes, Jonathan, ed. 1984a. The Complete Works of Aristotle. The Revised Oxford Translation. Princeton: Princeton University Press. ——, trans. 1984b. Aristotle’s Posterior Analytics. In Barnes 1984a. ——, trans. 1994. Aristotle, Posterior Analytics. 2d ed. Oxford: Clarendon Press. Gentzen, Gerhart. 1934 35. “Untersuchen über das logische Schliessen.” Mathematische Zeitschrift 39: 176 210. Heath, Thomas L., ed. and trans. 1926. The Thirteen Books of Euclid’s “Elements.” Cambridge: Cambridge University Press. Heiberg, I. L., ed. 1883. Euclidis Elementa. Leipzig: Teubner. Johnstone, Henry W., Jr. 1989. “Self-Application in Philosophical Argumentation.” Metaphilosophy 20.3&4:247 261. ——, 1996a. “Locke and Whately on the Argumentum ad Hominem.” Argumentation 10.1: 89 97. ——. 1996b. “The Rejection of Infinite Postponement as a Philosophical Argument.” The Journal of Speculative Philosophy 10.2: 92 104. Leblanc, Hugues, and William A. Wisdom. Deductive Logic .Boston: Bacon and Allyn 1972. Peirce, Charles S. 1982. Writings of Charles S. Peirce, A Chronological Edition. Ed. Max Fisch. Bloomington, IN: Indiana University Press. Price, Robert, 1962. “Arbitrary Individuals and Natural Deduction.” Analysis 22:94-96 Stocks, J. L., trans. 1984.”On the Heavens” (“De Caelo”). In Barnes 1984a. Tredennick, Hugh, trans. 1926. Aristotle’s Posterior Analytics. Loeb Classical Library. Cambridge, MA: Harvard University Press.

1 Psalms 8, 4 2 Those who have helped me in many ways all essential, but some more logistical or editorial than dialectical with the preparation of this brief article include Ray Ayoub, Barbara Bennett, Donna Black, Carl Hausman, Dale Jacquette, Mari Lee Mifsud, Dawn Osselmann, Robert Price, Stephen Wheeler, and especially David Engel.

133

Interdisziplinarität und Interpretation

ON CONTEMPORARY PHILOSOPHY

HANS LENK Interdisziplinarität und Interpretation Im ersten Teil meines Beitrages möchte ich auf die Verflechtung der verschiedenen Problembereiche in unserem komplexen und verflochtenen Zeitalter hinweisen; im zweiten Teil werde ich auf Typen der Interdisziplinarität und auf die Frage der Abgrenzung oder Abgrenzbarkeit der verschiedenen Disziplinen eingehen. Im dritten Teil werde ich meinen Ansatz skizzieren, den ich einem methodologischen Interpretationismus oder methodologischen Interpretationskonstruktionismus nenne; dieser scheint mir eine Brücke zwischen den Disziplinen zu ermöglichen. Zuletzt werde ich noch hinweisen auf bestimmte Tendenzen der Geschichte der Textinterpretation bzw. der methodologischen Hermeneutik, die in gewisser Weise im Zusammenhang mit diesem interpretationistischen Ansatz stehen. I. Zunächst beginne ich mit einem allgemeinen Überblick über Problemverflechtungen. Es ist weithin bekannt, daß die drängenden Fragen der Zeit sich heute nicht mehr in sauberer fachlicher Abtrennung oder disziplinärer Verpackung stellen, sondern interdisziplinär nur zu bearbeiten sind. So sind etwa die ökologischen Fragen nicht nur naturwissenschaftliche Probleme, aber natürlich auch nicht nur geistes- oder sozialwissenschaftliche oder kulturelle Probleme. Sondern wie bei allen unsere komplexen und übergreifenden auch natürlich auch die Nationengrenzen übergreifenden Problemzusammenhängen können wir eine innere Wirkungsverflechtung und eine Vermaschung der entsprechenden Bereiche feststellen. Wir brauchen mehr und mehr abstraktere fachübergreifende Methoden, Disziplinen und Erfassungstechniken, sozusagen generalisierte operationale Techniken, um diese Probleme überhaupt präzisieren und auch behandeln zu können. Das heißt, es gilt in gewisser Weise eine abstraktere und methodologische Sicht einzunehmen, die ich im zweiten Teil zu diskutieren versuchen werde. Das gilt übrigens auch für die Geisteswissenschaften – und auch für die Germanistik. Es ist deutlich, daß in den letzten

134

Hans Lenk

Jahren die Informationsverarbeitungsverfahren, die elektronische Datenverarbeitung und andere Techniken in den Geisteswissenschaften erheblich zugenommen haben und immer mehr Relevanz gewinnen. Es geht um Informationen, genauer: um gedeutete Informationen oder Interpretationen bzw. um Ergebnisse von Interpretationen. Man könnte von “Interpretaten” sprechen oder von “Interpretationskonstrukten”, die gleichsam zu einem neuen höherstufigen “Rohstoff” der Wissenschaften, in diesem Falle der Geisteswissenschaften und der entsprechenden angegliederten Informationsverarbeitungsdisziplinen, geworden sind. Zumal die Information und ihre Handhabung sind in den letzten Jahrzehnten dem systematischen technischen Zugriff zugänglich geworden. Man kann geradezu diagnostizieren, wie ich es bereits 1970 getan habe, daß die Informationsverarbeitung systematisiert wurde und zu einer Art von “informationsund systemtechnologischen Zeitalter” Anlaß gegeben hat. Das gilt natürlich sowohl für die Produktionsautomatisierung und die Einbettung in Systeme wie für die systemhafte operationale Abläufe generell. Hier kann man von einem Trend zu einer umfassenden Systemtechnik sprechen. Das gilt aber natürlich auch für die repräsentationalen, also für die Darstellungen in Informationsdisziplinen, Informationssystemen; auch hier sprechen wir von Informationsverbundsystemen, die immer mehr Bedeutung gewinnen. Der Einsatz des Computers ist hierfür natürlich auf allen Gebieten der wissenschaftlichen Darstellung, aber auch der Produktions- und Verfahrenssteuerung und -kontrolle einschlägig. Alle diese Trends sind Gesichtspunkte einer umfassenden Informations- und Systemrationalisierung in den hochindustrialisierten Gesellschaften. Ich habe schon 1970 davon gesprochen, daß das technische oder technisch-wissenschaftliche Zeitalter sich zum “informationsund systemtechnologischen Zeitalter” – oder kurz zum “systemtechnologischen Zeitalter” – gewandelt hat. Die Informations- und Systemtechnologien übergreifen die Grenzen einzelner Disziplinen. Man kann sozusagen von einer systemtechnologischen oder sogar systemtechnogenen Interdisziplinarität sprechen. Die scharfe Abtrennung in Disziplinen ist eigentlich nicht mehr in diesem Sinne haltbar – und zwar schon aus operationalen methodologischen und quasimethodentechnischen Gründen nicht. Das führt natürlich auch zu entsprechenden Herausforderungen einerseits der wissenschaftlichen Methodologen, andererseits natürlich auch der Sozialphilosophen und Moralphilosophen. Wir alle kennen die Probleme, die sich mit den Dokumentationssystemen, der Zugänglichkeiten von Daten, der Kombinierbarkeit von Daten in Hinsicht auf Datenschutzfragen, Datenschutzgesetze usw. ergeben. Man kann fast befürchten, daß wir so etwas wie eine Computerokratie erleben werden ; die Auffassung, daß dieses das unaufhaltsame Schicksal der industriellen Massengesell-


135

schaften im systemtechnologischen Zeitalter sei, ist weit verbreitet. Heutzutage gilt das natürlich ganz besonders und wird bedeutsam durch neue weltumspannende Informationssysteme, wie beispielsweise das INTERNET, World Wide Web und die direkten Zugriffsmöglichkeiten allerorten, die dann aber auch die Frage der ethischen Verantwortbarkeit für die darin enthaltenen oder eingespeisten und oft gar nicht mehr zuordenbaren Daten aufwerfen. Die humane Verantwortung auch für die Folgen und Entwicklungen in verwickelten Informationssystemen kann weder ethisch noch rechtlich von einem einzelnen getragen werden. Diese Fragen sind durchaus noch sehr offen und man kann sich hierfür durchaus noch keine Lösung vorstellen, wie eine operationalisierbare, eine greifbare und handhabbare Ethik bzw. Rechtsprechung oder Legislative bezüglich der weltweiten Informationssysteme aussehen wird. In gewissem Sinne sind wir natürlich aber immer für eine konkrete Humanität und deren Bewahrung im Umgang auch mit Informationen verantwortlich, ähnlich wie wir im Umgang mit unseren Handlungspartnern verantwortlich. Für die Fragen der Tradition und der Entwicklung der konkreten Humanität sind natürlich die philosophischen und geisteswissenschaftlichen Disziplinen zuständig, aber das ist eine Frage, die heute und hier nicht behandelt werden kann. Jedenfalls kann es nicht mehr angehen, daß die Problembereiche und Disziplinen voneinander abgeschottet werden. Die Probleme stellen sich nicht in einer schubladenartigen Auftrennung der Fächer, und insofern können natürlich auch die einzelnen Disziplinen nicht mehr unabhängig voneinander operieren, sondern sie alle müssen die interdisziplinäre Herausforderung aufnehmen. Das gilt zumal natürlich auch für die Geisteswissenschaften. Die Systemzusammenhänge im systemtechnologischen Zeitalter erfordern eine Anwendung von abstrakteren Verfahren und Verallgemeinerungen sowie von formalen und funktionalen Gesichtspunkten der Darstellungsweisen: Es wird immer wichtiger, eine fachübergreifende Erfassung, d. h. also eine Beschreibung und auch einen praktischen handelnden Umgang mit Gegenständen, Verfahren, Systemen, zu entwickeln, der nun “interdisziplinär” im echten Sinne genannt werden kann. Es gilt also eine praxisnahe und wirklichkeitsangemessene Methodologie, aber auch eine Erkenntnistheorie, Epistemologie zu entwickeln, die diese Herausforderung annehmen kann und die interdisziplinäre Verfassung und Verflechtung der Problembereiche angemessen berücksichtigt. Dabei ist es wichtig zu beachten, daß es in der Tat Tendenzen zu einer Verselbständigung der Systemoperationen und Systeme selbst, der Systemtechnokratie bzw. der Computerokratie gibt, die natürlich auch nur durch eine überfachliche Abstimmung, Sicherung, Kontrolle der vielfältigen Einflußgrößen und Gesichtspunkte, die über die fachspezialistische Einseitigkeit

136

Hans Lenk

hinausgeht, aufgenommen werden kann. Demzufolge brauchen wir außer der Teamzusammenarbeit unterschiedlicher Fachspezialisten und Experten aus unterschiedlichen Fakultäten auch Generalisten, die abstraktere Methodologien, methodische und operationale Ansätze beherrschen und in verschiedenen Gebieten anwenden können. Das gilt ferner auch für die “Spezialisten für das Allgemeine”, die Universalisten, die übergreifende Ziel- und Wertprobleme und die wissenschaftstheoretischen und methodologischen Grundlagen der entsprechenden Disziplinen mitbehandeln können. Diese interdisziplinäre Verfassung und Verflechtung der Probleme stellt sich in dem Kreuzungsbereich vieler klassischer Einzelfächer dar. Es handelt sich aber um überfachliche oder nur transdisziplinär zu erfassende Fragenkomplexe, also um die Notwendigkeit, interdisziplinäre und gar supradisziplinäre Ansätze zu entwickeln. Die klassischen Einzelwissenschaften sind bei diesen bereichsübergreifenden Problemen häufig überfordert. Die Vielfalt der disziplinären Perspektiven und die Widersprüchlichkeiten der Expertenurteile aus einzelwissenschaftlicher Sicht führen zu wichtigen organisatorischen und methodologischen Problemen für die Anwendung der wissenschaftlichen Ansätze, der Expertisen und Resultate, die nun über die Einzelwissenschaften hinausgehen, die sich also der klassischen Schubfächereinteilung entziehen. Hier gibt es eine Reihe von Beispielen auch von neuen Forschungsbereichen, die sich von vornherein in gewisser Weise interdisziplinär angelegt haben bzw. verstehen: einerseits beispielsweise – ganz aktuell – die Umweltforschung, oder man denke an die Wissenschaftsforschung, Wissenschaftswissenschaft (oder “Science of science”), die Peter Weingart eine “multidisziplinäre Aggregatwissenschaft” genannt hat, bei der weniger die Methode als der Gegenstand, eben der Problembereich, der Untersuchung zugrunde liegt. Die Teilforschungsbereiche beispielsweise dieser Wissenschaftsforschung sind etwa Wissenschaftsgeschichte, Wissenschaftssoziologie, Wissenschaftsökonomie, Wissenschaftspsychologie, Organisationswissenschaft, Planungswissenschaft, Teile der Politikwissenschaft und natürlich Wissenschaftstheorie, Methodologie, also Philosophie der Wissenschaft. Es gibt hier offensichtlich noch keine direkte Möglichkeit, eine wirkliche interdisziplinäre Theorie zu entwickeln, deshalb “Aggregatwissenschaft”. Wie kann man nun unterschiedliche Disziplinen und unterschiedliche Arten der Disziplinarität voneinander unterscheiden? (Vgl. das Diagramm.)


137

Unterscheidung der Disziplinen nach ihren/m/r 1. Gegenständen und Bereichen 2. Methoden(arsenalen) 3. Erkenntnisinteressen (Habermas) 4.“Theorien und deren systematischen und historischen Zusammenhängen” (L. Krüger 1987) 5. Theorie-Praxis-Verhältnis 6. Substantivität versus Operationalität versus Formalität der Theorien (Bunge) 7. Systemholismus versus Bereichsspezifik 8. Apriori-Analytik, Methodenformalität versus Empirik 9. Erklärungs- und Systematisierungsmustern (beschreibend versus erklärend, historisierend versus systematisierend) 10. Kognitivität und Normativität (deskriptive versus normative Disziplinen) 11. Fiktionalität und Sekundärrealitat (soziale “Geltung”) versus Primärrealität (“Imprägnation”) Die Disziplinen werden herkömmlich 1. nach Gegenständen und Bereichen voneinander unterschieden oder 2. nach ihren Methoden oder Methodenkombinationen, Arsenalen, 3. nach “Erkenntnisinteressen”, (Habermas) was immer das im einzelnen genauer heißen mag. 4. sind auch die Theorien und deren systematische und historische Zusammenhänge entscheidend und unterscheidend für die Disziplinen. Diese Zusammenstellung bis hier ist von Lorenz Krüger (1973, S. 111 ff.), dem leider verstorbenen Wissenschaftstheoretiker. Wichtig sind aber auch noch die folgenden Punkte: 5. die unterschiedlichen Bezüge zu Theoriepraxis in den unterschiedlichen Wissenschaften; man denke beispielsweise an komplexe Systemtheorien in der Mathematik und der Umweltforschung. 6. ist hervorzuheben – was ich für besonders wichtig halte und was bisher nicht genügend berücksichtigt worden ist – der Unterschied zwischen inhaltlichen, den sog. “substantiven Theorien” (Bunge 1967) und “operativen Theorien”, die eher Verfahren betreffen. Substantive Theorien wären beispielsweise die Gravitationstheorien nach Newton oder Einstein in der Physik, operative Theorien wären beispielsweise die Informationstheorie, die mathematische Spieltheorie oder generell formale Verfahren, analytische Instrumente, die in verschiedenen Wissenschaften angewendet werden können. Offensichtlich sind operative Theorien in der interdisziplinären Forschung besonders sinnvoll verwendbar ; die formalen Theorien sind dabei natürlich diejenigen, die nur formale idealsprachliche Konzepte erarbeiten, wie

138

Hans Lenk

beispielsweise mathematische Theorien. Ein 7. Gesichtspunkt wäre der Punkt “Systemholismus gegenüber Bereichsspezifik”. Systemorientierte Wissenschaften achten auf Gesamtzusammenhänge, auf “holistische” Problemstellungen im Gegensatz zu bereichsspezifischen Einzeluntersuchungen. 8. Ein Unterschied, der auch für die Trennung zwischen Disziplinen wichtig ist, ist die Apriori-Analytik, die Methodologie apriorischer (also von der Erfahrung bereits vorausgesetzter) Art, z. B. formale Methoden gegenüber empirischen, beschreibenden, beobachtenden Vorgehen. (Beispiele liegen auf der Hand, sagen wir einmal in der Logik einerseits und der Paläontologie andererseits.) 9. ist zumal in den Geisteswissenschaften dann bedeutsam und bekannt die Unterscheidung zwischen “erklärenden” und eher “historisierenden” Theorien. Es handelt sich um Erklärungs- und Systematisierungsmuster, also systematisierend oder theoretisch-begrifflich verallgemeinernde erklärenden Theorien auf der einen Seite gegenüber eher beschreibenden und historischen Ansätzen. 10. gibt es den Unterschied zwischen kognitiven und normativen Disziplinen. Kognitiv-deskriptive Disziplinen sind natürlich die Naturwissenschaften, während beispielsweise die Rechtswissenschaft als eine normative Disziplin aufgefaßt werden kann. 11. Fiktionalität und soziale Geltung, also sekundäre Existenz von “Gegenständen”, gegenüber einer Primärrealität, die in den Naturwissenschaften unterstellt wird. Ich komme darauf im zweiten Teil noch zurück. Ich nenne den Beitrag , den die Natur oder die Welt an sich ausüben, der einen begrenzenden Einfluß auf die durchaus aktive Theoriebildung hat, die Wirkung der Imprägnationen im Gegensatz oder Unterschied zu rein fiktiven produzierten Interpretationen im engeren Sinne. Das alles sind also Gesichtspunkte, die zur Unterscheidung der Disziplinen Anlaß geben, die aber dann auch zu verschiedenen Typen der Interdisziplinarität führen. Ich habe im folgenden Diagramm zehn unterschiedliche Typen der Interdisziplinarität aufgeführt: Typen der Interdisziplinarität 1. Interdisziplinäre Projektkooperation 2. Bidisziplinäres bzw. interdisziplinäres Forschungsfeld 3. Multidisziplinäre Aggregatwissenschaft (Sammeldisziplin) WEINGART 4. (echte) Interdisziplin (Bidisziplin) 5. Multidisziplin (multidisziplinäre theoretische Integration) 6. Generalisierte interdisziplinäre Systemtheorie(n) (“Allgemeine Systemtheorie”) V. BERTALANFFY


139

7. Mathematische Theorien abstrakter und komplexer dynamischer Systeme (“Chaostheorie”) 8. Supradisziplinäre angewandte Struktur- und Operationsdisziplinen (“Operations Research”) 9. Methodologisch-metatheoretische Supradisziplinen (Wissenschaftstheorie, Wissenschaftswissenschaft) 10. Philosophisch-methodologisch-erkenntnistheoretische Metadisziplin (“methodologischer Schemainterpretationismus”) Zunächst 1. die einfache bloße Projektkooperation interdisziplinärer Art ; man denke z. B. an die ursprüngliche Entwicklung in der Stadtplanung, wo beispielsweise Experten ganz unterschiedlicher Ausrichtungen mitwirken bzw. eine Rolle spielen. Ich habe das bei der Neuplanung eines Teils der Innenstadt von Karlsruhe (des sog. “Dörfle”) z. T. miterlebt. Dann gibt es 2. eine bidisziplinäre oder interdisziplinäre Verortung eines Forschungsfeldes oder 3. eine entsprechende multidisziplinäre Zusammenwirkung von Forschungsvorhaben, beispielsweise in der Umweltforschung, die jetzt eine Art von Sammeldisziplin geworden ist, evtl. 4. eine multidisziplinäre Aggregatwissenschaft (nach Weingart) ist im Unterschied zu 5. einer spezifischen echten Interdisziplin, wie beispielsweise der Molekularbiologie oder der Biochemie oder besser vielleicht noch der physikalischen Chemie, um eine traditionelle Interdisziplin zu nennen. Dann gibt es 6. die generalisierten interdisziplinären Systemtheorien, wie etwa die seit den 30er Jahren existierende Allgemeine Systemtheorie nach Ludwig von Bertalanffy. 7. sind die abstrakten rein formalen mathematischen Theorien der komplexen dynamischen Systeme anzuführen, die insbesondere in den letzten Jahren besondere Aufmerksamkeit erlangt haben, z. B. die sogenannten Theorien des deterministischen Chaos und die darüber hinaus auch erst zu entwickelnden probabilistischen Chaostheorien, die bisher praktisch noch nicht existieren. (Die Anwendung der Chaostheorie auf die Sozial- und Geisteswissenschaften steht also noch aus, denn es handelt sich hier nicht um deterministische Theorien.) 8.wäre zu nennen eine supradisziplinäre angewandte Struktur- und Operationswissenschaft bzw. eine Disziplin, wie sie etwa in den Verfahrenswissenschaften der Ökonomie in Gestalt des Operations Research schon lange betrieben wird. 9. gibt es dann methodologisch metatheoretische Supradisziplinen, durchaus auch solche traditioneller Art wie die Wissenschaftstheorie und die erwähnte Wissenschaftswissenschaft, der Wissenschaftsforschung. 10. wäre schließlich noch zu erwähnen und zu entwickeln eine philosophisch-methodologische oder metatheoretische Erörterung der entsprechenden Systemzusammenhänge bzw. der betreffenden Disziplinen unter einem gewissen Gesichtspunkt, nämlich dem

