Organizational Change and Innovation
For further volumes, go to http://www.springer.com/series/8833
Davide Secchi
Extendable Rationality Understanding Decision Making in Organizations
123
Davide Secchi Department of Management University of Wisconsin-La Crosse La Crosse, WI 54601, USA
[email protected] ISBN 978-1-4419-7541-6 e-ISBN 978-1-4419-7542-3 DOI 10.1007/978-1-4419-7542-3 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2010937275 © Springer Science+Business Media, LLC 2011 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
A Zia Brunella e Nonno Egeo Perché ogni decisione che prendo è inevitabilmente sottoposta, nella mia mente, ai test di realismo e umanità che mi avete insegnato.
Preface
Writing a book is not very much appreciated in today’s academic world and it is not popular by any means. Scholars (including myself) are oriented toward writing and publishing journal articles, that is where the newest advancements in the field are often found. For this reason, the number of academics that rely on books for their research is lower compared to those that lean on articles. According to Thompson (Chronicle of Higher Education, 2005, Vol. 51, Issue 41), in the 1970s the number of copies per monographs that US and UK publishing companies would print was somewhere between 2,000 and 3,000. Expectations were to sell the most part of them. That trend has changed dramatically. Today, publishing companies print 400 or 500 copies of a monograph and hope to sell most part of them. Some publishing companies print “on-demand” and make the text available as an e-book. The decline in monograph publishing is only limitedly attributable to budget cuts in libraries, financial independence for academic presses, or for business conglomerates. There has been an increasing lack of support for this kind of publishing from scholars. This is my guess on the evidence reflected in the numbers that Thompson mentions in his book and in an article that appeared in the Chronicle of Higher Education. You may disagree with my interpretation, but you cannot disagree about the data that show a declining trend for monographs and on their lack of popularity in today’s academia (general management and organization behavior fields are not excluded). Another factor that supports what is written above is that an academic career is usually not affected by book writing as much as it is by article writing. This element may vary depending on the discipline, but it has expanded from the hard science to all the remaining fields. On top of this, writing a book involves a type of activity that is very different from what it takes to write a journal article. The latter is usually based on one (hopefully) original and innovative idea that is empirically tested or validated through a theoretical model. The former contains several ideas, organized in a web of connections. They are two outcomes of research and creativity that require different efforts, time, and dedication. Of course, I am not concerned with those monographs that are collections or that rewrite previously published articles. These are not original and add little to the spirit of a research effort. Everyone that earned a PhD knows the difference between writing a dissertation—that is very close to a book—and writing
vii
viii
Preface
an article. The dissertation may well be a collection of articles, where the candidate explains what the links are among them. Even in this case, dissertations and books usually require more effort, if not more time. To sum up, monographs (1) are not supported nor considered for one’s academic career, and (2) they require more effort and time. It seems that my decision to write a book makes little sense. If I add to these reasons that I am an assistant professor at the early stage of my career, this decision makes even less sense. Was it a rational decision? Why did I write this book then? Didn’t I care about my career and the academic community I belong to? From what I wrote, it is apparent that the reasons for this book should be found elsewhere. In fact, this is an attempt to ask some questions on rationality and decision making. Given what I wrote above on monographs, my decision to write this book should in itself be analyzed under the lenses of rationality. Is it rational for a young scholar at the early stage of his career to write a monograph? Well, besides writing that I really hope so, in the following I present the reasons that brought me to the decision to write this book. The entire book may be used as a tool to understand, among many others, if the reasons I provide below are rational or not. The reason underneath this project is tied to my research plan. I needed an outlet to present and analyze how the well-known theory of bounded rationality changes when its assumptions are modified. This effort cannot be contained in a single journal article, nor can several together give the breadth or the continuity that are needed. This is the rationale for the book. Bounded rationality is the idea that individuals have (a) limited computational capabilities (i.e., internal bound) and (b) limited access to information (i.e., external bound). As far as I know, there are no (recent) books that discuss foundational assumptions of this theory. Attempts to find a different perspective or to improve the existing framework are limited. The book aims at introducing some concepts that have the potential to redefine bounded rationality. For almost five years Emanuele Bardone and I have been planning to write a book that could summarize most of our efforts in understanding rationality and developing the bounded rationality theory. Ever since I moved to the United States, this distant cooperation worked fine with respect to articles, but we found it hard to write a book. We still plan to do that in the future. In the moment I am writing, both of us are about to submit books to the attention of some publishing companies. When Emanuele and I started to discuss rationality and decision making, it took both of us awhile before getting a common understanding of what we were saying. He is a cognitive scientist, I am a management scholar. At first, we just couldn’t understand each other. Anyway, I hope that this work will be the first of many attempts to present these ideas, and that we can soon add more writings on our common and, we believe, very interesting and intellectually challenging projects. Second, it happened that one semester (spring 2009) I taught an MBA class at the University of Wisconsin–La Crosse together with Thomas Krueger, professor of finance. The course was “Decision Framing II.” I have no idea why somebody decided to use the word “framing” instead of “making;” anyway, this is not relevant here. What is important is that Tom was in charge of teaching students how to
Preface
ix
make decisions in uncertain financial environments, and I was in charge of providing students with an introduction to decision making. In particular, as a professor of organizational behavior, I had the idea to put all my efforts toward defining how people make decisions inside organizations. I thought this part should have been very easy, although I wanted to give students perspectives that were current (e.g., sensemaking, embodied cognition, distributed cognition, and more). I looked at books on this topic, but they ended up being too narrow, too broad, or too simplistic. I wanted something that fitted my needs. I couldn’t find anything. After that, I started collecting articles on the topic and soon I realized that what I wanted to teach stayed together with difficulty. Then I tried to mix chapters of books with academic papers, but still the material remained too heterogeneous. At that point I realized that it was hard for students to read from all of these different sources without ending up being confused. Problems were related to two major points: (a) academic papers are not the easiest thing to read (sometimes even if you are in the profession!) and (b) jumping from author to author makes it hard to follow a common line of thought. Although the first part presents a quasi-standard reading of bounded rationality, the second part is too explorative and far from mainstream research. If used in a class on decision making, for example, the book needs to be associated with more “standard” readings—as I personally do in the classes I teach. It is probably more suitable for doctoral seminars. Third, early in January 2009 I was stuck in a situation that I didn’t want and that I didn’t expect. I accepted a tenure-track position at the University of Wisconsin–La Crosse on January 7, 2007, and started working for that university that same year on August 27. While I was getting a few papers published, I found it difficult to study and write on new topics or even to continue with my older streams of research. The reason was that here at UW-L we teach three classes per semester and, if I include summer sessions, in the first year I taught seven classes, most of them for the first time. The following fall I was stuck with four classes, two of which I had never taught before. I like teaching, but not at this rate. When my third semester at UW-L finished the only thing I could do was to recharge my batteries. I asked myself “Is this why I moved to the States? Is this what I wanted to do in the profession? Teach?” The answer was crystal clear in my mind. I decided to enter this profession because I have always had a passion for research and I know I cannot live well without my usual studying, reading, thinking, opinion exchanges, and writing. I needed to change something, and I needed to do it quickly. I decided to push hard on research, no matter what. While the teaching load for the coming semesters hasn’t changed, my research plan for the year 2009 was to have at least one publishable outcome per month. The first two months of the year were particularly productive. I wrote a total of six papers between January and February, well beyond my plans and expectations. Although I have other projects going on and I know that a book is not the best outcome for a tenure-track guy, I also thought that I have never written for career purposes only; I write because I think I have interesting questions to answer. Moreover, sometimes there are concepts that don’t fit into one, two, or several papers. I thought a book should have been a good product to start with in March.
x
Preface
I ended the first draft in April. I was still tied to my writing and research plan; also I built some foundational concepts for the future. I know that this is probably something you usually don’t write in prefaces, but this has been a starting point in my decision making. Although I wrote this book having my students in mind, this is not a textbook. There are several aspects that have potential interest for organizational behavior and decision making scholars. In the first chapter I go into further detail and explain what is new with the approach to decision making you are about to read, and then conclude with a short overview of the book’s contents. As far as the “rationality” of writing a monograph is concerned, I believe that these few pages (and Chapter 1) help clarify why I came out with the unpopular and improbable decision to write a book. La Crosse, Wisconsin
Davide Secchi
Acknowledgments
Multiple sclerosis is a terrible disease that hit my family very hard. For this reason, all author’s royalties will be paid directly to Fast Forward, a subordinate organization of the National Multiple Sclerosis Society (USA) that “focuses on expediting the drug development process, bridging the gap between promising discoveries and the commercial expertise and funding to move them forward” (http://www.nationalmssociety.org/fast-forward//index.aspx). I wish to thank the publisher, Springer US—especially Nick Philipson—that handled this with extreme flexibility and had gentle and supportive words for me. Also, Timothy Coetzee and Carol Miller from Fast Forward were very kind and worked over the weekend to match the publisher’s deadline. There are many people that contributed, and in many different ways, to everything it took to write and publish this book. It goes without saying that, despite all the help, I hold myself responsible for shortcomings that still are in the book. Everybody at the Department of Management at the University of Wisconsin–La Crosse supported this effort, and I wish to thank every single colleague for slightly different reasons. William Ross read the first draft of the manuscript and gave me the most useful insights and suggestions that I ever received on this book. It is not easy to find colleagues that are so much dedicated to research that it does not matter if it is their own or somebody else’s. William is one of these amazing persons. Gail Gillis also read the original version of the manuscript and gave it back to me with her very useful notes on it. Given her very busy teaching schedule, I really have no idea when she found the time to read the book. For exactly this reason, I express double appreciation and a million thanks for her help. I wish to thank John Betton and Tom Hench for challenging discussions, carried over on a regular basis and on an incredibly wide range of topics. Leticia Pena offered me to go to Caen, France, where I had the idea and wrote the book. I have shared with Drew Stapleton some of the ideas that are included in this book and benefited from his comments. Tom Kuffel, former Chair of the Department, has done his best to let me have all resources that the department could afford in order for me to complete the book or whatever I was doing at that time. Also, our discussions were like fresh air in the desert! Lori Komarek did a wonderful job simplifying everything that could be simplified to ease my job at the university.
xi
xii
Acknowledgments
My friend and colleague Thomas Krueger deserves a special mention here. We teach together the MBA class where I talk about many of the topics that you will find in this book. During my part of teaching, Tom always sits in the class and provides me with useful and original insights in the form of questions. Some of the points raised by Tom have been included in the text. Tom has an inquiring mind, and it has been a real pleasure to work with him. Many others helped with their thoughts, words, and other issues that I happened to connect to the manuscript I was working on. Among the many that served this purpose are Bruce May, Bill Colclough, Kuang-Wei Wen. MBA students at both the IAE–Centre Franco Américain, Université de Caen Basse Normandie, and the University of Wisconsin–La Crosse were always very open to ideas that we discussed in class. They helped me significantly understand the strengths and weaknesses of the various topics. Among those that deserve special thanks for sending their comments back to me are Michael Keith, Roman Yeskov, and Chen Yi-Jui. Many of the ideas that are in this book came out during discussions with my friend Emanuele Bardone, mostly at lunches and afternoons when we were PhD students at the University of Pavia, Italy. Those incredibly rich and in-depth discussions that lasted hours are the solid base for this book. Also, I am indebted to Lorenzo Magnani for introducing me to the world of abduction and distributed cognition; I have been touched by his humanity and deep mastery. My thanks and appreciation go also to the editor of the book series on Organizational Change and Innovation, Ann Gilley, for her comments. I benefited significantly from suggestions and from notes by Nicolai J. Foss, who served as reviewer of the original manuscript. His insights on several parts of the book certainly led to an improved version. Nick Philipson, Executive Editor at Springer US for the area Business and Management, has always been very supportive and active in promoting the idea of the book within Springer and with the book series editor. I enjoyed very much our e-mail correspondence and his willingness to keep me updated and current at every single stage of the process. Charlotte Cusumano, Editorial Assistant at Springer US, was always there to answer all of the (sometimes silly) questions I might have had. Thanks so much for making this publishing experience very smooth and extremely positive! On a more personal basis, my fiancée Claudia deserves very special thanks for all her loving support and care for my person when I was too much involved in reading, thinking, and writing. She even gave up some time of her favorite activity— shopping—when we were in Caen, France, to stay with me while I needed to discuss some ideas. She also read every single word I wrote and carefully edited the manuscript. I could not have had the luxury to focus on the book the way I did if it wasn’t for her. Also, my mother—now Dean of the Faculty of Economics and Business in my home town, Cagliari—has always supported me in everything I wanted to do. Being herself an author of several books, her comments and suggestions on an early draft of the manuscript have been extremely helpful. My father has been a safe place
Acknowledgments
xiii
to land when I needed a break from work and other too-serious thinking. Thanks, dad. My two brothers contributed to this book too. As a PhD student in supply chain management, Enrico has always been very critical of whatever his older brother was naively thinking. His challenges are always very important to me. Marco helped me in a subtler way; his and Zio Ninni’s relentlessness, passion, and dedication when they started up what now is a very successful brewpub business—Il Birrificio di Cagliari—showed me actual problem solving and decisions in the making. They extended their rationality very often, especially during the first months, when uncertainty was incredibly high. Notwithstanding the fact that Marco and Zio Ninni were my “guinea pigs,” they pour their tasty and unique beer for me every time I visit them. For free! A special thanks goes to my most precious friends Patrick and Gamze Randolph, Plator Ulkinaqu, and Raffaello Seri for our endless discussions on this and many other subjects. Last but not least, I want to extend my thanks to those of you that decided to buy the book and are eventually reading it. I really hope you will enjoy it.
Contents
1 Introduction . What to Expect Book Structure Notes . . . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
1 2 5 6
2 Rationalization and Rationality Kinds of Decisions . . . . . . . . The Legacy of Herbert Simon . . Summary . . . . . . . . . . . . . Notes . . . . . . . . . . . . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
9 9 13 17 17
3 Bounded Rationality . . . What Is Bounded Rationality Summary . . . . . . . . . . Notes . . . . . . . . . . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
19 19 24 24
4 Maps of Bounded Rationality (I) Prospect Theory . . . . . . . . . This Is a Biased World! . . . . . . Summary . . . . . . . . . . . . . Notes . . . . . . . . . . . . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
27 27 31 38 38
5 Maps of Bounded Rationality (II) . . . . Heuristics . . . . . . . . . . . . . . . . . . Accessibility, Representation, and Framing Two Logics . . . . . . . . . . . . . . . . . Implications of Using One or More Maps . Summary . . . . . . . . . . . . . . . . . . Notes . . . . . . . . . . . . . . . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
41 41 48 51 54 56 56
6 Simon’s Error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Distributing Cognition . . . . . . . . . . . . . . . . . . . . . . . . .
63 64
Part I
Part II
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
The Limited Cognition
. . . .
. . . .
The Extended Brain
xv
xvi
Contents
How Bounded Is Rationality? . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
74 77 77
7 Stretching the Bounds (I) . . . . Through Doing Decision Making The Rationality of Change . . . . Summary . . . . . . . . . . . . . Notes . . . . . . . . . . . . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
81 82 89 93 93
8 Stretching the Bounds (II) The Others . . . . . . . . . Advice Taking . . . . . . . Passive Advice Taking . . . Summary . . . . . . . . . . Notes . . . . . . . . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
97 97 98 108 110 110
9 The “Docile” Organization . . The “Docile” Individual . . . . Levels of Docility . . . . . . . . Understanding Docility . . . . . What Is a Docile Organization? Summary . . . . . . . . . . . . Notes . . . . . . . . . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
113 113 119 123 129 131 131
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
135 135 137 138 141 143 144
Afterword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
147
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
149
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
159
10 Conclusions . . . . . . . . . The Point on Rationality . . . What Are We Mapping? . . . The Individual and the Group A Methodological Note . . . . Extendable Rationality . . . . Notes . . . . . . . . . . . . .
. . . . . . .
List of Figures
8.1 8.2 8.3 8.4 8.5
Information richness scale . . . . . . The shared meaning of communication Information richness fit . . . . . . . . A low richness medium . . . . . . . . A high richness medium . . . . . . .
. . . . . . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
102 103 105 106 107
xvii
List of Tables
3.1 4.1 5.1 8.1 8.2 8.3 9.1
A combination of processes and outcomes . . Common errors . . . . . . . . . . . . . . . . The “Gang of eighteen” . . . . . . . . . . . Judge’s options . . . . . . . . . . . . . . . . A judge-advisor system (one type of advice) . A judge-advisor system (two types of advice) Characteristics of individual docility . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
20 37 50 99 100 100 123
xix
Chapter 1
Introduction
The first chapter of a book should be on how and why the author (me) decided to work on this topic, and made the decision to start writing. Ironically, it should take a whole book to explain how decisions of this sort happen. And this is exactly the reason for this book: explain the how and why of decisions. The decision to write this book in particular relates to the need that, I believe, the academic community has of it. There are four major points that are addressed in the book:
1. In an article published in the Journal of Economic Psychology in 2003b, Nicolai Foss suggests that the idea of bounded rationality is “much cited and little used.” This implies that there is a limited interest in foundational questions among management scholars and behavioral scientists. The book provocatively questions the nature of bounded rationality and of decision making. The need for conducting this study is dictated by the fact that science, and especially management, has changed since Simon presented his theory of rationality more than 60 years ago (1947). Does this relatively old theory fit today’s management theories? Does it fit advancements in cognitive science? The book is an attempt to answer these questions. And the answer is negative: The theory needs to be updated. However, if the reader thinks that ideas presented in the book do not provide a plausible answer to these questions still, the book may be useful for raising questions on what theory of rationality suits recent advancements in science. Going back to Foss, I believe the questions raised in the book are “much needed, little expected.” 2. The book presents an application to management and organizational behavior of the distributed cognition approach. Given the fact that this is a new and emerging approach, this is the first book that tries to connect it to management. What Hutchins (1995) wrote in his book Cognition in the Wild may be helpful for management scholars, and it is an interesting starting point. However, Hutchins is a cognitive scholar, and this remains his approach throughout the book. An attempt to bring distributed cognition close to management has been recently done by Michel (Administrative Science Quarterly, 2007). This book takes Michel’s study one step further, trying to answer background questions D. Secchi, Extendable Rationality, Organizational Change and Innovation, C Springer Science+Business Media, LLC 2011 DOI 10.1007/978-1-4419-7542-3_1,
1
2
1
Introduction
and to provide readers with a broader picture of how this approach may change our perspectives of managerial rationality and decision making. 3. There is a limited number of studies on “docility” (see Chapters 7, 8, and 9). I use this concept to connect cognition to behavior. In doing that, the book is one of the few to address socially based decision-making procedures to management. It is very unusual that authors go beyond the atomistic and individualistic views of the decision maker. In fact, those who study socially based decision making end up studying group dynamics. One of the most significant arguments advanced in the book is that no decision is (can be) made in complete isolation. The consideration of advice giving and taking, through decision-making processes, and docility is an attempt to fill this gap. The need for this gap to be filled relates to the need for having theories that are closer to the ways individuals actually make decisions. 4. The book tries to connect bounded rational decision making to epistemology. While philosophers (epistemologists) are aware of psychology and managerial studies, we cannot say that the relation works the other way around. I decided to make this philosophical connection clear (e.g., logical fallacies, abduction). The idea of an “extendable” rationality emerges when the nature of bounds is considered: What if human rationality is shaped by external resources? What if there is no clear divide between internal and external bounds? How can we analyze decision making if rationality is “extendable?” What is the role of social resources (or channels) in decision-making processes? Are innovation and change processes in organizations better understood through the extendable rationality approach? What are the implications for organizations? The approach presented in this book aims at bringing to models of human rationality the adaptability and flexibility that can be observed when individuals face change and, especially, innovation. One of the paragraphs of a later chapter is dedicated to studying how high-tech innovations shape human rationality.
What to Expect For a large part, the work on decision making has two roots: (a) formal (e.g., game theory) and (b) behavioral (e.g., prospect theory). There are still few works that combine the two, although cross-disciplinary research is growing. These approaches maintain their abilities to explain decision making, to analyze how these processes work, and to provide multiple resources to forecast the outcome of decisions. Notwithstanding all of these significant advantages, there are many ways to enrich these “classic” approaches to decision making. The objective of this book is to specify what domains can contribute to this enriching of decision making and how they can do that. On the one hand, the formal perspective reached its highest levels with the subjective expected utility (SEU) theory of von Neumann and Morgenstern,
What to Expect
3
and continued to inspire and define theories of decision making after that. The rational choice approachi is related to that first attempt. The idea of defining a mental process strictly on the basis of mathematicsii has never left the decision-making field of study. Moreover, utility continues to be the concept on which most economic theories lean on. Decision trees, cooperative and non cooperative games, and many other approaches and theories shape the field of decision making at a very deep level. It is a shared belief that theories of bounded rationality also share with SEU some of its basic assumptions and especially its methodologyiii , which has remained unchanged for too many years. On the other hand, the behavioral root has started with a critique of the formal approach, trying to make decision making more human. Behavioral economics (and especially behavioral finance) provides descriptive theories of how individuals act in real-life settings. We now have an ever-growing number of experiments, field studies, and all kind of dataiv supporting the hypothesis that human beings are not even close to the calculator-like metaphorv . This field of study brings psychology and sociology into play so that the focus is on prejudices, biases, anomalies, mistakes, heuristics, and on the many imperfections that characterize human decision-making processes. This manner of picturizing the whole field of decision making is probably too simplistic. However, it offers many advantages. First, it allows us to define two approaches to the same problem: explaining how people make decisions. Second, although convergences can be found between the two roots—for example, Herbert Simon, one of the fathers of the behavioral approach, is also deeply immersed in the formal root—studies that focus exclusively on either the first or the second root are different indeed. On average, mathematicians, statisticians, engineers, cognitive scientists, and operations and production management scholars approach decision making from a formal perspective. Organizational behavior scholars and psychologists study the behavioral side of these processes, on average. Economists are divided among the two: They are agreeable enough to find good company everywhere! The fusion of these approaches brings new subfields, new research agendas, and new ideas. For example, neuroeconomics and social decision making are two subfields that have the potential to change the ideas we have of decision making one more time. This research allows us to think about economic processes in new ways. We have the opportunity to found the theory on observed neural mechanisms, having a more precise idea of what determines a specific feeling, for example, that leads to, stays together with, or comes after a certain decision. This gives scholars the potential to have more accurate descriptions and forecasts on how thinking and behavior are actually linked. Although very promising and interesting, this book is not about neuroeconomics. In recent years, the two perspectives have contributed greatly to the development of our understanding of decision making. However, they tend to focus too much on the way the respective frameworks support or reject specific hypotheses on human behavior and/or theoretical analysis. There is nothing wrong about this. Quite the contrary: This is the way science improves! However, in the last decades, there has
4
1
Introduction
been a lack of foundational questions. I refer to those questions that, for example, lead Herbert Simon to define the concept of bounded rationality. Who questions this concept today without falling into either the behavioral or the formal paradigm? It is far from the purpose of this book to pretend to present a “new” theory that has an impact similar to those of the past. All I can do is to ask questions. And this is what I will be doing in this book. In so doing, I present a different approach to decision making that integrates cognition and pro-social behavior (including neuroeconomics) to bounded rationality. The book presents bounded rationality and analyzes its limitations. The objective of this work is to show that, if we focus on bounds, we have a very limited and inaccurate representation of human rationality. If we think of great achievements that individuals reached in the past, what comes to our minds first? What intrigues us? It is a legend that when Gödel demonstrated his incompleteness theoremvi , he started writing in silence on the board during one of the most important math conferences of his time. When he was done showing his colleagues what was about to change the history of mathematics, the tribute of the hall was a very intense applause. Creativity and originality come together with Gödel’s theorem and the way he delivered it to his colleagues. I guess we do not think of his bounds and how limited his rationality had been in what he was able to achieve. What makes Gödel’s work so important and what allowed him to prove a theorem that nobody else had been able to prove before is based on his ability to think outside the box. Gödel’s brilliant demonstration comes from the ability to challenge one’s own limits, and to go beyond them. It is not about bounds, it is about what makes an excellent mind overcome its limitations. Overcoming one’s own limitations is not typical of excellent minds only. If you think about it for a second, you know that everyone of us goes past his or her limitations. How? Take a simple task such as to write a 1,000-word review of a book you have just finished reading (let’s say Extendable Rationality). Writing is a typical activity that requires external support. A personal computer is something that could help you with proper software that has a template that fits your needs, with help functions showing you how many words you have written, and with a vocabulary that supports your writing. Depending on the software you are using, you can activate a grammar and syntax assistant, a function that signals if the sentence you are typing is correct in standard American or British English. If somebody is around, you can also ask for comments and opinions on what you are about to write. What is this about? This is a way that you enrich and expand your rationality in order to reach a better outcome. This process is that of exploiting external resources, and it is what increases our chances of being rational. We overcome our limitations when we exploit external resources. What if there is no real divide between internal and external resources, between our brain and the tools and artifacts that we use? This book explores exactly this hypothesis, showing that if we remove this divide we can explain a significant part of human behavior (and thinking) and redefine rationality.
Book Structure
5
Book Structure I do not pretend to offer a shared interpretation of the concept of bounded rationality, nor do I think that what the reader finds in this text is close to what other scholars are thinking. I have found very limited evidence that other scientists are thinking about the nature of bounded rationality. The most common sporting activity is that of taking it for granted, on the basis of an ipse dixit of medieval flavor. I strongly believe that this is a luxury that scientists cannot be allowed to have and, for this reason, I support the idea that any critique and attempt to better understand limits as well as potentials of any theory are extremely welcome. I was particularly surprised that critics of bounded rationality only came from economists who were in love with neoclassical economics. Few behavioral scholars think that the understanding of its limitations is what makes the theory stronger. This work presents my interpretation of bounded rationality and some of the problems of this theory. However, my criticism is not designed to take the field of decision making backwards (i.e., neoclassic theory), but instead looks at the future. This doesn’t mean that I expect to be right with what you are about to read, but I hope that a debate on the meanings and perspectives of bounded rationality could start. The ideas presented in this book are an attempt to improve what defines bounded rationality. The book is divided into two parts. The first part is The Limited Cognition, from Chapters 2 to 5, dedicated to the analysis of bounded rationality. The second, The Extended Brain, Chapters 6 to 9, presents the idea of extendable rationality. Chapter 10 offers a summary of the second part and includes implications, a research agenda, and concluding remarks. What a rational process is and how it is connected to rationality is the topic of Chapter 2. The starting point is that of defining decision making on the basis of three types of decisions: (1) mechanical, (2) decisions that imply a choice, and (3) creative. In short, the question that I try to answer in this chapter is “How do we make sense of our decisions?” Individual processes of making sense of one’s own behavior and thinking start from there, and rationality can be analyzed as a distinct, but associated, concept. The thesis that will be presented is that the traditional idea of rationality is a good fit for mechanical decisions and for those that imply a choice, but creative decisions are less capable of being explained that way. In Chapter 3 I introduce the idea of bounded rationality in a classic fashion. The chapter answers the question: “What is bounded rationality, and why is it so important?” Concepts such as procedural and substantive rationality, satisficing versus maximizing, and the nature of rational bounds are presented and discussed in the chapter. Chapters 4 and 5 present an overview of selected studies that map bounded rationality. Starting with the notorious prospect theory, I present some of the most typical analyses, including biases, heuristics, representation, accessibility, and framing processes.
6
1
Introduction
Chapter 6, provocatively entitled Simon’s Error, introduces the reader to the concept of “distributed cognition”, and presents critiques of bounded rationality based on stylized and simplistic assumptions known as the “cognitive divide.” The question that I try to answer there is “What are the limits of bounded rationality?” The next two Chapters (7 and 8) then explore the idea of unstable bounds of rationality, and there I present a different perspective on decision making. They are based on the question “How can we move forward?” Chapter 9 continues with a concept that is close to both distributed cognition and extendable rationality, and puts these two into play when we analyze organizations. It is the concept of docility that is an answer to “Is there a theory of human behavior and extendable rationality?” Some chapters present an “epistemological corner.” This is a cognitive perspective on some of the problems I mention in the book, and helps the analysis and understanding of the concepts presented here. It gives a multidisciplinary perspective from a field of study, that of the philosophy of science, where some of the topics under analysis have a long tradition of study. The decision to skip and not read the “epistemological corner” does not hinder the ability to understand the book’s theme and follow its line of argument. If you are a student of bounded rationality, you may read directly the second part of this book with no prejudice to your understanding. Or, you can read the ending paragraph (summary) of each chapter that reviews materials there presented. However, I suggest starting from the beginning, because you may probably want to see what my starting points are and what my reading of bounded rationality is. Have a good trip.
Notes i. Simon (1979) provides a review of early rational choice approaches. See also Gilboa (2010). ii. This is true also for what concerns works on bounded rationality (Mousavi and Garrison, 1992; Patokorpi, 2008). iii. Zelený (2001), Mousavi and Garrison (1992), Patokorpi (2008) Peng (1992) Sent (1997, 2005) and Langley et al. (1995). iv. See Camerer (2007) for a survey of what is the recent evolution of behavioral economics, called neuroeconomics. v. The computer metaphor of the brain is analyzed and discussed in Patokorpi (2008). vi. An explanation to the theorem for non-mathematicians is offered in Nagel and Newman (1958).
Part I
The Limited Cognition
Chapter 2
Rationalization and Rationality
This is our starting point: Rationality. If you want to understand decision making, the first step is to define what a rational choice is. This has always been the first task for students of decision making.i However, it doesn’t have to be like that. Some can argue that decision making is not the same as rational decision making, and that many of our everyday decisions are not rational at all.ii Think of when you get into the grocery store to buy water and bread only. How could it be that every time you get out from the supermarket you find yourself with a seemingly infinite list of products? And, by the time you are home you realize that you didn’t need anything besides bread and water. Still, the next time you get into the store you will make the same mistake.iii We cannot say that buying unneeded products is a rational decision, but the process that is involved in making the decision has to deal with rationality. This is the link that is explored in this chapter.
Kinds of Decisions Every day, every one of us makes a number of decisions. These decisions are not always the same; they vary depending on circumstances, importance, events, involvements, and many other factors. The first step is that of making a distinction between different types of decisions. It is important to discern that not all decisions are equal, and that some of them differ from the others in terms of efforts of the decision maker and effects deriving from their implementation. In order to have a clear understanding of what I mean, we may find decisions that – are mechanical, automatic, or immediate; – imply a choice or are non-mechanical; – are “creative.”
Mechanical Decisions The first kind of decisions is those that happen “without thinking.”iv These are widely diffused in our everyday behavior and often we take them for granted. We D. Secchi, Extendable Rationality, Organizational Change and Innovation, C Springer Science+Business Media, LLC 2011 DOI 10.1007/978-1-4419-7542-3_2,
9
10
2 Rationalization and Rationality
never think about these decisions in a proper and specific way because they come out as usual. For example, I am pretty sure that you never make the decision to put some clothes on before going to school, work, university, or wherever you want to go. You assume that you have to dress and you do that. However, you don’t actually make the conscious, deliberate decision to get dressed, you just do it. This is a mechanical decision, where you don’t go through a decision-making process that is well-defined and distinct. In the example, the decision to get dressed has roots that go to a period of your life that is undefined and that you don’t remember. The thing is you continue to practise the way you learned: You get dressed. Other examples of mechanical decisions could be related to everything you do without any deliberate decision-making process, such as eating, the way you tip, how you write (or don’t write) the first line of your e-mails, what you say to people when you meet them, things you do everyday when you have breakfast. From a personal point of view, these are all examples of habits.v This doesn’t mean that there is no process here, it means only that the decision-making process is embedded in the way you behave or think. We have many examples of mechanical decisions in organizations too. A typical case could be that of sending invoices, or that of sending monthly emails with earning statements to employees. Everything that could be defined as an organizational routine falls into a mechanical decision-making process.vi The way to identify it is the same that we used at the individual level, but it has more complicated implications. A routine is something that deals with implicit knowledge and that is done on the basis of what has been done in the past and written rules. It is a convenient and efficient way to make quick decisions. Sometimes it is in the interest of the company to break the routine, especially when it emerges in areas where it is not needed. For example, routines are very welcome in call centers. Here, people don’t need to think too much,vii they need to provide information/solutions to customers according to corporate guidelines. Nothing outside the guidelines is admitted. In other terms, the routine (and standard operating procedures) is what makes the call center work. Other departments of the company suffer when too many routines are set up and when people don’t step out of the so-called ordinary business. Research and development teams as well as marketing departments are usually well aware that too many routines are not good ways to achieve excellent performances. This is to say that companies usually tend to avoid mechanical decision makingviii in key areas. Now, these are examples of mechanical decisions at both organizational and individual levels. In summary, a mechanical decision is a particular behavior or thinking attitude that people practice because they are used to it.
Decisions that Imply a Choice Decisions that imply a conscious, deliberate choice are of a completely different type. This means that we evaluate possible alternatives and then select one.
Kinds of Decisions
11
Evaluations that we bring in might range from trivial, or fairly simple to very complicated or complex. We may face decisions that imply simple alternatives. To stay with the example above, while the decision to get dressed is mechanical, that of how to get dressed, i.e., what clothes to put on, is not. The latter decision implies a choice. For most of us this is a very simple choice; it may become more complicated if we get into the domain of color matching (e.g., socks with belt, pants with hat)—where my fiancée Claudia excels and I make miserable mistakes. The difference between the two types of decisions is not related to alternatives, though. It is not that when we are involved in mechanical decisions we don’t have alternatives, or while in decisions that imply a choice we do. The point is that we don’t see alternatives in the first case while in the second the decision is based on alternatives. The choice implies that we select among different options; hence, the existence of these options is not a necessary condition for the choice to be made while it is a sufficient condition. In other words, it is the individual that defines the kind of decision, not vice versa. It is not that certain types of circumstances lead you to make a mechanical rather than a choice-based decision. It is your actual involvement in the process that defines what decision you are making. If you are involved in selecting, weighing, and choosing among alternatives, then you are in the second type (i.e., the decision implies a choice); if you don’t see alternatives and you just do what you are about to do without further thinking, that is a first type decision (i.e., mechanical). There is continuity between the two types of decisions. We can switch from the first to the second or vice versa as needed. And the point is that we need to do that, on some occasions. Let us consider a few examples. What is a decision that implies a choice for businesses? Goal-setting activities and hiring/firing decisions are examples of non-mechanical decisions. Or, are they? We all like to think of them as something that implies a choice, but the passage between mechanical and non-mechanical decision making is very weak.ix Goals may be set up in ways that replicate the past or imitate competitorsx , and firing may be an automatic expulsion of those with poor performance (see GE when Welch was CEO).xi Other examples may include scheduling and budgeting activities where previous ones may be simply “carried over” into the next period (thus treated as mechanical decisions) rather then examined carefully (thus treated as decisions that imply choices). Of course, you have never heard of any academic that is willing to support these activities as mechanical; and this is especially true when we are dealing with human resource management. However, this is what happens sometimes in organizations. Now, if we want to deal with it, I believe that the continuity rather than the opposition of concepts better serves the scope of defining how decisions are taken and can change. Therefore, we can assume that it is the power of individuals, groups, and organizations that take into consideration alternatives when making decisions. Again, if they are considering those alternatives, we have a decision that implies a choice; where these alternatives fall outside of the person’s will, then we are facing a mechanical decision.