140

Hans Lenk

methodologischen Interpretationismus, den ich im zweiten Teil diskutieren werde. Soweit also zunächst ein Überblick über zehn verschiedene Aspekte oder Möglichkeiten, Typen von Interdisziplinarität aufzustellen und nach gewissen Gesichtspunkten der Schärfe oder Stärke des Zusammenhangs zu unterscheiden. Bloße Sammeldisziplinen, die lose ein Forschungsfeld abdecken, sind etwas völlig anderes als eine exakte Interdiziplin wie die physikalische Chemie oder als eine mathematische operative Theorie wie die deterministische Chaostheorie. Hier muß man also meines Erachtens deutliche methodologische Unterscheidungen einführen. Für alle diese Gesichtspunkte gilt aber, daß Wissenschaftler, wenn sie sich dieses interdisziplinären Vorgehens befleissigen, so etwas wie eine Zweitkompetenz in der betreffenden oder wenigstens einer entsprechenden Nachbarwissenschaft brauchen. Hermann Lübbe sprach einmal von “Mitführkompetenz” in einer anderen Wissenschaft, so ist es z. B. ganz klar, daß ein Wissenschaftstheoretiker, der sich systematisch mit Methodenfragen der Biologie befassen will, auch den Stand der Biologie einigermaßen übersehen muß ; er muß zwar nicht ein produktiver Forscher in der Biologie sein, aber doch in der Lage sein, den derzeitigen Entwicklungsstand der Biologie zu beurteilen. Eine solche Mitführ- oder Zweitkompetenz müßte natürlich auch dann für die Ausbildungsgänge gefordert werden. Diese Ausbildung von “Mehrseitigkeiten” und Vielfachkompetenzen erfordert einerseits das Eingehen auf unterschiedliche Wissenschaften – das ist für den einzelnen natürlich nur begrenzt möglich – , andererseits aber auch die Entwicklung gerade der allgemeineren generellen Systemkompetenzen, die schön erwähnt worden sind, insbesondere der abstrakteren und formaleren Methoden der Generalisten und eben auch der Fähigkeit der Universalisten, über die spezifischen Fachorientierungen hinaus beispielsweise übergreifende Wert- und Normensysteme diskutieren, erforschen zu können. Die relativ beste Lösung, die man sich vorstellen kann, ist natürlich nicht diejenige, die nun dem enzyklopädischen Gehirn des Universalisten entspringt, sondern eben eine Gemeinschaftsschöpfung in Teamarbeit durch Wissenschaftler unterschiedlicher Provenienz, insbesondere soweit es unterschiedliche Disziplinen betrifft, die in einem Forschungsbereich relevant sind. Heinz Heckhausen (1987, S. 135) hat einmal gesagt, daß in den Geisteswissenschaften sowieso im wesentlichen intradisziplinär geforscht werde, weil eben alles historisiert werde und man im Grunde eigentlich immer nur unterschiedliche Textgrundlagen habe, aber die Forschungsmethode eigentlich einheitlich dieselbe sei. Ich denke, daß das heutzutage nicht mehr gilt. Was allerdings gilt, je stärker Kontexte, historische Traditionen usw. in Ansätze


141

eingehen, desto stärker wird natürlich auch dieser Gesichtspunkt der Beteiligung unterschiedlicher Disziplinen relevant werden. Es wird heutzutage in den Geistes- und Sozialwissenschaften ein besonders hoher Anspruch an interdisziplinäre Forschung und Lehre gestellt. Ich selber versuche in Karlsruhe seit fast drei Jahrzehnten in interdisziplinären Seminaren immer eine Zusammenarbeit mit den Fachexperten der entsprechenden Disziplinen durchzuführen. Meine Erfahrungen zeigen, daß interdisziplinäre Lehre – insbesondere an Technischen Universitäten – auf die Praxisnähe, die Problemorientierung und die Projektnähe achten muß ; es ist am besten, im Rahmen eines praktischen Forschungsprojekts oder eines begrenzten thematischen Operationsfeldes interdisziplinäre Veranstaltungen zusammen mit Experten der entsprechenden anderen Fächer und Fakultäten anzubieten. Das heißt also, die Zusammenarbeit muß eigentlich, wenn sie sinnvoll betrieben werden soll, auf Dauer gestellt werden und auch in fortwährender Konfrontation und Kommunikation projektbezogen und problemorientiert ausgefächert werden. Dies setzt bei den Beteiligten eine Art von Mitkompetenz voraus, zumindest eine Art von Bereitschaft, die Fachsprache der anderen Disziplin wenigstens verstehen zu lernen und sich damit vertraut zu machen. Man unterstellt dabei auch, daß es so etwas wie eine gemeinsame Basis der Methoden und Zugangsweisen, der Methodologie der Wissenschaften, also der Wissenschaftstheorie, gibt und auch, wie ich darüber hinaus sagen möchte, der “Methodologie des Handelns”, des handelnden Forschens. Das impliziert per se, daß es interfakultative und interdisziplinäre Ansätze und evtl. Institute, die möglichst praxisnah solche Gesichtspunkte einbringen. Das hat unsere eigene Fakultät nun auch versucht. Wir haben vor einem halben Jahrzehnt schon ein interdisziplinäres Institut für Angewandte Kulturwissenschaft gegründet, ein interfakultatives Institut für Kulturwissenschaft, das versucht, angesichts der unterschiedlichen Fragen der Kulturerforschung sich solche Zielsetzungen zu setzen und eine Art von Zusatzqualifikation, eine Art von “Begleitstudium” für Studenten anbietet, die z. B. insbesondere in die Medienpraxis gehen wollen. Unsere Fakultät hat neuerdings auch “berufsorientierte Zusatzqualifikationen” konzipiert, eine Art von kleiner Bestätigung von zusätzlichen Kompetenzen, welche die Möglichkeiten ergeben, daß jemand beispielsweise eine berufsqualifizierende Kenntnis erwirbt, die ihm als Geisteswissenschaftler beispielsweise gerade bei der Arbeit für Rundfunk und Fernsehen, bei der Praxis in den Medien, insbesondere Umgang auch mit Multimedia oder auch bei der interkulturellen Kommunikation eine Hilfe bietet. Entsprechende interdisziplinäre Initiativen finden wir beispielsweise auch in Straßburg, in der Université Louis Pasteur: nämlich im Centre de Recherche Transdisciplinaire sur les Sciences et les Techniques, das ja seit längerem ver-

142

Hans Lenk

sucht, für Naturwissenschaftler solche Fragestellungen zu diskutieren, die über die disziplinären Gesichtspunkte hinausgehen und ein interdisziplinären Diskurs zu pflegen und den interuniversitären Charakter der Interaktionen zwischen harten Wissenschaften einerseits und Disziplinen der Humanwissenschaften andererseits zu betonen. Es geht, wie Pierre Karli (1993, S. 172 f.) sagte, um eine “Entkapselung” der Einzeldisziplinen innerhalb der Universität selber sowie um eine größere Öffnung der Universität auch ihrer Umgebung, der sozialen Umwelt gegenüber. Man wolle sich Gedanken machen “über sich selbst”, d. h. über die Naturwissenschaften und Technikwissenschaften und über den “Sinn” des Beitrages zur Förderung von Wissenschaft und Technik. Das sind ganz parallele Entwicklungen in unseren Universitäten, die, so glaube ich, recht wichtig sind und die Transdisziplinarität der Problemstellungen auch explizit hervorheben und darstellen. II. Der zweite Teil hat die Aufgabe, theoretische Folgerungen aus der skizzierten Problemlage zu ziehen. Dabei möchte ich gerade auch die Implikationen für die Sozial- und Geisteswissenschaften kritisch unter dem Gesichtspunkt des herkömmlichen traditionellen Separatismus der Methoden zwischen Naturwissenschaften und Geisteswissenschaften diskutieren. Es handelt sich um die bekannte Zweikulturentrennung im Anschluß an C. P. Snow (1967), die nur zu begeistert von Profilneurotikern beider Seiten aufgenommen worden ist, obwohl Snow eigentlich gar nicht die naturwissenschaftliche “Kultur” gegenüber der geisteswissenschaftlichen “Kultur” gemeint hat, sondern von der naturwissenschaftlichen und der “literarischen” Kultur sprach. Er kritisierte nicht einen Gegensatz zwischen Wissenschaftsarten, sondern einen solchen zwischen allgemeineren Aktivitäten intellektueller Art. Es ist natürlich ein Unding – da hat Snow sicherlich recht – , daß nach wie vor der Zweite Hauptsatz der Thermodynamik von Geisteswissenschaftlern oder Literaten eher kühl und verächtlich als belanglos angesehen wird, wenn er überhaupt zur Kenntnis genommen wird, und daß ebenso umgekehrt die Naturwissenschaftler, nach ihrer Shakespeare-Lektüre befragt, meistens auch “passen”. Ich kann aus meiner eigenen Erfahrung und von unserer Fakultät an einer Technischen Universität Beispiele dafür anbringen: z. B. hat ein ehemals bekannter Mittelaltergermanist unserer Universität auf einer Party gesagt, es gehöre ja wohl nicht zur Bildung, die Einheit der komplexen Zahlen zu kennen. Solche Behauptungen zeigen Arroganz und Ignoranz derjenigen auf, die solches behaupten. Hingegen hatte ich selber in der Schule einen sehr gebildeten Chemielehrer in der Schule, der beklagte, daß der Mythos von Achilles und der


143

Schildkröte, jedoch nicht die Jod-Stärke-Reaktion zur Bildung zählte. Ich denke, daß solche Entgegensetzungen viel zu grob sind. Ferner fordern manche, daß man mindestens eine “dritte Kultur” (Lepenies), nämlich die der Sozialwissenschaften konstatieren müßte. Ich glaube, nicht einmal das reicht, um die genannte Polarität zu überwinden. Es gibt unterschiedliche Zwischenphänomene, Mischlingsdisziplinen und neuerdings eben sehr viele interdisziplinäre Gebiete, wie ich erwähnt habe, z. B. Aggregatdisziplinen, die eigentlich dieser herkömmlichen Einteilung glatt zuwiderlaufen. Man hatte ja schon immer und traditionellerweise Schwierigkeiten, z. B. die Logik oder die Mathematik einzuordnen, denn diese sind eigentlich Geisteswissenschaften, Präzisionsgeisteswissenschaften par excellence. Sie werden aber sozusagen”natürlich”, aber eigentlich fälschlich in den naturwissenschaftlichen Fakultäten verortet. Dasselbe gilt natürlich in gewissem Sinne ja für die Linguistik, die heutzutage z. T. auch als halb-mathematische Disziplin unter Rückgriff auf Halbgruppentheorie und Halbverbandstheorie betrieben wird ; auch bei ihr kann man nicht unmittelbar sagen, sie sei nur eine herkömmliche Geisteswissenschaft. Sogar in den Naturwissenschaften gibt es auch historische Disziplinen, wie z. B. die Paläontologie oder z. T. auch die physikalische Kosmologie bzw. die Kosmologie generell. Der herkömmliche Separatismus ist also sicherlich abzulehnen, faktisch und methodologisch falsch, überholt und eher “ideologisch” motiviert. Das gilt erst recht, glaube ich, für die traditionellen Unterscheidungen zwischen “Verstehen” und “Erklären”. Wollte man sagen, der Naturwissenschaftler erklärt nur, aber versteht nichts? Das kann man ja wohl kaum sagen. Und sollte vom Geisteswissenschaftler umgekehrt gelten: er “versteht” nur, aber er “erklärt” nichts? Das ist pauschal gesehen ebenso unsinnig. Das heißt, es geht also darum, diese unfruchtbare Dichotomisierung auf einer höheren Metastufe, auf einer methodologischen Metastufe zu überbrücken, zu überschreiten oder gar zu überwinden.Ich denke, daß eine solche Möglichkeit dadurch gegeben ist, daß wir eine allgemeinere konstruktive Theorie der Interpretationen oder, wie ich besser neuerdings sage, der Schemainterpretationen oder Schemaaktivierungen aufstellen, nämlich einen methodologischen Ansatz entwickeln, der gewisse Voraussetzungen der traditionellen methodenorientierten Philosophie mit bestimmten Überlegungen eben auch der Erkenntnistheorie der Naturwissenschaften verbindet oder kompatibel werden läßt. Dies sind Überlegungen, wie sie etwa z.T. von Kant vorgegeben sind, bei Peirce mit seiner Zeichen- und Symboltheorie sowie in Cassirers Symbolphilosophie zu finden sind – und insbesondere bei Nietzsches Auffassung der Grundtätigkeit des Interpretierens. Diese Wurzeln kann man, glaube ich, zu einem konsistenten Ansatz zusammenführen, den ich einen methodologischen

144

Hans Lenk

Schemainterpretationismus nennen möchte. Der Grundgedanke ist, daß alles Erfassen – und zwar sowohl das passive, das theoretische Ergreifen und “Fassen”, z. B. Wahrnehmen, wie auch das Handeln unter Ordnungsmustern, Schemata und in diesem Sinne unter Schema-Interpretationen steht, regelhaft ist. Schemata werden etwa in der Psychologie verstanden als hierarchisch strukturierte, relativ dauerhaft oder gar permanent gespeicherte Wissens- und Handlungsabrufstrukturen, die etwas repräsentieren können, die unter Umständen konventionell zustande gekommen oder erlernt sind, die insbesondere Konfigurationen und Merkmalskombinationen repräsentieren. Sie enthalten zu spezifizierende, d.h. durch konstante Größen oder Werte zu ersetzende, Schemavariable, die dann geeignet sind, diese Schemata auf spezifische Situationen zu beziehen. Die Schemata wurden eingespielt, sind interpretationsgebunden oder gar interpretationsimprägniert in dem Sinne, daß unter Umständen äußere Weltfaktoren (wie beim direkten Wahrnehmen) eine prägende, ja, determinierende Rolle mitspielen. In diesem Sinne ist also das Schemainterpretieren oder das schlichte “Interpretieren” (obwohl das etwas mißverständlich ist) etwas Allgemeineres als das spezifische Textinterpretieren der Hermeneutik. Textinterpretieren ist ein Spezialfall des Schemainterpretierens. Ich komme darauf noch zu sprechen. Bei diesem methodologischen Interpretationismus, genauer: Schemainterpretationismus, wird ernst gemacht mit der Geformtheit aller unserer Erfassungen, sowohl im Erkennen als auch im Handeln – und insbesondere natürlich bei symbolischen Repräsentationen und Erfassungen sprachlicher Art und auch anderer Schematisierungen biofunktionaler oder konstruktiver bzw. darstellender Art. Es handelt sich hier also um einen methodologischen Ansatz, der relativ abstrakt ist, genauer um eine Methodologie der Methodologie, also um Metamethodologie – wir hatten ja auch von Metakonzepten gesprochen. Man könnte den Ansatz auch als einen für Erweiterungen offenen, hypothetischtentativen (nicht dogmatisch fixierten) quasitranszendentalen Ansatz im Sinne Kants bezeichnen. Die Grundidee ist, daß alles, was wir erfassen können, unter der Bildung der relativen Stabilisierung, also Etablierung, Stabilisierung, und der Aktivierung oder Reaktivierung von bestimmten Schemata geschieht wie man sich neuerdings durchaus auch im Rahmen der Neuro- und Kognitionswissenschaften als durch die Einspielung von sog. Neuronenassemblies (Neuronenensembles) plastischer Art getragen vorstellen kann. Die Neurowissenschaftler sprechen ja neuerdings sogar von “Hirnkonstrukten” (z. B. Singer 1990, S. 8) und meinen damit plastische, aber durch wiederholte Aktivierung relativ stabilisierte Neuronenassemblies, die immer wieder reaktiviert werden

145


können. Zumindest kann man auf diese Weise eine “Wie-ist-es-möglich?”Erklärung der Bildung von Schemata verständlich machen. Damit ist natürlich die semantische Lücke zwischen der naturwissenschaftlich-physiologischen Beschreibung von Neuronenassemblies einerseits und der Deutung von eben semantischen Gehalten andererseits noch nicht überdrückt, aber die Korrelation ist etwas enger und nachkontrollierbarer geworden. Das ist ein recht schwieriges Phänomen, das ich hier leider nicht ausbreiten kann. Ich denke aber, daß in dieser Richtung eine Art von Ansatz und Kooperation interessant wird, der für Natur-, Geistes-, Sozial- und eben auch Neurowissenschaften eine Art von Überbrückung und gemeinsamem Ansatz liefert. Ich möchte noch ein wenig auf die schematisierenden interpretierenden Tätigkeiten, Aktivitäten eingehen und diesbezüglich einige Unterscheidungen oder Schichtungen vorstellen. Das läßt sich am besten anhand einiger Diagramme zeigen (vgl. Lenk 1993, S. 109 ; 1993a, S. 254):

INTERPRETATORISCH-SCHEMATISIERENDE AKTIVITÄTEN: (SCHEMA-) INTERPRETATIONEN konstituierende -> konstruierende -> Konstituieren, unbewußtes Auslösen, Aktivieren, Ausbilden, Entwickeln, Differenzieren, Stabilisieren

bewußtes Auslösen, Diskriminieren, Kontrastieren, Vergleichen, (Re-) Identifizieren, Darstellen, Auswählen, Verfeinern von Schemata

Entwerfen, Zuordnen, Aufprojezieren, Variieren, Kombinieren, Organisieren, Integrieren

rekonstruierende Aktivitäten Anwenden, Projizieren, Durchführen, Konstruieren, Repräsentieren, Vorstellen, Kog nizieren, Darstellen

(Re-)Identifizieren, (Wieder)erkennen, Unterscheiden, Zuordnen durch Einsetzen, Subsumieren, sortierendes Klassifizieren, Verstehen i.w.S. sukzessives Weiteranwenden

durch, mittels, mit nach, oder in Schemata (von Konstanzen, Formen/Strukturen/Gestalten, Gegenständen, Ereignissen, Prozessen, Fakten, Relationen, Kontexten) bei Textinterpretation: Re-Identifizieren Anwenden von Schemata (Wieder)Erkennen Verstehen i.e.S.

146

Hans Lenk

So steht das herkömmlich verstandene Interpretieren oder Schemaaktivieren unter dem Gesichtspunkt der interpretatorisch-schematisierenden Aktivitäten oder Schemainterpretationen. Dafür sind in der Abbildung Beispiele aufgeführt und ein wenig nach ihrem Konstruktcharakter geordnet. Man findet auf der linken Seite eher Tätigkeiten wie “Schemata etablieren” und “stabilisieren”, die Gegenstandserfassungen und – formierungen konstituieren; das sind also konstituierende interpretatorisch-schematisierende Tätigkeiten. In der Mitte des Diagramms sind dann im engeren Sinne, konstruktiv bewußt-planmäßig entworfenen Darstellungen oder Formen angeführt, und rechts die eher wiederholenden, wiederaufgreifenden, reaktivierenden, rekonstruierende Aktivitäten oder Schemaaktivierungen wie das Wiederlesen oder Erinnern. Um einige Beispiele zu nehmen: Zum Konstituieren gehört sicherlich das Stabilisieren eines Schemas bei einem Lernprozeß, zum direkten Wahrnehmen aber speziell auch das unbewußte Auslösen durch einen äußeren Reiz. Zum Konstruieren i. e. S. gehört das Variieren: z. B. eines Grundmusters von vorgegebenen Kriterien, das bewußte Variieren. Zum Rekonstruieren gehört das Subsumieren unter einen bekannten Begriff, das einordnende Klassifizieren oder Sortieren – und last but not least natürlich das Lesen, Erkennen, Deuten von Texten. Textinterpretation ist natürlich eine rekonstruierende interpretatorisch-schematisierende Aktivität. Insofern sind natürlich die Textinterpretationen oder die traditionellen Schemata der Hermeneutik, die auf das wiedererkennende Anwenden von Schemata und das Verstehen im engeren Sinne ausgerichtet sind, ihrerseits Spezialfälle dieser schematisierenden Aktivitäten. Darüber könnte und müßte man natürlich im einzelnen ausführlicher diskutieren und referieren, als das hier geschehen kann. Stattdessen möchte ich, um den methodologischen Ansatz und auch die philosophische Relevanz deutlich zu machen, auf einige Gesichtspunkte der Schichtung der Schematisierungen eingehen. Man kann nämlich: Stufen oder Ebenen des Interpretierens, des Schemainterpretierens unterscheiden.


147

(Ebenen) Stufen der Schema-interpretation IS1 praktisch unveränderliche Urschematisierung oder die Imprägnation (genetisch angelegte primäre Konstitution bzw. Schematisierung oder die Imprägnation in der direkten Wahrnehmung) ———————————— IS2

gewohnheits-, gleichförmigkeitsbildende Musterbildung und -reaktivierung (erlernte Muster – Schematisierung), habituelle Form- und Schemakategorialisierung + vorsprachliche Begriffsbildung

IS3

sozial etablierte, kulturell tradierte, übernommene konventionalisierte Schematisierung IS3a

vorsprachlich normierte “Begriffs”bildung und Schema-Interpretation durch soziale und kulturelle Normierungen

IS3b

repräsentierende sprachlich normierte Begriffsbildung (i. e. S.)

———————————— IS4

anwendende, aneignende bewußt geformte Einordnungsinterpretation und Unterordnungs- oder Einbettungsschematisierung (Klassifikation, begriffliche Subsumierung, Beschreibung, Artenbildung u. -einordnung; gezielte Begriffsbildung)

IS5

erklärende, “verstehende” (i. e. S.) rechtfertigende, (theoretische) begründende Interpretation Rechtfertigungsinterpretation, Einbettung in Folgerungs-, Argumentations- und Begründungs-“Schemata” oder Metaschemata

IS6

erkenntnistheoretische (methodologische) Metainterpretation der Interpretationskonstruktmethode

Ich unterscheide sechs verschiedene Stufen (vgl. Lenk 1993, S. 56 ; verändert Lenk 1995, S. 103): einmal praktisch nicht von uns veränderbare, aufgrund unserer biologischen Anlage Urinterpretationen (IS1) – beispielsweise bei einem Normalsichtigen die Unterscheidung zwischen “hell” und “dunkel”: das ist nicht vermeidbar. Das nenne ich also Urinterpretationen oder primäre Schematisierung. Zweitens – es wird von Schicht zu Schicht immer flexibler und plastischer – gewohnheitsbegründete oder gewohnheitsmäßige, gewohnheits- oder gleichförmigkeitsbildende Schema- oder Musterinterpretationen (IS2), die durch Habitualisierung, durch gemeinsames Verhalten vorsprachliche Begriffsbildung z. B. eingespielt worden sind, also durch Musterstabilisierung oder – eingewöhnung. Drittens dann – für uns natürlich

148

Hans Lenk

besonders interessant – die konventionellen Begriffsbildungen, konventionelle Schema-Interpretationen i. e. S. (IS3), die aufgrund von Verabredungen und Normen oder Normierungen durch die Interpretationsgemeinschaft (Kultur, Sprache, Gesellschaft) zustande kommen: Dazu zählen alle symbolischen Scheam-Interpretationen ; denn Symbole sind ja konventionelle Zeichen. Hier sind auch alle Verabredungen bzw. Interpretationen aufzuführen, also sozial etablierte kulturell tradierte und aus der Interpretations- oder Sprachgemeinschaft “übernommene” Begriffsbildungen. Hier muß man natürlich die Unterscheidung treffen zwischen solchen Begriffsbildungen, die aufgrund von Gesten oder Verhaltensweisen konventioneller Art, aufgrund von Normen oder Gewohnheiten, Sitten, Gebräuchen eingespielt worden sind (IS3a) – ohne schon die Sprache explizit zu benutzen – und den sprachlich repräsentierenden, explizit sprachlichen Begriffsbildungen (IS3b). Die nächste Stufe wäre, daß man das symbolische Repräsentieren auf bestimmte Musteranwendungen und -einbettungen anwendet: die vierte Stufe wären also die Unterordnungsschematisierungen und Einordnungsinterpretationen (IS4): Klassifikation, Subsumierung, Beschreibung, Artenbildung und -einordnung, gezielte Begriffsbildungen alle diese Aktivitäten wären also i. e. S. begriffliche, also verallgemeinernde schematisierende Einordnungsinterpretationen. Die nächsthöhere Stufe stellt dann die der Rechtfertigungs- oder Begründungsinterpretationen (IS5) dar, die argumentative Interpretation, wobei man eben erklärt, versteht, begründet, rechtfertigt. Und die letzte Stufe (IS6) leistet die Thematisierung der interpretatorischen Tätigkeit des Methodologen, des Interpretationstheoretikers, selber, nämlich die Metainterpretation der Interpretationskonstruktmethode selber. Denn das, was man als Methodologe beschreibt, ist ja ein Modell, das selbst auch dem Muster des Interpretierens und dieses Schematisierens und somit dieser Schichtung unterliegt. Insofern muß es auch so etwa wie eine höchste Stufe der erkenntnistheoretischen oder methodologischen Metainterpretation der Interpretationsmodelle und -methoden geben.1 Man sieht den Ansatz in dem wiedergegebenen Diagramm dargestellt, das einigermaßen selbstexplizierend ist. Man kann durchaus sagen, daß es durch solche Stufungen und Schichtungen der Schematisierungen und Interpretationen möglich ist, traditionelle methodologische und philosophische Probleme unter neue Sichtweise zu stellen und zu behandeln. Beispielsweise wird das Wahrheitsproblem sich anders stellen, wenn man einen solchen interpretationstheoretischen Ansatz

1 Das Vorgehen kann man natürlich wiederholen, das ist eine kumulative Art, man kann natürlich auch über philosophische Metainterpretationen wieder reflektieren – man müßte eigentlich eine Metastufe höhergehen, aber das ist also in diesem Sinne kumulativ zu verstehen.

149


durchführt. Man kann etwa das Wahrheitsproblem dann im Sinne einer Beziehung zwischen unterschiedlichen Interpretationsschichten deuten, z. B. als die Beziehung zwischen konventionellen IS4- und IS5-Interpretationen und eben den grundlegenden Urinterpretationen IS1 und IS2. Eine generelle Schlußfolgerung ist, daß das Aktivieren und Etablieren von Schemata auf der Grundlage von plastischen neuronalen Ensembles der gleiche oder ein gleichartiger Prozeß ist wie das Reaktivieren. Das Wiedererkennen ist ein ähnlicher Prozeß wie das Aktivieren von Schemata und das Ausbilden. Das Bilden, Etablieren, Stabilisieren und Reaktivieren von Schemata nenne ich Schemainterpretieren. Schemainterpretationen sind in diesem Sinne eben repräsentierende Aktivitäten. Es ergibt sich hier eine Brücke zwischen dem Konstruieren im engeren Sinne, dem Konstituieren von Gegenständen und dem Rekonstruieren, dem Wiedererkennen, dem Wiederaktivieren von Schemata, einfach deswegen, weil das Aktivieren das erste “Einspielen” und das wiederholte Aktivieren von entsprechenden Schemata, das “Wiederabspielen”, jeweils eine Aktivierung von Neuronenassemblies und somit cum grano salis derselbe Prozeß ist. Repräsentieren, Rekonstruieren ist gebunden an das Wiederholen von Aktivierungen, deren Ablauf schon einmal eingespielt worden ist. Das Metaschematisieren, das Metainterpretieren erfolgt auch nach diesem Muster. “Erfassen” – und d. h. Erkennen und auch Handeln (“Fassen” i. e. S.) – ist in diesem Sinne stets schemainterpretationsgebunden oder – etwa beim Wahrnehmen – sogar an äußere Weltfaktoren und deren Einfluß gebunden (“Imprägnation”). Alle diese repräsentierenden Aktivitäten sind schemaimprägniert im expliziten Sinne. Wir können hier von einem Grundsatz allen repräsentierenden “Erfassens” sprechen, dem Grundsatz der Schemainterpretationsgebundenheit und der Schemaimprägnierung – oder schlicht: einem Grundsatz der Schemainterpretation und Schemaimprägnation. Wir können nicht nicht schematisieren, wir können nicht nicht schemainterpretieren. In der Interpretationstheorie heißt es ja eben herkömmlich – auch in der hermeneutischen: wir können nicht nicht interpretieren. Das ist natürlich durchaus richtig – und das läßt sich wie angedeutet auch verallgemeinern. III. Ich denke übrigens, daß alles zuletzt Gesagte auch über die herkömmlichen hermeneutischen und sprachphilosophischen Disziplinen hinaus erweitert werden kann, daß man sogar das Wittgensteinsche Modell der Sprachspiele erweitern kann auf Schemaspiele (vgl. Lenk 1995), daß also in gewissem Sinne verallgemeinert werden kann über das Verbalsprachliche, über das Sprachliche hinaus und in Kontakt und Übereinstimmung gebracht werden kann mit der

150

Hans Lenk

Neurowissenschaft. Man kann direkt in Anschluß an und Analogie Kants Satz aus der Kritik der reinen Vernunft “Gedanken ohne Inhalt sind leer, Anschauungen ohne Begriffe sind blind” (KrV B 75) formulieren: Schemainterpretationen ohne Aktivierung, ohne Interaktionen und gar Interventionen sind leer und Interaktionen und Interventionen ohne Schemadeutungen sind blind. Interpretation, zumal Schemainterpretation, ist stets abhängig von Interaktion und Intervention – und umgekehrt. Das heißt, es ergibt sich eine neue Zusammenführung zwischen den traditionellen methodologischen Gesichtspunkten der Handlungstheorie und der Erkenntnistheorie. Ich denke, daß hier auch eine Brücke zwischen einem wissenschaftlichen Begründungsansatz und auch dem Alltagserkennen und Alltagshandeln gefunden werden kann, allerdings um den Preis einer abstrakteren, höheren Stufung. Das gilt übrigens auch für die Überbrückung zwischen den unterschiedlichen Wissenschaften, beispielsweise den Naturwissenschaften und den fast gänzlich bzw. teilweise hermeneutischen Geistes- und Sozialwissenschaften. Es gibt im methodologischen Zusammenhang eine Einheit der Erkenntnisformen auf höherer Stufe, die die traditionelle methodologische Trennung zwischen den Disziplinen zu überbrücken gestattet. (Der methodologische Separatismus bleibt (zu) vordergründig.) Das hier Entwickelte ist ein recht genereller, abstrakter, aber ein pragmatischer und umfassend anwendbarer, interdisziplinärer, ja, in gewissem Sinne supradisziplinärer methodologischer, ja: metamethodologischer, Ansatz, der die Theorie-, Begriffs-, Hypothesenbildungen der unterschiedlichen Disziplinen übergreift. Es gibt natürlich gewisse Unterschiede und Unvereinbarkeiten unterhalb der abstrakten, höherstufigen Einheitsebene. Das will ich nicht bestreiten. Beispielsweise konzentrieren sich ja die Geisteswissenschaften auf z. T. interpretationsproduzierte, fiktive Gegenstände – und die Sozialwissenschaften auch. Die Institutionen, der Staat oder was immer sind ja keine Dinge, sondern Fiktionen, die nur dadurch sekundäre soziale Existenz erlangen, die aber natürlich sehr wirkmächtig ist, dadurch, daß viele Menschen an sie glauben, bzw. auch durch Regeln zu dieser fiktiven Existenz dressiert werden. Der übergreifende gemeinsame Gesichtspunkt ist, daß Wissenschaft immer Menschenwerk ist. Man sieht, daß die Einheit der Wissenschaften auf einer höheren methodologischen Abstraktionsebene unter diesem Primat eines interpretationskonstruktionistischen Ansatzes modellhaft erkannt und analysiert werden kann. Insofern läßt sich sagen, daß man im methodologischen SchemaInterpretationismus eine höherstufige Brücke zwischen den zwei angeblich völlig getrennten “Wissenschaftskulturen” wie zwischen Erkennen und Han-


151

deln wiederherzustellen vermag. Das ist die Hauptbotschaft und man kann das natürlich auch (vgl. Lenk 1993, 1993a, Kap. 44) selbst an der Geschichte der Hermeneutik und der hermeneutischen geisteswissenschaftlichen Disziplinen exemplifizieren. Traditionell ist ja schon in der Geschichte der Hermeneutik ein Perspektivismus hervorgehoben worden, z. B. bereits 1742 von Chladenius, der die Erkenntnis nach Auszeichnung, Auswahl, Synthese, Komparation immer als von einem Standpunkt aus – er spricht von einem “Sehepunkt” (1969, S. 187) – auffassen wollte. Er sprach auch direkt von “Perspektiven”. Und auf diesen “Sehepunkt” oder perspektivischen Ansatz ist jeder Interpret in den Geisteswissenschaften in den Deutungswissenschaften, zumal in der Exegese, angewiesen. Dabei war auch schon Chladenius klar, daß dieses Einnehmen eines “Sehepunktes” und diese Interpretation unter diesem Gesichtspunkt konstruktiv ist (ebd. z. B. S. 518), eine interpretatorische Tätigkeit, eine Art von Kunst. Dieselbe Einsicht spielt dann natürlich bei Schleiermacher eine große Rolle, der ja ausdrücklich von einer schöpferischen Synthese spricht, von einem Begriff als einem intellektuellen Schema – in Anlehnung natürlich an Kant – , durch das ein Subjekt sich auf seinen Gegenstand beziehen kann und durch das auch bereits ein Gedanke als Ergebnis einer synthetischen und symbolisierenden Leistung dargestellt werden kann. Diese konstruktiv herstellende, poietische oder gegenstandsformierende konstitutive Funktion der Sprache ist also bei Schleiermacher in gewissem Sinne schon erkannt; er spricht auch vom “Schematismus” (natürlich wiederum im Anschluß an Kant) und von einer “Denkgemeinschaft” (Schleiermacher 1977, S. 443 ff., s. a. S. 29) innerhalb eines Sprachkreises, die zu solchen Formierungen führt. Bei Dilthey wird das Konstruktionselement noch deutlicher; er spricht von der “Nachkonstruktion der gegebenen Rede” (Schleiermacher 1974, S. 83) oder von der Nachbildung des ganzen Prozesses, durch den ein Werk “in Regelform” entsteht. Er nimmt Schleiermachers Satz auf: “Ich verstehe nichts was ich nicht als notwendig einsehe und konstruieren kann.” (Schleiermacher 1974, S. 31): ein Satz, der übrigens von Vico und Hobbes schon vorweggenommen worden ist. Dilthey (WW VII, S. 220) erweitert das, indem er sagt: “So entsteht der Sinn, indem das Unbestimmte durch die Konstruktion bestimmt wird.” Er ist sich also durchaus im klaren, daß die Konstitution konstruktiv ist, daß eine Interpretation eine Tätigkeit des handelnden Subjekts darstellt; und er kritisiert m. E. völlig zu Recht an Kant, daß Kant das nur als Problem der reinen Erkenntnis gesehen habe, als Kategorisierung im Sinne der Anwendung von reinen Verstandesformen. Stattdessen meint er, völlig richtig, müßte man auch die Handlungsregeln und die Handlungsformen, die Lebensregeln und die Lebensformen zur fundamentalen Konstitution der entsprechenden Weltorientierung hinzunehmen. Und das entspricht genau dem Vereinen von

152

Hans Lenk

Erkenntnis und Handeln, wie es auch in der neueren Philosophie eine Rolle spielt, nicht nur seit Peirce im Pragmatismus, sondern natürlich auch bei den Ideen des späten Husserl, in dessen Begriff der “Lebenswelt”, bei Wittgenstein in Gestalt der “Lebensformen” und im Sinne seiner Sprachspielkonzeption, die ja eine Art von handlungsbasierter Fundierung eben auch der sprachlichen Bedeutungen vornimmt. Die zentrale Idee des späten Wittgenstein ist ja, daß man Bedeutungen zurückführt auf Handlungsgebräuche, auf eingeschliffene, musterhafte, regelhafte Handlungsformen. Pragmatisierung der Semiotik und Semantik ist das Stichwort. (Das kann hier nicht näher behandelt werden; vgl. Lenk, 1998, Kap. 7.) Wichtig im vorliegenden Zusammenhang ist auch jedenfalls, daß die Funktion des verstehenden Erschließens von Gegenständen grundsätzlich und generell – und zwar schon vor aller geisteswissenschaftlichen Spezialmethodik – konstruktiv, entwerfend, schematisierend, interpretatorisch verfahrend ist oder implizit so geprägt, jedenfalls in vielerlei Sinn aktiv(istisch) ist. Das Verstehen ist also ein erschließendes Konstituieren und Rekonstruieren, wie es z. B. auch Gadamer (1960, 1986) etwa am Beispiel der Konstitution eines Kunstwerks bzw. eines Spiels durch deren Auffassung als Realisierung von bestimmten jeweils geregelten Vollzugsmustern exemplifiziert. IV. Wir haben gesehen, daß der methodologische Ansatz der Interpretationskonstrukte, des methodologischen Schemainterpretationismus, uns ein Werkzeug an die Hand gibt, die interdisziplinären Beziehungen und Arbeitsweise der symbolisch repräsentierenden Geisteswissenschaften unter einem einheitlichen, wenn auch höherstufigen Gesichtspunkt zu begreifen und eine neue, aber abstrakte, methodologische Einheit der Wissenschaften zu vertreten – und zudem Stukturierungsweisen (Schematisierungen, Schemainterpretation und -imprägnation) des Erkennens und des Handelns einzubeziehen. Diese Art von doppelter Supra- und Interdisziplinarität ist natürlich für die Geisteswissenschaften von höchster Relevanz. Eingangs dieses Beitrages wurde kritisch-fragend das Problem der ZweiKulturen-Trennung zwischen naturwissenschaftlichen und geisteswissenschaftlichen Ansätzen aufgeworfen und die expositorische Frage nach Überbrückungsmöglichkeiten dieser inzwischen überholt erscheinenden Kluft gestellt. Eine entsprechende Kluft scheint sich auch – vielleicht noch drastischer als zwischen naturwissenschaftlicher und literarisch-geisteswissenschaftlicher Intelligenz – zwischen der wissenschaftlich-technischen Theorienbildung und -anwendung sowie der hochkomplizierten computergestützten


153

Informationsverarbeitung einerseits und dem Alltagsverstehen bzw. Umgehen mit umgangssprachlichen Konzepten andererseits aufzutun. Die Frage war, ob die Erkenntnistheorie etwas zur Überwindung dieser Kulturentrennung beitragen kann. Hierzu und zur künftigen Entwicklung der Epistemologie unter diesem umgreifenden Gesichtspunkt ist noch in einem resümierenden Ausblick Stellung zu nehmen. Entgegen dem ersten Anschein einer Absolutheit und Unüberbrückbarkeit der genannten “Kulturen”-Spaltung hat sich im Verlauf der differenzierten Auseinandersetzung mit bisherigen Interpretationstheorien und interpretationistischen Ansätzen, mit wissenschaftstheoretischen Konzepten der Naturund Sozialwissenschaften sowie mit erkenntnistheoretischen Überlegungen zur Alltagserkenntnis ergeben, daß es eine abstraktere übergeordnete erkenntnistheoretische Ebene gibt, auf der diese so unüberbrückbar erscheinende Spaltung doch überwunden werden kann. Die “Brücke” dabei bildet der Begriff der konstruktiven Interpretation bzw. der Schemainterpretationen und der interpretatorisch-schematisierenden Aktivitäten sowie des Handelns und Erkennens mit Symbolen und symbolanalogen inneren Repräsentationen. Das Erkennen und Handeln sowohl im Alltag als auch in den Wissenschaften und den geisteswissenschaftlichen wie auch den philosophischen Disziplinen ist symbolvermittelt. Ernst Cassirers Einsicht (1944, dt. 1990), daß der Mensch zwischen die Welt und sich eine “symbolische Zwischenwelt” spannt, die ihm ein “symbolisches Universum”, seine symbolische Welt bedeutet, ist hier einschlägig. Der Mensch als das “symbolische Wesen” ist auf die Ausbildung eines “Symbolsystems”, “Symbolnetzes” angewiesen, das ihm erst die handelnden und erkennenden Zugänge zur Welt bzw. sogar deren Konstitution als Gegenstandswelt ermöglicht und in differenzierter Weise zu strukturieren erlaubt. Symbolverwendung und symbolische Darstellung sind nun für die unterschiedlichen Umgangsweisen und Darstellungen sowie Handlungen auf beiden Seiten der Kulturentrennung charakteristisch. Hier ist ein übergreifender Gesichtspunkt gefunden, der zur Überbrückung der Spaltung auf höherer, abstrakterer erkenntnisphilosophischer und methodologischer Ebene Ansatzpunkte, sozusagen die Brückenköpfe für den Überbrückungsversuch, bietet. Denn alle zentralen Begriffe der Erkenntnis und der Handlungsanalyse sowohl im Alltag als auch in der Wissenschaft und in den geisteswissenschaftlichphilosophischen Disziplinen sind auf Symbolbildung, -verwendung und deutung angewiesen, beruhen auf interpretatorisch-schematisierenden Aktivitäten, auf der Anwendung von teils evolutionär vorgegebenen und unbewußten, teils interaktiv mit der Umwelt entwickelten, teils konventionell aufgebauten bzw. im engeren sozio-kulturellen Sinne erlernten Schemata. Deren Bildung, Ausdifferenzierung, Weiterentwicklung und Anwendung ist in diesem

154

Hans Lenk

Beitrag als Interpretieren im weiteren Sinne aufgefaßt worden: genauer als Schemainterpretieren. (Das traditionelle, von der Hermeneutik untersuchte Verstehen von Texten wurde demgegenüber als Spezialfall, nämlich als Unterkategorie des Textinterpretierens verstanden.) Es zeigte sich, daß alle Arten der Darstellung und des Zugriffs zur Welt, zum erkennenden und handelnden Subjekt selbst und auch zur Situation der Person und des Menschen in seiner Handlungs- und Weltsituation (Lebenswelt) zutiefst interpretatorisch geprägt, von Schemainterpretationen strukturiert und durchdrungen, hinsichtlich ihrer Darstellung und Erfaßbarkeit generell unvermeidbar interpretationsimprägniert sind. Der Grundsatz des methodologischen Interpretationismus (Interpretationskonstruktionismus), daß alles Erkennen und Handeln schemainterpretationsgebunden bzw. -imprägniert ist und wir nur in zutiefst interpretationsabhängiger Weise überhaupt etwas erfassen, konzipieren, meinen, ordnen und komponentendifferenzierend (be)handeln können, steht jenseits aller Bezweifelbarkeit. Er ist der Fundamentalsatz der methodologisch orientierten Erkenntnistheorie der Interpretationen und Theoriebildungen. Selbst ein irgendwie differenzierendes, also Komponenten herausstellendes und unterscheidendes, auf Erfassung hin ausgelegtes bzw. Einzelerfassung ermöglichendes Wahrnehmen ist schon durch eine wahrnehmungsbahnenspezifische, aber erregungsleitungunspezifische Schemaanwendung in Gestalt einer Zerlegung der ursprünglichen Reizinformation, einer Verarbeitung und späteren Wiederzusammensetzung (cortikalen Synthese) gekennzeichnet. Schemainterpretation waltet hier überall, wie auch die neurobiologische neuere Gehirnforschung zweifelsfrei bestätigt hat, so wenig sie auch noch über die Integrations- und Syntheseprozesse der höheren Stufen empirisch gesichert aussagen kann. Darüber hinaus sind zentrale Begriffe des symbolischen Erfassens natürlich selber Teil eines interpretatorischen Ansatzes: Nicht nur ist das Modell des Interpretationskonstruktionismus selbst ein erkenntnistheoretisches Interpretationskonstrukt höherer Stufe, sondern auch die Grundbegriffe dieser Methodologie und ihres Verständnisses (wie auch des Alltagsverständnisses selbst) wie z. B. “Bedeutung”, “Information” usw. sind ihrerseits Interpretationskonstrukte – wie auch praktisch alle allgemeineren, unserer Welt- und Selbststrukturierung zugrunde liegenden Alltagskonzepte: Selbst die Grundbegriffe von “Ich” und “Welt”, der Unterscheidung von Subjekt und Objekt, von Erkennen und Handeln, von Form oder Struktur und Gehalt sind erkenntnistheoretisch-methodologische Begriffe von interpretativem Charakter. Zunächst ist der Interpretationskonstruktionismus als methodologischer Ansatz konzipiert und entwickelt worden. Er kann jedoch auch in der Traditi-