12
2 Rationalization and Rationality
Creative Decisions A particular kind of decision is that related to the creation of something. Creativity may be defined in many waysxii , and I don’t think the present discourse could benefit from entering this vast domain of knowledge. For the purpose of this text, creativity is an activity (a decision) that brings something new, something that was not experienced before. This third type of decision is close to the one mentioned in the previous section (e.g., choice-based); however, it is somehow different from that. What is a creative decision? It may be (a) the ability to find a new pattern to make a decision, or (b) the substantial newness of the decision. The latter relates to the outcome of the decision, while the former is about the process. Although a creative process may lead to a creative decision, this is not automatic and we cannot take it for granted. Unusual or highly creative decision processes can also lead to a decision that is ordinary (I have written more on this distinction in the following pages). There are several examples of what creativity can be. This ranges from a simple rereading of data you already know to find new insights, to a decision based on a newly generated and unexplored set of options. Creativity in decision making is the ability to generate alternatives that serve as a basis for your choice. The more you generate, the wider the basis for your choice. The better you generate, the greater your chances of success. It is not a matter of quantity, nor is it a matter of quality only. It is a general attitude that brings together quality and quantity. Creative minds are not limited to a single outcome of their creativity; quite the contrary. If you look at highly creative minds you find that these persons are also very prolific. How many books does a creative writer publish? Take Simenon, Christie, or Hemingway as examples. Not everyone is like Simenon, but we can see that the wave of creativity is not limited to one single shot. It might be, but if we experience that wave once, we tend to reproduce it for personal and group satisfaction and fulfillment. It enhances the chances of success.
Epistemological Corner I would like to make a very quick point on the processes that relate to each one of the three types of decisions. What is the logic behind mechanical, non-mechanical, and creative decisions? What are the processes that the mind carries on when we are involved in one of the three processes? Are these processes the same, or do they differ depending on the type of decision? Each one of the types described above relates to a different process. What follows here are speculations about what it could be, not on what it surely is. The first type, mechanical, leans on previous decisions. It is a replication of something that has been generated in the past and that continues to be. I should say that what happens is very close to induction, i.e., the ability to generalize starting from single events. In our case, the generalization is the fact that we tend to repeat the
The Legacy of Herbert Simon
13
same decision (e.g., we get dressed) because we always did in the past. Or, we deduce that this is the right thing to do or think. Put differently, the “universal rule” emerges from repetitions of single actions.xiii Individuals extrapolate the norm from repeated actions so that the point of reference becomes the norm, not the action. When the norm is followed, the decision is mechanical. In the example, to get dressed after waking from sleep and before going to work is a norm for the individual decision making that has become such because of social habits and repeated behavior over time. In the second type, the non-mechanical, we have a set of alternatives among which a choice is made. We analyze these alternatives and try to deduce what is the best choice to make. Therefore, and contrary to the first type of decisions, the major process involved in decision making is deduction. It is the way to infer something from assumptions using a logical analysis. This logic is close to what most studies on decision making follow. As already stated in the previous chapter, decision making has a significant computational and formal root that makes significant use of deductive logic to frame decisions. Creativity needs something different from both induction and deduction. The process of getting something innovative has induction, since it gathers conclusions from experience (i.e., generalizes through the collection of single events), and it has deduction, since there is a logic that arrives at conclusions from given assumptions. However, we should say that this description of creativity is poor and that we need something more sophisticated. We need a process that is able to describe how some general principle could emerge from assumptions that could not be directly and immediately related to it. This is what is called abduction, and it is a process often used by philosophers of sciencexiv to describe creativity. Since deduction and induction are part of the general scientific vocabulary while abduction is not, I think it needs further explanation. The term abduction was coined by Charles S. Peirce “to describe inference that involves generating and evaluating explanatory hypotheses.”xv For example, take the manager that explains the low response rate to the survey on the quality of the working environment due to its online delivery. This is not a deduction, rather an intuition, for many other variables could have affected the low response rate (e.g., insufficient efforts toward explaining its value, workers’ beliefs that the survey’s findings will have no impact, workers are dealing with a period of increased productivity and extended working hours, and more). The cause (hypothesis) that the manager finds appropriate (evaluates) is generated through a creative effort. As shown in later chapters, the importance of abduction is not limited to creative decisions only, but it has the potential to provide significant insights on the other types of decisions. For now, this introduction is sufficient.
The Legacy of Herbert Simon If we look at the past century in search for somebody who well represents the decision-making field, that person is Herbert A. Simon. I have always been fascinated by the studies and the legacy of this man, that I consider a sort of modern
14
2 Rationalization and Rationality
Leonardo da Vinci. It is very difficult to define his studies as confined to psychology, decision making, rationality, artificial intelligence, cognition, social psychology, sociology, organizational behavior, or economics. He made profound and insightful contributions to all of these fields. I stop here with the acknowledgment of his work because the very best way to get an idea of what I mean is to read one of his writings. Simon’s starting point has been that of criticizing the neoclassical model of rationalityxvi and of decision making. This model is used in economic models to describe and predict human behavior on the assumptions that, when making decisions, individuals have full access to information and their cognition works as a perfect computational device. Arguments on whether the neoclassical model is useful or not and to what extent we should switch to a bounded rationality model started more than half century ago and, believe it or not, the debate is still alive. However, we don’t need to summarize it here; for the economy of our discourse, we will recall concepts and ideas as needed. One of Simon’s major contributions has been on the understanding and analysis of rationality. We start from a classic point, that of “rationalization.”xvii
Rationalization Suppose you are asked the following question: How do you go to work? The answer may vary between car, bike, foot, tram, bus, metro, train, airplane, helicopter (maybe the last two apply only if you are a Stanford or Harvard student). Whatever means of transportation you use, most of the time you don’t make an active choice, but you just use the means that you always use. If you have a monthly ticket for the metro system in your city, you don’t make the decision to take the train since this is a typical mechanical decision-making process. Now, what if I ask you why do you take the metro-train? Here you can recall to your memory the reason why you do so. Here too, the hypothetical answers may vary between, for example, (a) it is the fastest way to get there, (b) I haven’t a car, and this is the only way for me to get to the university or to work, (c) it is the cheapest means of transportation, (d) it is where I always meet my sweetheart, (e) and the like. What are you doing when providing such answers? You are explaining the reason(s) why you take the metro-train, or you are offering sound reasons that support your choice. This attitude is widely diffused, and we use it every time it is needed, no matter if somebody else asks about our behavior or if it is ourselves asking for reasons. We use motives to put a rational emphasis on our behavior.xviii We explain our behavior (and choices, and way of thinking) through what is called rationalization. Rationalizing one’s behavior doesn’t mean that you put efforts to always find reasons that make sense. It means that your ability to think about your thoughts becomes real and useful. Rationalization is a mental process that is driven by selfawareness,xix something that lets you think about what you are doing. For example,
The Legacy of Herbert Simon
15
suppose that you provide a different answer to the question above on why you use a specific means of transportation. You go to the university by car but prefer walking, and it isn’t that you are that far from the university, it is that you are lazy. Moreover, you don’t like being lazy at all! Now, what is the rationale for taking the car? Apparently, according to this answer, there is no rationality in taking the car, but still the process of analyzing, explaining, and providing reasons for your behavior falls under the rationalization phenomenon. Once again, rationalization is the action of making sense of what you do or think, and it happens through your own interpretation categories. The last part of the sentence is particularly important. The fact that you or other people ask you to rationalize is important but not relevant in the argument I want to make. It is you that need to make sense of what you do. Of course, you can do this through social constructs or through your personal beliefs (when does the first end and the second start is a question that we address below) but the most important point is that you do take time to make sense of your behavior. When individuals fail to rationalize or find out that their rationalization is inaccurate, then a feeling of discomfort and/or distress emerges.xx There are several experiments and studies that point out how this mechanism works. Imagine that you have to rank 10 different music tracks.xxi Soon after that, you are asked to explain why you ranked tracks #5 and #6 in their respective order. To explain why #5 is better than #6, you may over rate it. Otherwise stated, you are trying to make sense of your choices through a rationalization process. What happens if the two alternatives are equally likable to you? This is the case when you listened to that music for the first time. It may happen that you do not have a preference and do not significantly prefer #5 over #6. Your difficulty (distress and physiological arousal) in finding a sound reason for your choice is called cognitive dissonance.xxii Put another way, “regardless of the significance of the decisions, people faced with equally attractive alternatives tend to experience cognitive dissonance and justify their decisions.”xxiii This is exactly the case of the example for most people: You have to make up an explanation to convince (yourself) and the person asking the question that your choice makes sense. Although very interesting and important for both our psychology and cognition, there is no point in continuing to analyze dissonance here. The only purpose of introducing cognitive dissonance is to highlight the fact that rationalization is a matter of utmost importance. In our everyday reasoning, when we (a) fail to rationalize, (b) realize that our rationalization is not consistent with the choices we made or that we will make, or (c) rationalization is poor for us or for other people important to us, then we experience distress and discomfort (i.e., cognitive dissonance). The tendency to see oneself as a rational individual that makes rational choices is an important part of the way people think of themselves.xxiv Another way to think of rationalization may be that of defining it as the tendency to avoid dissonance by activating sense-making processes. It is now apparent that rationalization is one of the major activities that comes together with decision making. We cannot analyze our decision-making attitudes or our choices if we cannot understand the underlying process that allows us to do so.
16
2 Rationalization and Rationality
Rationality Now that we have dealt with rationalization, we can try to understand what rationality is. Rationality is brought into the decision-making discourse by the process of rationalization, and it has always been a study of how good the decision (or the underlying process) has been or should be. While rationalization defines thinking activity in general, rationality defines its contents (both goals and procedures). To use Simon’s words, rationality “is concerned with the selection of preferred behavior alternatives in terms of some system of values whereby the consequences of behavior can be evaluated.”xxv This is a very classical definition of rationality, and many scholars don’t use it anymore. However, it is a good starting point from where we can build, modify, and add. For a better understanding of this concept of rationality, we can divide the definition into four parts. Rationality means 1. 2. 3. 4.
the selection of alternatives through a system of values (i.e., weights or choice drivers) that allows individual to make decisions and to make evaluations on potential and actual consequences of behavior (or actions).
According to Simon, rationality is what allows individuals to make decisions out of a pool of alternatives that are consistent with one’s values. We can also define a rational decision when the process of rationalization reveals a system that is structured according to the definition. For example, the answer of the lazy person—i.e., the one who prefers to use the car even when recognizing that this is not the best choice for him/her—is not rational. The choice in that case is not consistent with the evaluation of alternatives. There, the “walking” alternative has more expressed value than the “driving” alternative; however the former has not been chosen. Again, the selection of alternatives does not follow the system of values and suffices to define the decision as irrational. Or, we can find a different explanation for the example and make it rational. We can, for example, imagine that what that person is telling us is only a partial truth. In reality, there is a trade-off between the feeling of not being lazy and the comfort of getting to the university by car. In this case, the weighting function (i.e., the values) is different and, since the comfort is superior to the laziness feeling, the choice is consistent with the system of values. It becomes apparent from this simple example that the most important part of the definition is point four, the “evaluations on potential and actual consequences of behavior.” Put differently, the analysis of consequences that derive from a specific action. This last idea brings into the decision-making computation the future or expected impact of our behavior. There are many limitations in this definition of rationality. It is what we should call a definition of the process through which we actually make decisions. However, what kind of decisions? Are all three types of decisions—mechanical, choice-based,
Notes
17
and creative—well-represented by this idea of rationality? When we select alternatives through a system of values that allows us to try to forecast future consequences, we are making a choice. Apparently this definition of rationality supports the second type of decisions; those decisions are rational if they follow the process as described in the definition. Mechanical decisions never follow explicitly a process similar to that of the definition. They can be analyzed (rationalized) ex post, and a rational explanation of behavior can be added to them. Creative decisions are the most difficult to define in relation to this basic concept of rationality. The whole point here is that our decisions are rational if we follow this well-defined pattern, and that there are certain types of decisions that are more likely to be so. Those that imply a conscious mental activity directed toward a decision-making effort have more chances to fall under the rationality cap. But we are struggling between two different ideas of rationality: One looks at the process, the other at the goal. Is a decision rational when we find its rationality after it has been made? The next chapter addresses this and other points.
Summary In this chapter, we have learned of three types of decisions: (1) mechanical, (2) choice-based, and (3) creative. Alternatives play a crucial role in these three as they are overlooked in the first-type decisions, examined in the second, and newly generated in the third. We have also associated mental processes to each of these decisions (respectively, inductive, deductive, and abductive reasoning). After that we started with the analysis of how people make sense of their decisions (i.e., behaviors and thoughts) and explained that this is what is called rationalization. A decision could be rational or not, but individuals can always rationalize it. The chapter ends defining rationality as the selection of alternatives, filtered by each one’s values, that allows the individual to make the decision on the basis of the evaluation of potential outcomes associated with that decision.
Notes i. The work of H.A. Simon (1947, 1955) affected decision sciences greatly, and this may be considered an outcome of his work. However, I suspect that this link between rationality and decision making is more than old. Philosophers track down to Plato the first connection between decision making and rationality, and to Descartes for what concerns a modern approach to rationality. ii. This is the point that many behavioral economists make crystal clear; for example, see Ariely (2008). iii. This behavior is studied by marketing scholars and it is not something where I cannot be said to have any sort of expertise. However, an interesting reading can be Dickson and Sawyer (1990), in relation to the (secondary) role of price for grocery store shoppers. iv. The literature on decision making prefers the word “intuition” to define these decisions. A distinction between intuitive and controlled mode is offered by Daniel Kahneman (2003).
18
2 Rationalization and Rationality
v. This is what Verplanken et al. (2005) study in their paper on the measurement of habits. According to these authors, “given the prevalence of repeated over new behavior, there is good reason to pay more systematic attention to constructs like past behavior, repetitive choices, experience, routines, and habit” (p. 231). I agree with them and believe that mechanical decision making is a significant part of human behavior and that more attention should be directed to it by scientists of all disciplines. Unfortunately, it has been largely overlooked and has had only occasional scientific investigation. It is not in the economy of this book to explore habits or mechanical decision making in particular although I will refer often to this type of decisions. vi. Whether it is called “tacit knowledge” (Polanyi, 1966) or routines (Nelson and Winter, 1973). Although the notion of a tacit knowledge has potential to lead to creativity when made explicit, here we consider its connection to routines, as analyzed by Nelson and Winter (1973) (see also Foss, 2003a). vii. At least, this is what managers think is an effective way to conduct that business. As I show later in the book, excessive limitations on human discretional decision making is often not counterproductive for organizations. viii. There are several types of routines. In particular, some of them are “policies or actions that prevent the organization from experiencing pain or threat and simultaneously prevent learning how to correct the causes of the threat in the first place” (Argyris, 1986, p. 541). These are called defensive routines; organizations are not immune from creating these sorts of pain-killers but tend to avoid them when they become apparent obstacles to managerial goals. ix. Betsch and Haberstroh (2005), analyze routines in decision making. x. This is exactly what studies on bandwagons in innovation diffusion show (Abrahamson and Rosenkopf, 1993). xi. David Olive (2001) explains how Welch represents the bright and the “dark side of an era.” With his job-cutting strategy, he forced workers to better perform through the fear of losing their job more than commitment to the same cause, goals, values, or any other thing that motivational theories should suggest. It seems that Welch himself recognizes now—when it is too late—that “it was ‘a dumb idea’ for executives to focus so heavily on quarterly profits and share price gains” (Guerrera, 2009). xii. See Sternberg (Ed.), 1999. xiii. I am borrowing the definition of induction from Popper’s Logic of Scientific Discovery (1935/2002). xiv. Introduced by Peirce (1955), abduction has been put at the center of philosophic and scientific inquiries (see Magnani, 2001). xv. Magnani (2007, p. 224). xvi. This is what can be found in Von Neumann and Morgenstern (1944), Friedman (1953) and summarized by Wallister (2008, pp. 52–54). xvii. Rationalization is the process that leads people to “explain their own actions in terms of their alternatives and the consequences of those alternatives for their preferences. Similarly, they explain the actions of others by imagining a set of expectations and preferences that would make the action rational” (March, 1994, p. 3). xviii. This is what can be found in Festinger (1957); see also Kunda (1999, p. 216f). xix. Self-affirmation is also a by-product of this mental process; Kunda (1999, p. 220f). xx. Zanna and Cooper (1974). xxi. This is a variation of Heine and Lehman’s experiment (1997). xxii. Festinger (1957) has been the first to point out the importance and implications of cognitive dissonance. After his seminal work, a significant amount of studies has been conducted (see Kunda, 1999, Chapters 6 and 11). xxiii. Hoshino-Browne et al. (2005, p. 294). xxiv. As Heine and Lehman (1997) show there is a cultural difference in the way people experience cultural dissonance and use rationalization processes as a reduction mechanism. xxv. Simon (1997, p. 84).
Chapter 3
Bounded Rationality
The study of rationality has always dealt with mathematics. The analysis of and the solution to a problem have always been connected to the logical ability of the individual so that tests like IQs and GMATs or GREs are supposed to predict how successful a person will be in the real world.i Unfortunately, all of these measures fall short of such predictionsii and raise interesting questions on the limits of individual rationality. To make a long story short, one of the most interesting debates in decision making is the one that relates to why and how rationality is limited. I postpone discussing another basic question, if rationality is limited, until the last chapters of this book.iii An interesting idea on rationality has been introduced by Herbert Simon in 1947 and then again in 1955iv . It is the idea that rationality is bounded. This chapter is dedicated to the analysis of bounded rationality (BR), according to its father.
What Is Bounded Rationality The most important strength of the BR modelv is that it contributed immensely to understanding and overcoming the limits of the theory of a fully rational or unboundedly rational individual. In short, this theory assumes that individuals have perfect computational capabilities and make decisions on the basis of information that is complete, available, and unambiguous. This idea was (and still is) widespread in economics, finance, and in certain management domainsvi (e.g., strategy). Problems with this theory relate to the fact that theoretical models lacked empirical validationvii , especially when tested at the individual level. This is probably due to the nature of the assumptions and to the historical period in which they emerged. Models of the fully rational individual were theoretical first, and only years later have they been empirically tested.viii Students of bounded rationality, on the contrary, maintain that behavior and observation of behavior come first. They tend to define theory on the basis of observed evidence instead of offering the theory firstix . For this reason, the approach and the outcome (i.e., bounded rationality) come as descriptions of how people actually behave. D. Secchi, Extendable Rationality, Organizational Change and Innovation, C Springer Science+Business Media, LLC 2011 DOI 10.1007/978-1-4419-7542-3_3,
19
20
3 Bounded Rationality
This distinction may seem close to the older one, based on descriptive and normative theories. The full-rationality theories explain how human behavior should be, while the theories of bounded rationality describe the world as it isx . The question is what behavior happens to be prescribed in a world of fully rational individuals. And, has this prescription any relation at all to a world of boundedly rational agents? We see in the following that the world of boundedly rational agents has its own rules that could prescribe as well as describe individual behavior. Nevertheless, these points are hardly definitive. There is an interesting debate on human rationality that still goes on, and what you read here is my take on it. I believe that there is enough evidence to support the statement that a completely rational individual has no room in our world, not a normative or a descriptive role. Any theory of rationality and decision making needs to start from something close to actual behavior and thinking. The theory of bounded rationality describes with reasonable accuracy how individuals make decisions in the “real world” rather than in some idealized condition. And this is the reason why it still offers an excellent starting point to any study on decision making. Therefore, there is no treatment of neoclassical theories in this book. The following pages are dedicated to a description of some of BR’s most significant contributions.
Substantive and Procedural Rationality Bounded rationality is concerned with the process, first and foremost, and with the organization of resources. Assuming that the goal of the firm is profit maximization—it cannot be, according to BR, but this is an easy way to start—there are multiple ways to get to the same point. We don’t have an optimal configuration of resources here; instead we have different patterns that lead to a cluster of similar goals. This means, basically, that the same profit goal can be achieved through operational cost cuts, layoffs, supply chain optimization, improved quality control, outsourcing, tax shields, increased sales, and more. The world of bounded rationality is more complex than that of full rationality; it approximates the real world. Of course, when the process and the outcome are separated, we may have multiple combinations then. Table 3.1 presents these possibilities. Only cases one and three are consistent with the full-rationality theory: The use of resources leads to a result that is perfectly in line with it. Table 3.1 A combination of processes and outcomes
1
Process Rational
Outcome Rational
2
Rational
Irrational
3
Irrational
Irrational
4
Irrational
Rational
What Is Bounded Rationality
21
Broadly speaking, all results may be considered consistent with a broader idea of rationality. The most rational use of resources can relate to an irrational outcome. I realize that this is an understatement, but the process through which the Inquisition executed thousands of women (witch-hunt) over the past centuries follows a rational process that shows consistency between alternatives, values, and evaluation of consequences. However, we cannot see any rationality in the outcome of killing individuals based on prejudices, no matter if we are God or Goddess believers or not. On the other hand, irrational processes may lead to rational outcomes. The process though which a company installs and promotes the use of a new software among employees may be sloppy and inefficient. The outcome may be that of improving internal information processing anyway. Another example may be that of considering emotional choices or decisions we make when short of time. Even though we skip some of the logical passages we should follow, decisions are rational. In discussing decision-making processes, Simon placed great emphasis on the distinction between substantive and procedural rationalityxi . He described that difference stating, “We must give an account not only of substantive rationality—the extent to which appropriate courses of action are chosen—but also procedural rationality—the effectiveness, in light of human cognitive powers and limitations, of the procedures used to choose actions.”xii According to substantive rationality, the rational character of decision making is concerned with the result one could get following the “appropriate” actions, whereas procedural rationality points out the process by which people make decisions. According to Simon, bounded rationality belongs to the latter category because it does not look only at the result one could get, but at the way people make decisions.xiii The definition of rationality presented and discussed in the first chapter is exactly of this kind. The difference between substantial and procedural rationality is fundamental since it gives to the domain the same width that efficiency and effectiveness give to management. In a fully rational model these two aspects are mixed up, since there is no possibility that the same goal (substantial r.) could be obtained from an alternative use of resources (procedural r.). Otherwise stated, in a perfect world the process leads to the only possible optimal outcome. If the objective of a firm is that of profit maximization, this means that resources are allocated according to this goal. There are no different configurations of resources since the only possible outcome is the optimal level of profits for that firm. Therefore it doesn’t matter what is inside the firm because the rational exploitation of resources leads to the optimal result. Here is where the idea of the firm as a black box comes from. Table 3.1 presents only two levels of analysis—(1) process, (2) outcome—and I have voluntarily left unclear the relation between these two, and (3) goals. As far as the neoclassical model is concerned, there is no difference between outcomes and goals. The former is what results from the process, while the latter is something one wants to achieve, or the aim at which action is oriented. When there is one best solution, the way resources are organized lead to the outcome that is the goal. In the case where process and outcome vary, as in the case of BR, the outcome may be far
22
3 Bounded Rationality
away from the goal. And this gap is unavoidable, due to how human beings process information. What I am trying to define relates to the nature of bounds since “rationality is bounded when it falls short of omniscience. And the failures of omniscience are largely failures of knowing all the alternatives, uncertainty about relevant exogenous events, and inability to calculate consequences.”xiv In short, we can argue that there are external (informational) and internal (computational) limitations.
External and Internal Limitations Our knowledge of the external world is not perfect. We don’t know all possible alternatives to a given problem. This is the reason why, for example, in the problem with corporate profits as mentioned above, there are multiple ways to get similar results. Uncertainty is the word that describes how the external world affects our attitude to make sound decisions. It is very unlikely that managers and top executives spend their time collecting information on all possible variables that could potentially affect decisions and their consequences. It is more likely that these individuals get what they consider the most relevant information and then make a decision in a timely fashion. There, a second and internal limitation occurs: computation. This second is the limit to compute large amounts of data. That is to say that even if all possible data and variables related to a given problem were known, we shouldn’t be able to use them. This is because of the cognitive internal bounds that characterize our rationality. Take the case of competition. Suppose that you are looking for a convenient home insurance. You start surfing the web in search for information, then you start visiting companies. Imagine that the market is not perfect (this is easy even if your imagination is not that vivid!), so disequilibria occur, there are frictions and transaction costs. What is the difference if you get information from 50 or 100 companies?xv Unless you do the market price tester as a full-time job, you don’t have the skills or the time to handle all of this information. You would probably give up much sooner than needed. I believe that even price collectors and analysts need particularly sophisticated tools to compare market prices. Tools and skills that are not available to everybody. On average, for the individual it doesn’t make any difference if the information out there is complete and available because he/she cannot handle it.
Satisficing Internal and external limits define bounded rationality in terms of the procedure that is the most appropriate to make a decision, should a problem be given. This procedure is shaped by the limits to get the knowledge that individuals show in terms of cognition and access to information. The result is that people don’t get the
What Is Bounded Rationality
23
optimal result but only suboptimal ones. Therefore they do not maximize but obtain results that are only satisficing. There are important differences here from the traditional economic model of rationality and decision making. First, the result is not one but multiple outcomes to any given problem. Going back to the example of profits, this is quite understandable when we deal with a single goal since we can get to that point through many different pathways: There are so many ways to get to a given amount of profit. However, what a company gets is never the best result possible, the maximum amount of profit possible. There are multiple levels of analysis here, though. Let’s try to take a clearer approach to the problem. In relation to profits, our managers may have to deal with (a) defining the level of profit that is sustainable for the company (i.e., that is satisficing); (b) understanding the possible/viable (satisficing) alternatives between cost cutting, tax shielding, or sales incrementing strategies; and (c) making a choice that satisfies the company’s needs at best. The latter point is close to the maximum only by chance, since not all of the variables are known. As you can see, satisficing can be used in many ways when we deal with a problem. Some scholarsxvi suggest that satisficing can also be a search strategy. In this particular meaning, decision makers adjust their behavior according to whether the goal is getting closer or not. If the company’s managers set the goal to sell N units of the new product by the end of the sixth month of its introduction, they can increase their efforts toward the goal (e.g., ads, marketing, put pressure on retailers, TV commercials) if after the third month they sold only N/4 units. Otherwise, they can decrease their efforts if the units sold by the third month are more than N/2 units. This mechanism defines satisficing as a decision and search strategy. Broadly speaking, satisficing defines results and it could, on certain occasions, specify the process by which individuals achieve these results.
Bounds or Limits? Although Simon writes about “bounds” of rationality, the worldwide translation of this concept may indicate a different aspect of the problem. Translations in many languages use the term “limited” to describe the same idea as “bounds.” Here are a few examples: The French expression is rationalité limitée; the Italian is razionalità limitata; the German is begrenzten Rationalität; in Spanish, it is racionalidad limitada. Why do translations put such an emphasis on limits instead of bounds? Well, I don’t have the answer to this question, but I think that we should focus a little bit on the differences between the use of different words to describe the same idea/theory and try to understand which one of the two does a better job describing the theory. Some scholarsxvii studied the genesis of BR and highlighted the fact that the process through which Simon came to this terminology had been a refinement process. In fact, he passed through “approximate rationality” and “limited rationality” before settling for what became the label for his idea.
24
3 Bounded Rationality
James March, once very close to Simon and part of the so-called Carnegie schoolxviii , in his book A Primer on Decision Makingxix , uses the terms “limited rationality” to define the theory presented in this chapter. However, I don’t know if this is just a recognition of what is widely accepted outside the U.S. or if this is a readaptation to what the theory really means. As the inventor of this terminology puts it, You have to realize about the bounded rationality terminology that I began to use this as a label for the things that economists needed to pay attention to—and were not. It was never intended as a theory in any sense.xx
This came as a reading of what happened 40 years after the first time the words were used. It really seems that the choice of words has been made to state something different from what one should think, and it is not clear whether the two words— “bounds” versus “limits”— reflect any change in Simon’s original idea. The point is that the theory of bounded rationality has always been intended as a theory of limits and limited rationality. While bounds give the impression that something is tied up, restricted to a specific territory, limits are something stricter. Bounds can be modified while limits cannot. Think of the fact that we call boundaries those of a state, not limits; and we can eventually pass these boundaries. Limits give the impression of being stable; we have limits of a mathematical function, for example, and we know that these are fixed. Is this theory of rationality closer to limits or to bounds? This is what we will see in the coming pages.
Summary In this chapter we have defined bounded rationality in a classic fashion. The distinction between substantive and procedural rationality is useful to make the difference between goals (outcomes) and resources employed in the process, respectively. Individuals are boundedly rational in that (a) they have computational and analytical limitations (internal limits) and (b) they also cannot get all possible information from the outside environment (external limits). The combination of these limitations leads us to satisficing rather than maximizing decisions. Moreover, satisficing can be a useful search mechanism that proceeds through feedbacks that adjust the search toward the achievement of a given goal.
Notes i. In The Intelligence Controversy, Eysenck and Kamin (1981) exchange opinions on the nature versus nurture debate on intelligence. They discuss also intelligence tests and go over their history and limits. See also Richardson (2000). ii. A very interesting way to look at this is Gladwell’s Outliers (2008). iii. In his writing “Why bounded rationality?” Conlisk (1996) offers significant evidence of what are bounds of rationality.
Notes
25
iv. According to Klaes and Sent (2005) the first time the terms “bounded rationality” appeared were in Simon’s Models of Man, in 1957, page 198. “Herbert Simon, very likely the first person to have used ‘bounded rationality,’ tentatively applied a number of expressions that seemed available on the basis of associations with what he sought to express. In a series of writings between 1947 and 1957, he consciously refined and replaced concepts such as ‘approximate rationality’ and ‘limited rationality’ until settling for ‘bounded rationality’” (Klaes and Sent, 2005, p. 37). I refer to these attempts, since the meaning Simon intended to give to these expressions is the same; what changed was the label, not the content. v. I am not sure whether to call this model, theory or, more generally, approach to bounded rationality since there is no one theory nor there is one single model. However, the major assumptions of all theorization are clear and remain very similar for every scholar (Foss, 2003b). There are many models of bounded rationality (e.g., Rubinstein, 1998) but there is no single theory. Maybe, we can define it as a general approach to rationality that shares the belief that human beings have limits. vi. I recently read a very vivid debate on strategic management and objectives of the firm. Whether authors never mention or explicit any link with bounded rationality, I believe that Sundaram and Inkpen (2004) clearly refer to full rationality when they advocate for maximization of profits as a single-goal function for the firm while Freeman et al. (2004) implicitly refer to the limits of rationality when considering the stakeholder approach and multiple goals for the firm. vii. I will explain this later in this text; see Conlisk (1996). viii. The debate around Tversky and Kahneman (1974), reveals very little evidence of full rationality. ix. There are two streams of research in bounded rationality (BR). The old school, that of Simon (1955), started from the neoclassical theory of the fully rational homo economicus to build the theory of BR. The new school, that of neo-Simonians (e.g., Kahneman and Tversky, 1979; Gigerenzer and Selten, 2001), started from empirical tests on neoclassical theory to show limits in human rationality. x. In the words of Herbert Simon, theories of full rationality explain “how people ought to behave, not how they do behave” (Simon, 1959, p. 254). xi. See also Munier, Selten, et al. (1999, p. 234). xii. Simon (1978, p. 9); italics in the original text. xii. Simon (1978). xiv. Simon (1979, p. 502). xv. I use this argument in a paper (Secchi, 2009; unpublished paper) that criticizes the theory of perfect competition as Block et al. (2008) apply it to the banking and insurance industries. xvi. See for example March, 1994, Chapter 1. xvii. Klaes and Sent (2005, p. 37). See also Augier (2000). xviii. Augier (2004). xix. March (1994). xx. Simon (1999, p. 23); quotation retrieved in Klaes and Sent (2005).
Chapter 4
Maps of Bounded Rationality (I)
In the late 1970s, scholars started to study if and how bounded rationality works in practice. Experiments were set up and there started the ever growing literature on the subject. The first evidences were related to falsifying neoclassical approaches to rationality, e.g., the theory of expected utility (see below), while the latter efforts cover a wider terrain, focusing on heuristics, ethics, cooperation, and altruism (discussed in a later chapter). This chapter offers an overview of some of these studies. The objective is to show that the work on bounded rationality is well-alive among behavioral scholars, and that it has recently gained momentum. A general study of the effects of these theories on organizations has yet to come. I do not try to provide any organizational “transposition” of these theories in this chapter, but attempt to offer hints on how they can be utilized and what meanings we can get.
Prospect Theory With a series of experiments, Daniel Kahneman and Amos Tverskyi introduced a test of rationality directed at falsifying the assumptions of expected utility theory. This is the iconic neoclassical theory where individuals are supposed to make decisions on the basis of statistical calculations.ii Before getting into a foundational and theoretical analysis of their work, I believe that examples will serve the reader better. As Kahneman himself wrote, “[o]ur research attempted to obtain a map of bounded rationality, by exploring the systematic biases that separate the beliefs that people have and the choices they make from the optimal beliefs and choices assumed in rational-agent models.”iii Truth to be told, most of the constituents of this “map” of bounded rationality were introduced in some of their original pieces in 1974 and 1979. The topics reproduced below have been selected from the 1979 article.iv
The Certainty Effect Among behavioral studies, one of the most important developments of the last few decades has been prospect theory. This is the analysis of decisions under uncertainty D. Secchi, Extendable Rationality, Organizational Change and Innovation, C Springer Science+Business Media, LLC 2011 DOI 10.1007/978-1-4419-7542-3_4,
27
28
4 Maps of Bounded Rationality (I)
where people choose between contracts, where each alternative could materialize depending on its probability. For example, suppose you have an 80% chance to get $4,000 or that you can have $3,000 for sure (i.e., with probability 100%). Or, more formally stated, A: ($4,000, .80) or B: ($3,000, 1.00)
What do you choose? Well, I can tell you that most of the people choose to get $3,000 right away instead of betting on the higher sum of money. Now, consider the following two alternatives: C: ($4,000, .20) or D: ($3,000, .25)
Is your choice C or D? If you were a strictly rational (logical-mathematical) person you should have chosen D, the only choice that goes well along with your previous choice, i.e., B. If you were not using this mindset, mostly based on expected utility, you have probably contemplated C among your choices and gone for it! This is what the two scientists found in many of their experiments. People tend to prefer what is certain compared to what is not (this is why choice B is preferred to A) despite the fact that the expected value of the two choices differs. Choice A is worth $3,200 while choice B is only $3,000. The more risk-seeking individuals would have chosen A while more conservatives would have chosen B. The majority of people that took part in the experiments preferred B to A. However, this is not the most interesting part of the experiment. When presented with the second set of choices, where quantities stay the same and the relation between the two probabilities also remains the same, people are not consistent with that first choice. It is worth noting that 80/100 is the same as 20/25. On a strictly logical basis, if you prefer 100 to 80, you should also prefer 25 to 20. Or, don’t you? Well, people usually don’t use this strictly logical way of reasoning, they prefer a gain that is for sure to something that is uncertain. This is what is called the certainty effect. Another interesting way to look at this is to focus on the value of free goods.v Imagine you are attracted by two bowls full of chocolate. In one bowl you have Hershey’s Kisses while in the other you have Lindt Truffles. Once you get close to the bowls, you read that the price of Hershey’s is ¢1 while Lindt costs ¢15. You make your choice and eat the chocolate. Now that you have just finished with your chocolate, you go back to the same place and see that the two bowls are still there. You decide that you want another chocolate but, as soon as you get closer, you realize that something has changed. In fact, the price of Hershey’s is now ¢0 while that of Lindt is ¢14. I don’t know what you are thinking, but I know what I should do: Take the free chocolate! This experiment has been conducted among students/customers, and results were aimed at defining the value of a free good. It resulted that customers preferred Lindt (36%) over Hershey’s (14%) in the first round, i.e., when Lindt’s price was ¢15. In the second round (0 and 14), the preference was reversed and customers chose Hershey’s (42% vs. 19%). When researchers tried to lower the price of the higher-quality product to ¢10 maintaining the other chocolate on ¢0, the result didn’t change: Customers continued to choose Hershey’s (40%; Lindt 12%). The benefit of
Prospect Theory
29
choosing the free product overcomes the value of paying any money for a “superior” taste. Together with the authors of this experiment, I believe that the mechanism here is very similar to the certainty effect defined above.vi People exhibit different behaviors when exposed to something that is given for sure or that is available without any cost. A free good causes people to make decisions on that basis. The availability of a product makes it certain in the sense that it is given for sure, you can take it with no deprivation of any sort. This is what the experiment shows very clearly.
The Possibility Violation Another significant violation of the assumptions of a mathematically based mind (expected utility) is what can be expressed by the domain of the possible confronted with the domain of the probable. The following prospect is particularly helpful. Choose between A: ($6,000, .45) or B: ($3,000, .90)
Here, most of us choose prospect B where the probability to get the money is the highest. Note that the expected utility of the two prospects is identical since 45% of 6,000 is 2,700 and this happens to be the same number we obtain from the second prospect. Of course, the fact that the probability is greater in prospect B makes the difference. Now, consider the following problems: C: ($6,000, .001) or D: ($3,000, .002)
You may face a dilemma here since the probability to gain something is very close to zero for both prospects. However, the probability to get $3,000 is double that of getting $6,000; still you don’t think it really makes a difference whether you choose the first or the second. Your decision making is impaired: What is to be done? These two prospects maintain the same characteristics of the previous ones in that the expected utilities are identical for C and D respectively, and the relation between the probabilities of A, B and C, D is 1/2 in both cases. However, when the probability becomes a mere possibility, people go for the highest amount (i.e., they take prospect C) because they think that probability doesn’t make any difference and if they have to take a risk. . . well, it is better to risk for the highest amount! This logic finds application in a lottery, for example. There, the probability of winning something is very low, so low that an economist once called it “the fool’s gamble.”vii People usually choose to buy tickets of lotteries where the jackpot is the highest compared to the price of the ticket. This choice is consistent with the possibility violation. It is called a violation since it violates what an individual with perfect rational preferences should have chosen, according to expected utility theory. Of course it is not a violation of common sense and of what is a rule for everyday survival.
30
4 Maps of Bounded Rationality (I)
The Reflection Effect What happens when we lose instead of gain something? Do we react the same way when we are about to lose or when we are about to gain a certain amount of money? Imagine that your personal bankerviii just called saying that your portfolio with a market value of $120,000 needs to be renegotiated. Here is her proposal. You leave the money invested in the same stocks and, according to financial forecasts, have 80% probability to gain 20% in the near future, i.e., you get $24,000 on your original capital which makes a total of $144,000.ix This implies that you face 20% probability of getting nothing more than what you have right now. Or, you can invest nothing and get your money back with an $18,000 bonus on the $120,000 that you already have, i.e., $138,000. What is your choice? According to a simpler choice made on prospects, most of us choose certainty over the risk of getting nothing out of an investment. It is better to have something right now than to face the risk of having nothing (no gain, in this example you still have the principal). This is the basic certainty effect that we explored above. Now, try to picture the following scenario. Your personal banker proposes to keep your money invested in the same stocks, but financial forecasts outline that you have 80% probability to lose 20% of your investment, i.e., to have a loss of $24,000 that brings your capital to $96,000 total. However, you also have 20% probability to get your capital intact at that same near future date (which means $120,000). The alternative is to have your money back right now, but the bank should retain $18,000 from your account to pay for losses and various fees so that your money is $102,000 in this second option. The first alternative is uncertain: You don’t know if you are going to lose or not, and you still have the chance not to lose a single cent. The second alternative is welldefined so that your capital decreases by $18,000 for sure. The two options bring you the possibility of ending up with $120,000 still. At the end you decide to risk and try to get all of your money back. And, this is what most of the people should have chosen according to prospect theory. What is interesting here is that it seems that the certainty effect works on the positive domain while it doesn’t in the negative. When people face the chance to lose money they tend to do whatever it takes to retain what they have earned. A certain loss is still a loss, while a potentially avoidable loss is not a loss yet! Even when the probability that the loss will not materialize is low (20% in our example) people decide to take the risk. The two behaviors are contradictory, and this is the reason why this is called the reflection effect. Here is another example. Think of the people playing “pool” for money. Who is more likely to say “double or nothing?” after the first game? The player who lost! He is advocating a “risky” strategy in an attempt to eliminate losses.x From the reflection effect, it follows that people are risk averse when they operate in the positive domain, while they become risk seekers when they operate with losses.xi
This Is a Biased World!
31
The Point on Bounded Rationality This last paragraph brings into our discourse the fact that individuals are not cold computational machines, but they perceive differently gains and losses and/or they value differently gains and losses. To this respect, being rationally bounded means that when confronted with the same sort of variations that, for example, go from the positive to the negative domain, individuals aren’t perfect calculators. They are bounded in the sense that they cannot treat the same way alternatives that are fundamentally equal; the limits are internal computational limits in all of these cases. From this perspective, they are not consistent. They probably don’t have a clear idea of how to use statistics either.xii The possibility violation and the certainty effect point out specifically this, i.e., the fact that individuals are not able to explore to the fullest extent the alternatives that they face. On the contrary, they show limits in computational abilities; luckily, these limitations happen to be particularly useful to individuals’ everyday life and survival. The following pages and the next chapter explain why.
This Is a Biased World! If we were to find a mechanism that defines our rational bounds, this would be a bias. Psychologistsxiii isolate a list of biases so long that this topic cannot be completely described and analyzed here. An entire book is needed for that purpose. Nevertheless, biases and prejudices are what characterize many decisions, and strategies leading to their avoidance are among the most studied. Although I cannot analyze all biases and prejudices, I have decided to present a selection of those that represent and support the idea of bounded rationality more clearly than others. In the following pages you will find some of the most studied biases in economics, psychology, sociology, and decision making.
The Endowment Effect A thief took your wallet. You are desperate because you had placed in it (temporarily) your wedding ring. “What is the value of what was in the wallet, sir?” asks the policeman. The point is that you don’t know. It is not that you don’t know how much that ring cost (you remember that very well indeed); the fact is that you attribute to that good a value far higher that its market price. That is something you will never even consider selling. Its value is too high. This is a very specific and particular case of what is called the endowment effect,xiv the fact that something gains value only because of ownership. It is this effect that explains why certain goods have higher prices when possession is exercised on them. This is a pattern of behavior typical of one phase of the game Monopoly. At a certain stage of the game, close to the beginning, players can
32
4 Maps of Bounded Rationality (I)
negotiate on the cards they have bought in the very first part of the game. Suddenly, through the negotiation process you learn that the price of those “contracts” is never close to its book value (i.e., to what is written on the card). In fact, the selling price can be very high and far from its original. Of course, we should consider the economic trade-off here since, once sold, the property can become a treat to the seller that could land there and pay the sometimes expensive rent. A mix of contractual abilities, risk aversion, and the endowment effect make that price so high.xv This effect can be found within organizations too. For example, the information that you gained on a specific quality test is particularly relevant once it becomes of your knowledge. And, in many cases you “sell” it at a price higher than the one that should be. Another example could be that of acquisitions. When one company makes an offer to the other, the first price is never the right one (for many reasons, of course) but I guess that a not so small part of it is played by the endowment effect. Moreover, I believe that in negotiation, arbitration,xvi compensation, benefit administration,xvii and in many more areas of management we have the potential to find this bias at work.
The Status Quo Bias When you repeat mental or behavioral patterns that have an anchor to how things have been set up in the past (and/or still are in the present), you are following a status quo.xviii A status quo bias becomes apparent when individuals tend to make the same decisions that they have made in the past. This is very likely to happen when past decisions have been successful. Business managers tend to repeat successful decisions, realizing afterwards that the circumstances were not the same. The repetition, in Europe, of the same dynamics that were extremely successful in the U.S. lead Jack Welch to lose the opportunity to complete one of the most important mergers in history, the one between General Electric and Honeywell.xix The negotiation tactic that worked with the US Justice Department’s Anti-Trust Division miserably failed with the Directorate-General for Competition of the European Commission. Without the European market, the merger made no sense. Now, if one of the most acclaimed and successful managers can fall under a status quo bias, what about you and me? What about every one of us? The reality is that this bias is one of the most powerful and widespread. It is very easy to become overconfident if we have been successful with a tactic in the past. If you have always beaten your opponent with the “barber’s opening,” why change it next time you play chess? This is basically what happens when we lean on past experience. The more we have been successful with that formula in the past, the more we will lean on it in future occasions. However, circumstances change all the time, and this is the only thing we know for sure. A second application of this bias can be found when people don’t want to change the state of things. It is not that we apply our mental frames to any situation that
This Is a Biased World!
33
has something in common with past situations, but sometimes we do not want to change. This has very interesting effects when it transfers into organizations.xx People find themselves stuck in their ways of managing their everyday routines and elaborate sophisticated arguments to support this status quo. Universities are no exception. These years are crucial for universities to have students get online courses. Times are changing at a very fast pace, and the number of students taking online classes is growing dramatically in the U.S. and everywhere in the world.xxi The fact that universities need to provide at least a few online classes is becoming a requirement rather than an option. We can discuss, as academicians, overall value, educational outcomes, and teaching quality of online classes compared to traditional classes. However, there are more than sound arguments that leave no doubt that it should be better to act first (i.e., go online) instead of catching up later. The risk of losing terrain to competing universities is becoming higher and higher. Having said that, I know that everybody has valid personal opinions (and supporting data) on online classes, but people who argue against them are usually falling under the so-called status quo bias. Those who advance arguments such as “the value of traditional education” or “people come to our university because of the class experience” are particularly falling under this bias. If we, as educators, had to follow that line of argument in the past, I think that we would be teaching Socratic maieuticsxxii instead of having modern classes. Or, we would have small and elitist universities like Pavia, Italy, around year 1000 AD (what is more “traditional” than that?). To avoid this bias, the point should not be that of saying “yes” or “no” to online classes, but that of asking how to get the best from this new learning medium. The status quo bias works in many different ways; however, an effective way to summarize what this is about can be the following: We face a status quo bias when something changes and our ideas do not.
Anchor Bias Remember the experiment with Lindt and Hershey’s chocolates? The price of the first was ¢15, while the second was sold at ¢1. Now, let’s make a thought experiment. If I tell you that you can take the first or the second chocolate and you can name a price, what is this price? There is a significant probability that you will name a price that is close to ¢1 or to ¢15. If something like this happens then the numbers that I just recalled serve as anchors, and your rationality has been affected by the so-called anchoring effectxxiii or anchor bias. This bias has a huge impact in the way managers make their decisions. Think about how a meeting is usually set up. Especially important meetings, those where you are supposed to make decisions, are set up with fancy tables, handouts with tons of numbers, many words and effective presentations in small professional rooms set up with food and drinks. Well, not “real” drinks but coffee and water abound. These numbers have the specific function to constrain your reasoning and to let you
34
4 Maps of Bounded Rationality (I)
focus on something that will remain close to that. Of course, this is not something you do consciously. You always have the feeling that you are being very creative indeed. Unfortunately, you are not. Numbers that come out of those meetings aren’t really creative or innovative. The point is not that you can do better without those numbers. The point is that managers need to be aware that sometimes they are too close to original ideas, points of view, numbers, data, statistics, etc. available at the moment of the meeting. They are being subjected to the anchoring effect. You can stay “anchored”, but at least let it be your choice! A second aspect of anchoring happens when people “use their own beliefs and perceptions as a judgmental ‘anchor’ from which they adjust—usually insufficiently—to accommodate differences between themselves and others.”xxiv Studies in social psychologyxxv provide strong evidence that people anchor interactions with other people to their own assumptions and beliefs. In this respect there is a fundamental bias that operates when we approach and start a discussion with other people for the first time. When uncertainty on how that person will react to our comments is high, we tend to think that there must be something we share and interpret their thinking accordingly.xxvi Sometimes, when there is evidence that others do not react as we expect, we do not change our behavior or way of thinking and stay anchored to what we pretend to know for sure: how we are. This knowledge and identity-related anchoring has potential to explain misunderstandings, cultural unawareness, and some aspects of misbehavior.
Bandwagon Effect Have you ever jumped on a bandwagon? I am sure you have. It is easy to check this. If you answer positively to at least one of the following questions, you have. Have you ever (a) bought something (a pair of shoes, t-shirt, watch, car, book, etc.,) because you saw it on somebody else, (b) expressed an opinion that was someone else’s, (c) cited or quoted an idea, an opinion, a statement, a book, an article, a chapter because everybody knows it (or everybody cites it), (d) started listening to the music that everybody likes, (e) spoke out because somebody else did the same, (f) looked up because somebody in front of you was doing that, (g) clapped your hands when you realized that everyone around you was doing so, (h) laughed after others started doing it, (i) read this list because somebody told you to do so? The list could continue indefinitely, but these are very simple events and I am pretty sure that you answered “yes” to at least one of the questions. If you mindlessly followed somebody’s behavior then you have been part of a bandwagon.xxvii This is a peculiar phenomenon that attracts sociologists, economists, and management and marketing scholars.xxviii There are few studies on bandwagons within organizationsxxix , but there is a general understanding of what triggers that behavior and that mindset. Bandwagon is a bias since (1) it could prevent a decision maker from thinking in situations where this is exactly what is needed, and because (2) it is based on the
This Is a Biased World!
35
prejudice that what other people are doing is correct (worth imitating). Take how a financial crisis could materialize: A senior financial analyst of a well-known bank tends not to give importance to certain ratios, he overlooks some trends and focuses on others. Then junior analysts, still in the first part of their learning curves, tend to repeat the same mistake. Other analysts that try to make sense of those critical and overlooked trends are told that those data show only temporary anomalies that will be reabsorbed by the economy. Reports from this important and well-trusted bank become public, and analysts from other banks jump on this bandwagon, diminishing the value of data that nobody analyzes. Who wants to look foolish pointing a finger at something that is trivial to everybody? The answer is a quick and easy one. When the crisis is widespread, with a sad ex post feeling, people realize that those numbers were not so trivial indeed. Instead of jumping on a bandwagon, next time they will go through every single ratio that makes sense to them (we would hope!). Of course, short memory and positive feelings tend to delete the feeling of what happened. A new bandwagon could then arise.
Prejudices Prejudices are biases of a particular kind. The word means that your judgmental ability is given before even facing the situation where that judgment is required, i.e., your judgment operates a priori. Although the meaning of what a prejudice is could be very close to that of “bias”, it is usually associated with diversity and discrimination issues.xxx For example, during the apartheid regime in South Africa, you could say that white people were biased toward black people. However, this is not what you usually hear. What you hear is that people there operated on the basis of prejudices. This does not relate to racist regimes only, since our societies and organizations face problems of discriminationxxxi (or, you can use the politically correct word “diversity,” if you like) on a daily basis. What is different, if anything, between a prejudice and a bias? First of all, we are writing about a very specific decision making, the judgmental activity.xxxii This means that you are providing your opinion implying that something is right or wrong, according to your belief. This doesn’t happen every time you make a decision. Sometimes you make a decision without being judgmental (maybe you are so only at a very deep level). For example, when you make the decision to wear a yellow shirt, you are not being judgmental. When you express your opinions on how polluted the environment might be because of chemical agents that are in the yellow ink used by the clothing industry, this is the case when you are being judgmental. The outcome is that you probably will not wear (or buy) the yellow shirt. The difference to make, then, is that a prejudicial judgment is necessarily a biased judgment, but we cannot state the opposite (i.e., that a bias is always a prejudice). The bias could come up at the exact time you are making the decision while the prejudice is a priori. The former depends on experience and framing effects, while the latter needs certain circumstances to emerge but affects your judgment.xxxiii
36
4 Maps of Bounded Rationality (I)
Groups of students at Stanford, Penn-State, and Berkeley have been exposed to a series of pictures. Some of these pictures were about primates, some of these were about colored and non-colored human beings. After being exposed to the pictures, students were asked questions where it was possible to map their cognitive associations. Researchersxxxiv found that black people were more often associated with apes by non-colored students. This surprising result brings to our knowledge that prejudices could be very deeply embedded in people’s thoughts, something very well rooted in the hidden and unconscious beliefs. However, they could come out unexpectedly. Isn’t this an a priori judgment? Is this a different type of bias?
Errors Biases are nothing but errors, mistakes that we make while making decisions or analyzing specific problems. This is the reason why in the following I consider some of the most typical errors.xxxv As for biases, these errors are supposed to provide effective examples of individual bounds. Imagine that your company has an opening for a junior supply chain manager’s position. Your boss asked you to classify and analyze resumes so that you can evaluate candidates’ work experience and skills on the basis of a form that needs to be filled in each of its six sections for every candidate. You start your job, and when you are close to the end you suddenly realize that you are rating very high in every section of the questionnaire as those candidates that have no experience in the field. This is probably because when you entered the company you had no experience in that field, so you think that somebody with your characteristics may be a good colleague and be highly productive, as you are. But, you also remember your college studies and know that these kinds of errors are very typical and need to be avoided. In fact, you are making a halo and a clone error (also known as similar-to-me error). The former is the tendency to base a high evaluation on a single excellent factor, independent of the other characteristics. The latter is the attitude to prefer something which has the same characteristics as the decision maker. Romans recognized very well what this clone factor was all about when they stated simil cum similis, meaning that people that are similar to each other tend to get together (“birds of a feather flock together”). Of course, this is not only limited to the hiring process, but can extend to any kind of decision-making process. Now, let us assume that you redo your ranking to avoid those errors. When you are done with it something strange happens and you realize that you have fallen under the opposite error. You have penalized candidates that have not had previous work experience in the field. Moreover, you notice that together with this your ratings are all extremely low, nobody scored “excellent,” none are even close to “well.” Here we are, again! You have made two kinds of errors. One is the opposite of the halo error and it is called the horns error. It is the tendency to make poor evaluations on the basis of one single factor. The other one is the severity or “harshness” error, and it happens
This Is a Biased World!
37
when the decision maker consistently evaluates something or somebody below what is deserved. While the horn error is the opposite of the halo error, the severity is the opposite of the leniency error. There are other common errors based on biases. When you make a decision on something based on the impulse of the very first idea that comes up in your mind, this is known as the first impression error, also called the primacy effect. Many of the mechanisms that underlie certain shopping behaviors are based on these kinds of errors. You see something in a shop and you suddenly think that you cannot leave without it! You buy it based on this first idea, but you will realize later if the money spent was or was not worth the price. In other words, you will learn later if your first impression-based behavior was an error or not. This error has its opposite, the recency error, and it happens when your decision is based on the last impression you get from something or somebody. The tendency to systematically avoid extremes causes the central tendency error. It is generated when individuals never take into consideration extreme alternatives so that what they get is always on average. Again, a buying behavior directed toward the consistent avoidance of the cheapest and most expensive products is the effect of this error. The opposite error could be dangerous, especially when the extreme is made of the most expensive products. This is the extreme tendency error. Last but not least, when you decide not to do (or think of) something because of problems that it caused when you first did it, you are committing a spillover error. This seems to be generated by a status quo bias since it is anchored to something that happened in the past, i.e., it is a spillover of past memories. Table 4.1 provides a summary of all of these errors.
Table 4.1 Common errors
The error. . .
. . . And its opposite!
Halo error Central tendency error First impression error Leniency error Spillover error
Horn error Extreme tendency error Recency error Severity error
The simple fact that we are not omniscient leads to the implication that we make mistakes. The point here is that these mistakes are part of the common way to any reasoning activity. Of course, sometimes we could be better off without these errors, but it is crystal clear that we can never avoid them completely. They are part of the way our rationality is organized, and research shows that we are bounded. We can also use the opposite argument writing that we are bounded because of the mistakes we make. These studies on errors provide additional evidence that individuals are boundedly rational.
38
4 Maps of Bounded Rationality (I)
Summary There are many ways to show how individuals are rationally bounded. The chapter has started the process of drawing a map of bounded rationality. For the moment, we have explored bounds that deal with the failure to use mathematics and probability mechanisms (prospect theory), and the limits that derive from biases and errors. The chapter presented a selected review of the findings of prospect theory: (a) the tendency of individuals to prefer certain alternatives over uncertainty (certainty effect), (b) the attitude to divide what is possible from what is probable (possibility violation), and (c) the idea that risky choices are different when the positive or the negative domain is considered (reflection effect). Associated with these effects, there are several biases. We have presented those biases related to ownership (endowment effect), past experience (status quo), external resources (anchoring), and social imitation (bandwagon). Prejudices have been defined as a priori judgments and errors classified on the basis of their formation. We have tried to list many of the fallacies that can be found in organizational settings. The map is only half complete. Keep reading!
Notes i. Kahneman and Tversky (1979). ii. A foundational study is Morgenstern and von Neuman’s Theory of Games and Economic Behavior, 1944. iii. Kahneman (2003, p. 1449). iv. I prefer the vague expression behavioral “studies” to indicate all behavioral contributions, including psychology, marketing, finance, and economics. v. Examined in Shampanier et al. (2007). vi. Shampanier et al. (2007, p. 743). vii. This economist is Luigi Einaudi. viii. Contrary to prospects mentioned earlier in the chapter, this example is not taken from Kahneman and Tversky (1979). ix. A finance professor could argue that $20,000 in the future is different from $20,000 now. This is right but still, it is more than the $10,000 the banker offers you right away. x. I owe this example to William Ross, Professor of Management at the University of Wisconsin–La Crosse. xi. The literature on risk aversion and its anomalies is very extensive. For further readings, see Rabin and Thaler (2001), Post et al. (2008), Benartzi and Thaler (1999), and Thaler and Sunstein (2008). xii. This point has been made clear in many papers but it is very well described in Thaler and Sunstein (2008), and in Ariely (2008). xii. For example, see Bazerman (1994). xiv. Kahneman et al. (1990). xv. Better examples could be found in Neale and Bazerman (1991). xvi. Neale and Bazerman (1991). xvii. See Milkovich and Newman (2008). xviii. Silver and Mitchell (1990); Kahneman et al. (1991). xix. Chapter 4 of the book The United States of Europe by Reid provides an interesting summary of what happens to Jack Welch’s overconfidence. The title of that chapter is sarcastic and appropriate: Welch’s Waterloo.
Notes
39
xx. Change management is the subfield that studies determinants and implications of variability in organizations. From those studies, we understand that change is first and foremost a state of the mind. Many examples are provided by Martin (2007), in his book The Opposable Mind where he shows that one of the most important characters of successful leadership is related to the way these managers think about alternatives. xxi. In the U.S. “[o]ver 3.9 million students were taking at least one online course during the fall 2007 term; a 12 percent increase over the number reported the previous year” (Allen and Seaman, 2008, p. 1). The point with online education is that it also changes the boundaries of where a student is located. If teaching methods are corrected to fit the virtual environment, graduate and undergraduate programs can reach anybody, independent of the location. I am not impressed by the argument of the implicit inferiority of online classes compared to the face-to-face experience. Chapters 6–9 offer some ideas to think (or rethink) on these topics. xxii. This method is based on dialogue and conversation between teacher and students, where the role of the teacher is that of knowledge and learning facilitator. It is a learner-centered approach that is not based on lectures where all that the student happens to know comes from dialogue and confrontation. Plato wrote several dialogues that offer an idea of what this art of teaching looks like. xxii. The anchoring heuristic has been isolated and studied by Tversky and Kahneman (1974). In the book Nudge, Thaler and Sunstein (2008) exploit this effect a number of times and in many different domains. xxiv. Goldman and Mason (2007, p. 284). xxv. For example Nickerson (1999). xxvi. This is close to the curse of knowledge in Camerer et al. (1989). xxvii. See Leibenstein (1950); Granovetter (1978). A cognitive approach to bandwagons is offered in Bardone’s PhD dissertation (and forthcoming book). xxviii. Recently there has been a proliferation of studies on the bandwagon effect. Among the others, see Abrahamson and Rosenkopf (1990, 1993, 1997); Chiang (2007); Deephouse (1996); Laland (2001); Staw and Epstein (2000). xxix. An exception is Fiol and O’Connor (2003) and Secchi and Bardone (2009a). xxx. This is what social cognitive studies tend to show (Pennington, 2000). xxxi. Every management textbook has a section or chapter dedicated to diversity (e.g., Jones and George, 2009). xxxii. Usually judgment and decision making are considered together (e.g., Weber and Johnson, 2009). In the text I try to make a distinction instead. xxxiii. It is apparent that I am not making a distinction between conscious (or implicit) and unconscious (or explicit) cognition here (see Chao and Willaby, 2007), for prejudices may be both explicit or implicit. Although this is an interesting difference to consider, I believe that the latter is the most intriguing for organizational behavior since it is the most difficult to analyze and govern. xxxiv. Goff et al. (2008). xxxv. What follows is taken from an analysis of errors in performance appraisal, in Milkovich and Newman (2008, p. 333f).
Chapter 5
Maps of Bounded Rationality (II)
Heuristics What is the mechanism that leads us to a biased decision? How can we be so aware of the fact that something will happen when we have no evidence of that? How can we be so convinced that ours is the solution to a given problem when we know it is only a guess? Heuristicsi are “mental shortcuts.” They are the mechanism that let us jump to conclusions instead of going through detailed logical reasoning. The process that lets us cut corners is defined this way; but the conclusion may or may not be biased (or an error). Broadly speaking, the bias falls under the domain of substantive rationality, while heuristics is an analysis of procedural rationality. The former emerges in relation to the objective of our decision making, the latter is the underlying process. We can have a closer look at the way the status quo bias works. Suppose that a pharmaceutical company has a new drug that needs a few additional tests to be safely put on shelves. These tests are not standard for the company, and they are necessary for this product only, R&D managers suggest. The product is innovative and has the potential to be a blockbuster sale item. There is a 0.07 probability that the drug will end up having significantly heavy side effects for patients that use it above the average dosage; this suggests the need for further testing and analysis. However, the price per share of company stock is plummeting, and the market for the company’s existing products is decreasing. The choice is hard: take the risk and eventually start a new period of high sales and profits for the company, or have a break in the standard procedures and be sure that the drug is safe. The risk is very low, and the procedures the company will follow require extra legislation, too. The decision maker may not consider all of these implications but says a straight, “Yes, go for the market,” since this is what they have always done. This describes a status quo tendency, but also puts evidence on the fact that the decision lies on a logical flaw. Every new drug (or new product in general) is different from previous ones and needs to be treated accordingly. This is the logic behind innovation and, in our very short summary of the case, the decision maker jumps to a conclusion with no sound logic. The process that doesn’t help you in making sense of alternatives, situations, variables, etc., is a heuristic, while the result is a biased decision. D. Secchi, Extendable Rationality, Organizational Change and Innovation, C Springer Science+Business Media, LLC 2011 DOI 10.1007/978-1-4419-7542-3_5,
41
42
5 Maps of Bounded Rationality (II)
Nevertheless, not all heuristics end up being bad decisions. Many times when we use heuristics we end up with successful (or at least satisficing) solutions to problems. In the following pages I present a few studies that bring to our attention the fact that not being logical is not necessarily bad. On the contrary, it is what we do most of the time.
The Fast and the . . .Frugal! Gigerenzer and many others at the Max Planck Institute in Berlinii put incredible effort in to studying simple heuristics. According to their perspective, these mechanisms are very diffuse in our brain and “make us smart.” The point is that we have an “adaptive toolbox of simple heuristics” that is used depending on external conditions and internal dispositions. Bounded rationality is adaptive, and it is a toolbox. The idea that we have (or develop) many tools that help in our decision-making processes is not new. Studies on memory, capabilities, intelligence, traits, skills, etc.,iii are directed toward the understanding and mapping of cognitive tools of the human brain. What is new here is the focus on simple tools and not complex abilities (such as mathematics and logics). What is simple could be used more often and, as a matter of fact, is what helps our everyday decision making. There is something even more important than this. The idea that our brain adapts to the external situation is very well analyzed in evolutionary studiesiv , and it seems to be at the basis of the emergence of the homo sapiens.v Considering this attitude of change as typical of our brain is more than important; it is the core. How can we add this important new feature, i.e., adaptation, to bounded rationality? I follow the arguments that Gigerenzer and other co-authors present in their extensive and insightful publications.vi Simple heuristics are thought to be fast and frugal (called the “fast and frugal heuristics,” FFH). They economize in the time dedicated to the making of the decision and in the quantity (and quality) of information they use. A fast and frugal heuristic is, for example, what makes us think we prefer one piece of cake instead of the other when we see an indefinite number in a bakery. Think about this for a moment: You see an indefinite number of cakes. You don’t have the time to ask what each cake is made of; you lean on past experience and the appearance of the cake. Moreover, the bakery is overcrowded so you feel like you don’t have time. What will your choice be? You have very limited visual information (maybe a few words from the seller) and limited time.vii Your choice will be based on a fast and frugal heuristic. Mine doesn’t work very well because I always choose the wrong cake; it looks beautiful and tastes awful! Another example of a simple heuristic that helps us all the time is how language works. This may seem counterintuitive, but the activities of speaking and writing are made possible exactly because we use these sorts of mechanisms. The words we choose and their sequence are only partly dictated by a rational selection of choices. In fact, our speech and writing follow a well-established syntactical order
Heuristics
43
(when available), and a selection process that relies on which word is close to what we have in mind at that moment. While I am writing, I do not spend the same amount of time making decisions on every single word that goes in this book (I probably should!). Instead, I write sentences as they come, and I seldom think about the “right” word. This is faster because the choice is made in limited time, and the information I use is also limited. Sometimes a dictionary or thesaurus helps me find a better word, but I am never sure that my exploration and analysis of alternatives has been fully explored. This example becomes more significant when we think of a foreign language. What happens when we use a foreign language? There are different strategies that our brain uses, but the most common are two: (a) when you use sentences that you have already learned that come out without any effort (i.e., “How are you?” “I’m alright,” “What time is it?,” and the other survival-like sentences) and (b) when something unexpected happens and you have to quickly come up with something to say. This last case is more complicated and, in many cases, you have to think about which words to use and in what sequence. Suppose you are in France and ask for two tickets – you are about to visit the Musée du Louvre. Everything seems to go very smoothly but, at a certain point, the ticket seller asks you something regarding an audio guide. You were thinking about the Mona Lisa you are about to meet and are not prepared; you are not even sure that this is exactly what the seller just said. You have to make up an answer, saying something that is closer to what you should have said in your language. Of course, the alternatives of your mother tongue are quite infinite since you can choose among hundreds of expressions. You may have a polite answer such as “That is very kind of you but no thanks, I have my guide.” You should have told her a sentence trying to use humor, like “Wow, thanks. . . and do you have it with Dolby-surround?” Or, there is the quick answer that you never choose: “No, thank you.” In this case, the “Non, merci bien” seems very likely to occur, but the very first of your thoughts were on which are the right French words for the “dolby-surround” sentence. Time is extremely limited, the question is simple, people in line are waiting and you don’t have access to relevant information (i.e., you do not speak fluent French). Your fast and frugal heuristic helps you find an appropriate answer. And the result is successful: You don’t pay more money for an audio guide you don’t really need. This last example is particularly powerful regarding what we can learn on FFH. Studies on these mechanisms are recent and many questions remain unanswered. Research is ongoing, and we are confident we can get more from it soon; however, the idea that something so basic and widely used such as spoken language uses FFH suggests that these mechanisms are extensively used and are a good starting point for the study of decision making. These mechanisms are explicitly structured to let us make decisions in a timely fashion and without a significant amount of information. It seems that our knowledge structure and the mental frames we use rely dramatically on these heuristics. One last example that I can use is a fascinating experiment on bounded rationality. I have even tried this experiment in one of my classes. My results are not significant because I use it as a teaching tool, but students of FFH run similar experiments on the same subject matter. I report my experience relying on their data for
44
5 Maps of Bounded Rationality (II)
significance.viii I asked students which is the larger city, San Diego or San Antonio. Most of the students answer San Diego, based on their knowledge. When asked about this “knowledge” of theirs, they tend to raise interesting points. They say that they have been there once and they know how big the city is, they know somebody who lives in that city and that person told them how big the city is, they say that they don’t know the other city and they answered from what they knew. Basically, the answer is based on ignorance. I then asked if they knew of any data on population in the two cities; it became apparent that nobody really knew the answer to the question. What happened? The students used FFH to make a decision. They tried to retrieve information on the two cities from their memory and, when this retrieval ended up being related to one of these two alternatives, they tended to rely on the only data they had. This led to a discount effect on information they didn’t have. If you don’t know it, it must not be relevant! This is fast and frugal heuristics at work. What emerges from this experiment, that was not evident from previous examples, is that FFH are not always right. We can summarize that they are not accurate and uncertain in the appropriateness of their outcome. FFH can bring you to inaccurate answers that end up being not useful or not appropriate for the problem you are facing.ix FFH have not been fully analyzed in organizations. How do they affect managerial choices? How could personal experience, skills, capabilities, and other variables affect the development of fast and frugal heuristics? Under which conditions do organizations favor the emergence of FFH? Are FFH always successful in organizational settings? This is a list of questions that could and should be answered. Organizations are social environmentsx where individuals spend a significant amount of their lives. This means that organizations are shaped by individuals that work in them. Therefore, we need to consider if and how individuals using FFH affect organizations. In the following pages I present a few hypotheses (or thought experiments) on how FFH (a) emerge, (b) are embedded, and (c) are favored in organizations.