155

on der traditionellen Erkenntnistheorie als ein quasi kantischer transzendentaler Interpretationismus aufgefaßt und ausgebaut werden – gleichsam zu einer Erkenntnistheorie des kulturellen und symbolischen Wesens, das der Mensch ist. Darüber hinaus – und auch über Cassirer hinaus weiterschreitend – müßte dieser Ansatz zu einer Anthropologie des interpretierenden Wesens – genauer: des meta-interpretierenden Wesens – ausgearbeitet werden. Der Mensch ist nämlich nicht spezifisch genug durch die Charakterisierung als das symboleverwendende und -interpretierende Wesen gekennzeichnet, sondern gegenüber den durchaus vielfältige Symbole beim Wahrnehmen, Erkennen verwendenden und auch in primitiven, etwa in Labyrinthsituationen und Entscheidungsdilemmata, “logisch” schließenden höheren Säugetieren, zumal Primaten, ist der Mensch dadurch ausgezeichnet, daß er sein interpretierendes Erkennen und Handeln seinerseits wiederum zum Gegenstand einer höherstufigen Interpretation machen kann. Er unterscheidet, differenziert und interpretiert nicht nur in einer spezifischen Schicht der Interpretationen – etwa durch Konzeptualisierung unterschiedlicher Gegenstandsklassen, Eigenschaften, Relationen usw. – , sondern er kann das Interpretieren selbst wieder zum Gegenstand des Interpretierens machen – in prinzipiell unabgeschlossenen Übereinanderschichtungen von Interpretationsprozessen. Der Mensch ist also ein in spezifischerem Sinne symbolisches und handelndes Wesen, nämlich jenes, das seine Symbolverwendungen wiederum interpretierend zum Gegenstand machen kann, in Metastufen übersteigen, abstrahierend darstellen und manipulieren kann: Er ist das Metastufenwesen, das metaschematisierende und superinterpretierende Wesen, das symbolisch auch auf höheren Metastufen des Erkennens und Handelns Symbole über Symbole, Interpretationen über Interpretationen setzen, Interpretationskonstrukte von Interpretationskonstrukten bilden kann. Es ist diese Möglichkeit der Emanzipation von den Symbolverwendungen auf der Objektstufe bzw. in der Objektsprache, die den Menschen als das metainterpretierende Wesen auszeichnet. (Die Ausarbeitung einer derartigen Anthropologie des metainterpretierenden Wesens muß einer weiteren Arbeit vorbehalten bleiben.) Erkenntnistheoretisch gesprochen ist es jedenfalls deutlich, daß mit dem Konzept der Schemainterpretation und der interpretatorisch-schematisierenden Aktivitäten ein umfassender Ansatz gefunden ist, der sowohl die hypothetischen Theoriebildungen und -verwendungen der Naturwissenschaftler als auch die Deutungs- und Verstehenskonzepte der Geisteswissenschaftler (einschließlich der Philosophen und Methodologen selbst) wie auch der Alltagskonzeptualisierungen umfaßt. Natürlich können damit differentielle Unterscheidungen und Entgegensetzungen, ja, Unvereinbarkeiten im Sinne

156

Hans Lenk

einer jeweils wissenschafts- oder disziplinenspezifischen Erkenntnis- oder Wissenschaftstheorie nicht geleugnet werden. (Diese zu analysieren wäre die Aufgabe der speziellen Wissenschaftstheorien bzw. Hermeneutiken.) Insofern als generell Arten des Interpretierens beim Erfassen und Verstehen von Weltund Selbst-Strukturen und -Verhältnissen unabdingbar sind, hat die traditionelle Hermeneutik des Verstehens durchaus einen methodologisch richtigen und unverzichtbaren Ansatz gewählt (wenigstens was die Methodologie der geisteswissenschaftlichen Disziplinen im weitesten Sinne angeht, aber auch für die praktische Verwendung von Hypothesen und Theorien sowie Ordnungskonzepten generell in den Naturwissenschaften – etwa der Auffassung hermeutischer Methoden als “technologischer Kunstlehren” (Albert)). Darüber hinaus hat die philosophische Hermeneutik – andeutungsweise schon bei Dilthey beginnend, zumal aber bei Heidegger und Gadamer – die hermeneutische Geprägtheit der Welt- und Selbst-Vorentwürfe, der Konstitution und ihrer methodischen Konzepte und methodologischen Vorbedingungen, durchaus gesehen, indem in dem intentionalen Gerichtetsein von etwas auf etwas und dem Verstehen von etwas als etwas gewisse entwerfende Konstruktivitätsmomente erkannt werden, die als Andeutungen eines methodologischen Konstruktionsinterpretationismus aufgefaßt werden können. Jedoch verblieb das hermeneutische Modell noch allzu sehr im Banne des traditionellen Textinterpretationismus und des Leseparadigmas: Die Welt sollte sozusagen als Text interpretiert werden, ja, Handlungen selbst nur als Texte aufgefaßt werden. Die Universalhermeneutik blieb dem eingeengten Begriff der Textinterpretation verhaftet und konnte nur fallweise oder in Andeutungen zu einem interpretationistischen Konstitutionalismus hinlenken. Insbesondere vermochte es die traditionelle Hermeneutik wie auch die Universalhermeneutik nicht, auch auf neuronaler und biologischer Grundlage die schematisierenden Aktivitäten, die die Vehikel aller symbolisch-interpretatorischen Tätigkeit des Organismus sind, einzubeziehen. Demgegenüber hat die moderne Hirnforschung zweifelsfrei die Fruchtbarkeit der Schemabildungen und Schemaverwendungen sowie der Primärinterpretationen und ihnen nahestehender Deutungen als Schemaaktivierungen und -stabilisierungen zu erhärten vermocht. Unter dem Gesichtspunkt eines auf die Schematisierungen (Schemabildungen, -aktivierungen und bewußte -verwendungen bzw. -differenzierungen) ausgeweiteten Interpretationsbegriffs – also eines Interpretationskonzepts im weiteren Sinne – gelingt es, die erkenntnisund handlungstragenden biologischen Strukturierungen mit jenen der symbolischen und kulturellen unter einem begrifflich-theoretischen Dach zu vereinen und in gleicher Weise auf das wissenschaftliche wie auch auf das alltägliche Erkennen, Handeln und Konstituieren zu beziehen. Unter dem abstrakteren


157

methodologischen oder transzendentalen Gesichtspunkt des Interpretationskonstruktionismus läßt sich also eine übergreifende, verbindende (wenn auch mit einer gewissen Formalität und Abstraktheit erkaufte) Wiedervereinigung der Erkenntnis- und Handlungsdisziplinen unter dem Gesichtspunkt einer interpretationistisch-symbolistischen Anthropologie erreichen. Die Einheit der Erkenntnis ist auf höherer Metastufe, jener bestimmter abstrakt-interpretationistischer Formen, Regeln, Methoden, Erfordernisse und Ergebnisse der Interpretationen, wieder hergestellt. Auch dies ist für die Geistes- und Literaturwissenschaften wie für die Sozialwissenschaften von höchster Bedeutsamkeit!

Literatur Bunge, M.: Scientific Research : Bd. I. Berlin u. a. 1967. Cassirer,E.: Versuch über den Menschen. Frankfurt a. M. 1990 (Orig. 1944). Chladenius, J.N.: Einleitung zur richtigen Auslegung vernünftiger Reden und Schriften. Mit einer Einleitung von L. Geldsetzer. Düsseldorf 1969 (Nachdruck der Ausgabe Leipzig 1742). Dilthey, W.: Gesammelte Schriften. Stuttgart 1914 ff. Gadamer, H.G.: Wahrheit und Methode. Berlin 19865. Heckhausen, H.: ‘Interdisziplinäre Forschung’ zwischen Intra-, Multi- und Chimaren-Disziplinarität, in: Jürgen Kocka (Hg.) : Interdisziplinarität. Frankfurt a. M. 1987, S. 129-151. Kant, I.: Kritik der reinen Vernunft. Auflage B 1787. Hamburg 1956. Karli, P.: “L’experience du Centre de recherches transdisciplinaires sur les sciences et les techniques (C. R. T. S. T.) de l’Université Louis Pasteur de Strasboug. Erfahrungen des Zentrums für transdisziplinäre Forschung über Wissenschaft und Technik”, Universität Louis Pasteur, Straßburg, in: Arber, W. (Hg.): Inter- und Transdisciplinarität: warum – wie? Inter- et Transdisciplinarité pourquoi – comment? Bern 1993, S. 159-173. Kocka, J. (Hg.): Interdisziplinarität. Frankfurt a. M. 1987. Kreuzer, H.: Literarische und naturwissenschaftliche Intelligenz. Stuttgart 1969 (Tübingen, München 19872). Krüger, L.: “Einheit der Welt – Vielheit der Wissenschaft”, in: Kocka, J. (Hg.): Interdisziplinarität. Frankfurt a. M. 1987, S. 106-125. Lenk, H.: Werte als Interpretationskonstrukte, in: Hans Lenk : Zwischen Sozialpsychologie und Sozialphilosophie. Frankfurt a. M. : Suhrkamp 1987, S. 227-237. ders.: “Zu einem methodologischen Interpretationskonstruktivismus”, in: Zeitschrift für allgemeine Wissenschaftstheorie (Journal for general Philosophy of Science) 22 (1991), S. 283-302. ders.: Interpretationskonstrukte : Zur Kritik der interpretatorischen Vernunft. Frankfurt a. M. 1993. ders.: “Wissenschaftskulturentrennung und methodologische Wissenschaftseinheit im Blickwinkel des Interpretationismus”, in: Herbert Mainusch/Richard Toellner (Hg.): Einheit der Wissenschaft: Wider die Trennung von Natur- und Geist, Kunst und Wissenschaft. Opladen 1993, S. 195-225. ders. (unter Mitarbeit von E. Kaleri): Philosophie und Interpretation. Vorlesungen zur Entwicklung konzeptionistischer Interpretationsansätze. Frankfurt a. M. 1993(a). ders.: Schemaspiele : Über Schemainterpretationen und Interpretationskonstrukte. Frankfurt a. M. 1995.

158

Hans Lenk

ders.: Interpretation und Realität. Frankfurt a. M. : Suhrkamp 1995(b). ders.: Interpretation – Intervention – Interaktion. Buch-Manuskript i. Dr. von der Malsburg, Ch.: “Am I Thinking Assemblies?”, in:Palm, G./. Aertson, A. (Hg.): Brain Theory. Heidelberg, New York. 1986, S. 161-176. Roth, G.: “Kognition – die Entstehung von Bedeutung im Gehirn”, in: Krohn, W./Küppers, G. (Hg.): Emergenz: Die Entstehung von Ordnung, Organisation und Bedeutung. Frankfurt a. M. 1992, S. 104-133. D. E. Schleiermacher, F.: Hermeneutik. Herausgegeben und eingeleitet von H. Kimmerle. Heidelberg 19742. ders.: Hermeneutik und Kritik. Hg. Frank, M. Frankfurt a. M. 1977. Singer, W. (Hg. und Einführung): Gehirn und Kognition. Heidelberg 1990. Snow, C. P.: Die zwei Kulturen. Stuttgart 1967. Weingart, P.: “Das Dilemma: Die Organisation von Interdisziplinarität”, in: Wirtschaft und Wissenschaft 1974, 3, S. 22-28.

Causal Decision Theory

159

ELLERY EELLS Causal Decision Theory Philosophers concerned with the nature of rational decision making have distinguished two paradigms for rational deliberation. According to evidential decision theory, one should, when faced with alternative courses of action, perform an act whose performance would constitute the best possible evidence that one will get the good outcomes and not get the bad outcomes. On this paradigm, therefore, deliberation consists in trying to figure out which of the available actions is such that performing it should cause one’s beliefs to be altered in such a way that the good outcomes will be as subjectively probable as possible and the bad outcomes as subjectively improbable as possible, given constraints imposed by one’s prior beliefs. According to causal decision theory, on the other hand, one should be explicitly concerned with the causal relations between the available acts and the possible outcomes. On this paradigm, an act is evaluated according to the degree to which it can be expected to cause (rather than merely provide evidence for) the good outcomes and prevent (rather than merely provide evidence against) the bad outcomes. I will give a precise and formal version of each of these two kinds of theories below. Of course, one’s believing that an item A is a cause of an item B is not the only reason why A might be evidence for B: A could be evidence in favor of B for a person who believed that B causes A, or that there is a common cause of both A and B. In decision problems in which the only reason why an act would be, for the relevant agent, evidence for an outcome is that the agent believes that the act is a cause of the outcome, evidential and causal decision theory will agree about the appropriate course of action. Thus, many philosophers concerned with the nature of rational decision have concentrated on a class of decision problems, called ‘Newcomb problems’, in which acts provide evidence for outcomes in other ways. For some such problems, it seems that these two paradigms yield different prescriptions for action, the causal paradigm giving the correct prescriptions and the evidential paradigm the wrong ones. Some examples of this kind will be discussed below. It has been argued – for example, by Kyburg (1980), Eells (1980, 1982, 1984b), and Jeffrey (1981, 1983) – however, that (roughly), if a decision maker appropriately „monitors” the relevant aspects of his deliberation (e.g., he knows what his beliefs and desires are), then evidential decision theory will

160

Ellery Eells

deliver the correct prescriptions, the same as those given by causal decision theory. These arguments have come to be known as „metatickle” arguments; a metatickle can be thought of, basically, as a piece of information about one’s deliberation. I’ll describe three such approaches below, and note some of their potential shortcomings. Given these shortcomings, and the intricacy of metatickle analysis, it would seem that causal decision theory would have a great advantage over evidential decision theory if, unlike evidential decision theory, metatickle analysis were not necessary for causal decision theory’s delivering the appropriate prescriptions. I’ll argue, however, that there are decision problems in which even causal decision theory will give the wrong answers, if metatickles are not taken account of. This will indicate that many of the shortcomings of metatickle analysis as employed in defense of evidential decision theory are also relevant to the adequacy of causal decision theory. And it suggests that it is just as much in the causal decision theorist’s interest as it is in the interest of the evidential decision theorist to try to develop an adequate account of the impact of metatickles on deliberation.1 1. The Two Theories and the Prima Facie Conflict Evidential and causal decision theories give different ways of calculating the value of an act in terms of the decision marker’s subjective probability assignment (symbolized here by ‘Pr’) and desirability assignment for the possible outcomes of his decision (symbolized here by ‘Des’). Evidential decision theroy can be summarized as follows: Value (A) = SumiPR (Oi/A)Des(Oi), where A is any act and the outcomes Oi include (positively or negatively) everything the agent cares about in his decision problem.2 According to the subjective interpretation of probability, the sign and magnitude of the difference between Pr(Oi/A and Pr(Oi) is supposed to reflect the degree to which 1 The thesis that metatickles actually have a legitimate and important role in deliberation (both evidential and causal) is, incidentally, relevant to recent criticism by Frank Jackson and Robert Pargetter (1983, 1985) of the „tickle” and „metatickle” defenses of evidential decision theory. See Eells (1985) for a defense of evidential (and causal) decision theory against this criticism. 2 With the outcomes so understood, evidential and causal decision theory will agree about the values of the outcomes. Hence the occurrence of ‘Des’ in both formulas for expected utility. If the outcomes are not so specific, we should replace the function Des with the function Value, about whose values the two theories may differ. I think the difference between the evidential and causal paradigms is brought more clearly into focus when the outcomes are understood to be maximally specific, so that the difference between the two paradigms can then be characterized in terms of different ways of calculating the expected utility values of less specific things (such as acts) in terms of the functions Des (whose domain is maximally specific outcomes) and Pr, about whose values the two paradigms can find no room to disagree.


161

A would be evidence in favor of or against Oi. So the value of an act is high or low to the degree to which the act would be, for the agent, evidence for the good outcomes and against the bad one. The evidential value of an act A is also sometimes called its ‘conditional expected utility’, CEU (A).1 Since an act can be evidence for or against an outcome it doesn’t cause or prevent, and since causal decision theory wants to pay attention only to the degrees to which acts cause outcomes, the probabilities of certain items must enter into the calculation of causal expectation unconditional on the act. Here is one version of causal decision theory.2 Let Kj’s be propositions that say which of the Oi’s are outside the influence of which acts, which Oi ‘s are within the influence of which acts, and what the degrees are of the relevant acts’ influence on the relevant Oi’s. Another way of saying what the Kj’s say is this: each Kj specifies, for each available act, one set of causal propensities for the outcomes given that the act is performed.3 Then, according to causal decision theory, Value (A) = Sumi jPr(Kj)Pr(Oi/Kj&Des(Oi&Kj). (Given the above understanding of the Oi’s, Des(Oi&Kj) = Des(Oi).) This is the basic form of causal expectation. Different current versions of causal decision theory differ only with respect to the analysis of causal propensity. The most famous example of a decision problem which is supposed to exhibit a conflict between evidential and causal decision theories is Newcomb’s problem. You are given a choice between taking just the opaque box (act A1) and taking both the opaque and the transparent box (act2), where you can see that the transparent box contains $ 1,000 and you know that the opaque box contains either $ 1,000,000 or nothing, according to whether a fantastically accurate predictor predicted you would take just the one or you would take them both. It is assumed that your beliefs about the predictor are such that the probability of his predicting act Ai conditionally on your doing Ai is high, for each i. If these conditional probabilities are each equal to 0.9, and assuming monetary desirabilities, the evidential expectations of A1 and A2 are $ 900,000 and $ 101,000, respectively, so that evidential decision theory prescribes taking just the opaque box. By the lights of causal decision theory, however, this act is irrational, for taking just the opaque box is only sympto1 See Jeffrey (1965, 1983) for the full development of the theory. 2 This version of the theory is patterned after Skyrms’ statement of it in his (1982). See also Gibbard and Harper (1978), Skyrms (1980, and Lewis (1981). 3 Whether the acts A, the outcomes Oi, and such things as the Kj’s are considered to be „propositions” (as in Jeffrey 1965, 1983) or „states of nature” is irrelevant to the philosophical points at issue here.

162

Ellery Eells

matic of the $ 1,000,000’s being in the opaque box, and in no way causally affects the box’s contents. Also, no matter what’s in the opaque box, you’ll get $ 1,000 more by taking both boxes than you get by taking just the one. And the causal formula for expectation reflects just this. Here, our credence is divided between two Kj’s, corresponding to whether the $ 1,000,000 is stashed or not: Kj’s, which says that the causal propensity for receiving $ 1,000,000 given A1 and the causal propensity for receiving $ 1,001,000 given A2 are both equal to 1, and K2,which says that the causal propensity for receiving nothing given A1 and the causal propensity for receiving $ 1,000 given A2 are both equal to 1. What the exact values of A1 and A2 are, according to causal decision theory, depends on how we divide our credence between K1 and K2, but no matter how we do this, the causal value of A2 is $ 1,000 greater than the causal value of A1. Whatever one’s intuitions are about which act is correct in Newcomb’s problem, there are other, closer to earth, problems for which the prescriptions of causal decision theory are clearly correct, but where the evidential theory seems to give the wrong answers. They have the following form. Suppose that a prior state C tends to cause both act A and a very bad outcome O, where A causes a slight good, O’. The Fisher smoking hypothesis example is of this form, where A is smoking, O is lung cancer, O’ is the relatively small pleasure one gets from smoking and C is a genetic condition. If O and O’ are all the agent cares about in his decision problem, then it is clear that A is the rational act. And the causal expectation of A is greater than that of ~A by the amount of pleasure O’ brings. But if the causal relation between C, A, O, according to one’s beliefs, is such that A would provide strong evidence for C, and thus for O, then it seems that evidential decision theory will prescribe abstaining. Note, incidentally, that Newcomb’s problem is of the general form described above, if we assume that behind every good predictor is a common cause of both the event predicted and the prediction of the event, such as cold fronts causing both the rain and the falling barometer on which the weather forecaster bases his prediction of rain. Such decision problems constitute not only prima facie conflicts between evidential and causal decision theory, but also prima facie counterexamples to the former. 2. Three Defenses of Evidential Decision Theory I shall now describe three „metatickle” defenses of evidential decision theory, to the effect that theory gives the same prescriptions as the causal theory in Newcomb problems. Each defense in turn is supposed to accommodate potential shortcomings of its predecessor, the last thus being the most promising.


163

Before discussing the first real defense, note that, in the Fisher smoking case for example, if the way the bad gene caused one to smoke was by producing a certain kind of tickle in the taste buds, and one knew this and whether or not one had the tickle, then this knowledge would break the evidential connection between smoking and the gene, and evidential decision theory would then give the right answer: smoke. Both conditional on having the tickle (and thus the gene) and conditional on lacking the tickle (and thus the gene) and conditional on lacking the tickle (and thus also the gene), smoking only produces evidence for the small good, namely the relatively small pleasure one gets from smoking: in both cases, the vastly more important question of whether or not one has the bad gene is settled by the known presence or absence of the tickle. However, as Brian Skyrms has pointed out (1980), p. 131), there needn’t always be such a „tickle”, so we can’t count these observations as constituting a general defense of the evidential theory. However, the defenses described below all operate in a formally similar way: some analogue of the tickle – a „metatickle” – is sought that must always be present. The first defense I shall discuss (Eells 1980, 1982) looks to a rational agent’s beliefs and desires, i.e., subjective probabilities and desirabilities. If the gene, say, causes smoking in rational, expectation-maximizing agents, then it must affect either their subjective probabilities, ther subjective desirabilities, or their choice of expectation formula. For it is just these three things which determine the expectation-maximizer’s act. If we assume that the agent is rational and will rationally deliberate, then it seems to be a safe assumption that the gene will not affect the choice of expectation formula, for it is just this that models rational deliberation in light of one’s beliefs and desires. So the gene just affects one’s beliefs and desires, modeled decision theoretically by one’s subjective probabilities and desirabilities. But the agent must know what his subjective probabilities and desirabilities are if he is to apply any of the competing decision theories. Assuming that the agent, knowing some decision theory, fully believes all this, then his knowledge of his probabilities and desirabilities should render the act evidentially irrelevant to the presence or absence of the gene. In general, if C causes A and the only way it does this is by causing agents to have probabilities and desirabilities that rationally lead to A, then, if R is a proposition that specifies a particular set of subjective probabilieties and desirabilities, say our agent’s, we should expect our agent (assumed to know some decision theory) to have subjective probabilities such that Pr (A/R&C) = Pr (A/R&R ~C). The agent knows that the act will be based on the subjective probabilities and desirabilities specified in R, and that, beyond whatever role C or ~C had in producing these beliefs and desires, the prior state is now

164

Ellery Eells

completely inefficacious in affecting which act is chosen. So the agent’s subjective probability of his performing A should be based just on what R indicates to him about the chances of his deciding on A, C now being both causally and evidentially irrelevant to the act. By symmetry of probabilistic independence,1 it follows that Pr (C/R&A) = Pr (C/R& ~A), so that, conditional on one’s knowledge of one’s beliefs and desires, the act is evidentially irrelevant to the important prior state, and thus also to the outcome of which the latter is a cause.2 One may object (Eells 1980, 1982, Horwich 1983), however, that although the agent knows what his beliefs and desires are, he does not necessarily know just what to make of the probability of A or of C in light of them. Consider Pr (A/R&C) and Pr (A/R& ~C) for example. Considered as objective propensities, these two probabilities should be equal; but not necessarily equal when construed as subjective probabilities. For there are crucial features of the beliefs and desires described by R that the agent may not be aware of, but which he may need to be aware of in order for R to „screen off” the evidential relation between A and C. To the agent at an early stage of deliberation who does not yet know which act is rational in light of the beliefs and desires described by R, R&C might constitute better evidence than R& ~C that R describes beliefs and desires of the kind that rationally lead to A. This would be because R&C should be better evidence than R& ~C that R was influenced by the presence (rather than the absence) of C, where C causes beliefs and desires of the kind that rationally lead to act A. Thus, Pr(A/R&C) may very well be greater than Pr (A/R& ~C), and by symmetry, Pr(C/R&A) greater than Pr (A/R&P ~A). Perhaps we may still insist, however, that if the agent is strongly rational and experienced in rational decision making, he will ignore what is know to be no longer causally efficacious in controlling his decision (i.e., the presence or absence of C) and concentrate only on his beliefs and desires (i.e., on R) in gauging the probability that he will perform the symptomatic act, so that we in fact should expect that Pr(A/R&C) = Pr (A/R&P ~C), and evidential irrelevance of the act to the prior state, given R. 1 Henry Kyburg has argued (1974), pp 287-290, and 1980, p. 153) that independence (and thus also, by implication, positive and negative correlation) is not in general symmetric for the relevant kind of subjective probability. But see my (1984a, pp. 83-86) for arguments to the effect that Kyburg’s prima facie counterexamples to symmetry are not genuine. 2 If the causal „fork” from C to A and O is, subjectively, a Reichenbachian „conjunctive fork” (Reichenbach 1956, p 159), then the evidential irrelevance of A to O follows, by the laws of probability, from the evidential irrelevance of A to C. On the other hand, the fork might be „interactive (Salmon, e.g., 1978). For a discussion of this distinction and the case of interactive forks in decision theory, see Eells and Sober (1984), where the metatickle defense is extended and applied to cases of this kind in defense of the evidential theory.


165

In any case, it is desirable for a decision rule to be applicable to less sophisticated agents, and it was perhaps considerations similar to those just given that motivated Richard Jeffrey’s „ratificationst” approach (1981, 1983). On this approach, the metatickle is not just one’s knowledge of one’s beliefs and desires, but rather the opting (at least hypothetically, see below) for one act or the other. Suppose that opting for an act confers an extremely high subjective probability on the act, but short of 1 because of the realistic acknowledgment that one may „slip” at the time of action. This will allow for post-choice calculation of the expected utility of each act. Now consider Newcomb’s problem again. And suppose for now that what the predicator predicts is not, strictly speaking, what you will do, but rather what you will choose (more on this assumption below). Now suppose you have opted for A1. Since this confers a very high probability on A1, and because of the way in which you are confident in the predictor’s accuracy, this confers very high probability on the proposition, K1, that the predicator has predicted your choice of A1. Because the predicator predicts what you opt for and not what you do, the high probability of K1 should be the same conditional on your actually doing as it is conditional on your actually doing A2 – all given your choice of A1. So, when evidential expected utility is recalculated after you have made your choice of A1, act A2 will turn out to have the greater expected irrelevance of the acts themselves (as opposed to the opting for them) to the prediction. Also, if you opt for A2, then A2’s post-choice expected utility will again be greater than that our A1, again because we will have post-choice evidential irrelevance of the acts themselves to the prediction. This is what it means to say that a decision to perform a given act is ratifiable: the act has greatest expected utility relative to the post-choice probability function which reflects the knowledge (or at least supposition, see below) that one has opted for that act. A choice of A1 is not ratifiable since, as we have seen, relative to the probability function one adopts after having opted for A1 (which takes account of one’s having so opted)j, it is A2 that has the highest evidential expected utility. A2, however, is ratifiable. The maxim, then, is: „Make ratifiable decisions.” (Jeffrey 1983, p. 16).1 And the metatickle is the 1 As pointed out by Jeffrey, there are examples of decision problems in which there is no ratifiable action and problems in which there is more than one ratifiable action (1983, pp. 18-19). Jeffrey councils that „in such cases you do well to reassess your beliefs and desires before choosing.” (P. 19) For cases in which there is exactly one ratifiable act, it’s not clear to me whether the intent of Jeffrey’s maxim was that an act’s ratifiability is both necessary and sufficient for its being rational to opt for it. However, an example of Brian Skyrms’ (1984, pp. 85-86) strongly suggests that an act’s (unique) ratifiability cannot always be sufficient for its being rational to opt for it. Suppose that there are three shells before you. You can have what’s under the one you choose. If the predictor predicts you will take what’s under the first shell he puts 10 cents under it and nothing

166

Ellery Eells

knowledge (or the supposition) that one has made such and such a choice. Note that this defense of the evidential theory is not vulnerable tho the objection to which the previous defense, above, seemed open to. For, given that the agent has made a choice, the agent should no longer entertain an evidential connection between (in the notation used in the previous defense) R&C and the proposition that A is the rational act: presumably, when one has made a choice, one has made the choice which one has already deemed to be the rational choice in light of one’s beliefs and desires. There are two at least potential difficulties with this approach. The first was pointed out by Bas von Fraassen and reported by Jeffrey (1983, p. 20), in the framework of the prisoners’ dilemma. The analogue of this problem for Newcomb’s problem is this: suppose that , contrary to the assumption above, the predictor predicts, properly speaking, the act and not the choice, where the accuracy of the predicator holds up whether or not we succeed in performing the chosen act. In that case, we do not get evidential irrelevance of the act to the prediction of it, conditional merely on the choice of one of the two acts – for now the predictor is able to predict which way a „slip” between an intention and the act will go. But it seems to me that this problem is not as serious as the might seem, for the situation which the agent must envision for this kind of problem to have an effect on the ratifiability of A2 is one which is, to say the least , far from ideal for the application of standards of rational decision anyway, as explained blow. For i = 1, 2, let Ci be the opting for Ai. The following is a simple formulation of the assumption that the predictor’s accuracy holds up conditional on each choice (with the conditional probabilities going in the direction relevant to the calculation of evidential expectation, as explained above): Pr(K1/A1&C1) = Pr(K1/A1&C2), >> Pr(K1/A2&C1) = Pr(K1/A2&C2). Concentrating just on the leftmost probability on each side of the inequality sign, we get, by symmetry of (positive, say) probabilistic correlation, that: Pr(A1/K1&C1) > Pr(A1/K2&C1).

under the others. If he predicts you will take what’s under the second shell, he puts $10 under it, $100 under the third, and nothing under the first. If he predicts you will take what’s under the third shell, he puts $20 under it, $200 under the second shell, and nothing under the first. Assuming that it’s extremely subjectively probable, conditional on your taking any given act, that the predictor has predicted that act, then taking what’s under the first shell is the unique ratifiable act. However, it seems clear (as both straight evidential decision theory and causal decision theory agree) that this act is the worst.


167

And concentrating on the rightmost probabilities on the two sides of the inequality sign, we get: Pr(A1/K1&C2) > Pr(A1/K2&C2). For simplicity, let’s suppose, for i = 1,2, that there is a perfect correlation between Ki and the prior state of agents (or of whatever) which causes both the act Ai and the predictor’s prediction of that act. Thus, we may think of Ki as the prior state itself in the two inequalities just above. We see then that, according to the agent’s beliefs, no matter what he chooses to do there is still something about each prior state Ki that is causally positive for Ai: in case the agent (rationally) chooses the act that isn’t correlated with the prior Ki that actually obtains, that Ki causes (probabilistically) the agent to slip in favor of it.1 But it seems that to the extent to which the agent believes that his act is thus out of his control, he shouldn’t think it worthwhile to rationally deliberate anyway. For to the extent to which the agent believes this, he should believe that the probability is high that his act will be in agreement with the Ki that obtains no matter what the result of his rational deliberation (i.e., his rational choice) is. So, to the extent to which the agent should consider his problem to be appropriate for the application of standards of rational deliberation in the first place, it seems that he should believe that „slip” between the intention to perform a given act and the act itself should be „random” in direction, i.e., the direction of the slip (towards A1 or towards A2) is uncorrelated with factors that play other roles in his deliberation.2 If this first problem for ratificationism doesn’t seem so threatening under the assumption of an ideal setting for the application of standards of rational decision, the second problem isn’t so easily disposed of – if ratificationism is taken to be the doctrine that ratifiability should be used as the criterion of choice. Jeffrey explains that „we are to judge the ratifiability of each act by estimating the desirabilities of all acts form the point of view the agent would have upon choosing the one act whose ratifiability is in question.” (1983, p. 19). The problem is: How is the agent supposed to obtain that point of view? Supposing that an act A is the single ratifiable act, there is no guarantee that the agent will ever, at any point in his deliberation, actually choose it so that he will then be in a position to assess its ratifiability. But perhaps the agent can simply imagine that he chooses it and then assess its ratifiability from this hypothetical perspective. But what exactly is the agent supposed to imagine 1 This, incidentally, is in accord with the „contextual unanimity” model of probabilistic causality, as discussed by, for example, Cartwright (1979), Skyrms (1980), and Eells and Sober (1983). 2 For more details on this and another kind of „slippage”, and rational agents’ attitude about the possibility of so slipping, see Eells and Sober (1984).

168

Ellery Eells

here? Is he to imagine that he has chosen the act on the basis of its having maximal expected utility, or should he perhaps imagine that he has chosen it on the basis of its ratifiability? Suppose he’s to imagine the first of these two things, and that, on this supposition, the act has maximal expected utility, so that his (hypothetical) decision to perform the act is ratifiable. This alone would hardly seem to justify the act, for the act is ratifiable here only in the sense that it has maximal expected utility when calculated relative to the supposition that it has maximal expected utility: i.e., were it to attain maximal expected utility is recalculated relative to the knowledge that it has attained maximality. This would only seem to justify the act if, at some point in the agent’s deliberation, the act actually attained maximal expected utility, for it could only then actually retain its maximality, having attained it, and perhaps thereby be justified. But there is no guarantee that the act will ever initially attain maximality.1 But perhaps the agent is supposed to imagine that he chooses the act on the basis of its ratifiability. This is a natural suggestion if we suppose that the agent is, in the end, going to use ratifiability as his criterion of decision, and that he knows this. So now the question is whether the act is maximal on the supposition that it is chosen on the basis of its ratifiability. The question presents a vicious circle. For it is equivalent to: Is the act maximal on the supposition that it is chosen on the basis of its maximality on the supposition that it is chosen on the basis of its ratifiability, i.e., on the basis of its maximality on the supposition that it is chosen on the basis of its ratifiability, (and so on)? Indeed, on this reading, the definition of ratifiability is circular. It thus appears that, on any adequate notion of ratifiability, one cannot really assess the ratifiability of a choice until and unless the choice has been made, at least tentatively. But, again, there is nothing in ratifiability analysis that guarantees that one will actually finally make a ratifiable choice. Indeed, for all that has been said, it might require causal decision theory to deliver a choice which, after it has been made, can (only) then be seen to be ratifiable. What would seem to be necessary and sufficient for the rationality of an act is: ratifiability together with its prior, independently principled choice. What evidential decision theory needs is a plausible analysis of the prechoice dynamics of evidential deliberation; and if evidential decision theory is to prove adequate, such an analysis must show that evidential deliberation will eventually land the decision maker on an act that will turn out to be ratifiable. I shall next turn to some suggestions along these lines (see also Eells 1984b). In the examples we’ve looked at so far, the initial prescriptions of causal decision 1 Skyrms’ shell game (see Note 9 above) provides an example of a ratifiable, non-maximal act.


169

theory have all been ratifiable. The question arises of whether this is so in all cases, or whether the initial prescriptions of causal decision theory may in some cases be unratifiable, thus indicating that causal decision theory must also develop an analysis of the pre-choice dynamics of deliberation which will show that pre-choice causal deliberation will eventually land the decision maker on a ratifiable decision. In the next section, I’ll argue, by way of an example, that causal decision theory also stands in need of such an analysis. Suppose our agent always wants to maximize his conditional expected utility and he has to choose between two acts, B1 and B2, in some decision problem. Let B0 be the act of choosing between the two in a random way – plausibly the easiest way to decide. And let B1 be the act which, at this point in the agent’s deliberation, maximizes his CEU, where the agent doesn’t necessarily know at this point which act this is1 The agent does know, however, that CEU(B1) CEU(B0). Since the inequality might be strict, and since the agent wants to maximize his CEU, he should figure that it’s in his interest to calculate the CEU’s of the acts. Now the agent knows which act B1 is. Suppose the agent now comes into possession of some new information, e1. This can change the evidential values of the available courses of action: the new CEU of each act is equal to the old CEU of the conjunction of the act with e1, and, where B2 denotes the act, the agent knows not which, that maximizes the agent’s new CEU’s, the following inequality holds: CEU(B2&e1). And again, since this inequality might, for all the agent knows, be strict, and since the agent by hypothesis wants to maximize his CEU, he should want a recalculation. In general, if, for all i less than or equal to some number n, the agent learns ei after the i-th calculation of CEU, but before performing one of B1 and B2, then the following inequality should hold (for any such n): CEU(B n+1 &e1&en) CEU(Bn&e1&...&en). Since, for all n, the agent should know that the inequality holds, and since he desires to perform the CEU-maximal act, he should always desire a recalculation of CEU after coming into possession of new information en. Assuming that

1 „Classical” Bayesian epistemology requires a rational agent to assign probability 1 to any propositions known by him. This contradicts my assumption that our agent doesn’t know which act is CEU-maximal, since I assume that he does know what his subjective probabilities and desirabilities are. However, I believe that a „slightly more realistic” Bayesian epistemology is possible (see, e.g., Hacking (1967) and Garber (1983) for two tries), one that does not require (among other idealistic requirements of the classical theory) that the agent assign probability 1 to (or even entertain) each logical consequence of known propositions.

170

Ellery Eells

the cost of calculation is low enough (or that the differences in utility between the possible outcomes of his decision is great enough), he should recalculate.1 Note that the result of the very first calculation of CEU is itself new information. If it weren’t new, the agent wouldn’t have had to calculate CEU in the first place. Let us suppose that, in the course of the agent’s deliberation, he obtains no new information „from the outside” – the only new information he obtains in the course of his deliberation is the results of his various calculations of expected utility. Thus, let, for each i, ei be the information obtained from the i-th calculation of the CEU’s of the acts. It is thus plainly at least theoretically possible that even if the agent does not receive any new evidence „from the outside” in the course of his deliberation, it is nevertheless possible that the CEU-ranking of the acts should shift in the course of deliberation, for when new evidence is absorbed (be it from the outside or not), the CEU’s of the available acts can change. Given the possibility of such shifts in the CEU ranking of the acts, we can make sense of hypothetically supposing that such and such an act is CEUsuperior to the other, and vice versa: perhaps if the agent had deliberated longer, making more calculations, or had earlier cut off deliberation, the currently (actually) CEU-minimal act (the currently CEU-proscribed act) would have been CEU-superior (currently CEU-prescribed). Given the intelligibility of this idea, we can state a general ratificationism-like principle, actually a coherence principle for decision rules D, as follows: Principle of Ratification: IF (i) B1 and B2 are two of the acts among which the agent must choose in his decision problem, (ii) at any time in the course of deliberation, if decision rule D prescribed B1 (given of course, the agent’s beliefs and desires) and if B1 were chosen at that time on this basis, then post-choice, B2 would appear better than B1 relative to decision rule D (and the agent’s beliefs and desires, of course), (iii) at any time in the course of deliberation, if D prescribed B2 and if B2 were chosen on this basis, then post-choice, B2 would continue to appear better than B1 relative to rule D, THEN rule D must ultimately rank B2 above B1 for the agent; otherwise the rule is (at least for this reason) inadequate for the agent in his decision problem.2 1 I’m not assuming, or arguing, that the decision to recalculate can be justified on evidential grounds. Indeed, it seems t me that such decisions cannot always be so justified. 2 Note that Skyrms’ shell game example (see Note 9 above) does not tell against this principle. The principle pertains to pairs of acts, and the principle applies to none of the three possible pairs of acts taken from Skyrms’ problem. It’s possible, of course, that in a decision problem there will be several pairs of acts to which the principle is applicable. But note that the principle doesn’t say that either of the acts in any such pail should be chosen – only that one should be, in the final analysis,


171

I shall not argue for this principle, but merely assume it, since it seems clearly correct. I shall now argue that, given plausible and realistic assumptions about the effects of an agent’s multiple calculations of expected utility (another kind of metatickle), evidential decision theory will prescribe the two-box act in Newcamb’s problem – my working prima facie counterexample to evidential decision theory – and that it thus satisfies the above ratification principle in such situation. (In the next section, I’ll argue, by way of a different example, that causal decision theory also requires accommodation of metatickles if it is to satisfy principle.) As noted above, I agree with Jeffrey that, post-choice (no matter what the choice is), A2 should appear evidentially superior to A1 in Newcomb’s problem. Thus, (i), (iii) of the ratification principle above are satisfied. This means that, if evidential decision theory is adequate for our agent in the sense of the principle, then it must, eventually, after sufficiently many calculations of CEU, prescribe the two-box act and continue to do so. Also, if it can be shown that it does, then, I take it, it will have been shown that evidential decision theory does deliver the rational prescription in Newcomb’s problem – albeit possibly after a false start. The reason why A2 looks beter than A1 post-choice, no matter what the choice, is that the cohice, whatever it is, renders the acts evidentially irrelevant to the Ki’s. As deliberation proceeds, we can expect that the agent’s subjective probability of one of the acts will eventually get very close to 1. A plausible pre-choice version of the idea of post-choice evidential irrelevance of the Ai’s to the Ki’s is the following, where the subscript ‘t’ on ‘Prt’ indicates the time at which the value is taken: Assumption 1: As Prt(A1) approaches 1 (if it does), Prt(Ki/Ai) approaches Prt(Ki/A2), for i = 1, 2; and the same goes for the approach to 1 of Prt(A2). For some agents, the two conditional probabilities may approach the initial probability of the state on the act whose subjective probability is approaching 1.1 Assumption 1 is about the effects of changes in the probabilities of the acts. The next assumption says what causes the probabilities of the acts to change: ranked above the other. A simpler principle would apply only to decision problems in which there are only two available acts. But in the next section, I’ll want to apply the principle to a three-act decision problem. 1 Brian Skyrms’ model of „deliberation-probability dynamics” (1982) makes an assumption similar to assumption 1, which incorpotates this latter idea – both assumptions based on a point of Riffard Jeffrey’s (1981, p. 486).

172

Ellery Eells

Assumption 2: If CEUt(A1) > CEUt(A2), then Prt+1(A1) > Prt(A1); and this also holds when the two inequalities are reversed. This assumption merely states that the probability of the act that is CEUsuperior on the most recent calculation increases as a result of the calculation.1 The discovered current CEU-superiority of an act is evidence that it is the rational act (to CEU-maximizers), and is thus evidence that it will be chosen, so that this evidence should make the act subjectively more probable than before. Both assumptions are qualitative in the sense that they do not describe exactly what probability values are determined by the antecedent probability or CEU-values. This is desirable, I think, because it makes the analysis more general than otherwise. However, in order to carry the defense through, it will be necessary to be more quantitative about the idea in assumption 2. In the model of „deliberation-probability dynamics” investigated by Skyrms (1982), it is assumed that the rate of change of the probabilities of A1 and A2 is proportional to the difference between their expected utilities. This has the consequence that if the probability of A1 is low enough at the outset of deliberation, then evidential decision theory will eventually prescribe A2, and continue to do so until the close of deliberation; but for all other initial probabilities of A1, short of 1, evidential deliberation will leave the probability of A1 stuck somewhat sort 1 – deliberation thus not closing with a decision. (For details, see the reference cited, or Eells 1984b.) This need not be taken to be the fault of evidential reasoning itself, however. Instead, we may lay the blame on the way in which the probabilities of the acts are taken to depend on their CEU’s. Certainly if „mixed” or „ramdomized” acts (in which the agent is allowed to choose chances for the acts) are not allowed, then we decision theorists would not prescribe such deliberation-probability dynamics to a friendly CEU-maximizer. And a sophisticated CEU-maximizer who understood the implications of these dynamics would not embrace them in a Newcomb problem. It seems to me that the model investigated by Skyrms makes the probabilities of the acts sensitive to one thing to which they should not be sensitive, and insensitive to another thing to which they should be sensitive. As to the first point, it seems that the movement of the probabilities of the acts should not depend on the magnitude of the difference between their CEU’s, but rather only on the sign of the difference. For an agent should not perceive an act to be more or less rational according to how much greater its expected utility is perceived to be than its competitors’ utilities; rather, the agent should perceive 1 This assumption is also a part of Skyrms’ (1982) model.


173

an act to be more or less rational according to how clear his perception ist that, in the final analysis, its CEU will be (simply) greater than its competitors’ – where the perceived difference between the CEU’s should be irrelevant. And the agent’s subjective probability of performing an act should be based on his assessment of whether or not it will be evidentially superior in the final analysis, so this also should be based on the sign, rather than magnitude, of the differences between the relevant CEU’s. As to the second point, it seems that a rational agent should be sensitive to the fact that deliberation must eventually end and an act then chosen. That is, the agent should be sensitive to the approach of the time of action, when deliberation must end and action begin. This puts pressure on the agent to make a choice, and it seems that a realistic and plausible way to model an agent’s sensitivity to this factor is to constrain his subjective propbabilities of the available acts to more and more extreme values as the moment of decision approaches. Another rationale for such an assumption is that later calculations of expected utility reflect more information (namely the results of more numerous prior calculations) than do the earlier ones, so they should have a more dramatic effect on the agent’s assessments of the relative merits of the available acts, and hence also on his assessment of the probability that he will take them. There is a whole family of such constraints, some descriptive of decision makers who procrastinate taking a stand on which act is correct and others descriptive of more impulsive decision makers. In the following assumption, the constant v indexes this aspect of the decision mar’s personality, st is equal to 1 or – 1 according to whether A1’s CEU is greater than or less than that of A2’s at the time t-1,1 and T is the amount of time available for deliberation: Assumption 3: Prt(A1) = 1/2 + (1/2)st(t/T)v. The formula assumes that at the beginning of deliberation the agent is completely uncertain about which act he will eventually perform, so that Pr0(A1) = 1/2.2 The basic idea is that max {Prt(A1) = Pr0(A2)} gradually grows as deliberation proceeds. Because of the agent’s awareness of the fact that he must have a decision when the time of action arrives, the quantity approaches 1 as the time of action approaches. The value of v controls how fast the quantity grows at 1 A more general formula, for cases in which the subjective probability of A1 should continue to move in the same direction and at the same rate as it would if a previous inequality had been sustained. Or perhaps the subjective probability of A1 should move to 1/2, and then, inequality later obtains, the probability of A1 should then obey the assumption. 2 A more general formula, for cases in which the subjective probability of A1 doesn’t begin at 1/2, would be: Prt(A1) = r + qt(t/T)v, where qt = 1-r or – r according to whether CEUt-1(A1) is greater or less than CEUt(A2), and r is the initial probability of A1.