Organizational Heuristics The starting point is that individuals use FFH. I have described how individuals use FFH for personal decisions. Is this individual tendency also reflected in the working environment? Yes, these heuristics emerge in organizations as well as in other environments. Some decisions need to be quick and made with limited information available. There, individuals lean on what they know more than on what they should have known. The decision to set up a goal, select and recruit, launch a new product, or open a new branch or subsidiary leans on simple heuristics.xi Think of one of the processes mentioned above, recruitment and selection. There is no need to look at the entire processxii , but only at the beginning of it. The fact that the department has an independent budget doesn’t mean that every team is allowed to recruit without authorization. The process of getting authorization is complex, most of the times, since the head of the department needs to look at available funding and get
Heuristics
45
a first informal approval from the human resource department. During this process, whether the answer is “yes” or “no,” she will hear many motives supporting the choice. The fact is that, at the very end of the process, when recruitment is approved it could be for any reason. I am not saying that it is not rational, quite the opposite. It is rational because the choice relies on simple heuristics that make a decision possible. If for every position opening the process should be that of fully evaluating the case, no position will be easily opened in a few weeks. Shortcuts are used instead. The idea to let those departments or subsidiaries open positions may not be well-justified in terms of needs, but will be in terms of reward, for example, if that team usually shows an above average performance. Observing common and diffuse organizational facts like this one, we can assume that there is emergence of FFH. Some of these heuristics may also be embedded in organizational processes. Organizations work through formal and informal structures.xiii The informal (also called behavioral) is made of individual recurrent behavior, norms interpretations, and culture, while the formal structure (also defined as the normative structure) is defined by role expectations, values, and norms. It is this last form that offers the potential to incorporate simple heuristics. The logical passage is simple: If individuals use FFH, then it is probable that the social environments where they spend a vast amount of time (e.g., organizations) develop a structured way to facilitate the individuals to operate through FFH. When a behavior or a mental attitude becomes commonplace, then it becomes a social behavior or something which is favored by the community. This is the way FFH enters the organization. The next step is that it can be “formalized” and become a norm. This probably changes the nature of what is discussed here; however, this means that further work is needed in this particular domain of organizational behavior. Or, there might be a second way for FFH to become part of the organization. Instead of being integrated as part of a way to make decisions, simple heuristics may be accepted and favored. A comment pointing at a specific direction, for example, receives more interesting and supportive feedback than a comment pointing at another, less probable direction. The team supports and encourages your thoughts on that Return on Investments (ROI)—even if you don’t know the exact numbers that define it—because it is commonplace to discuss about that. When you take under consideration the Return on Assets (ROA), people don’t care about what you are saying even if you do know how this is generated in detail. Heuristics work in the first and in the second cases; however, the process/team supports heuristics when using ROI, while it does not when using ROA. Besides these financial measures, the point is that FFH sometimes needs the support of the team or organization where it emerges (more on this social dynamic in the second part of the book). When simple heuristics are considered, I believe that the most interesting element is that they can be used to explain a vast amount of mental processes. Can we categorize FFH? Can we classify them? How can we get more from these processes? Organizational behaviors that rely on simple heuristics have been overlooked in the literature but maintain a significant potential to explain individual thinking and behavior.
46
5 Maps of Bounded Rationality (II)
Emotionally Bounded In one of their works, Human Problem Solving, Newell and Simon offer reasons why they decided not to include emotions in their analysis: Our final choice is to exclude motivational and personality variables – what Abelson (1963) covered in part with the term “hot cognition.” We omit them by reason of convictions, not about the importance or unimportance of the phenomena, but about the order in which theory should develop. Many motivational and emotional phenomena operate through the lens of the cognitive system [. . .]. A plausible scientific strategy is to put our cognitive models in order before moving to these other phenomena. Our one explanatory foray in the direction of motivation and emotion was precisely in this vein [. . .].xiv
Emotions have always been excluded from the rationality discourse. For a long time decision-making scholars thought that feelings and emotions weaken and impair decision making. “Emotional” was opposed to “rational” decision making. This is not to criticize rationality scholars, since Platoxv stays at the very origin of this dichotomy with his myth of the black and the white horses. This is to say that the way these two worlds are opposed is the way the entire western philosophy and cognitive science have developed. And Newell and Simon are no exception. This ideal led thinkers of the past and present to believe that everything rational must be under mindful control and it can be computed by the human brain. The Leibnizian nihil est sine ratione (nothing is and happens without a rational explanation) served to define exactly this idea. According to that classical tradition of thought, if you can have an algorithm that describes any event after it happens, i.e., ex post, you can have one that explains the same phenomenon before it happens, i.e., ex ante. Even the most complex and random event can be reduced to a cause-effect relation using an ad hoc and ex post explanation. Mathematics is a very useful tool that serves this end, and it has been used many times to “uncover” patterns once the event has ended. Financial crises, economic collapses, rising stock prices, social trends, et altera, have been defined by econometric models that analyze how those events happened. However, this is not what we need to deal with here. It is not my intention to offer a literary review on advancements of mathematical modelingxvi in economics, decision making, or management.xvii Although this is a very stimulating debate, I would like to suggest a different perspective. My point is that the explanation of everything throughout the logic of mathematics (and reason for those thinkers who used it) fulfils the expectation that the scientist is willing to find in the observed phenomenon:xviii a pattern of relations. And this, in turn, supports the impression that you can find that same relation ex ante. Of course, if scientists were able to do that, we would be in a world of certainty; fewer biases and heuristics would be needed. What the (old ideas of the) cold world of mathematics also suggests is that analysis must be objective and it can be done in vacuum (i.e., in complete isolation).xix Instead, neurological studiesxx have found that emotions are associated with the process of making decisions. This means that we are not capable of making any decision without them. The example of mathematics is also interesting from this perspective. Every mathematician, or anybody who uses mathematics to find a
Heuristics
47
solution to a problem, enjoys the fact that she is close to a solution (when she is). These positive emotions are associated with different configurations of the brain and the release of chemical substances that help the search for a solution.xxi If we were rational in the sense that Leibniz (or Descartes) thought we were, we would never be able to find any solution.xxii Emotions trigger the ability to make decisions; thus they facilitate the solution. The approach taken here is narrower than the one mentioned above, though. I want to point out what emotions deal with bounded rationality. As Hanoch puts it,xxiii emotions work with mechanisms different from rationality, but they can help the way individuals process information. He isolates the areas where emotions contribute to rationality by “restricting the range of options considered (reducing the load on short and long term memory), [by] focusing on certain variables (certain stimuli receive higher ranking order), and [by] initiating and terminating the evaluation process (working as a satisficing mechanism).”xxiv There are two major contributions to rationality in Hanoch’s statement. The first is to consider emotions as a prioritizing mechanism: They help us to rank the alternatives operating within our bounded rationality. The second is a help focus on particular situations, events, facts, and circumstances. Emotions are studied within organizationsxxv with a growing emphasis in recent years. The outcome of many of these studies is that too much emotion is harmful in that it prevents social relations from forming, detrimental to productivity, impairs the ability to focus on specific issues, lets the worker misrepresent reality, etc. Few studies have been conducted on how emotions relate to bounded rationalityxxvi in organizational settings. We can speculate that it is very hard to find such evidence due to problems relating to experimental and field analyses. One solution to these problems may reside in the field of neuroeconomics.xxvii This new field studies how various parts of the brain are activated when individuals are confronted with economic decisions. When we apply this to emotions, we should be able to determine how emotional states are associated with decision making. Technology allows us to study this outside the laboratory. The fact that we use emotions to make quicker decisions is apparent. One of the most-used examples is that of the snake. Suppose you are walking in a forest and you see a snake close to your right foot. What should you do? I am sure you do not stay there and rigorously calculate which alternative is better, following the definition of rationality exposed in one of the previous chapters. What you do is jump as far as you can from it. You don’t even know if it is a dangerous snake or not! But, this is what you do: A quick solution that could potentially have saved you from a very painful bite. This is what emotions do: They help you prioritize among behavior alternatives.xxviii Now that we have defined a prioritizing mechanism, I would like to write—as a marginal note—that this example could not be very appropriate. If this is an emotional response, what is an intuitive one? What is the difference between emotion and intuition? Are intuitive decisions always emotional? And when can we say that a decision is more intuitive than emotional? With the next chapter I start answering these questions, but it seems to me that the task is not easy. For now, I suggest that we look at what intuition is. It is a
48
5 Maps of Bounded Rationality (II)
decision that comes out apparently from nowhere and uses knowledge already in our brains. Intuition is typical of improvisation for musicians; patterns are known, and the right notes come at the right time according to what the musician is able and wants to frame. Past knowledge is relevant and emotional involvement is important, too. Intuition can be defined as a set of fast and frugal heuristics that is often triggered by emotions. Isn’t this a bridge between the two areas?
Accessibility, Representation, and Framing Psychologists isolated three issues that define how people approach problems. These are of utter importance when it comes to defining our rationality bounds. The first is related to accessibilityxxix of data. How we are able to get information from a given data set is related to how it is accessible to us. Accessibility deals with the ease with which we get the data. A data set could be more or less accessible to us depending on its configuration and contents. Then comes the issue of representation of data. The way data are structured and the relations made between variables deals with the way we make sense of information. The third element is the mental framing of information. How we make sense of information depends on how we organize and analyze data. This, in turn, depends on our past knowledge, conditions at the moment the analysis is performed, and on our abilities. Think of a corporate report from a public company. This report is usually available online, which makes it accessible to most of us; almost anybody can download and get easy access to it. When you download an annual report, for example, you start with the letter from the CEO or sometimes from the chairman of the company. This is commonplace. The next section introduces you to the world of numbers with highlights from the balance sheet, cash flows analysis, and the profit and loss account for the last three fiscal years. If you are not an expert, it is hard to make sense of these numbers. This means that the representation of those data does not fit your understanding, and you cannot (or have a hard time) make(ing) sense of them. When you turn pages, you see figures of data where histograms and pie-charts represent the data under an easier and clearer perspective. There you have an immediate idea of how sales have grown, which costs have been cut, what the weight of labor cost is on total sales, how profits are stagnating, and the like. This representation of the same data is more helpful and actually makes the difference. Also, the way you frame information is different in the latter case. This is not the first time you see these kinds of charts, and this is the reason why data are more accessible here: They are represented in a way that fits your mental frame. These manipulations of data and information are often used by companies to trick their stakeholders’ understanding. Suppose you find the following statement written in the website, bills, annual report, of an energy company: “HIJ company is the first provider of clean energy in the nation.” The trick here is subtle and well-played
Accessibility, Representation, and Framing
49
since you are induced to believe that HIJ produces most of the clean energy they sell. However, they don’t state that; they write that they provide clean energy, but it is not clear whether they produce it or not. Of course, there is a significant difference here. A company that produces clean energy and is the first provider in the nation invests heavily in renewable sources and is willing to take the risks associated with innovation. Also, in-house investment plans may indicate that the company believes that becoming more environmentally friendly and socially responsible is part of what the highly pollutant energy industry should do. Instead, a company that buys clean energy is not taking any significant risk, nor is it showing that renewable energy sources are “strategic” for them.xxx However, if the stakeholders (customers, for example) limit their knowledge to what is accessible to them through the bill or the home page of the company’s website, they will be fooled by the tricky wording. Representation of reality is flawed, and framing somehow impaired. I believe this is an example of how companies’ PR play their cards, knowing that stakeholders deal with accessibility, representation, and framing effects. I believe that what emerges from the examples and from a more accurate analysis of the three factors of accessibility, representation, and framing is that they are intertwined so that it is not clear where the effects of the first end and the ones of the second and third start. Despite all attempts to study these factors in separation, I believe that a better job could be done with the study of their interrelations.
Epistemological Corner Heuristics, biases, and errors can be defined through classic logic as fallacies. A first and general definition of fallacy is that it is a common misconception, that is, a false proposition which is widely and confidently held, and which carries for the holder the strong appearance of truth. For a long time, logicians [. . .] bring to their investigations a somewhat more restricted conception, according to which, a fallacy is a common mistake, not about belief, but about reasoning.xxxi
Therefore, a fallacy is a procedural error in reasoning that gives the illusion of being right to the decision maker. To understand why they are relevant to our discourse, I suggest giving an explanation of the most relevant fallacies and then proceeding to link these to our analysis of decision making and rationality. One of the most used list of fallacies is known as the “Gang of Eighteen”xxxii (Table 5.1.) I found in Gabbay and Woods’ work that fallacies with asterisks appeared in a list that dates back to Aristotle. These fallacies support arguments that are (a) erroneous, (b) attractive, because they look correct, (c) universal, because they are employed very often, and (d) incorrigible, i.e., they tend to be repeated again and again.xxxiii This leads to the acronym EAUI, often used in that discipline. It is not my intention to define each one of the 18 members of the gangxxxiv , because many of them are not relevant in the economy of our discourse. However, a few of them are useful and strictly relate to concepts such as heuristics and biases.
50 Table 5.1 The “Gang of eighteen”
5 Maps of Bounded Rationality (II) 1. Ad populum 2. Ad hominem 3. Ad baculum 4. Ad misericordiam 5. Ad verecundiam 6. Ad ignorantiam 7. Secundum quid (∗ ) 8. Hasty generalization 9. Affirming the consequent and denying the consequent 10. Begging the question (∗ ) 11. Many questions (∗ ) 12. Equivocation and emphibody (∗ ) 13. Post hoc, ergo, propter hoc (∗ ) 14. Insufficient and biased statistics 15. Composition and division (∗ ) 16. Faulty analogy 17. Gambler’s fallacy 18. Ignoratio elenchi (∗ ) Note: Fallacies with (∗ ) were isolated by Aristotle (see Gabbay and Woods, 2007).
For example, the ad populum fallacy occurs when individuals support or accept an argument because it is what most people believe. Sometimes academicians are not exempt from falling under this logical trap; many papers and books are cited because they are popular, because everybody cites them. Of course, this does not say anything about the arguments presented in those books and papers. However, it seems that they gain their power on the basis of how many times colleagues cite those works. In the academia, this process is particularly harmful because it leads the attention of scholars far from the contents of our work and to the typical shortcut of leaning on a popular work. This is close to the way the bandwagon effect was defined in the previous chapter. Examples of this fallacy abound. In marketing, for example, sentences such as “This is the best-selling brand!” tell nothing about the product, only that it is popular. Close to the ad populum, there is the ad verecundiam (appeal to authority). This can also be defined as the ipse dixit of medieval flavor (trans. “he said himself;” used to support propositions that needed no proof because “he—then, Aristotle— said himself”). It works every time individuals advocate and take an argument as true because of the supposed knowledge of the one that first stated it. This is particularly diffuse among believers in all religions, and it emerges every time you believe that something is true because of the source from where it came. The argument is fallacious because you are not making any evaluation of its contents; what you are stating is that it must be true because of the position of the person who is saying that (e.g., Harvard professor, the President, the Pope, the Dean). Once again, marketing techniques seem to exploit this fallacy very well when advertising is based on the fact that a product is “the best” among certain categories of people (e.g., “#1 for the American Dental Association”).
Two Logics
51
An ad hominem proposition, in one of its forms, is based on the person and not on the argument. When individuals disregard information because of the reputation of the person instead of looking at its content, they are operating on the basis of a logical fallacy. Another way to explain how this works is that of using characterizations. It is like being against the suggestion to read this book because of the person who suggests it, the author for example, advancing the argument that there is a blatant conflict of interests. This argument has nothing to do with the content of the suggestion but it is a powerful way to invite people not to deal with it. The hasty generalization is the action of inducing a general assumption from one or a small number of events. The fact that you have visited the campus at Stanford University doesn’t mean that all American universities are that beautiful. Even when people visit more than one campus, let’s say five from east to west, they cannot make generalized assumptions based on what they have seen. This is not a significant sample, out of hundreds of US universities. Every time we make a poor use of induction and generalize on the basis of limited information, we are falling into this particular category of fallacies. Scientific works are not exempt from showing signs of hasty generalizations; think about how many studies are based on a limited number of students! We may argue that in this latter case, it is more abduction than induction at work (see Chapter 2). The fallacy composition and division is also very common in decision making. It is the faulty logic that leads you to the attribution of characteristics of the parts to the whole (composition) or, vice versa, to the attribution of characteristics of the whole to the parts (division). The latter happens when you deduce that the goal of each member of the A team is quality improvement, since that is the mission of that team as a whole. This lets you overlook that they should have divided quality into different aspects so that every member might have a specific subgoal. The division fallacy is very diffuse among neoclassical economics. The assumption that the goal of the bank is, for example, to maximize its profits makes each banker share the same goal.xxxv This has to be demonstrated. It is a false relation (a logical fallacy) unless somebody shows some correlation between the two utility functions, that of the bank and that of its employees. The former (composition) becomes apparent when you think that the team is a living creature because all of its members are living human beings.xxxvi What is interesting in the way the idea of fallacies is portrayed by logicians is that it could offer a sound basis to studies on heuristics, biases, and errors. More than 2,000 years of study is a significant tradition that can be exploited when analyzing dynamics and inherent logic of faulty mechanisms.
Two Logics How do all of these rational bounds work within organizations? This question is not easy to answer and it probably needs more than the end of this chapter to be answered. However, we try here a first explanation based on what Marchxxxvii calls
52
5 Maps of Bounded Rationality (II)
the two “logics.” Decision making can be thought of as a process where individuals answer questions and follow specific patterns. March’s idea is that we can identify a first “rationalistic” logic that follows a strict and well-defined pattern. This is the logic of consequence and it is very close to a decision maker that operates in vacuum, i.e., without an environment. The second is a logic of appropriateness, where the individual is embedded in a situation. In the following, I explain briefly what these two logics are about, posing a particular emphasis on the second which is more likely to explain why individuals make mistakes, use heuristics, fail in many circumstances or, in two words, why they are so boundedly rational. The goal is to provide a framework for everything presented above that maps BR (Chapters 4 and 5).
The Logic of Consequence In a strictly formal logic such as this, there are four questions that individuals (and theorists) ask. I consider these four very briefly below. The first question is that of alternatives. Followers of this logic need to define which are the courses of action that are possible and map them as best as they can. Only if you have this map you can start thinking of expectations, i.e., consequences associated with each alternative. This process is close to what Simon defines as rationalityxxxviii (step 4, see Chapter 2). The third element is that of preferences associated with the decision maker. Values and principles affect the way the decision is made; in many cases, values drive the decision-making process so that individuals choose according to what they believe is right. Since time is often more precious than the effort put on getting the best possible outcome, we need a decision rule. This is what allows us to stop the search, focus on alternatives and consequences that make sense to us (value), and decide. This logic well describes the original idea of bounded rationality that was originally introduced by Simon. Individuals make their choices selecting alternatives through a system of values. These preferences are evaluated in the potential outcomes they can provide to the decision maker. The fact that rational agents use heuristics, have problems with data accessibility and representation, are biased, and emotionally bounded, is part of how this process is shaped but does not change it in its major components. Still, with bounded rationality the logic of consequence works as well. Alternatives generated could be less than expected, preferences could be biased, decision rules subjected to emotions, and expectations defined on the basis of heuristics. This is to say that this logic works with individuals like you and me despite its appearance of perfect chain of events (alternatives-expectations-preferences-decision rule). Although the logic of consequence is powerful in what it defines, March believes that there is something missing. And this “something,” as you will see in the coming pages and chapters has the potential to change the way we understand and analyze decision making.
Two Logics
53
The Logic of Appropriateness The second logic is presented through the analysis of three questions. When decision making happens in organizations or in any social setting, individuals need to answer questions that pertain to those specific circumstances. This is the question of recognition that allows the decision maker to understand what situation he or she is facing. It is the analysis of conditions influencing the circumstances that affect or may be affected by the decision that is about to be taken. Together with this the decision maker needs to understand the role that he or she is playing in that situation. This is the problem of identity, and it is crucial when players don’t want to lose track of their potential to achieve the goal. Of course, when you are describing a situation where the decision is to be made in an organization, then the role and position of the decision maker may make the difference. For example, if you are about to make a decision on a new production line, your decision needs to be defined depending on the fact that you are in charge of that new line of production, you are an R&D manager, you are the CFO, or you are a middle manager. This is going to affect the way the decision is shaped. Roles and positions are also related to the rules that apply to the situation. Different roles often lead to different takes on the same problem. This may depend on multiple variables such as (a) personal attitudes, (b) abilities, and (c) possibilities of making the decision. This latter point depends on the power associated with the position, sometimes more than with the willingness and boldness of the decision maker. These three simple factors—recognition, identity, and rules—are supposed to define how the individual makes decisions in organizational settings. We can use an incremental logic to consider these two logics together. I don’t believe that the second excludes the first; that is, if you follow the second you cannot use the first. I believe that they depict different perspectives so that the logic of appropriateness includes that of consequence. The three factors posit questions that are inherently different from those used in the first logic. Of course, the problems that individuals face when they operate in social settings need to be framed through a more socialized context so that the decision maker that faces a problem may not be able to actually solve it because his/her position is not appropriate. The fact that a shop manager at a fast food restaurant recognizes that deliveries of meals at the drive-through window are not fast enough and finds a way to make them faster does not allow him to change procedures. The decision-making process follows a standard operating procedure that has to be the same for each shop in the restaurant chain. The decision is not appropriate according to the role of the shop manager. It is a direction that is more centralized and depends on somebody that makes that decision out of the power he or she has. The process can also be evaluated through expectations, alternatives, preferences and past the decision rules, and nevertheless it cannot be followed, according to the appropriateness logic.
54
5 Maps of Bounded Rationality (II)
Implications of Using One or More Maps In the last two chapters, I have offered an overview of those theories, models, and approaches that contributed to the legacy of Herbert Simon and to the idea of bounded rationality. What you have found in these pages is far from being an exhaustive and complete review of these studies, rather it offers a selection of some of the most influential works. Some concluding remarks on the impact of these works are probably needed. Besides the two logics that offer a framework through which decisions may be analyzed, there are two paradigms that seem to emerge. One was originated by the work of Kahneman and Tversky, and we may call it the biases paradigm; another is based on that of Gigerenzer and his colleagues, the toolbox paradigm. Although it may be implicitly or explicitly stated in this and in the previous chapter, in the following pages I describe what is the most significant import that these two paradigms add to the original idea of bounded rationality.
The Biases Paradigm First of all, I am not fully satisfied with the choice of the label used to define this paradigm. Of course, biases are not the only interest for this subfield. However, the choice falls to that word since this is what the two pioneers that originated the enormous amount of studies focused on. Having written that, as far as this book is concerned, I can summarize the most important contributions of this paradigm in two points. The first is that it fueled a revival of behavioral studies. The second is that it helped analyze and understand the computational limits of rationality. Behavioral studies. The “biases paradigm” gives to behavioral studies a significantly different origination compared to that of Simon. The father of BR used a deductive approach, founding his idea on critiques to a previously existent theory. Instead, the biases paradigm is based on empirical findings that reject assumptions of that same neoclassical theory from where Simon started. This appears as a tremendous change from the origins of BR, since it founds behavioral studies on empirical research instead of theoretical elaborations. Therefore, we have two traditions (now merged). The old behavioral research is concerned with the study of theories that explain how people think and behave. The new behavioral studies found their research on empirical findings and eventually build theory out of experimental results. The latter approach has become the standard for behavioral research, contributing enormously to find consistent empirical evidence of BR. Computational limits. What type of BR are these studies looking at? In order to make a significant contribution, the biases paradigm focuses on a particular aspect of bounded rationality. The connection between the idea of BR and the paradigm is that violations of neoclassical economics axioms are connected to computational limits of the human mind. And this is exactly one of the two sets of bounds that define
Implications of Using One or More Maps
55
rationality. Due to a wide range of studies, we are now able to define computational limits of the brain, have practical examples and applications, analyze implications stemming from these limits, and measure these internal bounds. This provides BR with a solid and measurable base.
The Toolbox Paradigm This second and more recent paradigm contributed—among the many other achievements—to defining two points that relate and improve the original BR. One is the concept of ecological rationality, a second is the focus on external information limits to rationality, and the last one is the corollary “less-is-more.” Information limits. Studies from this paradigm take an approach that is slightly different from what we have seen in the biases paradigm. The starting assumption for FFH is that “[f]or the most part, however, theories of human inference have focused exclusively on the cognitive side, equating the notion of bounded rationality with the statement that humans are limited information processors, period.”xxxix This is why they decided to start from the second set of bounds that relate to limited access to information. In so doing, the toolbox paradigm has been able to develop significant insights on how individuals analyze, frame, and collect information. Ecological rationality. Although Gigerenzer and others always see themselves as loyal followers of the work of Herbert Simon,xl they came out with a redefinition of rationality on the basis of its ecology. In short, their studies on heuristics (FFH) and human cognitive processes led them to highlight one of the features of BR, defined as dependent on availability of external information. The study of heuristics is a study of matching between these mechanisms and the environment,xli where information is located. There are adaptation and dependency patterns that these studies highlight. Less-is-more. This is a corollary that Gigerenzer and others have been able to derive from the major assumptions of the toolbox paradigm. The heuristics mechanism is based on the interaction between external information and the blink that allows individuals to make a decision. There are circumstances where the more information people get, the less accurate and efficient their heuristics become. This intriguing finding has many implications, one of them being, for example, the fact that it gives us a scientific explanation of the reason why, after hours of search on the Internet, we sometimes come out with answers that are not worth the time spent in the search.
Final Remarks It is now apparent that the two paradigms should be considered together since biases explain how computational limits work while the toolbox focuses on external bounds. And we know that these two together define BR.
56
5 Maps of Bounded Rationality (II)
The most important factor associated with using a map is that we do not get lost, and it allows us to find our way once we get lost. To some extent, this is what maps of BR are for. They help scholars and students of rationality find their ways to the bounds when analyzing and understanding human cognition. A map of BR points to these bounds and helps find them. Truth to be told, for the most part, “mapping” studies show exactly this, and highlight how limited our rationality is. However, maps are also sensitive to the tools at hand. If you happen to look at the map of a European city drawn during the Middle Age, you will find it very different from what we consider a reliable and valid map today. Even those parts of the city that have not changed that much are not easy to recognize. The same thing happens if you look at a geographical map of any area of the (known) world as it was seen in the twelfth century. It certainly does not look like the one we are used to. This analogy is useful to introduce two elementary concepts. The first is that we do not know how accurate these maps of BR are; so far they seem reliable, but how sure are we? What if our assumptions on cognition change as, in the analogy, the assumptions on the shape of the world changed? Accuracy of tools also plays a role in this process. The second is that doubt should always be at the basis of every scientific inquiry; this is how we can improve and advance. The next chapter starts instilling some doubts on the influential ideas reviewed in these pages.
Summary The maps of bounded rationality are now completed. In the chapter we have learned that heuristics allow us to make decisions. Especially simple heuristics, those concerned with limited time and information (FFH), seem to be the most widely diffused and an easy mechanism for our brain. Emotions are also an aid to bounded rationality because they function as prioritization and focusing mechanisms. Accessibility, representation, and mental framing define how individuals make sense of information. The epistemological corner opens a window on a very long tradition of studies that share something in common with it, and is particularly useful to analyze errors, heuristics, and biases. We have also pointed out that this theory of bounded rationality works under the logic of consequence, while organizational behavior needs to be tied to the logic of appropriateness. Finally, some of the most far-reaching studies that have been published in the last 60 years can be divided in the biases paradigm and the toolbox paradigm. Differences with the original idea of BR are then analyzed. With the next chapter I start telling a different story.
Notes i. Heuristics are usually defined as “rules-of-thumb that people often rely on rather than using more appropriate statistical rules” (Kunda, 1999, p. 53). ii. Gigerenzer et al. (1999), Gigerenzer and Selten (2001), Todd and Gigerenzer (2003), Goldstein and Gigerenzer (1996), and Gigerenzer and Brighton (2009).
Notes
57
iii. Evidence is provided in Gigerenzer and Selten (2001), and also discussed, on a different and cognitive perspective, in Thagard (2007). iv. See, for example, Humphrey (1976), and Odling-Smee et al. (2003). v. Gazzaniga (2008). vi. Payne et al., (1993) study the adaptive decision maker in the tradition of Herbert Simon. Their work has inspired Gigerenzer’s studies on heuristics (Gigerenzer and Brighton, 2009). vii. Todd (2001), offers a more detailed explanation of time and informational constraints leading to fast and frugal heuristics. viii. The example is taken from Gigerenzer et al. (1999). ix. A quick note on the population of the two cities. It is about the same; the last census (2000) showed San Diego leading with 1.22 million people and San Antonio with 1.14. However, 2007 estimates show San Antonio ahead with 1.32 and San Diego stuck at 1.26. Tricky question! x. Part V of Gigerenzer et al. (1999), is dedicated to social intelligence; we can consider this as a premise, although insufficient, for an organizational analysis of FFH. xi. Textbooks are usually optimistic and provide recipes of what managers should do to set up goals, plan, hire personnel, and perform other important managerial activities. However, there are too many cases when these processes are not followed thoroughly. Grove (1999) at Intel and Semler (1994) offer insightful revisitations of classical approaches to how managers—especially top managers and executives—make decisions. xii. For the full story, see Mathis and Jackson (2008). xiii. Scott (2003) offers a description of the social structure of organizations in the first chapter of his book Organizations. Rational, Natural, and Open Systems. xiv. Newell and Simon (1972, p. 8). xv. There are many translations of Plato’s Phaedrus; I used Penguin’s 2005 translation by C. Rowe. xvi. Books on mathematical modeling for social sciences abound because the needs of researchers can be very different in terms of what model can be used. I have recently used Shier and Wallenius (2000), Luce and Raiffa (1989), Meerschaert (1999), and Dreyer (1993). xvii. Formal models in mainstream management are not significantly diffused. It is not until recently that the Academy of Management Review published a special issue on formal mathematical modeling in management (vol. 34, No 2, 2009). Interesting insights on formal models of managerial and organizational behaviors are usually published in the journal Computational and Mathematical Organization Theory. xviii. Not exactly related to this is the problem of observer-phenomenon relations and the so-called second order cybernetics (von Foerster, 2002). xix. Ilya Prigogine and Isabelle Stengers would not agree with this statement. In their work (1986) they show that complexity is what drives scientific discovery and that the simplicity of relations is an illusion that we carry forward from the old approach to science. xx. I refer in particular to the study of Damasio (1994) and Bechara and Damasio (2005). xxi. This is close to what Antonio Damasio presents in his book Descartes’ Error and calls it the “somatic marker” hypothesis. He explains it using the following example: “Imagine yourself as the owner of a large business, faced with the prospect of meeting or not with a possible client who can bring valuable business but also happens to be the archenemy of your best friend, and proceeding or not with a particular deal. [. . .] When the bad outcome connected with a given response option comes into mind, however fleetingly, you experience an unpleasant gut feeling. Because the feeling is about the body, I gave the phenomenon the technical term somatic state (‘soma’ is Greek for body); and because it ‘marks’ an image, I called it a marker. [. . .] What does the somatic marker
58
5 Maps of Bounded Rationality (II)
xxii. xxiii. xxiv. xxv.
xxvi. xxvii. xxviii.
xxix. xxx. xxxi. xxxii.
xxxiii. xxxiv. xxxv. xxxvi.
achieve? It forces the attention on the negative outcome to which a given action may lead, and functions as an automated alarm signal which says: Beware of the danger ahead if you choose the option which leads to this outcome” (1994, pp. 170, 173). Needless to say, the somatic marker hypothesis works for both negative and positive prospects. One of the strengths of this approach to decision making is that it finds a way to reconciliation for positions since “[t]he partnership between so-called cognitive processes and processes usually called ‘emotional’ should be apparent” (p. 175). As Damasio puts it, “understanding neurobiological mechanisms behind some aspects of cognition and behavior does not diminish the value, beauty, or dignity of that cognition or behavior” (p. 176). This is clear from Damasio’s work (1994) where he describes patients with neurological impairments. Hanoch (2002). Hanoch (2002, p. 7). Almost 10 years have passed since the Journal of Organizational Behavior (2000, vol. 21) dedicated a special issue to the emerging role of emotions in work life. Both empirical and theoretical studies continue to emerge in this field. For a review, see Härtel et al., 2005. An exception is that of Ashkanasy et al. (2005). See Politser (2008) and Camerer (2007). Hanoch (2002), Maramatzu and Hanoch (2005). In one of their works, Hanoch et al. (2007), link emotions to bounded rationality and to fast and frugal heuristics and suggest that emotions are a mechanism that help individuals (older people, in the study) to come to a decision when information and time are scarce. I take the three factors from Kahneman (2003). This example is taken from a real case. Please contact me if you want to know more. Gabbay and Woods (2007). The Gang of Eighteen has been introduced by Woods, 2004. Here is a short explanation of the meaning of those fallacies with Latin names that are not considered in the text. The argumentum ad baculum fallacy refers to the use of force or to the menace of using external events to persuade one person to agree (e.g., “eat this soup or something bad will happen to you”). Baculum means stick or cane. The argumentum ad misericordiam is the fallacy of bringing emotional factors in praise of the argument (e.g., “we must give him tenure because of his precarious health conditions”). The ad ignorantiam fallacy is the assumption that a proposition is true because we have no proof that it is wrong and vice versa, a proposition is wrong because we have no evidence that it is true (e.g., this is typical of statistical procedures when we reject a hypothesis because we cannot prove that there is a relation). Secundum quid is the generalization of a particular part of a proposition even when circumstances have changed. It can be thought of as a particular type of hasty generalization. We have the fallacy post hoc, ergo, propter hoc when something happens after an event, therefore it must be because of that event (e.g., “every time it rains in Minneapolis, the Vikings lose the game!”). Ignoratio elenchi, “red herring,” or “straw man” occurs when a successful argument is held against one’s opponent, being the opponent not committed to that point. This is only an approximate explanation; therefore, for a full treatment of fallacies, I recommend reading Woods (2004) and Gabbay and Woods (2007). Gabbay and Woods (2007, p. 68). See Woods (2004) for a complete analysis of the Gang or Gabbay and Woods (2007), Chapter 1, appendix for a summary. This analogy could be found in Block et al. (2008). I found this fallacy the core concept of a management book, “The Living Firm” (Vicari, 1991); it may also fall under this same fallacy when adjectives such as “vital” are used to denote firms or organizational characteristics (Golinelli, 2005).
Notes xxxvii. xxxviii. xxxix. xl. xli.
59 March (1994), Chapters 1 and 2. Simon (1997). Goldstein and Gigerenzer (1996, p. 651). Goldstein and Gigerenzer (1996). Gigerenzer and Brighton (2009).
Part II
The Extended Brain
Chapter 6
Simon’s Error
In the previous pages I offered a vision of rationality and decision making based on bounded rationality and theories, models, and approaches that support its assumptions. All of these theories, and the original model of BR itself, explain how we have limits. Decision making, according to this perspective, is defined through the understanding of the limits that characterize our rational world. From this perspective, a map of all of our bounds is particularly helpful since it defines what limits (more than improves) our way of thinking. A theory of bounds is a “negative” theory in that it defines shortcomings, anomalies, or fallacies of individual cognition. The outcome is that rationality and decision making are more defined by what cannot be done than by what can be done. Approaches, biases, heuristics, emotions, etc., are mechanisms that limit what is intended to be the full potential of a brain. Bounded rationality is also “negative” because this is the way it was generated. Simon thought of it as an alternative to the subjective expected utility (SEU) model of full rationality,i widely used by economists. If we are not fully rational, we might be something-related-to-rational instead of irrational. The theory then opted for a somewhat rational individual, using the SEU model still as a benchmark. Today, there are papers that try to find room for boundedly rational agents in economic theoriesii and compare it to the mainstream model of full rationality. However, the point is that bounded rationality has been defined as a modification of an existing theory. As shown in the previous chapters, understanding and addressing the shortcomings of the SEU theory has always been a major concern for students of bounded rationality. These theories have been successful in their criticisms of the SEU paradigm. What is covered in this chapter is based on two different perspectives of rationality and decision making. First, the focus on the how of rationality is particularly useful for understanding and defining a context in which individuals make decisions. These theories show how individuals use bounded rationality to make decisions. However, the great absentee of this inquiry is the why. Why do people think the way they do? I voluntarily put the question in a vague format without including bounded rationality in it. This is because BR is a hypothesis and I prefer to treat it as such. It is very difficult to get too far away from the idea that rationality has bounds, but D. Secchi, Extendable Rationality, Organizational Change and Innovation, C Springer Science+Business Media, LLC 2011 DOI 10.1007/978-1-4419-7542-3_6,
63
64
6
Simon’s Error
still I believe that to fully understand the implications of a methodological construct it is better to consider it the way it is: a scientific theory. And scientific theories can be falsifiediii or supported. Raising the question of why our cognitive system is the way it is may reveal overlooked sides of our rationality. This is exactly where I intend to go in this chapter: How can we improve the theory of bounded rationality? Probably, we should start asking questions that seem to have simple answers. Second, I think it is time to get rid of the “negative.” Putting too much emphasis on the limits of rationality has prevented us from focusing on the study of potential of the human brain. Scholars were too involved in studying how the previous paradigm of the homo economicus could be falsified. Even though the old paradigm is in more and more trouble, we are still overlooking what we should call the “bright” side of rationality. The chapter offers a summary of the so-called distributed cognition (DC) approach that has been introduced by Hutchins in 1995 and has improved dramatically since then.iv The ideas that support this approach are provocative and do not constitute the mainstream in cognitive science. This is to state that what follows here is highly speculative, but if you follow the hypotheses that I am about to explain here, you will find a different rationality and a very challenging decision-making model. This is where you can also find Simon’s error.
Distributing Cognition How does our cognition work? This question, or variations of it, has been asked for thousands of years by western thinkers and philosophers, and many answers have been given. However, in recent times the answer has started to depart from usual philosophical frameworks. This is what we are considering here and, of course, it is well and deeply rooted in the same western philosophical traditions, although reinterpreted on a completely different basis.v Richardsonvi writes that there are three groups of theories that analyze intelligence, rationality, and cognition: (1) the computational group of theories includes thosevii who explain the human brain through the metaphor of the computer, where the study of symbols and algorithms mimics the functioning of the human and animal mind; (2) the connectionism movement is concerned with the study of neural relations, where symbols and algorithms have a secondary role compared to the way the elementary parts of the brain interactviii ; (3) the third groupix takes both computational and connectionist approaches one step further to include overlooked variables, especially those that belong to the social side of human cognition. Studies from this perspective put the human brain on an evolutionary track and explain that huge cognitive developments happen due to the web of social interactions that shape the life of human beings. Both the environment and human relations become part of the cognitive equation.
Distributing Cognition
65
The goal of theories from these three groups is to understand and analyze how the brain works. However, what all three perspectives share is the axiom that cognition is located inside the brain. In a recent work with Emanuele Bardone, we look at recent trends in cognitive science and “suggest that there is a fourth emerging category that consistently increases the explanatory power of previous theories.”x Once again, these theories take one step farther from previous ones and consider some of the distributed nature of cognitive resources. The resulting picture sees human cognitive processes shaped by external resources.xi Within this framework, the social environment provides a significant source for cognitive materials, although it is only one among many other sources. This idea of a distributed cognition (DC) approach can be related to a sort of “ecology of the mind,”xii but I maintain that there are significant differences between this and several other studies on the human brain and on rationality. The following pages are dedicated to what I believe constitutes the newness of this approach, and explore the extent to which it successfully explains the individual cognitive system in a way that seems to be particularly helpful for organizational studies.
An Irreverent Hypothesis on Cognition After the trauma of birth, our mother and father, together with people in the childcare unit of the hospital operate to further our knowledge and understanding of the world. In a few days everything will be different forever. You are not able to remember when you spoke your first word and you don’t know what that word was. For most of us it is “mom” or maybe just “moo. . .” That sound or, better, the recognition of that sound by somebody who was close to us in that moment started a process that our minds will never get rid of. We started to understand that some sounds are different from others. Up to that point we only heard strange sounds from our parents, relatives, and close friends. But the fact that the child can interact using the same sounds is amazing! The brain starts a process it is set up for. We have the largest brains compared to other primates and to the relative size of our body, and we need more activity to cope with our social environment.xiii Size and activity are related to the fact that we are social beings. Size is not all.xiv The interesting fact is that when we realize that those sounds are so important, we start to learn more of them. At the beginning our vocabulary is pretty poor and our use of grammar and syntax is awful, but we get better every day, and fast up to a point where we do not understand (or remember) how life was without words. Was it possible? Words, our language,xv become part of us so that we think we have always had them. We cannot live without words, we cannot think without them. Now, let’s run a mental experiment. Imagine that you have to think of something without words. What does your thinking look like? How is it formed? By what is it defined? How can you express what you feel, need, or want? It isn’t easy. The point is that language defines you and the way you think, and this started
66
6
Simon’s Error
at a specific moment of your life. Where does language (or the words you have been taught) end and start your “real” rationality? Is language something really yours? Or, has it been “imported” from somebody outside of you? Your language was external when you were a baby. You, and your parents, made it internal. But where is the internal divided from the external? Here is the idea. When you are learning words, your cognition is defined by this process. Learning your first words depends on the meaning you attribute to them, on reactions that you see on people around you, on the sound of your voice that you hear for the first time that way, from the meaning that is associated with the word you pronounce. Not only is your cognition shaped by this, but your behavior will also be soon affected. As a child you learn that there are words associated with specific actions so that, for example, “eat” or “hungry” means that you are about to be fed. Your cognition defines itself on the basis of these mental and behavioral processes. Again, where is the distinction between internal and external? When do words start to become “internal” facts of your cognition? Especially when they are associated with action, do they become “internal”? Child behavior and mental development are very peculiar, and I don’t intend to be so naive as to pretend that those processes can be associated or assimilated with those of adults. However, language provides a very interesting tool of analysis for adult minds too. When we speak, we are using something that is supposed to lie inside of us and we put it outside. Now, I would like to consider another example, an extension in the use of language. Everyone who has tried to write something knows very well that many of the ideas and the exact shape of your thinking come together with the act of writing. However, if your ideas come out when you write them, are they really inside your brain? Once again the same question: Where do they lie? The idea of rationality that Simon and others use lies on a separation between the brain and what stays outside of it so that, for example, individual capabilities are separated from external information.xvi Data stays outside while abilities stay inside. The assumption that stays at the basis of the distributed cognition approach eliminates this divide. Internal and external cognitive resources stay together in a “smart interplay,”xvii where cognition is dependent and shaped by these set of relations. A child that learns the first words and an adult that writes some text are common examples of how distributed cognition works. The written word is not inside the brain, but continues to play with it; there is a continuous shaping activity that goes on between internal and external.xviii The cognitive process includes the written words in such a way that it is very difficult to define where and how this external tool can be considered only an outside source. What we know is that it is a resource for the brain and, in many cases, cognitive processes change depending on the kind of resources we use. This is the reason why spoken words are different from written words, and the same written or spoken word has a different meaning in the cognitive process it enhances. This point can be framed with what we have learned from the previous chapters. Data representation allows individuals to give more or less sense to a given problem. However, what proponents of the distributed cognition argue is that this is only part
Distributing Cognition
67
of the process. It is not that the brain, with its well-defined mental abilities, makes sense of something, but that it is shaped by what stays outside of it.xix The cognitive process changes and sets different limits depending on the kind of external resource (a single word in our example) and the medium used to transfer that information (e.g., spoken or written resource). Our ability is not given a priori then, but it modifies together with the interaction that we have with the external resource. Our mind has plasticityxx because it is flexible and adaptable. This interaction between internal and external resources can be observed in many processes that define our everyday life. Anything that includes resources that are located outside of our brains can be analyzed through the distributed cognition approach. When you speak, write, or perform an action, your brain is shaped by these processes. Nevertheless, if I accept words such as to “shape,” “define,” or “structure,” I face the risk of falling under the same old internal-external dichotomy. I believe that we need to consider what the distributed cognition approach could contribute on a wider basis. Neural processes are affected by external resources in the sense that synapses (the way neurons connect to each other) start to emerge only when this external resource is considered.xxi Hence, there is no (or limited, or different) cognitive activity without external resources interacting with us. Even when we dream we picture these interactions. This means that the cognitive process is activated, maintained, or lost depending on this web of internal-external relations. This is what students of DC and I mean with verbs such as to shape or to define. The described process is so powerful that even if there is no actual handling of or contact with external resources, our synapses start to function. The so-called mirror neuronsxxii are supposed to mimic the actual activities that we perform, and in doing that they mimic the interaction between the brain and the external resources. For example, when we watch a basketball player, an ice skater, or a dancer that executes an incredible performance, our brains mimic that person, activating areas that are usually associated with movement. This is, neurologists tell us, the reason why we enjoy sports, music, films, and all other apparently brain-based activities. In short, we see ourselves playing, skating, dancing just the same as people that we are only watching. Isn’t it curious that to enjoy something, our brain needs to function like we are actually doing it? Using our vocabulary, the brain mimics an active distributed cognitive process.xxiii The point is that this looks like a distribution within the distribution since the player, the skater, and the dancer are external resources to our brain (distribution #1), and we make sense of them using a defined schema that sees us doing the same thing (distribution #2). One of the corollaries of this approach is that individuals perform differently when they exploit diverse external resources. Suppose that somebody asks you to draw and analyze a mathematical equation. You have two options: (a) use paper and pencil or (b) use a computer. If you choose option (a) you will probably need more time and, depending on your math skills, you will need to apply rules to study the function and get the solution; otherwise it shall take forever. Choosing option (b) is easier, it usually takes less time, and you can use software that helps you in the analysis (and I am not referring to the option to Google it). However, this works only if you know how to let a computer study an equation and solve the
68
6
Simon’s Error
problem for you. Otherwise, learning new software could be more painful than the paper-and-pencil option. What counts most is that independent of what the choice is, the cognitive processes associated with each option are different. Also, the two activities bring together mental patterns so different that the ideas and analyses they can suggest may be different. Ideas on what is most interesting in the function may vary together with the outcome of the study. Of course, a computer may be more accurate, while the drawing may suggest (stress) areas of interest in the function. Again, results have the potential to be different. The way cognition distributes these external resources defines the process. This also suggests that there is something that the previous model is not able to seriously take into consideration: the quality of external resources.xxiv Bounded rationality models are concerned with the quantity of information that can be handled. Those who attempt to map its bounds also fall under the same category.xxv
Boundaries of the Mind: The Through Doing Logic In real life, the choice between the two options—paper and pencil or computer— does not happen very often. People do not choose between painful or less painful approaches to a problem. They usually don’t like painful processes at all, unless they are forced or are masochists, so the choice goes in the preferred direction first. Let us hypothesize that the first medium chosen (external resource) is the computer. You start with your study function using the computer and draw the equation through a dedicated software. The answer to the question is immediate, but soon after that you try to study the equation, setting its maxima and minima. You need a couple of math passages and the computer won’t help. But, you want to go further with that: What do you do? Well, one should take paper and pencil and start with a partial derivative analysis. A few passages may be sufficient, but while you write a new world is open to you, so that you start this strange interaction between paper, computer, and your brain: “What happens if I change this coefficient so that it equals 1, and what if it equals 0?” And what if you transform the equation into two new ones, where the old one is just the sum of these two? Does this help your study? While you work on the paper, you check passages on the computer software to make sure you are using the math correctly. This is a work in progress. As you can see, in this example you are not using one single source as an aid to your work, you are using multiple sources at one time and modifying your cognition constantly. The point is that you are defining—and redefining—your cognitive processes “through doing.”xxvi This brings us to at least two implications. Implication 1. Your rationality is not well-defined in a specific moment in time, and neither are your capabilities. They change according to what you do; they are dynamic. A rational decision maker is one who adapts the cognitive processes at a fast pace. A rigid one-resource person has more limits that one who jumps from one to another. Yes, to solve a mathematical problem with paper and pencil is
Distributing Cognition
69
more complicated, and it looks like it indicates your potential. But, what about solving a mathematical problem with no external resources at all? No speech, no paper, no computer? Brain activity only. That might probably be a true “muscle” showing. However, things change when you have, as usual, access to multiple external resources. For example, suppose that (a) you have access to a computer with (b) an internet connection. Moreover, imagine that (c) a mathematician friend of yours is available and willing to help you, and (d) a specific software that analyzes mathematical problems has been installed on your computer. Now, what is the real “muscle” showing when it’s time to evaluate if you are being rational? Is it rational to do without the external resources available? Or, what “muscle” are you showing? Consider this second task: A manager of an important software company has been asked to make a decision on a new product. Financial and marketing managers think that the software can be sold as it is, since they know the needs of the company and that eventual bugs can be adjusted by offering free updates to customers. Software developers and engineers keep saying that there are too many bugs and that the company cannot afford to sell a defective product. The decision here comes out of multiple and contrasting resources. When the manager faces the alternative (if it exists) to make a decision with or without an appropriate consideration of these resources/opinions, he is interacting with these resources anyway. But let us take the first option. What if the manager goes for it: What should he expect? The point is that, even before the scenario-drawing activity, the manager knows that whatever decision is made will be redefined and fine-tuned once made. It is not new that, especially in uncertain and complex environments, managers rarely make decisions once and for all. Part of the usual cycle of any strategic decision-making handbook is double-checking that the action remains in line with what is planned. These feedbacksxxvii are (or should be) the essence of organizational decision making. They suggest that some decisions are “interactive” decisions, in that they change and develop together with their action, i.e., they develop “through doing.” The decision to come out with an unfinished product needs further decisions. It is like in a decision tree when you know that option “A” implies a second and maybe a third subsequent choice. In this case, the option implies corollary decisions that could end up modifying or changing or reverting, in the case of the recall, the original decision. The decision is taken with a high degree of uncertainty, where the manager is not willing to take any definitive action; that of version “waves” of the same software product seems to be one of the main strategies in that industry.xxviii Basically the software company starts selling the product while engineers and developers continue working on it. As soon as a new and improved version is ready, they sell it to the market without giving adequate communication so that the product is perceived as the same, even if it is not. When customers have problems with the first, the company can give them the new version for free or sell it as a software update. Another strategy is to push the innovative side of the company, having the product sold with progressive numbers where version 1.0 is usually the first, followed by 1.1, 1.2 or 1.01, 1.02. It is apparent that whatever the choice, the decision-making process happens in a time continuum, and that organizations (and their products) provide the basis for this continuity.
70
6
Simon’s Error
The second option, that of not selling the product, leads also to a “through doing” logic. The fact that the manager needs to routinely check developments of quality tests and software debugging is something that may lead him/her to a different decision day by day as well as to reinforce the initial decision. This process is based on the influence that external resources have on the manager, and the process can move these bounds far away from where they started. The ability to see a solution where there are none is typical of this function of our intellect.xxix We can think of all of these decisions as separate from the other but this is just a thought experiment that we can run for educational purposes since we know very well that the continuity of organizational processes provides a significant help to managers (and all of the participants). This way managers can track their (and others’) decisions where the next step is related to the previous ones. It is this way that people extend their limits and their rationality. Isn’t it true that we learn constantly? Isn’t it true that learning implies the fact that we can move our rational bounds? Implication 2. We learn from our mistakes as well as try to modify our behavior according to past experiences and, sometimes, creative and innovative thinking. These factors set variable boundaries to our minds. It is true that sometimes individuals are not successful in overcoming their mistakes but they do try to do that. The processes that lead to the “through doing” or “working” decision making follow this pattern. Managers, like every human being, make bad decisions sometimes, this is undeniable. Does this happen when they do not use their distributive cognition properly? Does this happen when they fail to exploit external resources? When they fail to learn? Or, is it just the nature of their rational bounds? More analysis is needed here. This is just to say that sometimes the boundaries of the mind are the ones that we put on it not the ones that it has. At the very end, the distributed cognition approach is something that makes this point apparent. The richness of a cognitive process lies on its interactions with multiple external resources at one time instead of believing that rational “muscles” are fixed and stable or that they can be defined by studying individuals that perform one simple task at a time.xxx Now, the question is this: Are organizations places where people perform one single task at a time or we are moving far away from this Taylorist conception?
Externalizing and Reprojecting We have seen that the divide between internal and external resources is not necessary—not even particularly helpful—to understand why our cognition works the way it does. We need to find a different explanation. We have also stressed that interactions (or interplay) between internal and external is worth studying. How does this work? Which are the components and determinants of the distribution process? Individuals have two ways to exploit external resources. The first is a process known as externalization.xxxi This is the tendency to disembodyxxxii part of one’s
Distributing Cognition
71
cognition to resources other than the brain.xxxiii For example, listening to the music, reading a book, writing, singing, or solving a logical problem come from cognitive processes directed toward making sense of our reality. Externalization can be active or passive. This is one of the most important distinctions to notice since it allows us to separate acts of will (voluntary behaviors) from involuntary behaviors. The latter includes making sense of music, for example. The music is something we listen to and it appears that processing information from the outside is internally based; however if we indulge on how this happens, we find that interaction stays at its basis. We need to make sense of this external activity (the sound) located outside. And we do it and process it with emotional reactions, among others. It is counterintuitive but this is what stays at the basis of the DC approach: Everything is not located here or there, it is the interaction that explains the cognition behind it. If we are musicians, then our externalization becomes a mix of active and passive. It becomes active when we start writing music or playing an instrument, i.e., when it is clear that we are putting things outside our internally defined brain. The idea that the mind needs something outside to work properly is apparent to anthropologists that explain our ancestors’ mural paintings as a big jump ahead in the evolution of the brain.xxxiv If this could be thought of as a first example of distributed cognition I leave it to experts but, the fact that there is a tendency to externalize is out there. The list of examples can be endless and facts of organizational life fall into this endless category. Financial, annual, social, environmental reports are a well-defined and structured way to externalize: meetings, standard operating procedures, contracts, negotiation and arbitration procedures, as also hiring or firing procedures, routines, roles expectations, and the like. Moreover, corporate culturexxxv can be included in these organizational externalizations since it is defined by behaviors that emerge from a group of people that take part in organizational activities. This process comes with another one that is not less important. What happens when people engage in an active externalization? What happens when the above mentioned musician writes the music or plays an instrument? This person is doing two things at the same time: She is disembodying part of her knowledge over an outside tool and making sense of it constantly. This means that once this music is externalized on a piece of paper, for example, the process is not over yet. The cognition continues to lean on what comes to be an external resource now and this fact changes the perspective together with the process that is ongoing. That note is not the abstract idea that the musician had in her mind anymore, now it is something written on a piece of paper that makes sense together with its context (other notes) or that assumes a different meaning when it is there. It could be that the single note sounds better in a different context or that this is no more what the musician intends to do with that music. In other words, the externalization leads to a reprojecting activity.xxxvi This is essentially what happens every time we externalize and every time we subsequently exploit an external resource. The cognitive process is made of the combination of these two processes. Again, where the external and the internal part end and/or start is not relevant; the interaction is what cognition is all about.
72
6
Simon’s Error
We can read the above-mentioned distinction stating that representations are external, as they are described by the externalization process, and they are internal, as they appear in the reprojecting phase, but we may fall into the same old cognitive divide. An example may help clarify these points. How is the annual social report of a company created? Companies define and report their socially responsible behaviors in so-called social reports. There are many guidelines on how to write effective social reportsxxxvii , and it is not the purpose of this book to present or analyze them. However, there is usually an organizational “motto” or mission with colorful (and reassuring or “careful”) pictures that leads to a table of contents. After this, there is a letter from the CEO, the President, or the Chairman of the Board of Directors. In the next pages, the report starts with all of its sections. The distributed cognition works from the side of the reader, as well as from that of the writers. The letter from the chairman, for example, has been drafted many times and it has been modified from that written by the original writer and many other individuals from her team, as well as PR specialists. Sometimes, external service companies are used to write and contribute to the writing of the report. These function as another source of fine-tuning of the original message. All of these activities of writing, reading, changing, rewriting, etc., are externalizing and reprojecting processes that shape the cognition of individuals involved and the decision-making process too. Without these processes, there will probably be no letter from the chairman. I know that some of you are thinking that this should probably be a better outcome, but let’s stay focused on the DC process. First, when the chairman writes the message, he experiences externalizing and reprojecting activities together. The choice of words may come from a “stream of consciousness” at first, but it can also be a very complex activity made of writing-reading-rereading-changing-writing-etc. When the chairman puts words into word-processing software, she is externalizing; when she is evaluating, thinking of, and choosing words, the chairman is reprojecting. As you can see, these two processes are intertwined, one affects the other. All other processes are made of passive externalizations (other people that read the draft) and reprojecting activities. When we read a word, we make sense of it depending on the fact that we have that word in our memoryxxxviii and on the basis of the way it relates to the other words in the sentence. This is a very common and widely studied phenomenon. Take the sentence “she is very eclectic since she is interested in many different activities.” The sentence can be understood even if you don’t know what the word “eclectic” means because the other words suggest a meaning for it. The reprojecting happens all the time, and we have to make sense of things. The fact that people read the report means that they have to attribute to the signs that form the message a meaning that may be different from what they have learned before. Moreover, when reading something from a given source, e.g., when you know that you are reading something from your boss, it also implies that you frame the role, ideas of that person, beliefs, and all that you think of that person and of the situation you are facing (logic of appropriateness). Therefore the reprojecting may go well beyond the simple word-related meaning of words. This last point, for example, is particularly meaningful if we look at the reader of the report that is not part of the organization. The process this individual goes through is similar to that of an insider,
Distributing Cognition
73
except that she/he is reading an externalization that she/he perceives as being related to the company as a whole. How far can we go with these concepts when we analyze decision making in organizations? What can we get from the use of a DC framework to organizational decision-making processes? The next two chapters (especially Chapter 8) offer some analytical hypotheses.
The Extended Mind What I have tried to present here is apparent in the work of many students of the distributed cognition approach. In summary, the most relevant points that emerge from this approach are different from the old paradigm on which bounded rationality is based. Some of the achievements are that (a) the divide between external and internal cognitive resources has a limited explanatory power, and (b) we can have a better understanding of human cognition if we abandon that paradigm. The idea is that to understand cognition, we need not focus only on its internal bounds. As Wilson suggests, Accepting the extended mind thesis means holding that the mind is not physically bounded by the body but extends into the environment of the organism. [. . .] The extended mind version of externalism represents a stronger and more striking view of the mind than do earlier forms of externalism. It is a view that embraces the claim that technological and cultural artifacts may be physically constitutive of cognition.xxxix
This is exactly the take I have supported in this chapter: Our mind extends to external “aids” that define and shape cognition. To this respect, the mind is extended. And this is the reason why that in order to understand cognition—and rationality as part of it—we must take into consideration all the constituents of our cognitive activity. These constituents are not only internal, nor are they entirely separable from the outside variables. In a part of his work dedicated to learning, Hutchins writes that [t]his heavy interaction of internal and external structure suggests that the boundary between inside and outside, or between individual and the context, should be softened.xl
This is very close to the idea of the “soft self” that Clarkxli explains in his work: There is no self, if by self we mean some central cognitive essence that makes me who and what I am. In its place there is just the “soft self”: a rough-and-tumble, control-sharing coalition of processes—some neural, some bodily, some technological—and an ongoing drive to tell a story, to paint a picture in which “I” am the central player.xlii
The DC approach is not a complete break with past tradition of thoughts. Wilson’s discussion of collective memory is particularly helpful: For between the individualistic approach and the collectivist approach are extended mind views that, in some sense, borrow from both. From individualistic approaches, they accept that remembering is an activity that is done by individuals, and from collectivist approaches they take the idea that this activity is not bounded by what goes on in the head of the individual, and so encompasses commemorative objects and practices, mnemonic devices and strategies, external symbols and structures.xliii
74
6
Simon’s Error
We can go on with many other examples of what I have tried to evoke in this chapter: Bounds are not what bounded rationality students think they are. For the economy of the discourse, I believe I should stop here with examples from the literature and move forward. Here comes the need to spend a few words on bounded rationality.
How Bounded Is Rationality? Now, this is the question: How bounded is rationality? What the distributed cognition approach suggests is that a theory of the limits of our cognitive activities is helpful, but it falls short when it has to define the dynamic and plastic functions of human intellect. Do we solve problems within our rational bounds, or do we solve them because we are able to move these bounds? Do we make decisions depending on the way we are bounded, or do we make decisions based on a variation of these bounds? I believe these are not trivial questions and they need to be addressed. The first question is about successful decision making (problem solving), while the second is broader since it includes all kinds of decisions.
Cognition and Rationality Before addressing these questions, there is one point that needs to be considered. What is the relation between cognition and rationality? Until now I have used these two as synonyms with a certain degree of vagueness. Although there is a common ground between the two words, they maintain their differences. While cognition refers to every mental process,xliv rationality gets involved when these processes follow sound and consistent procedures and have effective outcomes (see previous chapters). It is clear that rationality is a specific type of cognitive process. However, the study of rationality in the last decades has emphasized irrational behavior and flawed decision making.xlv Moreover, many studies on cognition and decision making have appeared to explain and make sense of rationality. Right now, rationality is grounded in cognitive science for decision-making students, and different meanings of it have started to appear and be used by the scientific community. The interchangeability of rationality and cognition misrepresents the complexity of the issue, but, since there is no understanding of rationality without cognition, I believe that we need to define which cognitive hypotheses stay at the basis of any rational thought. This short note is useful for providing guidance in the use of the terms and in fact the answer to the question that entitles this section can be answered only if we look at cognitive theories. And the answer is clear, if DC is preferred to other approaches.
Problem Solving The starting point is the first, which is the easiest question: Do we solve problems within our rational bounds, or do we solve them because we are able to move these bounds? Bounds to human cognition depend on how individuals exploit external resources. One point here may be of stating that this is still a problem
How Bounded Is Rationality?
75
that pertains to individual cognitive capabilities. I am not arguing, nor are scholars and proponents of the DC, that all human beings have equal potential or the same capabilities. Individuals are different and they sometimes react differently to similar threats or problems; however, this is but a secondary point. This is because everybody is different, their minds work in a slightly different way so that an external resource can affect their cognition differently. This deals in part with the representation, accessibility, and framing processes. Why do they matter? They are important because of the way cognition interacts with external resources. This is not an effort to falsify previous attempts to make sense of rationality; the aim is to improve and redefine them on the basis of what I believe is a more accurate and broader approach to the problem at hand. The externalization process explains how bounds move. When a manager finds a solution to a problem, what are the factors most important to his or her cognitions? They are related to how the manager has been successful in moving his or her bounds, how he or she overcame the limits that impaired his or her decisionmaking process. These were limits of cognition that could move when the manager uses an externalization strategy (without even knowing what he or she is doing from the cognitive perspective). Of course, not every externalization process leads to a successful decision. Externalizations describe the ways our cognition works so that it shows how dynamic the process is and not how limited it is instead. Every successful decision passes through a process of externalization and reprojecting where interplay of resources causes our cognition to work more smoothly. Therefore, the answer to the question cannot be a straightforward statement where we deny the first part of the question and affirm the second. Human rationality is bounded, but these bounds move consistently with the nature of the problem and with the type of resources that are exploited. Are we still bounded then? Yes. Do we move these bounds? Yes. The point is that the way rational bounds move redefines cognition and rationality. Its plasticity and dynamism should be used to define human problem solving. Think of the way individuals adapt themselves (reset their limits) when asked to make decisions and solve problems in different social environments (e.g., office, family, church), or when different behaviors are in play (e.g., shopping, playing sports, studying). The individual is the same, but his cognition varies. Different problems call into play diverse bounds and cognitive processes, and the same problem may call for that bound adapting to the solution. What happens when you solve a problem because of new data found online, of advice from a colleague, of a reading? That impression that you are “connecting the dots,” that something is materializing in your brain, is the way our cognition works: “eppur si muove!”xlvi
Moving Bounds The second question is more general and includes all decision-making problems: Do we make decisions depending on the way we are bounded, or do we make decisions based on a variation of these bounds? Sometimes failure to solve problems can be described as failure to adequately exploit external resources or the lack of enough resources to exploit. If the understanding and analysis of a
76
6
Simon’s Error
problem depend on the three above mentioned activities—representation, accessibility, and framing—then individuals (and managers) should look more at access to information than to have more. The same piece of information should mean different things to different people, and sometimes two pieces of information mean the same to different individuals. This means that the external resource is what matters most in this analysis. People could fail in their decision making (i.e., make bad decisions) precisely because there is a failure in this exploitation process. The idea is that bounds move with the cognitive process. And if bounds are not stable and they are not the point in describing rationality, what is the concept of bounded rationality then? The distributed cognition provides us with an innovative take on decision making and rationality; the next chapter is an attempt to define how the concept of rationality changes as far as this approach is concerned. Before that, we still need to unveil Simon’s error, although it should be apparent at this point.
What Error? I do not pretend that my review of the DC framework is exhaustive, nor do I believe that this chapter presents the distributed cognition approach. I have tried to show some of the most relevant concepts that many of the DC approaches share, with an emphasis on those ideas that seem more significant to the discourse of rationality and to their transfer to organizational behavior. What this chapter presents can hardly be defined as mainstream cognitive science, but ideas described here capture something of what seem to be an incoming trend in that field, as of today. I believe that the reader now has an idea of what I mean by “Simon’s error.” And it is the stable divide between internal and external cognitive resources that bounded rationality shares with neoclassical economics (and expected utility theory). In the face of distributed cognition, this divide brings the theory to a critical point that has not been considered by recent BR developments and studies. Truth to be told, now that you have an overview of DC hypotheses, we can consider more than one “error” that can be associated with the idea of bounded rationality. Some of them have been implicitly or explicitly considered in this chapter, while others have not. These errors—it would be better to abandon the provocative word “errors” and switch to “critiques” or “challenges”—can be classified into three categories: 1. Neoclassical flavor. The idea of BR shares some of its assumptions with the theory Simon wanted to criticize. In particular, it maintains a strict relationship between means and ends (teleology),xlvii it uses a brute-force computational strategy to mimic the human brain,xlviii and, as we have seen, it does not abandon the “cognitive divide.” 2. Computer metaphor of the brain. BR is based on Simon’s belief that the human brain works like a computer so that symbols and serial processing of data constitute what needs to be studied.xlix This has also been called Simon’s paradox:
Notes
77
the extreme high esteem of computer powers and the lack of belief in human capabilities.l 3. In vacuum. The boundedly rational agent behaves in a social and resource vacuum.li This means that the decision maker operates in an environment where social, cultural, political, technological, or demographic variables are not relevant to the efficiency and effectiveness of his/her decision. There are many corollaries that stem from these three assumptions. As far as what is presented in this chapter is concerned, the most significant corollaries are that (a) information is only quantitatively relevant to the decision maker (i.e., quality, sources, mediums, content are not relevant), and (b) individuals have stable cognitive capabilities and preferences over time. The overwhelming number of studies that map BR takes most of these assumptions for granted. It is very unusual to find somebody, among these scholars, that questions or criticizes some or even one single assumption on which BR is founded. In the next chapters I suggest how the DC approach urges to start a discussion on these assumptions that may eventually lead to corrections and to a remodeling of BR.