174

Ellery Eells

the beginning of deliberation compared to at the end of deliberation. Assumption 1 and 3 have the consequence that if the agent make sufficiently many calculations of expected utility before the time of action, evidential decision theory will, in the end, prescribe the two-box act and continue to do so for all further calculations of expected utility. It’s worth noting that the exact form which assumption 3 takes is irrelevant to the force of this metatickle defense: as will become apparent below, all that is necessary is that the greater of the probabilities of the two acts gradually grows and approaches 1 as the time of action closely approaches. Initially, A1 will be evidentially superior, but, as the probability of A1 increases, both acts become less and less evidentially relevant to the Ki’s. And when the acts are sufficiently evidentially irrelevant to the Ki’s, A2 will be evidentially superior and as a result take on a hig probability. If the acts become evidentially irrelevant to the Ki’s as fast when the probability of A2 gets high, then, after the initial leap in the probability of A2, this probability will not turn around: it will continue to get higher so that the acts become more and more evidentially irrelevant to the Ki’s and evidential decision theory will continue to prescribe A2. See Figure 1. If, for some reason, the acts don’t get evidentially irrelevant to the Ki’s as fast when the probability of A2 gets high as they do when the probability of A1 gets high, then the probability of A1 may bounce up and down several times – but in the end, when the probability of A2 is high enough that the acts are sufficiently evidentially irrelevant to the Ki’s for A2 to be evidentially superior when A2 has high probability, A2 will remain evidentially superior as the probability of A2 continues to grow into the time of action. See Figur 2.

I admit, of course, that there will be some agents for whom the assumptions of this defense of the evidential theory will not hold true, in particular, the proviso that the agent recalculates expected utility sufficiently many times. But it is arguable that this is not the fault of the theory: indeed, perhaps all we can conclude is that evidential decision theory is simply not adequate for such


175

agents, be that the fault of the agent or the theory. In the next section, we’ll see that causal decision theory stands in the same relation to similar agents. 3. Metatickles and Causal Decision Theory We have just seen that the results of calculating the evidential expectations of the available acts is itself evidence (a metatickle) that can sometimes require reassessments of the evidential relevance of the acts to the outcomes. Indeed, it seems that for some agents, evidential decision theory would prescribe the wrong course of action were it not for this phenomenon. Turning to causal decision theory, are there situations in which the results of calculating the causal expectations of the available acts will be evidence (a metatickle) that will require reassessments of the causal relevance of the acts to the outcomes? More interestingly, are there situations in which causal decision theory will deliver an inappropriate prescription if such reassessments are not carried out? In what follows, I will describe a decision problem for which the results of causal calculations do require reassessments of the causal significance of the acts for the outcomes? More interestingly, are there situations in which causal decision theory will deliver an inappropriate prescription if such reassessments are not carried out? In what follows, I will describe a decision problem for which the results of causal calculations do require reassessments of the causal significance of the acts for the outcomes. Decision problems for which just this is true have already been described – e.g., Gibbard and Harper’s „Death in Damascus” problem (1978, pp 157-159), Skyrms’ „mean” and „nice demon” problems (1982, pp. 705-706), and Jeffrey’s „green-eyed monster” and „triumph of the will” problems (1983, pp. 18-19). But for these problems, it is clear, because of a kind of symmetry in the problems, that there is no unique appropriate action. The decision problem which I shall describe, however, is one for which causal decision theory can give the wrong answer initially, and one which brings the ratification principle of the previous section to bear on the causal theory, thus suggesting the need for causal expectation maximizers to listen to their metatickles. The decision problem for which this is true is built up from two others. The first – call it ‘decision problem I’ – is just Newcomb’s problem, described above. The second problem, which I’ll call ‘decision problem II’, is one in which the agent has control over the contents of the opaque box. In problem II, the „predicator wits until after your choice before filling, or not filling, the opaque box with the $ 1,000,000. In problem II, if you take just the opaque box (act A1), then the predictor will put the $1,000,000 in, but if you take both boxes (act A2), he leaves it out. Thus, the one-box act leads, you know, to your

176

Ellery Eells

getting $ 1,000,000, and the two-box act leads, you know, to your getting just $1,000. Finally, decision problem III is combined from problems I and II as follows.1 You have three available acts in decision problem III, A1 and A2 as above, and A0: take no boxes, the invariable result of which is receiving $999. Call the predictors of problems I and II ‘predictor I’ and ‘predictor II’, respectively. In the third problem, there is a third predictor, predictor III, who predicts whether or not you will do A0. If he predicts A0, then he turns you over to predictor II (essentially putting you in decision problem II, except that you have that third alternative), and if he predicts that you won’t do A0, he turns you over to predictor I (essentially putting you into decision problem I, Newcomb’s problem, except that again you have that third alternative). Your subjective probabilities are such that it is extremely probable, conditional on your choice, that predictor III has predicted that choice, and it is extremely probable that predictor I has predicted your choice if it is one of A1 and A2;2 „predictor” II, you believe, is infallible. Letting F0 and ~F0 be the prior states of predictor III’s predicting A0 and ~A0, respectively, and K1 and K2 the prior states of predictor 1’s predicting A1 and A2, respectively, the causal expectations of the three available acts are as follows, assuming monetary utilities: Uk(A0) = 999; Uk(A1) = Pr(F0)1,000,000 + Pr( ~F0&K1)1,000,000; Uk(A2) = Pr(F0)1,000 + Pr( ~F0&K1)1,001,000 + Pr( ~F0&K2)1,000. Let r = Pr(F0) and p = Pr(K1). And assume that F0 and K1 are independent of each other. (Say predictor I makes a prediction no matter what predictor III does, based on what the thinks you would do were you to choose one of A1 and A2. Or perhaps predictor I believes you are in decision problem I rather than in decision problem III.3 Then 1 For a simpler example of this, see Eells (1985 b). 2 Actually, the agent’s believing predictor I to be accurate in the above sense, or even at all, is not essential for my purposes. Indeed, this predictor, for the purposes for which I shall use decision problem III, could decide in some random way which act to predict, as can be seen from the calculations that follow. Decision problem III is complex enough as it is, however, and it is easier to describe it (and perhaps also to grasp it) in terms of decision problem II and the already familiar Newcomb’s problem, in which it is assumed that the predictor is accurate in the above sense. 3 As pointed out in Note 20, the causal ranking of the acts, given by the calculations below, would be unaffected if predictor I decided what to predict in some random way, e.g., by flipping a coin, in which case the plausibility of K1’s subjective probabilistic independence from F0 would be even more plausible.


177

Uk(A0) = 999, Uk(A1) = 1,000,000r + 1,000,000(1-r)p = 1,000,000r + 1,000,000p – 1,000,000rp; Uk(A2) = 1,000r + 1,001,000(1-r)p + 1,000(1-r) (1-p) = 1,000,000p – 1,000,000rp + 1,000, so that Uk(A1) > Uk(A2) if and only if r > 1/1000. And note that no matter what the values or r and p are, Uk(A2) is at least 1,000, so it is always greater than Uk(A0). Thus, it is quite plausible that, on the initial calculation of causal expectation, A1 will be the prescribed action. Suppose A1 is chosen on this basis. Because of the way in which we have assumed predictor III to be accurate according to the agent’s beliefs, this should cause the agent’s subjective probability of F0 to decrease. Suppose it gets lower than 1/1000. This means that, postchoice, A2 has a causal expectation greater than that of A1. Thus, clause (ii) of the antecedent of the ratification principle of the previous section is satisfied. Now suppose that A2 is chosen on the basis of causal decision theory’s prescribing it. This implies that at that point r > 1/1000. And it seems that the effect of the choice of A2 on the probability of F0 can only be to lower that probability, in which case, post-choice, causal decision theory will continue to prescribe A2 – over both A1 and A0, the latter of which, as we have seen, has a lower causal expectation than A2 no matter what the values of r and p are. Thus, clause (iii) of the antecedent of the ratification principle is also satisfied, as is of course, clause (i). Thus, according to the principle of ratification of the previous section, causal decision theory must, in the end, rank A2 over A1 – or else be inadequate for the agent in this decision problem. Indeed, A2 is the only one of the three available actions that one should like to carry out once one has decided to do so, where if one were to have decided on any of the other actions, one should then prefer to carry out A2 instead. (Actually, a choice of A0 should make A1 optimal, though A2 would still be better than A0. Act A0 is evidence that one is in problem II, in which A1 is best. But, as we have seen, if the agent is then lowed to change his choice to A1, he would then regret is choice, wishing he had chosen A2 instead.) On this basis, it seems that A2 is the unique rational act in this decision problem. How can causal decision theory avoid prescribing A1? We could build in a ratification clause, which would allow the decision maker to change his choice if it is not ratifiable. But adding such a ratification clause would, it seems, result in the agent’s viewing his initial choice as only tentative. And if the

178

Ellery Eells

agent views his choice, of say A1 in decision problem III, as only tentative, then he may not view the act as sufficiently probable to make F0 sufficiently improbable to render A2 causally superior, whereas a choice viewed as final would. (In the case of evidential decision theory, a tentative initial choice in a Newcomb problem may not immediately render the acts sufficiently evidentially irrelevant to the outcomes to render the dominant action evidentially superior, whereas a choice viewed as final should.) Also, as in the case of evidential decision theory, there is nothing in ratifiablitiy analysis that instructs us how always to arrive at a ratifiable choice, one that we should view as final. And, in addition, there is that viscious circularity problem, described in the previous section, for ratifiability as itself a criterion of choice. On the other hand, we could build into causal decision theory a sensitivity on the part of the agent to metatickles of the kind discussed last in the previous section. Then, after the first calculation of expected utility, on which act A1 has superior causal expectation, the probability of A1 goes up – perhaps only a little though, the agent being wary (the value of v of the previous section’s deliberation-probability dynamics is greater than 1). The second calculation still gives A1 the highest expectation, say, and the probability of A1 goes up a bit more – as does that of ~F0. Eventually, though, the probability of A1 is sufficiently high for the probability of F0 to be sufficiently low (so that it becomes sufficiently probable that one is essentially in the original Newcomb problem) for the causal expectation of A2 to become superior. At this point, the agent’s subjective probability of A1 may have become quite high, though, so that the probability of A2 is still low. Here, just as in the case of the third metatickle defense of evidential decision theory, it might be necessary for the agent’s sensitivity to the approach of the time of decision to result in a leap in the probability of A1, from very close to 1 to very close to 0. For, in the first place, decreases in the probability of A1 in small increments will not do if this happens to the benefit of the probability A0 as well as that of A2: when the probability of A0 is not too low, the probability of F0 will be high enough ( 1/1000) so that A1 again becomes the causally superior act. And in the second place, the agent might be running out of time for deliberation, so that, without such a leap, the probability of the rational act, A2, might not get close enough to 1 for its subjective probability to reflect a choice of that act.1

1 One might be inclined to suggest that a jump in the probability of A2


179

4. Conclusion If evidential decision theory is to deliver the appropriate prescriptions in Newcomb problems, then the agent must be sensitive to metatickles of some kind. But the intricacy of metatickle analysis, and the fact that it is at a relatively primitive stage of development (with respect to both formal rigor and intuitive plausibility), does not constitute an advantage of causal decision theory over the evidential theory. For we have seen that there is a class of decision problems in which causal decision theory can also give wrong answers, unless metatickles of some kind are taken into account. And it seems that the effects of metatickles must be very much the same in the application of the causal theory as they must be in the application of the evidential theory if either theory is to give the right answers – albeit in different kinds of decision problems, and after false starts. It thus appears that a final assessment of the relative merits of the causal and evidential paradigms of rational deliberation must await further refinements in the analysis of the role of „metatickles” in rational deliberation – refinements both on the level of intuitive plausibility and on the level of formal rigor (paying attention to both prescriptive and descriptive adequacy). Metatickle analysis is essential to the „rescue” of both evidential and causal decision theory, and it appears that the fate of each will turn on the developments and refinements of metatickle theory to come. I thank the University of Wisconsin-Madison Graduate School for financial support, and Elliott Sober for useful comments on an earlier draft. Rep. From PSA 1994, Vol. 2, pp. 177-200, With permission of the Philosophy of Science Association.

References Campbell, R. And Sowden, L. (eds.) (1985): Paradoxes of Rationality and Cooperation: Prisoner’s Dilemma and Newcomb’s Problem, University of Britisch Columbia Press, Yancouver. Cartwright, N. (1979): „Causal Laws and Effective Strategies,” Nous 13, 419-437. Eells, E. (1980): Newcomb’s Paradox and the Principle of Maximizing Conditional Expected Utility, Unpublished Ph.D. Dissertation, University of California, Berkeley, University Microfilms Publication Number ADG80-29383. —— (1982): Rational Decision and Causaltiy, Cambridge University press, Cambridge, Entland, and New York. —— and Sober, E. (1983): „Probabilistic Causality and the Question of Transitivity,” Philosophy of Science 50, 35-57. —— (1996): „Common Causes and Decision Theory,” Philosophy of Science 53, 223-245. —— (1984a): „Newcomb’s Many Solutions,” Theory and Decision 16, 59-105. —— (1984b): „Metatickles and the Dynamics of Deliberation,” Theory and Decision 17, 71-95. —— (1985a): „Reply to Jackson and Pargetter,” in Cambell and Sowden (1985) 219-223.

180

Ellery Eells

—— (1985b): „Weirich on Decision Instability,” Australasian Journal of Philosophy 63, 473-478. Garber, D. (1983): „Old Evidence and Logical Omniscience in Bayesian Confirmation Theory,” in J. Earman (eds.): Testing Scientific Theories, University of Minnesota Press, Minneapolis, 99131. Gibbard, A. And Harper, W. (1978): „Counterfactuals and Two Kinds of Expected Utility,” in Foundations and Applications of Decision Theory, Vol. 1 (C.A. Hooker, J.JU. Leach, and E.F. McClennen (eds.): The University of Western Ontario Series in Philosophy of Science, Vol. 13. Reidel, Dordrecht, 125-162. As reprinted in Ifs. W.Harper, R. Stalnaker, and G. Pearce (eds.)(1981): The University of Western Ontario Series in Philosophy of Science, Vol. 15, Reidel, Dordrecht, 153-190. Also rep. in Campbell and Sowden (1985), 133-158.) Hacking, I. (1967): „Slightly More Realistic Personal Probability,” Philosophy of Science 34, 311325. Horwich, P. (1983): „Decision Theory in the Light of Newcomb’s Problem,” unpublished manuscript, MIT. Jackson, F. And Pargetter, R. (1983): „Where the Tickle Defense Goes Wrong,” Australasian Journal of Philosophy 61, 295-299.

Authors

181

AUTHORS Prof. J. Anthony Blair, Department of History, Philosophy & Political Science, University of Windsor, Windsor, Ontario, Canada N9B3P4. Prof. Ellery Eells, University of Wisconsin - Madison, Department of Philosophy, Madison, WI 53706, USA. Prof. Leo Groarke, Wilfrid Laurier University, Department of Philosophy, Ontario, Canada. Prof. Henry W. Johnstone, Jr. Department of Philosophy, The Pennsylvania State University, 2450 Sparks Building, University Park, PA 16802, USA. Prof. Ralph H. Johnson, Philosophy Program, Windsor, Ontario, Canada N9B3P4. Prof. Dr. Hans Lenk, Universität Karlsruhe, Fak. für Geistes- und Sozialwissenschaften, Kaiserstraße 12, D-76128 Karlsruhe, Germany. Prof. Manfred Kienpointer, Karl-Franzens-Universität, 8910 Graz, Austria. Prof. Dr. Dr. Dieter Mans, J.W. Goethe-Universität, FB: Gesellschaftswissenschaften, 60054 Frankfurt am Main, Germany. Prof. Robert C. Pinto, Philosophy, University of Windsor, Windsor, Ontario, Canada N9B 3P4. Dr. habil. Gerhard Preyer, Privatdozent, J.W. Goethe-Universität, FB: Gesellschaftswissenschaften, 60054 Frankfurt am Main, Germany. Prof. Chris Tindale, Department of Philosophy, Catharine Parr, Traill College, Trent University 310 London Street, Peterborough, Ontario, Canada/K9H 7P4. Prof. Douglas Walton, Department of Philosophy, University of Winnipeg, Winnipeg R3B 2E9, Canada. Prof. John Woods, Department of Philosophy, University of Lethbridge, 4401 University Drive, Lethbridge, Alberta Canada T1K 3M4.

182

Imprint

IMPRINT PROTO SOCIOLOGY An International Journal of Interdisciplinary Research ISSN 1434-4319 Editor: Gerhard Preyer Johann Wolfgang Goethe-University Frankfurt am Main, Department of Social Sciences Editorial Staff: Georg Peter Editorial of the Vol. 13 1999: Georg Peter Project: Frank Siebelt, Dieter Mans, Mathias Boes Layout und Druckvorstufe: Georg Peter Editorial Office: PROTOSOCIOLOGY, Stephan-Heise-Str. 56, 60488 Frankfurt am Main, RFA, Phone: 069-769461, Email: [email protected], [email protected] Bank: Dresdner Bank AG, Frankfurt am Main, Kto: 44 121 168 01, BLZ: 500 800 00 Wir danken debis, concept and products für ihre freundliche Unterstützung.

Die Zeitschrift soll 1/2jährlich erscheinen. Die Anzahl der jährlich erscheinenden Hefte und Sonderhefte bleibt jedoch vorbehalten. Copyright: Die in dieser Zeitschrift veröffentlichten Beiträge sind urheberrechtlich geschützt. Alle Rechte sind vorbehalten. Übersetzungen, Nachdruck, Vervielfältigung auf fotomechanischem oder ähnlichem Weg oder im Magnettonverfahren, Wiedergabe durch Vortrag, Funk- und Fernsehsendungen sowie Speicherung in Datenverarbeitungsanlagen, auch auszugsweise, sind nur mit Genehmigung des Herausgebers möglich. Für eingereichte Beiträge wird keine Haftung übernommen. Weitere Publikationsrechte von Artikeln bleiben vorbehalten. Zweitpublikationen seitens des Autors werden eingeräumt. Bei einer Zweitpublikation ist das Heft (Nummer, Titel, Erscheinungsjahr) der PROTOSOCIOLOGY zu zitieren. Für unaufgefordert eingesandte Manuskripte wird keine Haftung übernommen. Gerichtsstand ist Frankfurt am Main. Copyright: All rights reserved. This publication may not be reproduced, stored or transmitted in any form or by any means without the prior permission in writing of the publisher. Additional publications of the articles are reserved. The authors retain the personal right to re-use their own articles. Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients is garanted by PROTOSOCIOLOGY, provided that the base fee is paid directly to VG Wort, Goethestr. 49, 80336 München RFA.The publisher accepts no responsibility for submitted manuscripts.