Summary In this chapter I have presented the distributed cognition approach. This theory is based on the “divide removal,” i.e., internal and external resources are in constant interplay, and cognition is defined and shaped by these activities. Boundaries of the mind are then redefined along with the two representation processes: reprojecting and externalizing. The result is an “extended” mind that challenges several of the assumptions on which bounded rationality is based. The final part of the chapter offers the moving bounds hypothesis and presents an overview of Simon’s error(s).
Notes i. This is clear in Simon (1979), where he compares bounded to full rationality; see also Simon (1997). ii. This approach to bounded rationality is very diffuse among students of behavioral economics. It seems that Simon inaugurated a trend that suggests to scholars how to introduce this idea. Or, it is more likely that these scholars tend to be “soft” and pay a tribute to mainstream economics, decision making, organizational behavior, etc. See for example, Aumann (1997), Camerer (1998, 2007), Conlisk (1996), Foss (2003b), Gigerenzer et al. (1999), Grüne-Yanoff (2007), Hanoch (2002), Kahneman and Tversky (1979), Klaes and Sent (2005), Knudsen and Levinthal (2007), March and Shapira (1987), Mumby and Putnam (1992), Patokorpi (2008), Rothschild (2001), Rubinstein (1998), Selten (1998), Shakun (2001), Todd and Gigerenzer (2003), and Zafirovski (1999). iii. Yes, I am referring to what you are thinking: Popper (1935/2002).
78
6
Simon’s Error
iv. Besides Hutchins (1995), there are several works that offer a framework of what the distributed cognition approach is; among many others, a starting point could be found in Clark (2003, 2008), Clark and Chalmers (1998), Thagard (2007), Magnani (2007) and Magnani et al. (1999). v. I cover this domain in Bardone and Secchi (2009) and Secchi and Bardone (2009b). vi. Richardson (2000). vii. von Neuman and Morgenstern (1944) and Simon (1947/1997, 1955). viii. This may derive from the interaction between external variables and individual capabilities, so that cognition can be described in terms of some interactive process. This approach is commonly used in organizational theory; see Beach (1998), Laroche (1995), Starbuck (1983), and Walsh (1995). ix. Axelrod (1976) and Cannon-Bowers and Salas (2001). x. Bardone and Secchi (2009, p. 184). xi. Clark (2003), Clark and Chalmers (1998), Hutchins (1995), Magnani (2001, 2007), Norman (1993), Perry (2003) and Wilson (1994, 2004). The distributed cognition (DC) approach has its enthusiastic fans and it has, of course, its opponents. Among the others, I have found arguments by Adams and Aizawa (2008) particularly detailed and I suggest reading Harris (2004), for a different take on the topic “language and DC,” also discussed below in the text. Andy Clark, one of the founding fathers of this approach, also offers some critical remarks in the last chapter of his book, Natural-Born Cyborg (2004). A good theorist always tries to find limits in the theory; and Clark undoubtedly is an excellent scientist. xii. This approach to human cognition takes a step further to the “ecological” approach of the mind as it is presented by Gigerenzer et al. (1999) and Todd and Gigerenzer (2003), among many other writings of these authors. xiii. Gazzaniga (2008). xiv. The social brain hypothesis by Dunbar (1998), explains this point. See also Dunbar and Shultz (2007). xv. Better examples on how language could be related to distributed cognition are in Love (2004), and Sutton (2004). xvi. This stays within the definition of bounded rationality (see Chapter 3) and it is not questioned by all of those that have provided maps rationality (e.g., Goldstein and Gigerenzer, 1996; Gigerenzer and Brighton, 2009; Kahneman, 2003). xvii. Clark and Chalmers (1998). xviii. Hutchins (1995, p. 287f). xix. Clark and Chalmers (1998) and Magnani (2007). xx. For details on plasticity, see Clark (2003). xxi. The central chapters of Hutchins (1995), can be read under this assumption. xxii. This is explained in Gazzaniga (2008), Chapter 3. xxiii. See Magnani (2007). xxiv. Mousavi and Garrison (1992). xxv. There is no explicit discussion of the quality/quantity dichotomy in BR studies. Mousavi and Garrison’s (1992) critique holds. xxvi. An exemplification of thought doing rationality in the case of morality is provided by Magnani (2007); see also Magnani et al. (2006), on the same topic. xxvii. See March (1994), end of chapter 1. xxviii. I don’t know if this strategy is deliberate so that my reading of it is a guess on what is available through the media. I am using abduction here! xxix. It is a typical mental leap, as the ones described by Holyoak and Thagard (1995). xxx. This is also the meaning of Hutchins’ (1995) “in the wild.” xxxi. Wilson (2005), Clark and Chalmers (1998) and Zhang (1997). xxxii. However, as Gibbs (2005), shows in his work, this remains a “tendency,” an attempt only, since all cognitive activities are embodied. xxxiii. See Magnani (2006).
Notes
79
xxxiv. Or that explain how artifacts such as a statue half-man/half-lion could possibly exist. That mythological figure has no room inside or outside the brain, it is the interplay between the two that made it possible. See Magnani (2006), and Mithen (1999). xxxv. Hutchins’ Cognition in the Wild is a study of the organizational culture of the ship Palau. xxxvi. In one of his writings, Magnani (2006) makes this point clear: “I maintain that representations are external and internal. We can say that (a) external representations are formed by external materials that express (through reification) concepts and problems that do not have a natural home in the brain; (b) internalized representations are internal re-projections, a kind of recapitulations, (learning) of external representations in terms of neural patterns of activation in the brain. They can sometimes be “internally” manipulated like external objects and can originate new internal reconstructed representations through the neural activity of transformation and integration” (italics in the original text, p. 350). See also Magnani (2007). xxxvii. For example, some of these guidelines are the SA8000, Global Reporting Initiative, AA1000. xxxviii. Otherwise, we use yet another external resource: a dictionary. xxxix. Wilson (2005, p. 230). xl. Hutchins (1995, p. 288). xli. Clark (2003, p. 130f). xlii. Clark (2003, p. 138). xliii. Wilson (2005, pp. 231–232). xliv. Thagard (2007). xlv. It is the work by Kahneman et al., (1990; 1991) that I am referring to here. Ariely (2008) offers an overview of behavioral studies on irrational behavior. xlvi. Translation: “. . .nonetheless, it moves!” Statement that Galileo Galilei is supposed to have pronounced after the trial held by the Inquisition when he was forced to deny that the Earth moves around the Sun. xlvii. This aspect has been analyzed by Mousavi and Garrison (1992). xlviii. Zelený (2001). xlix. In his own words, “thinking is governed by programs that organize myriads of simple information processes—or symbolic manipulating processes if you like—into orderly, complex sequences that are responsive to and adaptive to the task environment and the clues that are extracted from that environment as sequences unfold” (Simon, 1960, p. 81). See also Simon (1993a), and some works that highlight these points, e.g., Sent (1997) and Patokorpi (2008). It is also interesting to notice that this second error (computer metaphor) has strong relations with the first (neoclassical flavor). As some authors put it, “in so discrediting economic rationality, Simon nonetheless remained true to the broader but no less conventional notion of what might be labelled cerebral rationality, that decision making is a cognitive process that can be decomposed into a sequence of simple, programmed steps” (Langley et al., 1995, p. 262). l. Zelený (2001). li. “Simon’s psychology of economic rationality not only assumes a socially decontextualized mind whose primary function is to carry out Turing machine calculations, he also assumes a mind that is decontextualized from the material conditions of the environment involved in the agent’s habits and impulses that give substance to economic need and desire” (Mousavi and Garrison, 1992, p. 152). See also Langley et al. (1995).
Chapter 7
Stretching the Bounds (I)
What remains of Simon’s idea of rationality after the distributed cognition (DC) approach is considered? The DC approach points out that one of the most important activities of human cognition is that it is shaped by what is found in the environment, i.e., external resources or artifacts. What is or is not rational assumes a completely different meaning within this framework. The idea that the brain works in isolation and that information is taken from the outside, then computed, and subsequently expelled as a solution to a given problem largely overlooks what happens during the simplest cognitive process. One thing is clear: If we accept the DC framework, we can no longer rely on this input-process-outputi idea of rationality and decision making. First, there is no simple input since every external resource is part of the process and second, the output is a process. Externalizing and reprojecting are aimed at explaining that there is no divide between these stages in the cognitive process. The idea of bounded rationality (BR) is based on this input-process-output as it appears clearly from the first chapters of this book and from the last paragraph of Chapter 6. While studies and empirical assessments of this concept have emerged in the last few decades, I agree with Nicolaiii Foss that bounded rationality is “much cited and little used.” In fact, there are few studies that reconsider Simon’s original idea and try to explore its theoretical fallacies and/or interpretative shortcomingsiii . As far as I know, the book you are reading remains one of the few works aimed at getting a different idea of rationality in organizationsiv . It is apparent from the previous pages that I do not believe the concept of bounded rationality needs to be abandoned. Quite the contrary: There is enough evidence leading to a better definition of it, on the basis of a more realistic and, maybe, counterintuitive approach. However, these last points really don’t matter if the DC approach provides a better understanding of what we study. The analysis whether the old BR and DC are inconsistent or not needs to be developed further. I only gave the negative critique without going further on the positive proposition. This and the next chapter serve exactly this scope. They present an attempt to define how rationality extends over its bounds and where the analysis should be focused when a theory of distributed cognition is considered. This is the core of the book and explains its title. When referring to rationality, I use the term extendablev to highlight its instability and plasticity, to define how rationality tends D. Secchi, Extendable Rationality, Organizational Change and Innovation, C Springer Science+Business Media, LLC 2011 DOI 10.1007/978-1-4419-7542-3_7,
81
82
7
Stretching the Bounds (I)
to modify itself to cope with the challenges of decision making and problem solving. Why extendable? The idea of extendable rationality is related to the following: (a) human cognitive processes have the potential to modify themselves according to the context (external resources) and/or to the problem at hand; this makes rationality “extendable” in the sense that sometimes it does extend, sometimes it does not; (b) rationality is defined through its fitness to a given situation, it varies depending on both individual and environmental factors, and how the two fit together in this respect the modification, dynamic, plasticity is an extension (or, better, a redefinition) of individual bounds to reach a better fit with the context; (c) the outcome of a rational process is not suboptimal—what is the benchmark for suboptimality?—or satisficing (individualistic approach), but it is workable, i.e., it is appropriate in respect to the situation and/or to the problem at hand. If rationality is considered in its extendibility, then we need to start analyzing what makes this extendibility possible. This and the next chapter present two areas where the old idea of bounded rationality had explanatory problems and the extendable rationality comes well into play instead. The first area includes the so-called through-doing decision processes; these are dynamic contexts where the decision maker changes and refines the process of decision making together with a first decision and the change in the context. The interplay with external resources (including those externalized) is crucial. The second area (Chapter 8) includes those processes that come into play when other people are considered, i.e., the social side of rationality (see previous chapter). There, I analyze how other people affect the way we make decisions, considering two specific examples (advice giving-taking and bandwagon). This latter point is largely overlooked in studies of bounded rationality but, as we will see, it is fundamental for any human decision-making process.vi
Through Doing Decision Making This category includes decisions that are not well defined when the decision maker starts the cognitive processvii . These are in fieri decisions, meaning they progress as the situation evolves and resources become available. Some of them are intentionally made to follow up with other decisions (exploratory), while others are thought to be well-defined and “last” decisions that reveal the need for further elaboration while the decision is being made. The typical feedback processviii of a decision is part of the decision-making process itself, and happens at the same time the decision is being refined. This defines the through doing or emerging decision-making process. There are several cases that should be made to explain how through doing decision works. In the following pages, I consider two of them as paradigmatic
Through Doing Decision Making
83
examples: emotions and morality. Other examples of through doing decision making that are not considered here but seem interesting to explore include maintaining leadership, making sense of altruism, and personal and professional development, understanding corporate culture, adapting to the informal (behavioral) social structure of an organization, and more.
Emotions and Decision Making In this book I have considered emotions when mapping bounded rationality. Studies on emotions show how individuals are boundedly rational since they are together an aid and a threat to decision makingix . However, we have also seen that neurological studies have found that a rational decision maker cannot come to any conclusion without emotions. Emotions are needed to make decisions, they help us when quick decisions are needed, and they trigger heuristics, specifically fast and frugal heuristics. What emotions and heuristics in general help us do is that they prevent us from looking at our bounds and provide us with the solution to a given problem. This solution may be a certain behavior, a thought, an action, or anything that could help us with a given problem. The decision comes out of something that remains related to the way we usually process information, but sometimes it comes “out of the blue,” to use an expression everybody is familiar with. It is not crucial here to define conditions under which emotion can enhance or endanger the outcome of a decision-making process, although these are very relevant points in recent researchx . I would like to focus more on the meaning of emotions for cognitive processes: Why are they so important that we cannot do without them? One thing is that emotions are embedded in humans so that, at least up to a certain point (primary emotionsxi ), they are part of our genetic heritage. In human beings, these primary emotions are associated with complex thoughts or secondary emotionsxii . When individuals make decisions, the consideration of eventual consequences of that decision may lead to particular emotions when the decision is about to be made. How many times do you check an important e-mail before sending it out? And what does your wet hand feel like? Emotions are associated with specific thoughts, and they can be a way to externalize or to have bodily reactions to these specific thoughtsxiii . Now, the body is something outside the brain, it is an external resource according to the definition (everything that is not the brain is an external resource). We can argue that this “external” is very close to the brain but still external. There is a continuity between body and brain, and this is the interesting part with emotions. As soon as the emotion comes out, the body reactsxiv . We know this for sure when we are in love, dissatisfied, experience fear of something, sue a company, call a friend, or anything else. Our body reacts or represents emotions in many different ways; this is what I should call the externalization of emotions. It is in the external resources and body-brain interaction that emotions can be defined as through-doing activities. This becomes apparent when we witness how emotions evolve within a specific amount of time.
84
7
Stretching the Bounds (I)
Take the case of change. When change is disruptive, individuals usually react with emotional arousalxv . When a company introduces significant changes in procedures, or technological innovation is brought into the production process, individual responses are framed through their emotionsxvi . In the case of mergers and acquisitions, for example, the new board of directors and CEO usually attempt to get the acquired company closer to the acquiring one. Layoffs, changes in the chain of command, higher turnover, new standard operating procedures, and more, are likely to be implemented. Many people that passively witness all of these changes react with a mix of emotions. Let us take the case of someone being removed from the charge of leadership and replaced with a new boss who comes from the acquiring company. The discharged individual would pass through different states, including but not limited to the (a) shock/surprise (e.g., “What is going on?”) together with refusal or denial (e.g., “This isn’t possible!”), (b) defense, such as an attempt to rationalize (e.g., “This was the right thing for the company”), (c) discarding or acceptance of the new situation (e.g., “I am okay with this”), (d) adaptation, that involves learning (e.g., “How can I do my new job?”), and (e) internalization (e.g., “This is working well!”; “I have a new role in this team.”)xvii . Of course, each one of these stages is associated with emotions. As models of individual emotional reactions to change show, these need the body to be felt. Once they are out there, individuals can interact with them on a more conscious basis. I am not trying to understand which one comes first between cognition or emotions (affect)xviii ; my intention is to show that they usually stay together and that they need a process that is based on (a) externalizations and (b) continuity. This second condition means that emotions are not stable over time and they modify together with the exploitation of resources available. They evolve through these resources. Less apparent is the reprojecting of our emotions. The fact that they are externalized leaves room for taking these as external and to start thinking about them. For example, the fact that your heart starts beating at a faster rate when you first met your loved one forces you to consider behaviors associated with that state of mind when it goes too far. Sometimes it is too late and you have already said something stupid or that you didn’t mean. The point is that there is a process of redefinition of your cognition that increases when emotional states are more intense. This lets you realize the emotional state you have run into. A recent idea on intelligence can help us here. A recent approach to human intelligence is that it is composed of many different “areas.” There is the traditional (a) mathematical or logical intelligence that deals with the ability of the individual to solve formal problemsxix , the (b) cultural intelligencexx associated with the ability to understand and cope with people of different cultures, the (c) moral intelligencexxi that is related to the ability to behave well in different ethical contexts, (d) social intelligencexxii a concept we will deal with in the next chapter, and the (e) emotional intelligencexxiii that is about the ability to deal with one’s own and other people’s emotions. I believe that this last area of intelligence is nothing but the ability to gain self-knowledge of the distributed process of cognition that deals with emotions. The idea is to connect emotional intelligence to the DC approach. In particular, reprojecting seems to be very close
Through Doing Decision Making
85
to this idea of intelligence. A research question associated with this could be inquiring if this is a typical trait of people with high emotional intelligence quotients, that of having a better understanding of their (and others’) cognitive processes. While the self-detection and understanding of emotional intelligence is somehow integrated to theories of bounded rationalityxxiv , what concerns other individuals is not. If we externalize emotions, everybody does the same. In a social environment where bodily reactions (e.g., a forced smile, a twinkle of an eye, crossed arms or legs, arms down the body) are widely diffuse, we have learned to decode these expressions and to adjust our behaviors according to perceptions. Or, we use these as external resources and take them as major bases for our decision making. For example, it is easier to ask for information when the person at the info-center greets us with a smile. Otherwise, we opt to pick up a brochure and do whatever we need to do on our own. This is a typical through-doing process. The decision to ask a question or not may pass through the (light) emotion that encourages us to adapt and modify our behavior as we go. Our everyday decision making is affected by these simple events; we can think of a work environment where the same dynamic is in place. It is easier to ask advice from a colleague or from your boss when she greets you with a smile or when you “feel” from her bodily expressions that it is a pleasure for her to help you. This “external” altruistic-sided face of emotionsxxv is less considered and studied in organizational behavior literature, but it remains important. How many of us make simple decisions—of potentially high impacts, especially if repeated—on the basis of these perceptions? How could this affect your solution to a given problem? This is a practical note on distributed cognition that needs to be analyzed and developed further in future works.
Morality In his book Morality in a Technological World, Magnanixxvi presents an innovative idea of morality that is consistent with the assumptions of the distributed cognition approach and with the idea of extendable rationality. What ethical theory always postulates is that ethics is something fixed, determined by the rules that guide individual behavior in our societies. De George points out that [i]n its most general sense, ethics is a systematic attempt to make sense of our individual and social moral experience, in such a way as to determine the rules that ought to govern human conduct, the values worth pursuing, and the character traits deserving development in life.xxvii
It is apparent from the definition that we should probably be able to distinguish between ethics and morality, the latter being the behavioral (descriptive) and the former being the normative sides of the same coin. Even if we define the two sides this way, it is particularly difficult to divide the two into separate domains of study. Morality evolves with societies and so does ethics. If we think of racial issues,
86
7
Stretching the Bounds (I)
women’s rights, temperance, animal rights, or many other ethical topics, we soon realize that ethical rules and practices change. What was labeled as outrageous by the majority of people only a few decades ago is now a moral standard. The framework that the distributed cognition provides us with is that of a human being that exploits external resources in a way that redefines her cognition. While this is easy to figure out when we consider material external resources, it can be less apparent when we deal with constructs of the mind, such as ethics.xxviii Consider a work setting. The leader of the team asks one of the members to perform the following task: Check if the outcome and quality control reveal that the plant needs to renew machines involved in that production process and submit a report. One week after a first deadline, the employee has not submitted any report, but the leader knows that the employee’s mother had spent the last two days in the hospital for a heart attack. However, the employee takes this allowance for late report submission as something that the leader should give him, as something for granted. The leader doesn’t like it and decides to write an e-mail to this employee, explaining what the policy of the company is and how good-hearted she has been in overlooking this inefficiency. The e-mail is there and the leader is satisfied with the lesson she is about to teach this employee. She is ready to send the message. She clicks on the “send” button, and a pop-up window associated with spell-check appears: “Are you sure you want to send the e-mail?” She never runs a spell check, so the window is just annoying. Not this time. She has had time to rethink that e-mail, the problems that the employee just had with his family, the fact that she wrote the e-mail on the basis of an impression. What if she is wrong? She rereads the e-mail and decides not to send it. What happened? The points I would like to stress here are two: (a) What seems appropriate at first wasn’t so appropriate when the second thought came into play; (b) technology served as a facilitator to this process. A diffuse moral rule (e.g., the “golden rule” of Jesus) states that you must treat everybody as you want to be treated. This was what was going on in the case of the e-mail. However, the thoughts of the team leader and the final outcome changed “in the making.” The decision was finally made through doing (writing the e-mail). This helped the leader change her moral interpretation of the case. After all, was it ethical to give the employee a warning after her mother had that serious health problem? The second point is also important since it tells us that resources can mediatexxix morality. In the example, the pop-up window served as a mechanism that helped the leader change her mind. This mediation is a corollary of the distributed cognition approach, since every resource that individuals exploit may be interpreted as a mediator of a particular meaning. This is what morality also is: knowledge that helps individuals in their decision making. External resources—the e-mail software in this example—support cognitive processes and help the individual make decisions. This simple example offers a basic idea of how morality is distributed. Our rationality adapts to situations, making the solution more appropriate; i.e., something that suits our needs. A final remark on the interpretation of morality is needed here. What does this cognitive understanding bring to the study of morality? As Magnani points out, the
Through Doing Decision Making
87
distributed cognition approach shows the reasons why most individuals do not stick with the same rule but adapt its interpretation depending on the situation. This also has the potential to provide a cognitive explanation of why norms that have been written several thousand years ago still maintain their impacts on people. In short, what DC implies for ethical norms is that their understanding is situational because it adapts to cognitive interactions. This means that ethical norms may be stated once and for all; it is their framing that changes. A significant role is played, in the process, by external resources. In the example that we used above, the boss could have sent the e-mail. In that case, the result of sending the e-mail and its interpretation would have been different. Maybe a rationalization process might have prevailed, with the boss trying to make a rational explanation of the reason why her ethical principle (e.g., the golden rule) needed to be adapted. The e-mail worked as a mediator of that ethical principle; its absence made the individual replace that mediator with other mediators—e.g., the e-mail itself, thoughts, other reinforcements such as the memory of the company policy on internal deadlines or the code of conduct. There is a need to anchor one’s cognition to external resources. And morality is no exception to this process. For this reason, understanding ethics in practice is better off when using a through-doing logic.
A Third Logic: The Logic of Adaptiveness In Chapter 5 I introduced the two logics that help frame rationality and decision making. One is the logic of consequence and the other is that of appropriateness. As you may recall, these two logics describe how decisions happen, and they do so through a set of four and three questions, respectively. The decision maker (or the analyst) answers these questions and makes sense of the process. The logic of consequence is based on a strict logical progression of arguments, and it is particularly useful for interpreting decisions on the basis of the old paradigm of rationality. Therefore, it analyzes alternatives, expectations, and preferences, and defines a decision rule. The logic of appropriateness is instead more concerned with recognition, identities, social roles, and decision rules that apply to a given situation. This second logic is close to the socialized view of the individual and of rationality. Does DC fit one of the two logics? Where does it fall? If one of the two logics has to be picked up, I would go with appropriateness. That logic invites researchers to make sense of decisions through the cultural, social, and political environments, where individual identity is questioned and involved; it also deals with social external resources and internal sense-making activities. However, how can you include externalization and reprojecting? How do you define the non-definitive component (i.e., through doing) of every decision? How can you distinguish among resources? In other words, there is something that the two logics do not catch and that we can try to include in a third logic. Consistent with the previous logics, I would like to proceed with questions that help frame the decision-making processes. What follows here is a description of
88
7
Stretching the Bounds (I)
what I call the logic of adaptivenessxxx that indicates that decisions should be studied on the basis of resource availability, their actual exploitation, and pertinence to the problem at hand. There are three questions to this logic: 1. The question of context: What resources am I exploiting? 2. The question of fitness: How good is the use of resources? 3. The question of change: How fast is the situation evolving? The first question relates to the context: What resources am I exploiting? The analysis of the decision passes through the careful consideration of the type of resources the decision maker has (implicitly or explicitly) chosen to use in his/her cognitive process. As we know from the DC general assumptions, what stays inside the brain depends on what is outside such that “resources” available are those that are involved in the cognitive effort. Potential use of resources should be separated by actual use of resources. The second directly affects the cognitive process, while the first only indirectly shapes it. Although more on external resources is written in the next chapter, I anticipate here that there is a distinction between social and non-social. The second question is that of fitness: How good is the use of resources? The degree of success, efficiency, or effectiveness of a decision may depend on the interplay between what is inside and what is outside the brain. This implies that not every resource is good for everybody. Instead, there should be a good fit between the decision maker and resources at hand. Consider the following simple example. The director of a PhD program in Human Resource Management asks one of his students to write a review of an industrial economics paper. The author of the article presents and discusses a mathematical model of firm-supplier relations. Our PhD student feels incompetent and has no idea how to write the review. In this simplistic case, there are two resources available. The first is to carefully read the paper and try to make sense of the strange language through his own knowledge, Google, and other library or online resources. The second is to carefully read the paper and then bother a friend or another PhD student at the economics department. The first option may be cognitively much more expensive than the second; however, there are time constraints. The review may not take forever, and the deadline is the next week. Is the first option worthwhile? How many papers like this will the PhD student review in his entire career? It seems that the second option would work better because its fitness (internal-external interplay) is less likely to produce uncertain results than the first. Although this case probably needs to be treated more carefully, the fitness question is of utmost importance and should be asked every time the decision maker is about to use a cognitive external resource. What is the cognitive effort needed to make a regression analysis with software you don’t know? This may be another example of low fitness. However, fitness is subjective, and the analysis should discount personal matches of internal-external resources. In the case of the PhD student, we may argue that for some the first option—make sense of the article on his own—may result in a greater fitness. This adds complexity to the picture and forces the analysis to consider the diversity of each decision maker.
The Rationality of Change
89
The third question for the logic of adaptiveness is that of change: How fast is the situation evolving? This point is crucial for every decision maker. Changing conditions imply a decision that could be (a) subject to fine tuning, (b) definitive but irrelevant in the long run, or (c) part of a chain of subsequent and distinct decisions. Fast-pace environments usually imply uncertainty and ambiguity. This, in turn, implies that the decision one comes up with is also uncertain. Moreover, it may become ambiguous because of the changing context. A way to reduce uncertainty/ambiguity and increase the likelihood of coming out with a rational decision is that of distributing the resources used in the cognitive process. We can make an analogy with investment finance. One of the most important criteria to decrease the risk of a portfolio is that of increasing the number of stocks. Otherwise stated, traditional finance theory investors should follow the principle of diversification. Bringing this analogy back to decision making, in order to decrease uncertainty, a rational decision maker should “diversify”—distribute—his/her cognitive resources on a wider range of alternatives. It means that, when the environment is undergoing a significant and fast change, the decision maker should diversify online resources used; take more advice, suggestions, comments, information from coworkers; hire consultants; etc. The level of affordance changes depending on how badly change is related to the situation. Put differently, distribution should not continue ad infinitum, but should rely on the condition of fitness. The logic of adaptiveness is not an attempt to substitute the other two logics, quite the contrary. This is an effort to create a third logic directed toward highlighting those elements that are not included in the previous two. Therefore, the logic of adaptiveness integrates the other two.
The Rationality of Change The logic of adaptiveness points out that change is particularly important when analyzing the cognitive process of decision making. This aspect of the “new” logic is particularly interesting because it reveals that DC may help understand how individuals cope with the changing resources surrounding them; it is a rationality of change. Before getting to comments on change, a few words on technology may help introduce the topic.
Innovative High Technology In the last 20 years or so, several innovations have changed the way human beings interact, think, live, work, rationalize, and make decisions. It is probably trivial to point this out, but anyone who is required to solve any problem, whether it is personal or job-related, checks on the Internet to find some benchmark or help. Today, this “help” is everywhere, even on your phone (I use it a lot!). Or, maybe you are reading this book in its electronic version via your Nook, Kindle, iPad, or other
90
7
Stretching the Bounds (I)
device. You may have noticed that these devices are not only capable of making a significant part of your library portable. They also come with a set of tools that allow you to highlight and take notes—and we are in the imitative-substitutive domain still. They give you the opportunity to browse through a book as if you were looking for something in your computer, to check the meaning of words using an integrated dictionary, to buy other books on the same subject, to adapt the text on the basis of your eyesight needs, to have the device read for you, and more. Why are these technologies so helpful? Why are humans that quick in making their commercialization so successful? I would argue that successful technology is highly compatible with human cognitive processes. In particular, devices such as those mentioned above help individuals have better externalization processes. Their cognition is distributed on tools that favor more effective cognitive processes. This is similar to what can be found in Clark’s Natural-born cyborgsxxxi , where the author makes the point that individuals are so embedded in their distributive processes that these devices are successful to the extent that they become part of their cognition. Put differently, the continuous process toward more externalization and distribution is favored by new technological innovation. This seems to be a constant trend in human cognitive evolution.xxxii However, the availability of technology is not the only factor that we should consider in this analysis. In fact, there are costs associated with the use of these devices. These costs (marginal costs) decrease with the increasing use of technology. This is to say there are cognitive costs that are costs of learning. New externalizing devices need periods of time where we understand and learn how to exploit them in order to fit them into our cognitive processes. The more we use technology, the more our cognition is affected and shaped by them. One could make the case that it is always the individual that decides how to use technology, so there is control of that ongoing cognitive process. Such a statement is very hard to analyze. I mean that it is hard to have a precise idea of where confines are in cognitive processes involving high-tech products. The boundaries remain undefined. Take a simple task, that of storing a phone number on a cell phone. The device works as a memory storage for us. We store the number so that we can “remember” it when we need it. It does not make any difference who (or what) remembers the number; it is “ours.” If somebody asks me if I know Tom’s phone number, my answer is “Yes, I do.” However, with the traditional cognitive model of BR my answer should be “No, I don’t know it but it is stored in my cell phone.” The reason why we don’t say that is because we use these devices—e.g., the cell phone—as extensions of our cognition. Is the number in our brain? No. Is the pathway that leads us to find that phone number on our portable device in our brain? Hard to tell. There are intuitive designs in the phone that “lead” you to find the number. It is based on interaction, not on somebody remembering every single step that leads you to retrieve the phone number. A user-friendly interface leads you to the number. The process may become automatic (mechanical), but in that case it is a waste of cognitive capabilities because the device may use your intuition to find the number. You don’t have to use “internal” memory, you can externalize it as well as the process. Who controls this process? Is it you that
The Rationality of Change
91
pushes buttons on your phone to retrieve the number, or the software engineers that designed the process that way? I think the answer lies somewhere in between these two alternatives. What matters here is that the cognitive process comes out through interactions. To support the idea that asking the “question of control” may be misleading, I invite you to think on the fact that everybody seems to have a slightly different way to use even simple devices such as a cellular phone. We have just discovered that DC makes it easier to consider every cognitive process different and, for this reason, to value the diversity of each human being. Today, technology stays at the core of any business. We can no longer imagine businesses—even farms—without computers, for example. If we follow what we have been arguing above, we may come to the conclusion that managers have the opportunity to get the best from people. If we extend this paradigm to the next level, we should probably lean on companies that auto-regulate, where control is a tool of the past. However, I am going too far here. Chapter 9 helps to get a better idea of some of the implications of DC for management and organizations.
The Process of Change It is common sense that people do not like change. This is what is usually found in management textbooksxxxiii , and what we tend to believe due to our own personal experience. During the 2008 presidential campaign something strange happened, though. The two candidates, Mr. John Mc Cain (Republican) and Mr. Barack Obama (Democrat) had very different, if not opposite, campaigns. It was Mr. Obama that campaigned under the slogan “change we need.” That and many other decisions proved to be quite successful, as he became the 44th President of the United States about America. Americans voted for Mr. Obama for several reasons—and I do not know enough about American politics to comment on this—although some of them explicitly liked his campaign, and wanted change, they were looking forward to new “management.” The question is why do people seem to be willing to address and support change in a situation like the elections, and are more reluctant when that change is brought into the company where they work. What triggers repulsion in this latter case and approval in the former? DC may help understand this phenomenon better. Change has been addressed earlier in this chapter. We connected change to emotions and to how they follow a through-doing logic (that we may now extend to the logic of adaptiveness). As many scholars point out, change processes are related to learning mechanisms.xxxiv Following Lewin’s modelxxxv , there are three stages through which change is analyzed: (1) unfreezing, (2) changing, and (3) refreezing. The whole theory is based on the idea of a previous freezing state, where stability is prevalent. That seems to be the “state of nature” for Lewin. This natural state may change and be set on a different level depending on individual thinking and behavior. However, this means that a break in the continuity of the “state of nature” needs to be made (i.e., the crisis). The problem becomes that of defining stability and what perturbs it so to make it unstable and then stable again. Recent
92
7
Stretching the Bounds (I)
interpretations of this modelxxxvi suggest that unfreezing happens when the status quo is disconfirmed. And this is also a fundamental process that occurs whenever learning mechanisms are elicited. Many consider Lewin’s as a paradigmatic model of change; in the following, I try to elaborate how it relates to DC and extendable rationality. DC provides a significant way to reframe change. Is change something that occurs in isolation? Is it detached from the environment where it is happening? Is it in the head of the one that is experiencing it? The answer to these questions is negative. Whether change is natural, planned, or unplannedxxxvii , it occurs in a context where external resources mediate it. As in the case of the manager that is about to send an e-mail as a reprimand to one of her employees, we can argue that individual as well as organizational change do not happen in isolation. There are always resources that mediate what change could mean. In order to frame change from the DC perspective, understanding the role of these mediators becomes crucial. This is very similar to what Schein considers “cognitive redefinition:” (1) semantic redefinition—we learn that words can mean something different from what we had assumed; (2) cognitive broadening—we learn that a given concept can be much more broadly interpreted than what we had assumed; and (3) new standards of judgment or evaluation—we learn that the anchors we used for judgment and comparison are not absolute, and if we use a different anchor, our scale of judgment shifts.xxxviii
By rephrasing what Schein wrote under the umbrella of DC, we are able to argue that change is not a mere process of adaptation to external circumstances. It is something that needs to be cognitively processed. It is not a cycle that functions out of input-transformation-output process, but needs a strict interplay with outside and inside resources. These external or outside resources may represent changing conditions, depending on an individual’s ability to frame them. Without this matching, no change could ever happen. The function of recognition or “cognitive broadening” may happen only if individuals extend their minds, learning how to distribute their cognitive resources in more effective ways. This means that bounds of rationality are moving. Moreover, the unfreezing/changing/refreezing (UCR) process mimics what is postulated by the distributed cognition approach, although (1) it is now clear how individuals could stop performing UCR as a typical cognitive mechanism, and (2) there is no such a thing as a stable state or absolute freezing. Therefore, UCR processes never cease to happen; what the DC approach brings to the change discourse is that UCR is a matter of intensity. When intense states of UCR happen then, change may be recognized by external observers, Otherwise, in a case of low UCR intensity, we are witnessing the regular change process, typical of a distributed cognition. For this reason, the DC is a rationality of change. The next step of the study of how distributed cognition affects organizational and individual changes may be that of analyzing what resources lead to perceptions of change or stability and why. The intriguing part of such a study lies in the fact that a need to analyze shared meaning of resources becomes fundamental to understand decision-making dynamics in both individuals and groups. The next chapter provides more clues on this approach to organizational studies.
Notes
93
The most important factor of change and learning mechanisms is that they are processes that lead individuals to set themselves on different and new levels of conceptualization. Put differently, they are mechanisms that we use to go beyond particular bounds that limit our cognition. With learning and change, bounds of rationality move and adapt to the new internal-external conditions.
Summary The idea of extendable rationality has been introduced. It is based on the hypothesis that rationality (1) has the potential to change according to the context, (2) is defined through its fitness to a given situation, and (3) has workable outcomes, i.e., they are appropriate in respect to the situation and/or to the problem at hand. In this chapter we have analyzed two particular aspects that relate to the bounds of rationality. Among the ways individuals have to “stretch” their rational bounds, there are through doing decision processes and change processes. Although through doing processes are pervasive, we have considered two examples: emotions and morality. That analysis brought us to define a third logic for decision making that integrates those of consequence and appropriateness: It is the logic of adaptiveness. Three questions define this logic: the questions of context, fitness, and that of change. The final pages of the chapter have been dedicated to (a) innovation technology and its fitness to distributed cognitive processes, and (b) a further elaboration of how change could not be relegated to outside conditions that affect individual thinking but needs to be considered as part of the interplay occurring between internal and external cognitive resources.
Notes i. This is typical of neoclassical and mainstream economics (Von Neuman and Morgenstern, 1944) but it is also consistent with bounded rationality, as Patokorpi (2008) evidences. ii. His point is that “Simon’s Grand Theme of bounded rationality, first, has been rather incompletely absorbed in the economics of organization; second, does not constitute a necessary part of theorizing on economic organization and mostly serves a rhetorical function; and, third, that part of the reason for this is that Simon and those who have followed his lead did not develop a distinct, affirmative program for incorporating bounded rationality in the economics of organization that would be satisfactory to most economists” (Foss, 2003b, p. 246). iii. Most of the criticism comes from students that try to switch back to the traditional paradigm; this is in part the reason why those who introduce and use bounded rationality need to point out that it is a substitute for the full-rationality model (see above and specifically Camerer, 2007, on how the contrast is still alive). However, many scholars don’t use the concept at all (making assumptions such as flawless and uniform access to information or uniform cognitive abilities of players; e.g., Gilboa, 2010) or mention it as a rhetorical construct (e.g., Nelson and Winter, 1982). Examples of the latter are in Foss, 2003b, and any microeconomics textbook offers a picture of what I mean (see Mas-Colell et al. 1995). One of the seminars organized every month by the
94
7
iv. v.
vi.
vii. viii. ix. x.
xi.
xii. xiii. xiv. xv. xvi.
xvii. xviii. xix.
Stretching the Bounds (I)
Economics Department at the University of Wisconsin, La Crosse, featured a young and freshly graduated PhD student that presented findings from his dissertation. Presenting his complex and elegant model he explained that learning was not considered as a rational characteristic of the players and that the model was a better fit for the data if learning was excluded from individuals’ cognitive capabilities. Of course, we should be better off without learning! This is not only against any BR, it is also against any observation of real-life situation, including the one that brought that brilliant scholar to create his model. Among those who criticize BR from a non-neoclassical perspective are Langley et al. (1995), Zelený (2001), Patokorpi (2008), Sent 2005, Mousavi and Garrison (1992). I am not sure of this word. I believe that a better word can be found to define the dynamics and instability of human rationality. However, words such as dynamic, unstable, plastic, debounded, moving when describing rationality all seem to emphasize one single aspect of what I mean while extendable gives a potential modification at will and it is related to the idea that rationality is an attempt to fit a given context. An exception of this overlooked domain—i.e., the social side of rationality—may be represented by studies on the anchor bias. This bias needs external resources to come into play and sometimes it cannot exist if a social actor doesn’t point out the data, symbol, word, or else, that function as the anchor for the decision maker. See chapter 4 for references and further details. One may argue that all decisions pass through a phase of “through doing,” at least at the beginning of the process when the decision maker has no clear idea how to frame it. March (1994), explains how a feedback mechanism works when the decision maker uses a satisficing method for search. Hanoch (2002); Hanoch et al. (2007). The role of emotions in decision making is considered under many circumstances; there are studies in transient emotions, see Andrade and Ariely (2009), emotional labor, Bono and Vey (2005), justice and emotions, Weiss et al. (1999), Barclay et al. (2005), leadership, Emmerling and Goleman (2005), and many more (e.g., Härtel et al., 2005). The seven attributes for defining a basic emotion are described in Panksepp (1982, 1992); in the discussion of disgust in relation to basic or primary emotions, Panskepp writes that “[p]rimary emotional systems, as far as we know, are intrinsic within brain tools for allowing animals to generate complex, dynamically flexible instinctual action patterns to cope with specific environmental enticements and threats. [. . .] Their arousals are not restricted to narrow stimulus-driven survival issues, but ones that can be related to fairly large-scale organismic survival concerns arising from many environmental opportunities and exigencies” (2007, p. 1821). See Damasio (1995), on this point. On the primacy of emotions or cognition an important debate took place 25 years ago between Zajonc (1984), and Lazarus (1984). See Damasio (1995), and Gibbs (2005). Lazarus (1991); Liu and Perrewé (2005). For example, Liu and Perrewé (2005) offer a model to study individual emotional reactions to change. They argue “that, in a planned organizational change, individuals go through a cognitive–emotional process, in which they try to make sense of the change, struggle with their emotional tensions, and choose their ways of coping.” Moreover, “a change process typically involves an emotional episode that has four sequential but distinguishable stages. Following Lazarus (1991), the first three stages are termed the primary appraisal, secondary appraisal, and coping stage, respectively. The last stage is termed the outcome stage of the planned change” (p. 265; italics added). The coping cycle of change is taken from Carnall (1990), chapter 7. See the note above. This is the debate between Zajonc (1984), and Lazarus (1984). This is the traditional idea of intelligence; see Eysenck and Kamin (1981).
Notes
95
xx. This is an emerging concept; see Earley (2002), Earley and Mosakowski (2004), Thomas and Inkson (2003). xxi. An example of this can be that of Werhane (1999). xxii. Albrecht (2006) offers an idea of social intelligence that is only in part similar to what we intend here. The next chapter offers an explanation of how this concept relates to rationality. xxiii. Emotional intelligence became popular after Goleman (1995). It has been recently discussed in its potential by Ashkanasy and Daus (2005), and limitations by Locke (2005). xxiv. See, for example, Hanoch (2002). xxv. There is something in Damasio (1994), on this particular relation between altruism and emotions. xxvi. The book was published in English in 2007; the idea has been presented in many other writings by the same author and in one working paper that we wrote together with Bardone (2006). xxvii. De George (1999), p. 19. xxviii. Thoughts on thoughts are sometimes called meta-cognitive processes. xxix. Magnani (2007). xxx. Payne et al., (1993) are entirely dedicated to the adaptive decision maker. This logic shares with them the intentions to capture the dynamic and flexible adaptability of human decision making; and their work is a landmark contribution in the field. However, I believe there are also significant differences between our approaches. Besides what is in the text, Payne et al., share with Simon all assumptions typical of BR. They write that their “view of decision making as the application of a series of operators to knowledge states is not unique to them. [. . .] More generally, their view of decision strategies is closely related to views of problems solving as the application of a sequence of mental operators (see, e.g. Newell and Simon, 1972; [. . .])” (Payne, Bettman, and Johnson, 1993, p. 11, italics added). Later in the book, they specify that “[t]he idea that strategy selection is the result of a compromise between the desire to make a correct decision and the desire to minimize effort [. . .] fits directly into the concept of decision making as bounded rationality” (Payne et al., 1993, p. 73). They build on BR although their analysis leads somewhere farther than Simon’s original concept. xxxi. Clark (2003). xxxii. See, for example, how Magnani (2006) interprets findings by Mithen (1999). xxxiii. For example, Jones and George (2009). xxxiv. See, for example, Schein (1999), or Carnall (1990). xxxv. Lewin (1951). xxxvi. Schein (1999). xxxvii. Schein (2002). xxxviii. Schein (1999, p. 61).
Chapter 8
Stretching the Bounds (II)
The previous chapter set the logic for analyzing cognitive processes in the making, and defined change and high technology within the framework of distributed cognition. This chapter looks at a broader set of variables in an attempt to broaden the picture. It is about the unavoidability of the social context. After presenting a classification of external resources, a model of information sharing among individuals is considered. Then, a proposal is advanced on how to define richness in information mediums. The chapter ends with a few notes on bandwagon.
The Others One of the most important, or probably the most important improvement that a theory of distributed cognition brings to the idea of rationality is that of “others.” In many of the examples used in this book I haven’t omitted the typical environment that characterizes our life in and outside organizations. This environment includes many external resources of which the most widely diffused, in terms of availability and importance, are other individuals. It is almost impossible to think of a decision that happens without the occurrence of another human being. Even when we think of a decision in isolation, such as those of experiments, we need to discount the experimenter’s biasi or the fact that individuals have been asked to be part of the experiment by another human being and behave as if they are being observed. Or, we can consider a decision in isolation similar to that of passing through a supermarket lane to buy milk and sugar, and ending up grabbing bread, caramel, and apples for instance. These look like decisions in isolation but they are, in fact, affected by other people and many other stimuli:ii people that we observe buying that bread, or people that set those ads with colors so bright that they are impossible to overlook! Most of what we are affected by is social in that it has been set up by another human being with the explicit objective of finding us receptive to it. This leads us to a first classification of external resources. According to what is stated above, external resources can be classified on the basis of the type of relation that they establish with other people. Therefore, they can be D. Secchi, Extendable Rationality, Organizational Change and Innovation, C Springer Science+Business Media, LLC 2011 DOI 10.1007/978-1-4419-7542-3_8,
97
98
8 Stretching the Bounds (II)
(a) social, when there is an explicit willingness from other human beings to communicate with other people; (b) non-social, when the message is not intended to be public or when activities of human beings are not crucial for interpreting the information embedded in the medium. We can, for example, tell that a newspaper article, a label on a bottle, drug directions or warnings, a book, signs, etc., are all external social resources, while a river, a certain behavior that is not intended to be communicative, your private notes—taken when you are not in public—from a conference/lecture/meeting, and the like are non-social resources. There is one caveat that needs to be made explicit. In the social world where we live we cannot really state that there is something non-socially tied. The river of the example is not completely non-social, and a tree is not really non-social.iii The words that we use to define a river or a tree or to understand that what we see is a river or a tree come as social constructs to our mind. Even if that process of learning is so remote that we cannot remember anything (i.e., when we learned those words), we know that somebody – our parents, close relatives or somebody else – taught us that the “woody perennial plant typically having a single stem or trunk growing to a considerable height and bearing lateral branches at some distance from the ground”iv is a tree. From here we can see that every single piece of knowledge is social because we can always define it through another human being. Even when we create a new piece of knowledge, we try to set it up in a way that it can be socially transmissible. The limit of this approach to external resources is that it does not help to explain differences or to classify them. And, if it does not help to classify resources so as to clarify our concepts, it is probably not worth using. The difference in the proposed classification stays in the word’s “willingness” to communicate something. In the following I try to describe two cases where the DC approach may be considered and applied to decision making and rationality. The first is an attempt to consider active and mindfulv exchange of advice as tools for decision making; the second relates to passive or mindless imitation of other people’s behaviors.
Advice Taking People take advice. This activity is widely diffuse in daily decision-making processes as well as in one-time decisions. It happens with simple as well as with complex decision processes. The answer to the question “why people take advice” can be easily connected to the fact that these are simple, often free, and available resources for people. This particular kind of external resource becomes available when there is somebody willing to give advice that meets a request made by another person. It goes without saying that in order for this social exchange of giving-taking advice to take place, there must be a minimum of two people involved.
Advice Taking
99
The Judge-Advisor System The two-person system is called the judge-advisor system, or most commonly JAS.vi The advisor is the one that provides the other with the advice, while the judge is the decision maker, i.e., the one who is supposed to make a decision on the basis of the received suggestion. The system is defined in a way that the judge is about to make a decision, and is offered with the opportunity to ask and/or take another person’s advice. Now, how many situations do you know that can be defined through a JAS framework? I believe there are more than we can think of. And the reason for this is very simple. Since our childhood we have been taught to lean on advice from our parents and other important family members and friends. This system continues to work every day of our lives. The system is so diffuse that advice is offered even when it is not solicited. When a window pops up and suggests updating your software now (this happens especially with antivirus, for those who are still fighting against viruses), that is typical unsolicited advice. Also, this is offered independent of the importance of the situation you are facing and the decision you are about to make. Whether you are choosing your dessert or have decided to marry your girlfriend or boyfriend, you will find that there is always somebody providing advice. Moreover, advice comes to you even if you don’t want to listen to it. When you decide to quit your job and look for a new one, people tend to suggest what to do. Sometimes you are not in the appropriate mood to take or even listen to the advice. I have isolated three factors of advice, then, since they (1) don’t require explicit requests from the judge, (2) are not related to the importance of decisions, and (3) are not always wanted. Of course, when advice is given, the judge faces two alternatives. Suppose the CEO of an important firm that is facing a product recall has the option to go to the press and deliver a message or to not go and opt for a standby tactic (e.g., she is ready to deliver the speech as soon as the supplier responsible for the defective product delivers its official version to the press). The CEO has been advised by the CFO not to deliver the speech. In this case, she has two options: (a) follow the advice; (b) don’t follow the advice. The situation can be easily depicted by the following Table 8.1: Table 8.1 Judge’s options
Option (a) Option (b)
Advisor
Judge
Decision
Not Not
Not Delivery
Consistent Inconsistent
The decision of the judge (the CEO) is consistent with that of the advisor (our CFO) when they both agree that the standby tactic is the best. The decision is inconsistent when advisor and judge have different ideas about what to do and the judge ends up delivering the speech.
100
8 Stretching the Bounds (II)
This simple model can be modified to fit a wide set of different events. For starters, I consider n advisors situation and the judge’s opinion at time zero. This latter situation is the standard starting point for judges. Except when people face extremely uncertain or new situations, they often have an opinion on what the preferred decision is. This opinion is going to affect the outcome of the final decision depending on how strong the advice (and the advisor) is perceived by the judge. If we consider the above mentioned example, new and different combinations become available. Table 8.2 A judge-advisor system (one type of advice)
A B C D
jt0
A
jt1
Decision
Not Delivery Not Delivery
Not Not Not Not
Not Delivery Delivery Not
Consistent Inconsistent Inconsistent Consistent
Note: jt0 judge’s opinion at time zero; a advice; jt1 judge’s opinion at time one.
The judge may have two ideas on what to do, and the final decision may be consistent with the advice or not (Table 8.2). Here we have two cases: case A in which the judge doesn’t change her mind and this is also consistent with the advisor, and case B, where the CEO stays consistent with her prior opinion and disregards the advice. The third option C is less likely to happen: It considers a situation where the advisor is not consistent with the advice and a change of mind has occurred (maybe because of the advice!). The final option D represents the case when the judge changes her mind and conforms to the advisor. Consider the case where we don’t know what the advice of the CFO is. In this case, the Table 8.3 below includes four more options, E, F, G, and H. This picture is more complete and includes all the possible outcomes of the decision. Now, what is the probability that the judge changes his/her mind? What is the probability that he/she does follow the advice? Research on advice giving and Table 8.3 A judge-advisor system (two types of advice)
A B C D E F G H
jt0
a
jt1
Decision
Not Delivery Not Delivery Not Delivery Not Delivery
Not Not Not Not Delivery Delivery Delivery Delivery
Not Delivery Delivery Not Delivery Not Not Delivery
Consistent Inconsistent Inconsistent Consistent Consistent Inconsistent Inconsistent Consistent
Note: jt0 judge’s opinion at time zero; a advice; jt1 judge’s opinion at timeone.
Advice Taking
101
taking suggests that decision makers’ personal opinions strongly affect the final decisions.vii However, a better answer to these questions requires an additional step.
Variables Affecting Advice Taking It may happen that a judge has many advisors.viii In the example above, the CEO may have advice coming from the CFO, somebody from the supplier company, a union representative, and an external consultant. Now, the question is when advice is too much. Is more advice always better than one single piece of advice? And, how can we discriminate between them? Which one counts the most? Which is the most important? Which can be disregarded and which cannot? Research has indicated a set of variables that affect advice taking. I consider the three that have the highest explanatory potential when we analyze JAS in organizations. The first is the extent to which the advisor is expert on the subject matter. It is intuitive that the advice has more chance to be taken when the person providing it is knowledgeable and/or has experience in that situation.ix Suppose that, in our case, the CFO happens to have more experience than the other advisors because he is not facing the situation for the first time, having worked for 25 years in that company, knowing the supplier well, or similar factors that make him an expert. If this is the case, then the CEO is more likely to trust him and end up with a consistent decision. However, the experience of the advisor needs to be discounted with what the judge thinks his/her own expertise is and with whether the advice is required, by the judge or not. When the advice is required, the probability of a consistent outcome is higher. The second variable is the social relation between the two (or more) players of the JAS.x With this I mean the fact that the advisor is close to (or is considered close by) the judge plays a crucial role in the final decision. If the CFO has never given advice to the CEO and this is the first time this happens, while the relation between a top manager at the supplier company is stabler, the advice with the highest weight will be that of the latter, not the one coming from the CFO. Now, when is a social relation stable? This is a point that is not easy to analyze, but I believe that we can bring a bit of sociology into this.xi The social relation must be important; it doesn’t matter if it is based on the fact that the two share the same country club membership, meet at Sunday masses, play tennis together, graduated from the same university, or have established a sound professional relation. As far as advice taking is considered, what matters is that this judge-advisor relation is stronger than other relations, and it is so strong that the advisor (e.g., the supplier) affects the judge’s (our CEO’s) decision. The last variable that I would like to consider in this short analysis of advice taking and giving is the medium that is used to let the judge get the advice. Unfortunately, research in this particular domain is not developed enough to present its results.xii However, there are studies that support the idea that the medium through which information is transmitted makes the difference. I try to present a few hypotheses on how the medium affects decision making in the following pages.
102
8 Stretching the Bounds (II)
Information Richness First of all, we need to find a criterion to define the sources of information based on how well they transfer the data they carry. The degree of information that a medium can carry is called information richness (Fig. 8.1). To be more precise, a medium is “high or low in “richness” based on its capacity to facilitate shared meaning.”xiii Scholars define a scale of information richness that allows us to distinguish between mediums so that at the lower part of the scale there is an unaddressed document and the highest part is where face-to-face is located.xiv
Unaddressed documents
Written, addressed comm.
Telephone
Face-to-Face
LOW
HIGH
Fig. 8.1 Information richness scale Source: Daft et al., 1987, p. 358.
Very little specific research has been conducted on media information richness and advice taking.xv Also, most of the experiments on advice giving and taking do not use direct human advice but computer-based. This is a very interesting fact because, instead of seeing it as a limitation—as those who run experiments do in their writings—I believe it is a significant and valuable aspect. The argument is simple and related to the question: How do people communicate the most? What is the medium that most of us use when at work? The answer is simple, as I anticipated. It is the computer that shapes the way we stay in touch with people, we communicate, we feel about our job, we ask for advice, etc.xvi There is an entire universe of free and proprietary software out there that shapes our way of communication. Most of our information passes through these media. By the way, the information richness scale was defined in 1987 by Daft, Lengel, and Trevino (Fig. 8.1). This means that the authors had the first thoughts about it in the years before the so-called IT revolution, and I guess it may have been 1985 when we consider publishing time. In those years the computer was not as diffuse as it is right now, and it is understandable that their scale does not include e-mail, chat software, video conferencing, web sites, blogs, social networking websites, and the like. Advice giving and taking is something that heavily depends on the medium used to share the information. The definition of an up-to-date information richness scale (also called the Media Richness Theory, or MRT) is particularly needed and useful, although there are studies that suggest it should be abandoned when related to IT.xvii A few caveats apply here. According to the definition, information richness is defined by and associated with a given medium. This is not entirely correct. How rich a medium is depends on the information it could potentially contain and on the two (e.g., human or computer) parts involved in the exchange. Therefore, we should
Advice Taking
103
write that the scale is about the “potential” information richness that a medium could carry, given the fact that how rich that information is can be defined by the recipient only. Emotional states, knowledge on how to handle and use a given medium, frequency and past history of usage (confidence), and other variables could affect the scale. Consider the unaddressed document, an advertisement that you find in your mailbox. Most of the time they are completely useless, but not when you come home really hungry and your eye cannot do anything but stare at the pizza delivery ad you just found in your mailbox. In that case the medium provides you with poor but extremely useful information. Or, in that case, the information contained in that medium is rich in the sense that you didn’t need more than that information; that one is appropriate. This is to say that richness depends on the fitness between the individual and the external resource available. Can we associate an absolute value to any given medium? Is information richness all things stated once and for all, independent of the individual that uses that medium? It is worth noting that, according to our definition of social resources, all of the mediums in the scale are social resources. Therefore, it is apparent that the points on information richness are nothing but a corollary of DC and the relative idea of rationality and decision making. The ability to “facilitate shared meaning” lies in the interaction between individuals and the medium, not in the individual, not in the medium. The ability to facilitate shared meaning is thus related to how individuals are willing to exploit that medium so that they can share their knowledge. How can we redefine this scale in order for it to fit the JAS model? Three factors affect the scale: (a) advisor, (b) judge, and (c) medium. We are interested in the fact that richness depends on two individuals willing to share their knowledge. This brings us to the theory of distributed cognition.xviii Individuals make sense of everything through their own introspectiveness; however, part of the meaning of any given problem is common to more than one individual, i.e., it is shared. Introspection is the key that individuals use to get acquainted and informed of any given problem, resource, person. It is their personal way to access the problem. In the following Fig. 8.2, the green circle represents the common understanding of the problem while the yellow corners are the different perspectives (or framing access) that every individual (A through D) has on it. The four arrows have double directions because we assume that the problem at hand shapes the cognition of these four individuals. This is a very approximative and (hopefully) effective representation of what happens when individuals share the meaning of a specific piece of information. It is helpful to point out that if a medium
Individual A Individual B Individual C
Fig. 8.2 The shared meaning of communication
Individual D
104
8 Stretching the Bounds (II)
is something that helps individuals share any part of their cognition then it is not the medium that appears to be rich, but this richness depends on the interactions (or interplay) between resources. Moreover, mediums are not rich in vacuum (i.e., in isolation), and this is the reason why, as already stated, we should call the original scale “potential” richness information scale. The context and/or the situation provide the actual level of richness for a given medium. However, any medium has the potential to be extremely rich, depending on who is using that and why. This leads us back to the interplay between external and internal resources, once again. Consider e-mail writing. You are not very good in English and you need help. A close friend of yours is there and helps you; you also use a dictionary and the Web. How many resources help the transfer of meaning? What is the upper limit where one more resource decreases information richness? There are no easy answers to these questions, and studies on advice taking have barely one. It seems that we are coming out with a way to classify media and advice taking, though.
Information Richness and Advice Taking: A Proposal The analysis of the medium should be carried out on the side of the advisor also, not only on that of the judge, as it has been with the previous discussions. Choosing the right medium could prevent the fact that the advice is taken or even considered by the judge. For the medium to be effective in the JAS system, it needs to meet the judge’s needs and the advisor’s expectations. The judge needs information that has some degree of usefulness, while the advisor has expectations that the judge finds the shared information useful. These two, i.e., needs and expectations, can be considered as the common motivation for advice giving and taking. Also, this means that the medium must meet the condition of appropriateness in order to be effective. And that piece of information is appropriate when both judge and advisor find a common ground where it is possible to share it. This does not mean that the judge will agree with the advisor, but it highlights the fact that the medium is transferring information well. Put differently, that medium becomes exploitable from a cognitive standpoint, i.e., it becomes an available external resource. Within this system, information richness can be measured by the successful use that the judge makes of the advice. If the information is used then the tool (medium) selected has been rich enough to help; if the information has not been correctly categorized in the medium and hence transferred, then the result will be poor. The concept of shared meaning that Daft et al., use is absolute while it must be relative. Therefore, I propose to explain the richness scale according to Fig. 8.3. The two variables in the figure are distributed cognition (dc) and motivation or predisposition toward the advice (mo). The latter may be negative—a negative motivation to take advice—while the former is only positive, indicating that individuals may distribute or not, but it is hard to think of a negative exploitation of resources. Different levels of dc may be associated with mixed values of mo. What we should
Advice Taking
105
Fig. 8.3 Information richness fit
consider as the probability that the judge has to take the advice is the area included by the y axis and the curve. Depending on motivation, the curve may move up or down. An overall tendency not to take advice may have the curve move down such that the negative area increases (i.e., taking the advice becomes not very useful, therefore it is perceived as not needed by the judge). The parable may also become wider or narrower, depending on the dc potential attitudes of the decision maker. For any given level of dc there is a variability of mo so that the use of the external resource (advice) through the medium may be probable or improbable (depending on the area). The more the medium and the advice fit the dc system of the judge, the more likely it is that the external resource will be exploited. The four curves that you see in Fig. 8.3 are examples of different mediums and of their fitness (or appropriateness) to the judge’s distributed cognition. The more a medium employs dc—meaning that it is easier to get for the decision maker—the more likely it will be used. Motivation is also higher in that case. In a world where we can choose, the advisor selects the medium depending on how comfortable she is with that medium and on how she thinks the medium is “good” for the advice to be accepted (e.g., writing e-mails, if she believes this is the appropriate medium for that kind of advice). She expects that the advice (information) is easily transferred to the judge, i.e., enters in the cognitive process and helps decision making. Put differently, it gets “shared.” In our hypothesis, there are four alternatives (in the Fig. 8.3 they are m1, m2, m3, and m4). However, the advisor will not choose a medium that takes too much of a cognitive effort; that is to say, that is not likely to become part of how the judge distributes cognitive resources. For example, instant messaging on the cell phone usually takes more effort if you are older while it is easy, usually fast, effective, and widespread for the young. The use of text messaging depends on who you are (in most of the cases). If what I argue holds true, then the advice is more likely to be taken when cognitive efforts are on
106
8 Stretching the Bounds (II)
their average and expectations that the judge will make good use of it are reasonably high (it is m4 in Fig. 8.3). When cognitive efforts are too high, then according to this hypothesis, it is more unlikely that the advice hits the target. The cognitive effort has to be commensurate with the motivation (i.e., expectation) that leads the advisor to give advice. The example of text messaging with the cell phone shows what the proposed model means. If texting is too much of an effort for you, it means that you are not confident enough with the medium you are using. It is likely that your perceptions on the level of uncertainty surrounding this medium are high, and you might prefer to use another medium. Although the perspective of the advisor is very important, Fig. 8.3 shows what happens to the judge. The judge’s needs are described as motivation in the graph. This is because there may be motivations different from the need or utility of the medium, such as expectations or goal attainment. In the example, motivation becomes more and more positive as the compatibility with the cognitive (distributed) decision-making process increases. The negative/positive domain may indicate the likelihood of the judge to take the advice. Let us try to explain this dynamic with a simple example. The judge is not comfortable with short text messages taken by the phone. It takes a while to read them and they are, according to the judge, not reliable because they are too short and impersonal. Therefore, the likelihood for this judge to take the advice is given by what shown in Fig. 8.4.
Fig. 8.4 A low richness medium
The grey area represents the probability that our judge may take the advice. The more the medium selected gets closer to what the judge thinks appropriate, the more a positive motivation toward the advice arises. As I keep repeating, this also means that the medium is fully integrated in the cognitive processes of the judge. It is rich because it fits the judge’s needs; it is more appropriate. We can hypothesize that, for the judge, a small chat may be the optimal way to communicate. In that case, the medium is m1 in Fig. 8.4. It is worth noting that, if we use this approach, the
Advice Taking
107
same medium (a) maintains an area of substantial inappropriateness (the negative domain of mo), and (b) its degree of richness may be different when we consider other individuals. The first point (a) takes into consideration the intrinsic ambiguity related to any sort of communication. The medium, even that considered significantly “rich,” may not be chosen by the judge (the negative domain in the Fig. 8.5). The second point may be explained by considering how a text message could be accepted by another judge, one that is more familiar with that kind of communication—cognitive fitness—and highly motivated to accept the message (Fig. 8.5). The grey area (m1), in this case, is significantly greater than that of the previous Fig. 8.4, and extends the richness of the medium and the likelihood that the advice may be taken.
Fig. 8.5 A high richness medium
To sum up, the medium used conveys information and is rich when it is (a) needed by the judge, (b) cognitive efficient for the advisor, and (c) the advisor retains decent expectations that the advice will be useful.
Perspectives on Advice Taking There are several limits on the studies of advice giving and taking. One of the typical criticisms is that almost all of the findings derive from experimental settings and field studies are missing. This limitation applies to all recent model/theory developments based on experiments. Field studies, i.e., observation and interviews on site, usually in organizations, are time consuming, difficult to set up because of the increasing skepticism of managers toward research, and have variables that are harder to control. However, they are richer than experiments and give the observer a clearer idea of mechanisms that are in play in the working environment.xix
108
8 Stretching the Bounds (II)
A more stringent limitation is that of measuring the quality of the advice and the extent to which the advice has been taken. This is not a secondary topic as far as advice is concerned. Sometimes advice content is only partially considered, and the quality of it is not taken into consideration. The richness scale that I proposed in the previous pages can be a useful tool for defining this last point. Studies have focused on the successful solution of a given problem when advice has been used by the judge, and they have taken this as a measure of the quality of the advice. This works for task-oriented advice only but not with life, behavior, and other similar kind of advice. If my personal advice to you is to continue reading at least the next chapter of this book because you will have food for thought, you may follow only half of that advice and read a few pages at the end of that chapter. However, I will never know how good this “food” has been for your thoughts unless I ask you a specific question on this point. And even if you answer that question, I expect your opinion to be fairly well-motivated and grounded in your knowledge (both newly acquired and previous knowledge). How can we measure this? How can we consider this complex aspect of information sharing and byproduct of advice taking? This is not clear yet from research, and one of the reasons why it is not so relates to one major weakness of the above-mentioned models. A general theory of advice giving and taking is missing. What scholars do in this domain is collect data on how people react under specific circumstances. The implication of this is that we have several ad hoc explanations of how people behave when advice is provided, but we never get the whole picture. In the next chapter I will attempt to provide a theory of individual behaviors in organizations that includes advice taking as one of its major points.
Passive Advice Takingxx As defined above, bandwagon is a fallacy, something that is “bad for logic but (sometimes) good for life,” that results in a bias. In particular, this ad populum fallacy (see Chapter 5) describes the situation that people face when, for example, they have to make decisions under uncertainty, are short on time, seek legitimacy in the group, don’t see other alternatives, and the like. Broadly speaking, whether they use a rational calculation or feel under social pressure,xxi biases, prejudices, and rational boundsxxii work to make the choice of bandwagon more likely. These shortages of cognitive capabilities can be, and sometimes are, used by companies, organizations at large, and governments to suggest (or nudge, as Thaler and Sunstein put it)xxiii a preferred alternative. Bandwagon is a social imitative behavior that becomes widespread or popular. The idea is that it works as a threshold. Granovetterxxix explains that bandwagon could emerge at different times and on different base for different people. What triggers mine can be different from what triggers your imitative behavior. How many people do you need to see wearing orange before you start thinking that they look nice and that you want to dress in the same color? And how many people do you
Passive Advice Taking
109
need to see with ice cream in their hands before deciding “Yes, I would like to get one”? How many people must discount information coming from that turnover ratio before you decide that it is not affecting the low performance of the company? It is the case of the employee stealing pens or copy-paper from work because “everyone is doing it.” Another example might be the idea that you cannot ask that colleague to retrieve data for your report, because nobody does and “there must be a good reason for that.” These are very simple and quite meaningless cases, but think of employees disregarding one, two, three, and then n symptoms (i.e., data) of an upcoming crisis (reasons for sales increase/decrease, shareholders claims, environmentalist protests, etc.,) because everybody thinks they have no meaning and “It makes you feel like a fool if you think they are.” And the crisis arrives. Think of the number of cases where this could be a simple but accurate explanation for what happened. Are you thinking of subprimes, the crisis companies are facing right now? Well, I do not have the data and do not pretend bandwagon can be the sole cause for the crisis, but it is fair to consider that it could have contributed. You decide the extent to which it is important in such cases; however, the point is that bandwagons are an important part of organizational behavior and life.xxv What stays at the basis of bandwagon? Let us study the simple example: stealing pens for personal use from the company stockroom. The employee tells himself or herself that these pens are for his or her child as a justification—which might happen to be true; however, that is not the point here. The idea of thresholdxxvi is particularly useful. The employee decides to steal pens only when he or she observes that a certain number of other people behave that way. This number of people is his or her threshold. Then, the employee imitates what other people do. Bandwagon is about imitation. As biologist Kevin Laland suggests,xxvii imitation is related to social learning and it is very common in primates and other animals. However, when we refer to human beings, what kind of learning is this apparently “mindless” behavior? If you steal a pen from the stockroom because you imitate other people, are you learning anything besides how to steal? Bandwagon is one of those phenomena that managers would like to minimize in their company. Especially in conditions like those of the western countries where economic services are key, companies tend to ask people to think, to use their brains, to try not to follow the mainstream. There are plenty of companies that are in desperate need of people that give their unique contribution, that develop attitudes like creativity and continuous learning. Bandwagon facilitates integration of a new hire, because it is based on the same mechanism cognition works: it leverages external (social) resources. However, these cognitive processes are cheap, they do not involve active choices but lean on mechanical processes instead. This fact, when not recognized, becomes diffuse in too many areas of one’s behavior and could undermine performance, social relations, and individual contributions to the company. Bandwagon is a passive phenomenon that equals advice taking since the decision maker takes other people’s behaviors as (sometimes) involuntary advice giving. You see people stealing pens from the stockroom, therefore it might be okay to do that. It is a surrogate for asking “Is it okay to steal pens from the company’s stockroom?” The behavior provides the
110
8 Stretching the Bounds (II)
answer to that question, with no need for you to ask it. As many cognitively passive phenomena, it has drawbacks that need to be fully addressed. This topic is not new for management scholars; however, most studies focus on bandwagons between companies or as an individual phenomenon.xxviii Imitation has been studied in strategic management to understand how innovation passes from company to company. The integration of bandwagon in organizational behavior has yet to come. It should attach to the study of culture, routines, corporate norms, and other typical elements that shape human behavior in organizations. It is relevant here because it is close to cognitive processes described in this chapter. The idea of extendable rationality includes the fact that a seemingly sound behavior is effective and appropriate to a particular situation. Nevertheless, an idea of rationality based on the distributed cognition approach highlights that this extension lacks the crucial mechanisms that make human decision making so complex. In the next chapter I will be back on bandwagon and compare it to a similar but more “distributed” phenomenon: docility.