ProtoSociology im Internet: http://www.protosociology.de http://www.rz.uni-frankfurt.de/protosociology

183

On ProtoSociology

ON PROTOSOCIOLOGY

“Protosociology occupies an important position in the European intellectual scene, bridging philosophy, economics, sociology and related disciplines. Its volumes on rationality bring together concerns in all these topics, and present an important challenge to the cognitive sciences.” Donald Davidson, Berkeley (USA) “Protosociology publishes original papers of great interest that deal with fundamental issues in the human and social science. No academic library is complete without it.” Nicholas Rescher, Pittsburgh (USA) “Protosociology has been remarkably successful in publishing interesting work from different tradition and different disciplines and, as the title signals, in giving that work a new, eye-catching slant.” Philipp Pettit, Canberra, Australia “Protosociology is a truly premier interdisciplinary journal that publishes articles and reviews on timely topics written by and for a wide range of international scholars. The recent volumes on rationality are remarkable for their breadth and depth. Protosociology would be a great addition to any library.” Roger Gibson, St. Louis (USA)

184

Volumes

PROTOSOCIOLOGY An International Journal of Interdisciplinary Research

VOL. 12, 1998 — Special Edition AFTER THE RECEIVED VIEW — Developments in the Theory of Science Gerhard Preyer, Georg Peter, Alexander Ulfig (Eds.) in memoriam Wolfgang Stegmüller Content Introduction: Developments in the Theory of Science Gerhard Preyer, Georg Peter, Alexander Ulfig LOGICAL OPERATIONALISM — SIGNIFICANCE AND MEANING Wilhelm K. Essler Truth and Knowledge. Some Considerations concerning the Task of Philosophy of Science Gerhard Preyer The Received View, Incommensurability and Comparison of Theories — Beliefs as the Basis of Theorizing Robert Schwartz Reflections on Projection Jeffrey E. Foss The Logical and Sociological Structure of Science STRUCTURALISM — MEANINGFUL MEASUREMENT — THE CONCEPTION OF PHYSICAL LAW C. Ulises Moulines Structuralism vs. Operationalism Nicholas Rescher Meaningless Numbers

185

Volumes R. I. G. Hughes Laws of Nature, Laws of Physics, and the Representational Account of Theories James R. Brown Einstein’s Principle Theory

INDUCTIVE INFERENCES — INTERPRETATION OF PROBABILITY — GAME THEORY Kevin T. Kelly, Cory Juhl Transcendental Deductions and Universal Architectures for Inductive Inferences Howard H. Harriott R.A. Fisher and the Interpretation of Probability Brian Skyrms Evolution of an Anomaly PROPERTIES — UNDERDETERMINATION — SCIENTIFIC REALISM George N. Schlesinger Degrees of Characterizations Carl A. Matheson Observational Adequacy as distinct from the Truth about Observables Thomas R. Grimes Scientific Realism and the Problem of Underdetermination Paul C. L. Tang On Paul Churchland’s Treatment of the Argument from Introspection and Scientific Realism RATIONALITY — METAPHORS — VALUES IN SCIENCE David Resnik Scientific Rationality and Epistemic Goals Aldo Montesano Rationality in Economics: A General Framework Joseph Agassi Science Real and Ideal: Popper and the Dogmatic Scientist Michael Bradie Models and Metaphors in Science

David Gruender Values and the Philosophy of Science

186

Volumes

ProtoSociology An International Journal of Interdisciplinary Research VOL. 11, 1998 COGNITIVE SEMANTICS II — EXTERNALISM IN DEBATE Content RADICAL INTERPRETATION, ONTOLOGY AND THE STRUCTURE OF LANGUAGE Gerhard Preyer, Michael Roth On Donald Davidson’s Philosophy: An Outline Richard Manning All Facts Great and Small Barbara Fultner Of Parts and Wholes: The Molecularist Critique of Semantic Holism Louis Goble Re-Evaluating Supervaluations David Simpson Interpretation and Skill: On Passing Theory Wulf Kellerwessel Katz on Semantics and Pragmatics EXTERNALISM AND THE INDIVIDUATION OF CONTENT Ron Wilburn Knowledge, Content, and the Wellstrings of Objectivity Anthony Brueckner Content Externalism and A Priori Knowledge Consuelo Preti The Irrelevance of Supervenience Michael Liston Externalist Determinants of Reference Arnold Silverberg Semantic Externalism. A Response to Chomsky Gerhard Preyer Interpretation and Rationality: Steps from Radical Interpretation to the Externalism of Triangulation

187

Volumes

ProtoSociology An International Journal of Interdisciplinary Research VOL. 10, 1997 COGNITIVE SEMANTICS I — CONCEPTIONS OF MEANING Content MEANING, TRUTH AND THE USE OF LANGUAGE Herman Cappelen, Ernie LePore Semantic Theory and Indirect Speech

Jeffrey King The Source(s) of Necessity

Kirk Ludwig The Truth about Moods

Filip Buekens The Genesis of Meaning (a Myth)

Louise Röska-Hardy Language, Use and Action REFERENCE, INDEXICALS AND SPEAKER MEANING Robert Hanna Extending Direct Reference Peter Ludlow Semantics, Tense, and Time: a Note on Tenseless Truth-Conditions for TokenReflexive Tensed Sentences

Gerhard Preyer Verstehen, Referenz, Wahrheit. Zur Philosophie Hilary Putnams Reinaldo Elugardo Descriptions, Indexicals and Speaker Meaning

RATIONAL DIALETISM OF ANTINOMIES Mark Sainsbury Can Rational Dialetheism Be Refuted By Considerations about Negation and Denial? CONTEMPORARY DEBATES IN THE PHILOSOPHY OF MIND Klaus Sachs-Hombach Philosophy of Mind: Die Simulationstheorie IN RETROSPECT Joseph Agassi Wittgenstein – The End of a Myth

Kent A. Peacock, Richard Feist The Einstein-DeSitter Controversy

188

Volumes

ProtoSociology An International Journal of Interdisciplinary Research VOL. 8/9, 1996 RATIONALITY II & III INTERPRETATION, INTENTIONS AND PROPOSITIONAL ATTITUDES David K. Henderson: Epistemic Rationality, Epistemic Motivation, and Interpretive Charity; Roger F. Gibson: Stich on Intentionality and Rationality; Alfred Mele: Rational Intentions and the Toxin Puzzle; John Heil: The Propositional Attitudes; Volkmar Taube: Exemplifikatorische Darstellung. Zu den Grundlagen einer kognitiven Ästhetik NATURALIZED EPISTEMOLOGY, RATIONALITY AND NORMATIVITY Paul K. Moser, David Yandell: Against Naturalizing Rationality; Harvey Siegel: Naturalism, Instrumental Rationality and the Normativity of Epistemology; Ralf Naumann: Internal Realism, Rationality and Dynamic Semantics; Axel Wüstehube: Noch einmal: Rationalität und Normativität RATIONAL EXPLANATION, REASONING AND JUSTIFICATION Philip Pettit: Three Aspects of Rational Explanation; Keith Lehrer: Rationality and Trustworthiness; Alexander Ulfig: Stufen der Rechtfertigung CONCEPTS OF PRACTICAL RATIONALITY Peter A. French: Rationality and Ethics; Pierre Kerszberg: Feeling and Coercion: Kant and the Deduction of Right; Raymound Boudon: Une éthique est-elle possible en l'absence de croyances dogmatiques?; Raimo Tuomela: Rational Cooperation and Collective Goals; Wolfgang Welsch: Vernunft heute REVIEWS STATE OF THE ART Marcus Birke: Externalismus in der Philosophie des Geistes SCIENCE AND HISTORY Joseph Agassi: Die gegenwärtige Rolle des Technik- und Wissenschaftshistorikers 'RXEOHYROXPHSDJHVSULFH'0RULQFOXGLQJFRXUWDJH GLJLWDOYHUVLRQSULFH'0RU

http://www.rz.uni-frankfurt.de/protosociology

189

Volumes

Forthcoming

ProtoSociology An International Journal of Interdisciplinary Research PARSONS-SPECIAL EDITION CLASSICAL SOURCES AND CURRENT USES OF TALCOTT PARSONS’S THOUGHT edited and with an introduction by Bernard Barber and Uta Gerhard This volume wishes to cut across the vast array of misunderstandings, misreadings, and misinterpretations which have marred the proper recognition of Parsons’s work for the last thirty years. If the question is asked why such recognition would be important, the answer is twofold. For one, Parsons’s work is a classic that reaches from the l930s into the l970s of the twentieth century while sociology’s now arguably uncontroversial classics - like Max Weber, or Emile Durkheim - were only able to analyze society until about the end of World War I. Therefore, Parsons’s oeuvre has a special modernity in that his work takes into account the New Deal, National Socialism, McCarthyism, and the Student Rebellion, to name but a few timely events which are part and parcel of Parsons’s sociological analysis. Furthermore, his conception of social order and anomie, especially since they seriously discuss the problematic viability of utilitarian economics, are of particular importance today. Introductionary Essay: The Parsons Agenda Bernard Barber, Uta Gerhardt PART 1: Revisiting The Structure of Social Action Mark Gould: Neoclassical Economics, Social Order, and “The Structure of Social Action” Uta Gerhardt: National Socialism and the Politics of “The Structure of Social Action” Richard Münch: The Problem of Order: Sixty Years After “The Structure of Social Action” PART 2: Understanding Parsons’s Thought Bryan S. Turner: Sacricity and Solidarity in Talcott Parsons’s Economic Sociology Edward A.Tiryakian: An Emergent French Connection: Revisiting Parsons’s Durkheim Bernard Barber: The Nature and Functions of Religion and Beyond PART III: Evaluating the Scope of Sociology Daniel Bell: The Limits of General Theory in the Social Sciences Hans Jürgen Puhle: (Title to be announced)

190

Bookpublications

%RRNSXEOLFDWLRQV

PROTOSOZIOLOGIE IM KONTEXT „Lebenswelt“ und „System“ in Philosophie und Soziologie Gerhard Preyer, Georg Peter, Alexander Ulfig (Hrsg.) in memoriam Paul Lorenzen Einleitung: „Lebenswelt“ und „System“ in Philosophie und Soziologie, Gerhard Preyer, Georg Peter, Alexander Ulfig

ZUM BEGRIFF DER LEBENSWELT: Ernst W. Orth (Trier, Deutschland): Lebenswelt als unvermeidliche Illusion? Husserls Lebensweltbegriff und seine kulturpolitischen Weiterungen, Walter Biemel (Aachen, Deutschland): Gedanken zur Genesis der Lebenswelt, Alexander Ulfig (Frankfurt/ Main, Deutschland): Lebenswelt und Reflexion. Anhang: Lebenswelt als Fundament der Wissenschaft, Gerhard Preyer (Frankfurt/Main, Deutschland): Hintergrundwissen: Kritik eines Begriffs, Hubert A. Knoblauch (Konstanz, Deutschland): Soziologie als strenge Wissenschaft? Phänomenologie, kommunikative Lebenswelt und soziologische Methodologie

LEBENSWELT — BEGRÜNDUNG — WISSENSCHAFT Jürgen Mittelstraß (Konstanz, Deutschland): Das lebensweltliche Apriori, Peter Janich (Marburg, Deutschland): Die Rationalität der Naturwissenschaften, Jürgen Mittelstraß (Konstanz, Deutschland): Rationalität und Reproduzierbarkeit, Elisabeth Ströker (Köln, Deutschland):Lebenswelt durch Wissenschaft, Paul Janssen (Köln, Deutschland): Lebenswelt, Wissen und Wissenschaft, Richard T. Murphy (Concord, USA): E. Husserl's Phenomenology of Reason LEBENSWELT/LEBENSFORM — SPRACHE Pierre Kerszberg (Pennsylvania, USA): Lifeworld and Language, John F.M. Hunter (Toronto, Canada): The Motley Forms of Life in the Later Wittgenstein, Peter A. French (San Antonio, USA): Why did Wittgenstein read Tagore to the Vienna Circle? Georg Peter (Frankfurt/Main, Deutschland): Die Nebenbeschäftigung der Symbole. Zu Wahrheit und Funktion der Metapher

SYSTEM — SOZIALSYSTEM — GESELLSCHAFT Niklas Luhmann (Bielefeld, Deutschland): Die Lebenswelt nach Rücksprache mit Phänomenologen, Niklas Luhmann (Bielefeld, Deutschland): Observing Re-entries, Gerhard Preyer (Frankfurt/Main, Deutschland): System-, Medien- und Evolutionstheorie. Zu Niklas Luhmanns Ansatz, Richard Münch (Düsseldorf, Deutschland): Autopoesis per Definition, Hans Zitko (Frankfurt/Main, Deutschland): Ästhetik als Soziologie der Kunst, James Bohman (St. Louis, USA): The Completeness of MacroSociological Explanations: System and Lifeworld, Göran Ahrne (Stockholm, Schweden): Outline of an Organisational Theory of Society, Anhang: Karl Otto Hondrich (Frankfurt/Main, Deutschland): Zu Göran Ahrnes Ansatz 392 Seiten, 68,- DM

Verlag Königshausen & Neumann, Würzburg 1996

191

Bookpublications

STRUKTURELLE EVOLUTION UND DAS WELTSYSTEM Theorien, Sozialstruktur und evolutionäre Entwicklungen Gerhard Preyer (Hrsg.) Einleitung:

Gerhard Preyer: Strukturelle Evolution und das Weltsystem: Theorien, Sozialstruktur und evolutionäre Entwicklungen ZENTUM UND PERIPHERIE — INSTITUTIONELLE ENTWICKLUNG — ASKRIPTIVE SOLIDARITÄT Shmuel Noah Eisenstadt: Social Division of Labor, Construction of Centers and Institutional Dynamics: A Reassessment of the Structural—Evolutionary Perspective, Bernhard Giesen, Kay Junge: Strukturelle Evolution, Gerhard Preyer: Soziale Gesetze und askriptive Solidarität. Eine Skizze zu den Grundlagen der Gesellschaftstheorie, Anhang: Die modernen Gesellschaften verstehen. Richard Münchs Entwicklungstheorie moderner Gesellschaften, Erwin Rogler, Gerhard Preyer: Relationslogische Darstellung der sozialen Gesetze, Gerhard Preyer: Die modernen Gesellschaften „verstehen“. Zu Richard Münchs Entwicklungstheorie moderner Gesellschaften, Dieter Claessens: Bemerkungen zur Entstehung der modernen Ökonomie: Das Organistionsproblem, Richard Pieper: Strukturelle Emotionen, elementare Strukturbildung und strukturelle Evolution DIE EVOLUTION POLITISCHER ORDNUNGEN Rainer C. Baum: Parsons on Evolution of Democracy, Mathias Bös: Zur Evolution nationalstaatlich verfaßter Gesellschaften, Konrad Thomas: Das Ethnische und das Staatliche, Volker Bornschier: Die westeuropäische Integration als Gesellschaftsmodell im Zentrumswettbewerb ZUR SOZIOLOGIE DES WELTSYSTEMS Roland Robertson: Globalization and reflexive Modernization, Immanuel Wallerstein: Evolution of the Modern World-System, Christopher Chase-Dunn, Thomas D. Hall: The Historical Evolution of World-Systems: Iterations and Transformations, Albert Bergesen: Postmodernism: A World System Explanation, Richard Münch: Modernity and Irrationality: Paradoxes of Moral Modernization, Walter L. Bühl: Transformation oder strukturelle Evolution? Zum Problem der Steuerbarkeit von sozialen Systemen STATE OF THE ART: Michael Schmid: Soziologische Evolutionstheorien

Suhrkamp Verlag, Frankfurt am Main 1998, Suhrkamp Taschenbuch Wissenschaft

192

Bookpublications

LEBENSWELT — SYSTEM — GESELLSCHAFT Konstruktionsprobleme der „Theorie des kommunikativen Handelns“ von Jürgen Habermas *HUKDUG3UH\HU

I DIE ENTWICKLUNGSLOGIK VON WELTBILDERN 1. Die Rationalisierung von Weltbildern 2. Strategien der Analyse von Weltbildern 3. Folgeprobleme und Kritik II GESELLSCHAFT ALS LEBENSWELT UND SYSTEM 1. Die Strukturen der Lebenswelt 2. Allgemeine Bezugsprobleme der soziologischen Evolutionstheorie 3. Die Verständigungsformen 1. Weltbilder und soziale Integration 2. Zur Durkheim-Interpretation 4. Kommunikationsmedien und generalisierte Kommunikationsweisen 5. Folgeprobleme und Kritik III DIE FORMAL-PRAGMATISCHE BEDEUTUNGSTHEORIE 1. Die sprechakttheoretische Grundlegung 2. Interpersonal geltende Bedeutungskonventionen 3. Regelbewußtsein und Handlungskompetenz 4. Die Expansion des semantischen Gehalts 5. Verständigung und die Herstellung interpersonaler Beziehungen IV DER ERWERB DES MORALISCHEN BEWUßTSEINS 1. Die sozial-kognitive Grundausstattung 2. Diskurs und moralisches Bewußtsein 3. Folgeprobleme V KONSTRUKTIONSPROBLEME UND KRITIK 1. Rekonstruktionshypothesen 2. Zu den Konstruktionsproblemen ANHANG: Max Webers Religionssoziologie als eine Typologie des Rationalismus 290 Seiten, DM 28,– , auch als CD lieferbar

Digital bookpublication Humanities-Online, Frankfurt am Main http://www.humanities-online.de Email: [email protected]

193

Bookpublications

Die globale Herausforderung Wie Deutschland an die Weltspitze zurückkehren kann Gerhard Preyer Der Begriff Globalisierung hat bereits Schlagwortcharakter bekommen und ist in aller Munde. Globale Märkte haben die Grundsituation wirtschaftlichen und politischen Handelns grundsätzlich verändert. Das heute entstehende globale Weltsystem und der Medienverbund auf dem es beruht, fordert uns alle heraus. Was bedeutet Globalisierung? Worin bestehen die globalen Herausforderungen, die zu bewältigen sind? Darauf wird eine Antwort gegeben.

NEUE MEDIEN, WISSENSCHAFT UND TECHNIK IM ZEITALTER DER GLOBALISIERUNG Was heißt Globalisierung? – Die neuen Medien: Eine Kopernikanische Wende – Der Überlebensimperativ: Technologieentwicklung – Die Zukunft der deutsche Universitäten VERÄNDERTE KONSTELLATIONEN Globale Wirtschaft und globale Ordnung – Die Globalisierung der Finanzmärkte – Auf dem Weg zur virtuellen Organisation – Das Ende des klassischen Arbeitsmarktes – Zu einer neuen Wirtschaftspolitik EUROPA IM ZEITALTER DER GLOBALISIERUNG Zur Ausgangssituation im Zentrumswettbewerb – Frankreichs Zentralismus: Grenzen der Marktwirtschaft – Das Überlebenssystem Italien – Groß-britannien zwischen Tradition und Modernisierung – Deutsche Stärken im Umbruch – Neu gemischte Karten – Die Wiedergewinnung des Standorts Deutschland DIE EVOLUTION DES MITGLIEDSCHAFTSCODES Zur Soziologie der Grenzziehungen – Gesellschaft, Organisation und Interaktion – Gesellschaftsinterne Globalisierung und Entwicklungstrends – Die globalisierte Gesellschaft 290 Seiten, DM 72,–

Frankfurter Allgemeine Zeitung/Gabler Edition, Frankfurt am Main 1998

194

Bookpublications

INTEGRIERTES MANAGEMENT Was kommt nach der Lean-Production? Gerhard Preyer/Jakob Schissler Es gibt heute im Weltmaßstab kein Unternehmen, das seine Umstrukturierung nicht an den Bestandteilen des japanischen Modells der schlanken Produktion und des schlanken Managements, der Lean-Modelle, orientiert. Die Lean-Modelle sind zu einem integrierten Management fortzuentwickeln: dem über Vernetzung integrierten segmentierten Unternehmen, mit durchlässigen, ständig wechselnden Grenzen zwischen Unternehmen, Lieferanten und Kunden. Integriertes Management ist das Unternehmensmodell, das eine erhöhte Qualifizierung der Mitarbeiter und die Fähigkeit, in sich selbst steuernden Teams mitzuarbeiten, erfordert. Das bedeutet aber auch die Entwicklung eines neuen Führungsstils. Mit integriertem Management kann jedes Unternehmen sofort anfangen. Dazu wird in diesem Buch ein Weg gezeigt. Es wird eine wichtige Funktion als Impuls für weitere Entwicklungen haben. Inhalt DIE NEUEN HERAUSFORDERUNGEN 1. Vom japanischen und amerikanischen Management lernen? 2. Information, Wissen und Entscheidungen KREATIVITÄT — ORGANISATION — FÜHRUNG 1. Managementfähigkeiten 2. Innovative Organisation 3. Management und Führungsstile STRATEGIE — MARKT — KULTUR 1. Unternehmensstrategie 2. Controlling 3. Marketing 4. Zur Unternehmenskultur STICHWORT Was kommt nach der Lean Production? 164 Seiten, 32,- DM

Frankfurter Allgemeine Zeitung, Verlagsbereich Wirtschaftsbücher, Blickbuch-Wirtschaft, Frankfurt am Main 1996

195

Bookpublications

ÜBER DAS KÄMPFEN Zum Phänomen des Kampfes in Sport und Gesellschaft Axel Binhack

Kämpfen ist stets als ein Beziehungsphänomen, gleichsam als „Fortsetzung der Kommunikation mit anderen Mitteln“ zu verstehen. Anhand von sechs Strukturmerkmalen wird eine „Prototypik des Kampfes“ entwickelt. Ihre zusammenhängende Dynamik ist gesellschafts- und kulturunabhängig in vielfältigen Konfliktformen zu erkennen. In dem vorgelegten interdisziplinären Anschnitt wird die Struktur der Kampfbeziehung von vergleichbaren Sozialphänomenen abgegrenzt und am Beispiel des Kampfsports in Form des „Zweikampfes“ analysiert. 0HUNPDOHHLQHU)RUPDOVWUXNWXUGHV.DPSIHV Der duale Beziehungsaspekt: Antagonismus und Ambivalenz — Der duale Trägeraspekt: Entscheidungsorientiertheit und tendenzielle energetische Totalität — Der duale Inhaltsaspekt: Zweckgerichtetheit und riskante Offenheit 'LH6WUXNWXUPHUNPDOHGHV.DPSIHVLP9HUJOHLFKPLW$UEHLW6SLHOXQGGHQ %H]LHKXQJVSKlQRPHQHQ7DXVFK)OLUWXQG0DFKW Der „antagonistische“ Unterschied — Freiheit und Zweckgerichtetheit — „Zeitenthobene“ innere Unendlichkeit und „zeitbeschleunigende“ Entscheidungsorientiertheit — Die formalstrukturelle „Folgenhaftigkeit“ des Kampfes =XP.DPSIVSRUW Kampfsport als Radikalisierung des sportlichen Kampfes — Der Kampfsport als kulturelles Phänomen — Der Kampfdiskurs als Symbol männlichen Agierens — Ästhetische Perspektiven 'LH.DPSIVSRUWDUW.DUDWH'RDOV%HLVSLHO Historische Entwicklungslinien des Karate-Do — Der Schwertkampf und die „Leere Hand“: Zum karatespezifischen Aspekt der Waffenlosigkeit — Die Kultivierung des Kampfphänomens im Karate-Do — Der Kampf als Lehrmeister im Sinne des Zen

Campus Verlag, Campus Forschung Bd. 768, Frankfurt am Main 1998

196

Bookpublications

System der Rechte, demokratischer Rechtsstaat und Diskurstheorie des Rechts nach Jürgen Habermas hrsg. von Werner Krawietz / Gerhard Preyer Editorial: Zivilgesellschaftliche Assoziationen und spätmoderner Rechtssstaat (Werner Krawietz)

I JURISTISCHE ENTSCHEIDUNG, BEGRÜNDUNG UND RECHTFERTIGUNG DES RECHTS IN DER PERSPEKTIVE DISKURSIVER REFLEXIONSTHEORIE Enrique P. Huba: Standortbestimmung in der zeitgenössischen Rechtstheorie - Rawls, Dworkin, Habermas, Alexy und andere Mitglieder der modernen ‚Heiligen (Rede-) Familie‘, Thomas McCarthy: Legitimacy and Diversity: Dialectical Reflexions on Analytical Distinctions, Gerhard Preyer: Rechtsgeltung — Argumentation — Entscheidung II ORDNUNG, RECHT UND RATIONALITÄT IM JURISTISCHEN DISKURS Karl-Heinz Ladeur: Rechtliche Ordnungsbildung unter Ungewißheitsbedingungen und intersubjektive Rationalität, Ulfrid Neumann: Zur Interpretation des forensischen Diskurses in der Rechtsphilosophie von Jürgen Habermas, Ota Weinberger: Diskursive Demokratie ohne Diskurstheorie III KONSENSUSTHEORIE VERSUS SUBJEKTPHILOSOPHIE? VORAUSSETZUNGEN UND FOLGEN DES JURISTISCHEN DISKURSES IM DEMOKRATISCHEN RECHTSTAAT Michael Pawlik: Die Verdrängung des Subjekts und ihre Folgen. Begründungsdefizite in Habermas’ „System der Rechte“, Uwe Steinhoff: Probleme der Legitimation des demokratischen Rechtsstaats, William Rehg: The Place of Consensus in Democratic Legitimation: A Recommendation

Rechtstheorie Zeitschrift für Logik, Methodenlehre, Normentheorie und Soziologie des Rechts 3 1996, Habermas-Sonderheft (erschienen 1998) Duncker und Humblot, Berlin

197

Bookpublications

INTENTION — BEDEUTUNG — KOMMUNIKATION Kognitive und handlungstheoretische Grundlagen der Sprachtheorie Gerhard Preyer, Maria Ulkan, Alexander Ulfig (Hrsg.) Einleitung: Zu kognitiven und handlungstheoretischen Grundlagen der Sprachtheorie, Gerhard Preyer, Maria Ulkan, Alexander Ulfig

I INTENTIONEN UND KOMMUNIKATIVE HANDLUNGEN Maria Ulkan: Kommunikative und illokutionäre Akte; Georg Meggle/Maria Ulkan: Grices Doppelfehler. Ein Nachtrag zum Griceschen Grundmodell; Jan Nuyts: Intentionalität und Sprachfunktionen II INTERPRETATION UND BEDEUTUNG Gerhard Preyer: Kognitive Semantik, Anhang Sprechaktsemantik: J.L. Austin, J.R. Searle, H.P. Grice, P.F. Strawson; Louise Röska-Hardy: Sprechen, Sprache, Handeln; Frank Siebelt: Zweierlei Holismus. Überlegungen zur Interpretationstheorie Donald Davidsons; Peter Rothermel: Semantische Implikaturen; Volkmar Taube: Referenz und Interpretation. Zur Theorie nichtsprachlicher Symbolisierung; Georg Peter: Zu Richtigkeit und Interpretation der Metapher: Kognitive Funktion und rekonstruktive Schemainterpretation III KLASSIFIKATION VON SPRECHAKTEN Maria Ulkan: Informations- und Aufforderungshandlungen; Dirk Hartmann: Konstruktive Sprechakttheorie; Volkmar Taube: Bildliche Sprechakte IV KOMMUNIKATIVES HANDELN UND INTERSUBJEKTIVE GÜLTIGKEIT Jürgen Habermas: Sprechakttheoretische Erläuterungen zum Begriff der kommunikativen Rationalität; Karl-Otto Apel: Illokutionäre Bedeutung und normative Gültigkeit. Die transzendentalpragmatische Begründung der uneingeschränkten kommunikativen Verständigung; Peter-Paul König: Kommunikatives und strategisches Handeln. Kritische Bemerkungen zu zwei zentralen Begriffen der „Theorie des kommunikativen Handelns“ von Jürgen Habermas; Alexander Ulfig: Präsuppositionen und Hintergrundwissen. Eine Kritik am formalpragmatischen Präsuppositionsbegriff V DIALOGSTRUKTUR UND ARGUMENTATION Wilhelm Franke: Konzepte linguistischer Dialogforschung; Franz Hundsnurscher: Streitspezifische Sprechakte: Vorwerfen, Insistieren, Beschimpfen; Dieter Mans: Argumentation im Kontext, Exkurs: Zu Christoph Lumers „Praktische Argumentationstheorie“ 408 Seiten, 58.- DM

Westdeutscher Verlag, Opladen 1997

198

Bookpublications

Wittgensteins Spätphilosophie Analysen und Probleme Wulf Kellerwessel, Thomas Peuker (Hrsg.)

Die Spätphilosophie Ludwig Wittgensteins entfaltet nicht nur in der gegenwärtigen Sprachphilosophie ihre Wirkung, sondern auch zunehmend über den Bereich der Sprachphilosophie hinaus. Zwar macht dieser einen Schwerpunkt des Wittgensteinschen Denkens aus, aber dennoch verdanken die Philosophie des Geistes, die Erkenntnistheorie und die Philosophie der Religion der Wittgensteinschen Sprachphilosophie wichtige Impulse. Diese Schwerpunktsetzung zugunsten der Sprachphilosophie wie auch ihre Zusammenhänge mit den genanten anderen Disziplinen der Philosophie zeichnet der vorliegende Band aus.

Inhalt W. Kellerwessel, T. Peuker: Einleitung, A. Ofsti: Methodischer Solipsismus, Metasprachen(problem) und Dexis. Einige Überlegungen zum Privatsprachenproblem, T. Peuker: Die faktische Öffentlichkeit der Sprache. Zu Wittgensteins Privatsprachenargument, P. Niesen: Gemeinschaft, Normativität, Praxis: Zur Debatte über Wittgensteins Regelbegriff, A. Berndzen: Einer Regel entsprechen - einer Regel folgen. Zu einer Kontroverse in der Interpretation von Wittgensteins ‘Philosophischen Untersuchungen’, G. Preyer: Sprachbedeutung ohne Regelbefolgung, C. Stetter: Sprache und Schrift bei Wittgenstein, H.J. Schneider: Mentale Zustände als metaphorische Schöpfungen, W. Kellerwessel: Zum Begriff der Gewißheit in Wittgensteins ‘Über Gewißheit’. Ein Kommentar, M. Kroß: „Glaube Du! Es schadet nicht“. Ludwig Wittgensteins Vermischte Schriften zur Religion - Anhang.

302 Seiten, DM 86,–

Verlag Königshausen & Neumann, Würzburg

199

Bookpublications

Donald Davidsons Philosophie Von der radikalen Interpretation zum radikalen Kontextualismus

Gerhard Preyer Für Bruce Aune EINLEITUNG

I RADIKALE INTERPRETATION, LOGISCHE FORM UND EREIGNISSE 1. Donald Davidsons Philosophie: Ein Überblick 2. Wahrheit, Bedeutung und radikale Interpretation 2.1 Radikale Interpretation als ein radikaler Externalismus (1) Der Zirkel zwischen Überzeugung und Bedeutung: Die Asymmetrie von RI (2) Die Strukur von Inhaltssätzen: sagen, daß ... (3) Der Grundsatz der Nachsicht (4) Die Autonomie der Bedeutung (5) Radikaler Externalismus: Der innertheoretische Schritt (6) Die Ontologie von RI 2.2 Von der Idiolekttheorie zum dritten Dogma des Empirismus (1) Ausgangs- und Übergangstheorien der Interpretation (2) Die Demontage eines Mythos (3) Die epistemischen Beschränkungen von Verstehen 3. Die logische Form von Handlungssätzen und die singuläre Kausalaussage 3.1 Logische Form und adverbiale Modifikation 3.2 Kausale Beziehungen 4. Wahrheit und Überzeugung 4.1 Zum Haupteinwand 4.2 Zur empirischen Offensichtlichkeit von Einstellungszuschreibung 4.3 Rationalität als normativer Begriff

II KÖRPERBEWEGUNGEN UND HANDLUNGEN 1. Das logische Verknüpfungsargument

200

Bookpublications

2. Primäre Gründe und die Identitätsthese: Die synkategormatische Fassung von Absichten 3. Basisakte 4. Flucht vor den Körperbewegungen 4.1 Thalbergs Handlungstheorie und der „Akkordeon-Effekt“ 4.2 A.I. Goldmans Kritik an der Identitätsthese 4.3 H.L.A. Hart: Zuschreibungen 4.4 Körperbewegungen als Bestandteil von Handlungen

III EINE RADIKALE THEORIE DES HANDELNS 1. Der Begriff der Einstellungsrationalität 1.1 Die Homogenität der Interpretation 1.2 Bewertende Einstellungen 2. Handlungsbeschreibungen und Handlungsverursachung 3. Eine Handlungserklärung 3.1 Überzeugungen, Absichten, Situationen 3.2 Praktische Gedanken 3.3 Handlungsgründe 3.4 Zu Hempels Ansatz 4. Praktische Schlüsse 4.1 Zur Gültigkeit praktischen Schließens 4.2 Entscheidungen und die Ausführung von Absichten 4.3 Überzeugungen und willentliche Handlungen 5. Radikaler Kontextualismus 6. Individualismus versus Holismus 7. K.R. Poppers Kritik am Historizismus Literatur 220 Seiten, DM 29,– Digital Publication: Order from: http://www.rz.uni-frankfurt.de/protosociology

201

Bookpublications

Language, Mind and Epistemology On Donald Davidson’s Philosophy Gerhard Preyer, Frank Siebelt, Alexander Ulfig (eds.) Introduction: On Donald Davidson's Philosophy (Gerhard Preyer, Frank Siebelt, Alexander Ulfig) PHILOSOPHY OF LANGUAGE Jerry Fodor, Ernie Lepore (New Brunswick, USA): Meaning, Holism, and the Problem of Extensionality; Olav Gjelsvik (Oslo, Norway): Davidson's Use of Truth in Accounting for Meaning; Wilhelm K. Essler (Frankfurt/Main, Germany): Was ist Wahrheit? Arend Kulenkampff (Frankfurt/Main, Germany): Eigennamen und Kennzeichnungen EPISTEMOLOGY Roger F. Gibson (St. Louis, USA): Quine and Davidson: Two Naturalized Epistemologists; Eva Picardi (Bolongna, Italy): Davidson and Quine on Observation Sentences; Ralf Naumann (Düsseldorf, Germany): Events and Externalism; Dorit Bar-On (Chapel Hill, USA): Conceptual Relativism and Translation; David K. Henderson (Memphis, USA): Conceptual Schemes after Davidson; Frank Siebelt (Frankfurt/Main, Germany): Singular Causal Sentences and two Relational Views PHILOSOPHY OF MIND AND THEORY OF ACTION Louise M. Antony (Raleigh, USA): The Inadequacy of Anomalous Monism as a Realist Theory of Mind, Louise Röska-Hardy (Darmstadt, Germany): Internalism, Externalism and Davidsons Conception of the Mental, Marcia Cavell (Berkeley, USA): Dividing the Self, Ralf Stoecker (Bielefeld, Germany): Willensschwäche - Wie ist das nur möglich? Klaus Puhl (Graz, Austria): Davidson on intentional Content and Self-Knowledge, Johannes Brandl (Salzburg, Austria): Sharing Beliefs and the Myth of the Subjective, Kirk Ludwig (Gainesville, USA): First Person Knowledge and Authority, Gerhard Preyer (Frankfurt/Main, Germany): Rationalität: Absichten, Primärgründe und praktisches Denken. Donald Davidson: Dialectic and Dialogue 445 Seiten

Kluwer Academic Publishers, Synthese Library, Dordrecht

202

Bookpublications

Reality and Humean Supervenience Essays on the Philosophy of David Lewis Gerhard Preyer, Frank Siebelt (eds.)

Preface Gerhard Preyer, Frank Siebelt Reality and Humean Supervenience - Some Reflections on David Lewis’ Philosophy

MODAL REALISM Phillip Bricker: Island Universes and the Analysis of Modality, John Bigelow: Time Travel Fictions, Peter Forrest: Counting the Cost of Modal Realism, Paul Teller: Lewis's Defence of Counterpart Theory, Harold W. Noonan: The Case for Perdurance,

PHYSICALISM, CAUSATION AND CONDITIONALS Daniel Bonevac: Naturalism for the Faint of Heart, D. M. Armstrong: Going throug the Open Door again: Counterfactual vs. Singularists Theories of Causation, Jonathan Bennett: On Forward and Backward Counterfactual Conditionals

REDUCTION OF MIND Terence Horgan: Multiple Reference, Multiple Realization and the Reduction of Mind, Michael Tye: Knowing what it is like: The Ability Hypothesis and the Knowledge Argument

Rowman and Littlefield Publishers, Lanham, USA

203

Bookpublications

LOGICAL FORM, LANGUAGE AND ONTOLOGY On Contemporary Philosophy of Language and Linguistics Gerhard Preyer, Georg Peter (eds.) One of the central topics analytical philosophy and especially the theory of language is concerned with is the concept of logical form. As typically understood the concept of logical form not only covers investigations into universal logical features underlying languages. Conceived that way the logical form of, say, arguments, sentences, and other parts of language may be the focus of linguists and semanticists. However, from Frege and Russell on logical form analysis were not confined to such a narrow linguistic perspective. For them investigating the logical form of languages always followed the wider philosophical perspective of trying to understand the structure of language as our principle means for representing the world. From Russell’s theory of definite descriptions up to Davidson’s truth-theoretical analysis of adverbial modification, citation and reported speech, to lay open the logical structure underlying language was always seen as a means to reveal the structure and features of the thereby represented world. Following such a broader philosophical perspective, the book contains several new essays which discuss for example: Russell’s understanding of logical form analysis, the relational structure of belief-sentences, the descriptivist view of that-clause-structures in languages, the logical sources of intensionality, the logical content of de se attitudes, the relation between ontological questions and questions regrading logical form. The collection brings together work by philosophers from diverse points of view, and as such, it illuminates the lively and ongoing debate the concept of logical form still arouses within contemporary philosophy. Contributors: Mark Richard (Medford,USA), Barry Schein (Los Angeles, USA), Bernard Linsky, James Higginbotham (Oxford, Great Britain), Norbert Hornstein (Maryland, USA), Marga Reimers, (Tucson, Arizona), Ernest Lepore, Kirk Ludwig (New Brunswick, Gainesville, USA,), Richard Larson, (Stony Brook, New York), James Tomberlin (Northridge, USA), Robert May/R.F. Fiengo, (Irvine, USA), Robert J. Stainton, (Ottowa, Canada), Ray Elugardo (Oklahoma, USA), Peter Ludlow, (Stony Brook, USA), Stephan Neale (Berkeley, USA ), Roger Schwarzschild, (New Brunswick, USA), Thomas Baldwin, (York, Great Britain), Robert Mattews, (New Brunswick, USA), Mark Crimmins, (Ann Arbor, USA), Jeffrey King, (Davis,USA) Forthcoming For further Information see: http://www.rz.uni-frankfurt.de/protosociology

204

Subscription

PROTOSOCIOLOGY ISSN 0940-4147 Subscription I subscribe PROTOSOCIOLOGY 30,- DM (15,- Euro) each issue, abroad: $ 30 (including courtage); payable by cheque. No price fixing for Double-Volumes.

Date, signature ................... Separate order I order PROTOSOCIOLOGY volume Nr. ....... Single issue: 35,- DM (17,50 Euro), abroad: $ 35 (including courtage); payable by cheque. All prices plus postage and packing

Date, signature ............

Your address: Name ....................

Address ....................

Send the order to: PROTOSOCIOLOGY ISSN 0940-4147 Editorial office: Stephan-Heise-Str. 56, D-60488 Frankfurt am Main 90, E-Mail: [email protected], Tel. 069/769461 Can be ordered directly from: Bockenheimer Bücherwarte Bockenheimer Landstr. 127 60325 Frankfurt am Main, RFA Tel.: 069-771088 PROTOSOCIOLOGY participates in Poiesis Online Service, Philosopical Dokumentation Center, Bowling Green State University, Bowling Green, OH 43403-0189, http://www. bgsu.edu/pdc/ Editor: Gerhard Preyer, Johann Wolfgang Goethe-Universität Frankfurt am Main, Department of Social Sciences. Editorial staff: Georg Peter. http://www.protosociology.de

Digital Publications – Special Offer

205

Digital Publications – Special Offer: ProtoSociology publishs digital versions of its volumes. Vol. 11 is a free download from our homepage under “Special Service”. Vol 13 – Reasoning and Argumentation Vol 12 – After the Received View – Developments in the Theory of Science Vol.11 – Cognitive Semantics II – Externalism in Debate (free download!) Vol 10 – Cognitive Semantics I – Conceptions of Meaning Vol 8/9 – Rationality II &III (double volume) Single digital version (Journals only): 7,50 Euro or $ 15.-

Special offer: Three digital Journals for the price of two: 15.- Euro or $ 30. (including courtage) Payable by cheque. E-Mail posting! Date, signature ............ Your address: Name ....................

Address ....................

Send the order to: PROTOSOCIOLOGY, editorial office: Stephan-Heise-Str. 56, D-60488 Frankfurt am Main 90, Tel. 069/769461 Order via E-Mail: [email protected] or [email protected]

Editor: Gerhard Preyer, Johann Wolfgang Goethe-Universität Frankfurt am Main FB 3: Department of Social Sciences. Editorial staff: Georg Peter. Project: Frank Siebelt, Dieter Mans, Mathias Bös KWWSZZZU]XQLIUDQNIXUWGHSURWRVRFLRORJ\

206

Was ist DENKBAR?

Alles ist DENKBAR!

Ziel der DENKBAR ist die Förderung von Philosophie, Kunst und Wissenschaft. Sie bietet zum Beispiel Experten wie auch Studenten die Möglichkeit, in Ihren Räumen wissenschaftliche Vorträge zu halten.In diesem Sinne versteht sich die DENKBAR als ‘philosophische Werkstatt’ und Denkraum für Kunst und Kultur. Wenn Sie an den Zielen der DENKBAR interessiert sind, laden wir Sie sehr herzlich ein, sich mit uns auf das “Abenteuer der Vernunft” (Kant) einzulassen

Denkbar Schillerstraße 26 60313 Frankfurt am Main Öffnungszeiten: täglich – außer Sonntag – von 18.00 bis 23.00 Uhr

207

.,1'(5'25) gibt Kindern eine Zukunft

3(58 Schenken Sie mit uns Straßenkindern eine Zukunft

Meine Vision Das Kinderdorf Peru ist meine Antwort auf ein soziales Problem in Lima/Peru. Nachdem ich selbst, vom Schicksal begünstigt, 30 Jahre Erfolg in meinem Frankfurter Modegeschäften erleben konnte, möchte ich dem Leben etwas von dem Glück und Überfluss zurückgeben, mit dem es mich bedachte. Die Not und das Elend der Straßenkinder in Peru, die ich auf einer Reise in Südamerika kennenlernte, bewegten mich zu dem Entschluss, helfen zu wollen und Not zu lindern. Weil Helfen Freude macht, entschloss ich mich Ende 1995, meine Firma zu verkaufen, das Kinderdorf Peru zu planen und hierfür einen Förderverein zu gründen. Inzwischen verbringe ich mehrere Monate des Jahres in Lima, um dort vor Ort den schwächsten Mitgliedern der Gesellschaft, den Kindern, zu helfen. Ein Teil meines privaten Vermögens habe ich bereits in die ersten Projekte investiert. Zur Erstellung und Unterhaltung des Kinderdorfes jedoch bitte ich um Ihre Mithilfe, da dies meine finanziellen Möglichkeiten übersteigt. Ich würde mich freuen, wenn Sie meine Vision teilen könnten und Ihre Hilfe den Kindern geben würden. Robert Lemli

Mehr über das KINDERDORF PERU können Sie auf unserer Hompage erfahren:

http://www.kinderdorf-peru.de Email: [email protected] Kinderdorf Peru e.V., Tannenwaldstr. 6, D-61389 Schmitten, Tel.: 06082-930273

ProtoSociology An International Journal of Interdisciplinary Research Vol.13 - 1999 Reasoning and Argumentation P Editors of the volume: Gerhard Preyer and Dieter Mans P Editorial of the Vol. 13 1999: Georg Peter P Layout and Technical Conception: Georg Peter P Editorial Office: ProtoSociology, Stephan-Heise-Str. 56, 60488 Frankfurt am Main, RFA, Phone: 069-769461, E-Mail: [email protected]

Proto Sociology VOL. 13, 1999 - REASONING AND ARGUMENTATION

Proto Sociology Vol. 17, 2002 - Semantic Theory and Reported Speech

Fundamentals of Critical Argumentation (Critical Reasoning and Argumentation)

PROTO SOCIOLOGY VOL. 8 9, 1996 RATIONALITY II & III

PROTO SOCIOLOGY VOL. 11,1998 COGNITIVE SEMANTICS II — EXTERNALISM IN DEBATE

Argumentation - the study of effective reasoning

Fallacies and Argument Appraisal (Critical Reasoning and Argumentation)

Fallacies and Argument Appraisal (Critical Reasoning and Argumentation)

PROTO SOCIOLOGY Vol. 15, 2001 - On a Sociology of Borderlines: Social Process in Time of Globalization

Grantville Gazette Vol 13

Argumentation, the study of effective reasoning - guidebook 1

Proto Sociology Volume 20, 2004 - World-System Analysis

Proto Sociology Vol. 16, 2002 - Understanding the Social: New Perspectives from Epistemology

Argumentation and Social Epistemology

Proto-Kartvelian and Proto-Indo-European Plant-Names

The Grantville Gazette Vol. 13

Progress in Optics, Vol. 13

Falaise 1944 (Militaria, Vol. 13)

Biotechnology Annual Review, Vol 13

Novels for Students Vol 13

Practical Argumentation

1999

№1999

Handbook of logic in AI and logic programming, vol.3: nonmonotonic reasoning and uncertain reasoning

Annual Review of Psychology, vol 50 1999

Reasoning, Argumentation, and Persuasion with Special Application to Hebrew Wisdom Literature and Hebrew Wisdom Scholarship

Proto Met Problemen

Am Proto-Tor

Proto-Indo-European Syntax

Proto Sociology VOL. 13, 1999 - REASONING AND ARGUMENTATION

Proto Sociology Vol. 17, 2002 - Semantic Theory and Reported Speech

Fundamentals of Critical Argumentation (Critical Reasoning and Argumentation)

PROTO SOCIOLOGY VOL. 8 9, 1996 RATIONALITY II & III

PROTO SOCIOLOGY VOL. 11,1998 COGNITIVE SEMANTICS II — EXTERNALISM IN DEBATE

Argumentation - the study of effective reasoning

Fallacies and Argument Appraisal (Critical Reasoning and Argumentation)

Fallacies and Argument Appraisal (Critical Reasoning and Argumentation)

PROTO SOCIOLOGY Vol. 15, 2001 - On a Sociology of Borderlines: Social Process in Time of Globalization

Grantville Gazette Vol 13

Argumentation, the study of effective reasoning - guidebook 1

Proto Sociology Volume 20, 2004 - World-System Analysis

Proto Sociology Vol. 16, 2002 - Understanding the Social: New Perspectives from Epistemology

Argumentation and Social Epistemology

Proto-Kartvelian and Proto-Indo-European Plant-Names

The Grantville Gazette Vol. 13

Progress in Optics, Vol. 13

Falaise 1944 (Militaria, Vol. 13)

Biotechnology Annual Review, Vol 13

Novels for Students Vol 13

Practical Argumentation

1999

№1999

Handbook of logic in AI and logic programming, vol.3: nonmonotonic reasoning and uncertain reasoning

Annual Review of Psychology, vol 50 1999

Reasoning, Argumentation, and Persuasion with Special Application to Hebrew Wisdom Literature and Hebrew Wisdom Scholarship

Proto Met Problemen

Am Proto-Tor

Proto-Indo-European Syntax

Recommend Documents