Summary In this chapter we classified external resources into (a) social, when there is an explicit willingness from other human beings to communicate with other people, and (b) non-social, when the message is not intended to be public or when activities of human beings are not crucial to interpret the information embedded in the medium. This distinction led us to focus more on a human typical communication process, that of advice giving and taking. We have defined the judge-advisor system as a useful model to analyze whether advice is consistent between two individuals or not. Moreover, we have learned that the effectiveness of advice depends on actual or perceived expertise, social relations between judge and advisor, and the medium. A discussion of information richness and a proposal for a new model to frame it followed that analysis. According to the model, information is rich when it is appropriate to the circumstances that affect the judge-advisor relation, i.e., motivation, cognitive processes, and their fitness. The chapter ends with the proposal to consider bandwagon as passive advice taking.
Notes i. One of the first times it has been found in management was during the famous Hawthorne studies conducted by Elton Mayo. This is also known as the “interviewer bias” (Groves et al., 2009, p. 292f). ii. The behavioral economics literature on cues and cueing supports this non-isolated hypothesis (e.g., Politser, 2008, p. 64f). Thaler and Sunstein, 2008 in their book Nudge, offer several examples of how cues affect behavior. iii. A similar point of view is expressed in Maturana and Varela (1987). iv. New Oxford American Dictionary.
Notes
111
v. The two words “mindful” and “mindfulness” are related to the work of Langer (1989). See also: Fiol and O’Connor (2003) and Levinthal and Rerup (2006). vi. See Bonaccio and Dalal (2006). vii. See, for example, Gardner and Berry (1995), Harvey and Fischer (1997) and Yaniv and Kleinberger (2000). viii. This and the following points are treated in the extensive literature review by Bonaccio and Dalal (2006). ix. There are studies that have isolated this phenomenon of the impact of expert/novice advice to the judge; e.g., Harvey and Fischer (1997) and Sniezek et al. (2004). x. This is highlighted by Bonaccio and Dalal (2006, pp. 145–146), and in Slaugher and Highhouse (2003). xi. See for example Granovetter (1973), and Chiang (2007) on a particular kind of social relation known as bandwagon. xii. I was able to find Hedlund et al. (1998), on the difference between face-to-face and computer interactions in advice giving and taking. xiii. Daft et al. (1987, p. 358). xiv. Daft et al. (1987). xv. There is some research on negotiation and media richness. See for example Purdy et al. (2000), Drolet and Morris (1995) and Moore et al. (1999). xvi. I take a slightly different perspective from that of Kock (2005) here since he suggests that face-to-face is still what people prefer in business communications. xvii. El-Shinnawy and Markus (1997) and Dennis and Kinney (1998). xviii. This particular assessment of distributed cognition overlaps with what Cannon-Bowers and Salas (2001) call shared cognition. xix. Social network analysis is particularly useful and helpful when it comes to analyzing advice taking. See Knoke and Yang (2008) (at page 13 they explicitly refer to advice taking). xx. I owe the idea to treat bandwagon here to Emanuele Bardone, friend and coauthor of many articles. His PhD dissertation (2008) is a clear example of how the multidisciplinary approach to cognitive science is the past and especially the future of this field of research. This section of Chapter 8 is a summary of the working paper “A Model of Organizational Bandwagon.” xxi. Abrahamson and Rosenkopf (1993). xxii. Kahneman (2003). xxiii. Thaler and Sunstein (2008). xxiv. Granovetter (1978). xxv. Chiang (2007) and Granovetter (1978). xxvi. Granovetter (1978). xxvii. Laland (2001). xxviii. See Fiol and O’Connor (2003).
Chapter 9
The “Docile” Organization
When I visited the Acropolis at Athens, I remember that beside the wonderful Parthenon there is a smaller temple called Erechtheum, which is dedicated to Athena, goddess of knowledge, war, arts, and justice. This small temple has a worldfamous characteristic: Its columns are feminine statues that carry the weight of the roof. They are the caryatids. They serve a very useful purpose, that of preventing the temple from falling down and, in doing that, allow everybody to get into it. It is not in the power of these beautiful statues to decide who can and who cannot enter the temple. The previous chapters are like the caryatids since they bring the theoretical infrastructure up but cannot help state what can and cannot be a future theoretical outcome. They are premises and allow multiple options to become available. This chapter presents a theory of decision making in organizations that uses extendable rationality and the distributed cognition approach in a way that brings us to a better understanding of organizational dynamics. It is the theory of docility.
The “Docile” Individual The idea of extendable rationality represents a connection between the theory of bounded rationality and the distributed cognition approach. The link between the two explains what happens with the “through doing” and the social aspects of decision making. In the following pages I try to generalize these assumptions. Is there a trait that defines human beings on the basis of distributed cognition? What makes us behave as rational beings? Ethics, bandwagon, emotions, and advice taking all have a social basis. The distributed cognition approach seems to support hypotheses on the social functions of our intellecti , as well as those on the social brain hypotheses,ii and on ultrasociality.iii This gave the input to the late Herbert Simon to explain the concept of “docility.” This idea was not new since he mentioned it in one of his first writings, left it undercover for almost 40 years, and then published two important articles on that in 1990 and 1993. The first paper was published in the prestigious Science, D. Secchi, Extendable Rationality, Organizational Change and Innovation, C Springer Science+Business Media, LLC 2011 DOI 10.1007/978-1-4419-7542-3_9,
113
114
9 The “Docile” Organization
while the second was published in the American Economic Review. Notwithstanding these important outlets and the Nobel, the ideas exposed there and the implications were never the subject of any subsequent scientific papers until 2003, when Knudsen published the first paper dedicated solely to docility after Simon. Considering what I am about to write, this is quite impressive. Why did this happen? Why did the scientific community overlook this idea? I have two distinct explanations. Before getting into these explanations we need a definition. Docility is the tendency to depend on suggestions, recommendations, persuasion, and information obtained through social channels as a major basis for choice.iv
It is a typical human trait that emerges when we make decisions on the basis of information coming from other human beings. The fact that Simon defines it through the terms “suggestions, recommendations, persuasion, and information” allows us to relate docility to advice giving-taking and to emotions in a fairly direct way. Docility is a trait that unveils the way our cognition works; it is a behavior consistent with the way our cognition is distributed. As we state elsewhere in this book, social resources are major sourced of cognitive activity, and there is no surprise that we can find a specific human trait that defines how we handle this information. From this specific trait, behavior comes out. What is typical of human beings in social settings? They exchange opinions, tend to learn from each other, and make decisions on the basis of information they gather from other members of that team, group, organization, or social environment. Now that we have a first idea of what docility is, we can go on with the following points on why the scientific community overlooked it. The first is that the idea of docility was not consistent enough with that of bounded rationality. Even if Simon brings docility into consideration because of bounded rationality—he states that we are docile because we are boundedly rationalv —the passage is not clear enough. The fact that our rationality is bounded doesn’t lead directly to docility. It is probably a sufficient condition, but it is not necessary. It is sufficient because without limits to rationality we should never rely on external resources since we are omniscient in that case. If rationality is limited, then we get information from the outside where social channels supply part of it. However, this condition is not necessary since it is not clear why humans should lean on social channels more than other resources. It is not a need that we have to rely on external social resources; broadly speaking, the fact that we are bounded doesn’t necessarily lead to a particular kind of external resource to overcome our limits. Therefore, I believe this makes the connection sufficient only. Simon never made this point clear, and I believe this didn’t encourage scholars. The second argument is that the scientific community has had difficulty using the concept. Simon presented a simple model where he applied a Darwinian approach to a community of human beings showing that non-docile individuals do not survive and instead go into extinction. This strict interpretation of social Darwinism came late in the scientific debate, at a time when fallacies of these applications had already been identified.vi Therefore it was too hard for an idea like that of social Darwinism to enter the debate again. What students of decision making, management, and economics overlooked was the deepest meaning of “being docile.” As I
The “Docile” Individual
115
explained in one of my articles,vii I agree with most of the critics that any form of social Darwinism is too deterministic to work. I also believe that we have “softer” theories that explain how two or more species could survive in the same nicheviii (or environment), and we can argue that social dynamics are complex to a point where it is not clear how a certain attitude (e.g., thought or behavior) has really gone. This is only one of the possible explanations of why students overlooked the idea of docility. Another could be related to the fact that distributed cognition theory had yet to come in the early 1990s. It is only a few years later, but we now have the emerging trend of studies pointing out what external resources are for cognition. And I believe this makes the difference when going backward to interpret an idea that wasn’t successful in the past. We didn’t know how to deal with it. Now, we know. A third explanation can be that the concept is not worth studying because it has a limited explanatory power. Well, even if we had one reviewer making exactly that argument one time, on the basis that there is insufficient literature to support the argument, Bardone and I disregarded this last option since it is a typical ad populum fallacy.ix It is too early to dismiss this concept given its matching with the DC approach and extendable rationality. To get to the point, we need to discuss a few more assumptions underlying the model and the theory. In the following pages, I discuss fitness, altruism, and selfishness, docility as a theoretical background for advice taking, and conclude with decision making.
Fitness I would like you to run a thought experiment to understand docility. Imagine a world based on individuals that are 50% docile and 50% non-docile.x This means that half of them lean on external resources (suggestions, recommendations, etc.,) to make their decisions, and half of them do not. It may be difficult to understand how a human being could exercise even the basic function of thinking without this sort of interaction. But this is a thought experiment, and it is called that way because you can hypothesize the craziest things. So, imagine these two kinds of persons exist. Who will survive in a social world? Who will be the fittest? This is a typical question that evolutionary biologists ask, and while it has no simple answer in most of cases it does in our experiment. To help you get a better idea of what we are about to do, I suggest you think of the docile individual as an altruist, and to imagine the non-docile as selfish. The docile individual is open to the social side of human relations while the nondocile is not; in other words, he/she is capable of experiencing emotions and makes decisions to please other people. The docile individual is willing to support the cost of information sharing while the non-docile is not. When individuals decide to put something in common there always is a cognitive effort underlying that attempt. Again, the docile agrees that there are rules that regulate social life while the nondocile has a hard time understanding them and, in most of the cases, these social
116
9 The “Docile” Organization
rules are secondary to their utility function. To make a long story short, what defines the difference between non-docile and docile is that the latter understands the social dimension better. Thus, these individuals fit the social environment while the nondocile has a hard time doing that. The result is apparent and obvious if we look at it through the glasses of real life. Although economic models attempt to show you that there is no room for social beings, the field is moving forwardxi and these models look like applications of an old nineteenth century ideology. Our experiment is not concluded yet. Does the non-docile individual disappear? Stop reading. Look around (I hope not in the mirror) and think. Have a look at the cover page of the newspaper again; I am sure you will find something on the top executive, war master, politician, or man/woman of the day. Do you think that selfish individuals do not exist? Do you think there is no room for them? Although I don’t intend to make hasty generalizations, there is a possibility you think that one out of the many individuals you find cited in the newspaper is selfish. In fact, if we want the theory to be consistent with our experience, we cannot allow them to disappear in our experiment. The assumptions that we have made lead to a model where non-dociles do not disappear, show a low fitness, and are part of our hypothetical world. This is where this thought experiment differs from Simon’s model, and it is far from social Darwinism.xii Here comes another critical thought, since sometimes we meet people that constantly and consistently behave selfishly (i.e., are non-docile). There are organizations where the number of non-docile people is particularly high and persistent over time. How can we explain this phenomenon?
From Evolution to Social Relations What if Simon presented his idea of docility for purposes other than those that I have in mind here? What he wanted to do with his later work was to show that docility is a human trait that allowed evolution to go the way it did. One of the byproducts of this mindset is altruism, as I anticipated. This way he showed that neoclassical economics portray an idea of human beings that should be rejected because it does not match the reality of human evolution. The objective of those papers was, once again, to show how descriptively poor the economic models have been. His simple model makes this point very clear. However, and this is my take here, docility may probably be more useful than that if we use it to understand decisionmaking trends in social environments, like organizations for example. Here is the idea. Docility is the tendency to lean on other people’s dataxiii to make decisions. This happens all the time during our lives. When we are children, we lean on our parents’ and strict relatives’ knowledge; when we grow up, the circle extends to include teachers, friends, teammates, etc. The adult individual has a lot of people that provide data that he/she uses to make decisions. Since this is what happens in every domain where the individual lives, we need to make some restrictions if we want to understand and analyze docility: (a) consider a limited social environment (e.g.,
The “Docile” Individual
117
a company, an organization), (b) focus on social relations. How does our idea on docility change if the “world” of our thought experiment becomes an organization? How can we analyze docility in that case? I believe that the choice of a restricted environment, such as an organization (vs. society as a whole), and the idea to study it in strict relation to decision making (i.e., closer to its definition) make it easier to study and, at the end, enhances its practical use. Let us focus on what docility is about. We know that it involves a socially based retrieval and use of data that allows the making of a decision. How does this happen? It happens through social external resources which constitute the basis for social relations to emerge. Remember that a social channel (Chapters 7 and 8) is something that transfers information that somebody wanted you to have. What makes a resource (or channel, here the two are interchangeable) social is this willingness to provide data. Mediums may be different, but as far as a social transfer is involved, we are dealing with a social relation. This happens all the time. You can have contact with another human being or may be using a computer to get information. What counts from this perspective is that you are leaning on other people’s data, and are using that data to make decisions. You are being docile. If you do this in an organizational setting and make decisions that affect your working life, you are being docile in that environment. This discourse leads to three implications: (a) we can measure the extent and the degree to which a person is being docile; (b) docility attitudes may vary depending on the context/environment; (c) docility attitudes may vary depending on time. Degrees of docility. The extent to which one person depends on social resources defines the individual degree of docility. There are many ways to measure this degree of docility. First, there is a personal belief, or what one person thinks of himself or herself in terms of how many times he or she takes data from other people, on average. I tend to believe that people, especially when at work, think of themselves as open to comments, suggestions, and information, and thus think they are highly docile. This is a positive attitude to have about oneself, isn’t it? However, this idea may be different from what other people think of the same person. The manager may think (of herself) she listens to comments, suggestions, etc., coming from her subordinates, although these people may have completely different ideas.xiv This contrast leads us to the second measure of docility: what others think of that person. The environment. The previous argument leads us to define docility in relative terms. What we experience is that people’s behaviors and ways of thinking vary greatly depending on their environment. If you will, this derives also from the DC approach and can be explained through the idea of rationality presented in this book. Individuals develop different cognitive processes depending on the interactions they are exposed to. It may well be, for example, that an individual shows docile trends when volunteering for the Cancer Society, while the same individual is completely non-docile at work. Moreover, this attitude may change within the same organization, depending on the team, project, group of people, leadership, relation with the
118
9 The “Docile” Organization
boss, subordinates, coworkers, rules and regulations, financial and economic situation the company is facing, etc. We cannot expect a constant level of docility from the same individual in different contexts. When I refer to the level of docility of an individual then, I can see it as average or as related to a given situation. Another interesting point emerges here. Different organizations usually present diverse levels of docility. This is what Bardone and I call the docility effect.xv Organizations, in other words, support docile behaviors when they enhance organizational performance, well-being, working lifestyle, and when they help to smooth relations, provide support to useful practices, help build up one’s self-esteem, lead to creativity, and the like. Time. Docility is not only space-based, it depends on time. The fact that we all use our cognition everyday incessantly gives us a high degree of variation in the extent to which we would like to use it. Although we cannot stop using external social resources—I do not provide explanations here but it is worth considering—we can vary the intensity with which they affect our cognition. Moreover, we can vary the type of social resources used. For example, we can start to gather suggestions from a new hire instead of continuing to use our usual advisor. Or, we can opt for computer-based resources instead of “hard” social interaction. These changes happen at a given time and support our behaviors. When you are not motivated to stay in your position at work and you decide that you need to start looking for a new job, your behavior changes dramatically (even if you tend to hide it). In particular situations you diminish the quantity, and especially the quality, of social interactions in the organization you want to quit. Your docile attitudes tend to slow down to a point where you are no longer docile to those around you. You start thinking of your new job, and maybe you implicitly think of your new sources of docility.
Active and Passive What is very strange in Simon’s definition is that if you stick with it you cannot describe what happens within an organization or in any social environment. Consider the last example of an individual that wants to quit the job and look for a new one. Docility is defined only on the basis of information, suggestion, comments that the individual receives from other people. Is this sufficient? Do we only receive data from external sources? In other words, do we only reproject? Or, do we externalize also? This is how the original concept of docility can be modified through the distributed cognition approach. For the concept to be useful and to gain more explanatory power, it needs to include an active aspect. There is a Latin origin of this word (late 15th Century). Docility means “apt or willing to learn,” from Latin “docilis,” deriving from “docere,” i.e., to teach. This is very important for defining what we mean by the word “active” docility, since what we are talking about here is the attitude that people have to share information with each other. And, as anthropologists explain, this is a core attitude for human beings.xvi The definition of docility becomes then
Levels of Docility
119
Docility is the tendency to depend on suggestions, perceptions, comments, and to gather information from other individuals on the one hand, and to “provide” information on the other.xvii
It is this active-passive process that defines what docility is about. It is rare or uncommon that people lean on social channels without exchanging with them. Of course, it depends on many variables related to the medium, the individual, and the basis of knowledge transferred. It is apparent to me that this attitude needs to be considered in its entire process, not simply limited to a one-sided relation. The argument that could emerge here is whether we stop considering docility and talk about learning. I believe there is a chance that we end up confused if we do this. Docility is not learning; it defines a cognitive trait that is apparent in specific behaviors. While we may say that docility can lead to learning, it is not clear whether the latter is always related to the former. Moreover, docility is strictly connected to the DC approach while, as far as I know, there is no theory of learning that is related to that same approach beside what Hutchins considers in one of the last chapters of his landmark work.xviii Docility has the advantage of bringing together cognitive processes (decision making) and human behavior when they are based on social resources.
Levels of Docility Using the active and passive framework, we can define different levels of docility depending on the intensity they have in the organization. The more individuals show both active and passive docility, the more the organization supports these behaviors, and the more intense docility becomes within that organization. It is a loop. The idea that I would like to present here is an analysis of what docility means within an organization. What are the byproducts of docility? What behaviors stem from individuals being docile? This section is dedicated to the analysis of docility in respect to decision making in organizations, bandwagon effects, socially responsible behavior, and advice taking and giving. I believe that the definition of these aspects could help to better understand what the contribution of this concept is to rationality and decision making.
The Prerequisites of Docility Docility is defined as a decision-making activity where external social resources are involved in the process. What happens to the organization and to the relation between individuals when the social environment is analyzed through these lenses? There are three basic conditionsxix that support the way docility emerges in any given environment between individuals. These conditions are fundamental to describing docility in organizations since the stress on the first, the second, or
120
9 The “Docile” Organization
the third element can change the quality and quantity of socially based decisions. Docility depends on (a) the community, (b) the standards that are used to encrypt information, and (c) the public dimension of information. A community is necessary because docility emerges only if and when individuals share something, this being the place where they live, a goal, a thought, an ideology, or more. The sense of being part of a community enhances sentiments of trust and facilitates the sharing of information. Sometimes an organized community presents a structure for power where people are supposed to share information depending on specific rules, loyalty, group dynamics, etc. The underlying idea is that it is very unlikely people are docile in a community of strangers, i.e., outside of their “world.” In other terms, and all other conditions being equal, docility emerges in social environments where there is something to share: communities. The example is a travel abroad. Suppose you are in Chile and people do not speak English. Besides the problem of communication that relates to the language, you do not feel part of that community and tend to make decisions based on information sharing that takes place with people that travel with you, for example. This is because that is your temporary community and you are being docile with them, not with Chileans. Let me state it once again: Docility is community-based. The second condition for docility to emerge is the existence of a standard. This seems an obvious condition that is the basis of any communication whether it is social or not. However, by standard I do not mean only the “code” that we use to share information; it is not only the language (as in the case of Chilean people). It is about a specific pattern, behavior, or way of thinking that people of a given community use to share their information. For example, the community of mathematicians is one of the most docile because they work together on the basis of a shared formal “code”, which is mathematics. It is not only the language that they use in their articles, and it is not only the symbols that they operate with. There are rules that they must follow to demonstrate theorems and to advance in their discipline. They strictly follow a standard, defined as the rules of mathematics. If you think about this, you realize that there are many standards in our society. Culture is one of these standards that has many substandards. Docility emerges when people use the appropriate media, methods, behaviors, and follow the (formal and informal) rules that facilitate a decision based on information coming from social channels. In other words, social information conveys and drives decision only when a standard has been fulfilled. For example, there are many different ways you can use to give a suggestion to a friend that is—in your opinion—about to make a mistake. In the US culture, it is all right to be straightforward and tell the friend “I believe you are wrong, don’t do that!” The same sentence pronounced, for example, in a Mediterranean and/or Arab country has a completely different meaning and cannot be delivered as I wrote it. Even US culture has different approaches to this simple point, depending on who the two friends are and where they are. Standards for communication in a bar are extremely different from those you have at home, in a work environment, or at a funeral. A failure to stay within those standards transforms into a (potential) lack of docility which is, in turn, a threat to individual decision making.
Levels of Docility
121
The public availability of information you share is the third condition for docility to emerge. When people exchange information in a community using the appropriate standards, different people can access that information. Of course, it is the individual that decides to whom the information is made available, but still the public dimension is concerned. This condition is particularly binding when we consider an organization. There are many social channels that are used to share information with a significant amount of people (e.g., annual reports, press releases, invoices). Some other interpersonal exchanges are less public and may involve a limited number of people (e.g., confidential communications, board of directors meetings, communication to employees). The meaning of this public dimension is that people externalize their thoughts, and this is what makes docility possible. It is apparent that the public dimension is partly unavoidable, especially if we consider that behavior is something “publicly available.” However, what counts here is a public dimension that involves a conscious willingness to externalize it—this is consistent with our definition of external social resource—and to let others use it as part of their decision-making processes. If only one of these three conditions is missing, docility is less likely to emerge. Docility in organizations depends on the fact that one, two, or all three conditions/dimensions prevail. The following explains it better.
Docility in Organizations As I mentioned in the previous pages, we can define individuals on the basis of the extent to which they show docile behavior (and cognitive attitudes). With this simple idea in mind, we can isolate people who show significant levels of docility from people who don’t. People that fall into the latter category (non-docile) do not use significant data from other people when they make decisions, while people that fall into the former category do. Among docile individuals there is a divide between people who are only ordinarily docile from those who are exceptionally docile.xx I believe that we can find these three categories—non-docile, ordinarily docile, and exceptionally docile individuals—in every organization; what is interesting here is fully understanding the implications of considering docility a human trait. The first step is to better define who these people are. Ordinarily docile. By definition these people make decisions considering social channels (specifically other individuals) as a major basis for their choices. They like to exchange information, to give and receive comments, suggestions, and advice in general. If you pick a company, any company, these people are the ones that lean on standardized social practices, rules, routines, and bylaws. Ordinarily dociles (oD) do not make a distinction between social channels such that any social information is useful for them to make a decision. The identification with their jobs and with the company is solid for these individuals such that they tend, on average, to trust standardized practices and procedures as they are available in the organization. Routines are what they are very good at. It is apparent that I am describing somebody who
122
9 The “Docile” Organization
develops docility on the basis of the prevalence of one of the preconditions. All these three are in play when docility exists, but individuals may stress one more than the other. This is what happens with ordinarily dociles. These people use docility in its passive side most of the time; they take information more than they provide it. What I mean is that oD expresses also active attitudes, but this is a secondary strength for them. Making decisions on the basis of the information they gather from social channels is the core of their cognitive and behavioral abilities. Exceptionally docile. If every community has individuals that show close-toaverage levels of docility, we can also have outliers. This is the case for upper-level outliers. These people are the ones that show the highest levels of docility in a given community. Active and passive sides are both present, and they operate at their best. Exceptionally docile (eD) individuals are the ones who make the exchange of information, suggestions, comments, and advice the true basis for most of their decisions. One of the most significant differences is that data are clustered around them in a way that makes them very knowledgeable and proficient in their jobs. The information they provide to other people in the organization is of high quality and particularly appropriate. This is related to the fact that they (implicitly or explicitly, consciously or unconsciously) understand very well all of the three conditions that allow docility to emerge and expand in an organization. An exceptionally docile person should provide coworkers with useful comments and insights on what they need to know, and should listen to their coworkers very carefully. It is worth noting that I am not writing about accuracy of decisions done by eD people, but of particularly rich and appropriate decisions. For example, suppose a bank is in the process of developing and changing the software they use to analyze corporate information.xxi They need a very limited pool of people to run the transition. These people should be able to meet the employees (at least the managers, if the bank is large), explain to them the motives for this change, how the new software works, how to read the outcomes, how their customers will benefit from this analytical tool, and other related issues. Most important, these people should listen to critics and comments coming from the employees so that the software can be changed according to their users’ needs and annoying bugs can be eliminated (or reduced). The objective of these people is to have a better company, an organization that could offer superior services to customers, and let the employees be confident that the tools they use fit their needs. This is an example of how exceptionally docile individuals operate; they make decisions for the sake of the company and lean on other people’s information. Non-docile. I already described how these people (nD) behave in organizations in the previous pages. I can recall here that these individuals are not satisfied with the company’s goals, with their team’s goals, and probably with their personal objectives fitting the organization. It is likely that these persons are looking for a position outside the organization. nDs are not cooperative and do not see any value in sharing information with the rest of the people in the organization. They behave like the tourist that travels alone and arrives in a foreign country where he doesn’t know the language and doesn’t want to know, nor does he want to be there. As a consequence, he doesn’t trust anybody and tends to avoid social contacts.
Understanding Docility
123 Table 9.1 Characteristics of individual docility Prevalent behavior
Prevalent precondition
Info quality
Exceptionally docile (eD) Ordinarily docile (oD)
Active and passive Passive
High Medium
Non-docile (nD)
None
All three Standards, or community, or public availability None
Low-null
Table 9.1 offers a summary of what I presented for each type of individual. If we follow these assumptions, we will find that individuals in organizations make decisions on the basis of exchanges that exploit social channels. A limited number of them are exceptionally docile, and a very limited number are non-docile individuals.xxii DC supports the idea that individuals are social beings and that social resources are among the most exploited. This idea of docility also helps define the potential for the expansion of rationality. Social channels help people make decisions that better fit the organizational needs and define the extendibility of rationality.
Understanding Docility A significant contribution to the understanding of docility is given by the description of “what docility is not.” This is not an easy task since when you find an interesting tool (such as docility) that explains some basics of organizational behavior, you tend to focus on these positive aspects instead of seeing its limits. However, a theory grows exactly because of the deep understanding of its limitations, as well as of its potentials.
Bandwagon Versus Docility Sociologist Mark Granovetterxxiii studied a phenomenon that falls under the name of bandwagon effect. It is the concept of conformityxxiv to a crowd, to a group of people, or to any other apparent behavioral or thinking pattern. I already introduced this concept in the previous chapters; however, here I need to focus on a different aspect of the same phenomenon. Imagine students sitting in a class and the professor asks a question. What is the students’ attitude? If nobody raises a hand or speaks, it is less likely that anybody will do that. If you are in the class, would you rather say something or wait until somebody else does it? If you don’t care about being the ice breaker then we would say that your threshold is very low; if you wait until everybody else in the class has had their say on the point raised by the professor, then your threshold is particularly high. How many students do you need to listen to before deciding to speak? This is a measure of bandwagon, i.e., the tendency to repeat what others do without a specific reason or motive. Bandwagon is very close
124
9 The “Docile” Organization
to imitation and it does not require a consciousxxv understanding of the fact that those who jump on it realize that they are doing so. Is this behavior related to docility? How? I believe that docility and bandwagon are two different aspects of human behavior, and they are related in that the latter can be associated with levels of docility. However, I don’t see a pattern of causality between the two: (a) docility may facilitate the emergence of bandwagons but it is not their cause; (b) bandwagon does not support docile behaviors. In order to specify these two statements, I need to get into the bandwagon phenomenon more closely.xxvi What explains and fosters bandwagons? There are two major variables explaining bandwagon. The first is organizational culture.xxvii A stable and wellestablished culture is a prerequisite for imitation to emerge and grow. When change is not significant or it is perceived as not significant in a given social environment so that people’s behavior is based on shared and common values and beliefs, then individuals tend to reproduce the same behavior the majority of the other people are showing. This is the “church” effect. There are certain rhythms and regularities that shape a mass, for example. If you are a non-believer or if you believe in a different god(s) or goddess(es) and you happen to be in a church during a religious function, you’ll do whatever other people are doing: you stand up when they do, sit down when they do, even sing (if you know the song) and pray if you can or want. You follow the bandwagon. A mass is made of regularities; regular churchgoers share the same beliefs, the same values, and have built a common way to “live” the religious experience. They have what we should call here a culture. Of course, we are not giving to this word the full array of meanings that anthropologists do. However, we are characterizing a specific way of doing things or thought processes that are shared and define any given community. This “church effect” is very powerful, and shall happen to anybody who becomes part of that community, even if it is for a few hours. This could also happen in a company or in any organization. When there is a common and well-established way of thinking or behaving, that is how bandwagon is more likely to emerge. It doesn’t matter if it is a trivial behavior, like having lunch in front of your computer, always smiling at customers (does anything like that exist? Oh yes, McDonald’s), never talking about personal problems with colleagues, or if it defines how people approach a problem, such as the “right” software to use, what outcomes to use, from what source comes the “most relevant” information, and the like. Bandwagons can be very pervasive, and this can be related to the stickiness of a culture. A strong culture may also support change as a major variable; nevertheless, if this culture of change is strong enough people shall still jump on bandwagons. A strong organizational culture can be defined as a system of shared values, beliefs, and rules that affect people’s behaviors, ways of thinking, and expectations such that it becomes pervasive and somehow predictable. An example could help explain this better. How hierarchy is perceived and practiced has to do with organizational (and national) culture. Relations could be based on a friendly approach to power and authority, where subordinates are allowed to speak frankly and openly to their bosses and to anybody who is in charge of something. The
Understanding Docility
125
presence of this aspect is related to the way people within the organization think about this “egalitarianism” in organizational relations. Sometimes it does not matter what the corporate bylaws (if any) prescribe on these top-down or bottom-up relations; it is the openness and actual behavior that helps establish a positive and supportive climate. A bandwagon is easy to follow with respect to power and authority relations. Especially for new hires, the rule is to do whatever other people are doing, i.e., follow what seems to be a common and shared pattern of behavior. They will conform to what they see. When the culture is structured, formalized enough, it is possible that people do not even think they can behave differently.xxviii It is worth noting that few studiesxxix have addressed bandwagon at the organizational behavior level of analysis. There are clues to state what I have in these pages; however, an empirical evaluation of what I am hypothesizing here has yet to come. So far, what was written above can be summarized by the following: The stabler and stronger the culture, the more likely people are to jump on a bandwagon. The second determinant of bandwagons is the nature of social relations. This passive and mindless imitation usually emerges when the imitator has a limited and superficial relation with people around him. The whole point is that when individuals know each other well, they talk, exchange opinions, comments, advice, exchange information, etc. Especially when at work, and depending on organizational and national culture, people may be reluctant to share their ideas on a specific behavior or to ask if something is appropriate. In this case, they tend to adjust their behavior to what seems prevalent.xxx The significant implication of this is that between friends or close coworkers it is more likely that behavior derives from high levels of trust, shared ideas, commonality of intentions, and sympathy. The basis for bandwagon may look similar to what is needed for docility to emerge. I have hypothesized that loose social relations and strong organizational cultures foster bandwagons; how does this relate to docility? The two phenomena lean on similar elements; however, docility facilitates close social relations due to the fact that it needs some level of trust to emerge and become prominent in any given social environment. Moreover, it favors the creation of organizational culture, up to a point. The point being that too much docility may create instability and make a culture lose its strength. It seems to me that docility is based on distributed mindful interactionsxxxi between organizational members while bandwagon is not. In the previous pages I presented what I called the (pre)conditions of docility. Let me recall them in brief here; they are (a) community, (b) standards, and (c) public availability of information. At a first superficial look, these three conditions may seem to work well for bandwagon too. However, the three elements here are always consciously and deliberately chosen with the individual’s docile tendency. The first element, the community, is not needed when we analyze bandwagon, for example. The tendency to imitate other buyers’ choices is very high, but that does not mean that the two buyers are part of the same community. Imagine that you bought a Jimi Hendrix’s CD because, while at the store, you saw two people buying the latest remastered edition of Experience. The first person was about your age, the second was a young kid. Are you part of the same community? Maybe the only thing you
126
9 The “Docile” Organization
share is Hendrix’s music. Does this suffice to call this a community? This same pattern could be found in organizational behaviors too. Especially in large companies, the fact that you share the same place where work does not mean a community is out there. However, to the extent to which there are common interests, culture, and goals, the organization is a community. It just doesn’t seem to be an element that defines bandwagon since this could emerge even without it. The second element, standard, is not needed either. The only significant pattern that imitators are willing to follow is something that is recognizable. The standard here is not in the way people communicate or in the appropriateness of that communication process; it relates to something that is ordinary behavior or way of thinking. It is what people do on average; it really doesn’t matter if they are following any communication standard. Imitation can emerge with or without standards. The third element, public availability of information, is relevant for both phenomena. Bandwagon is based on the fact that information (i.e., a behavior or thought) is observable, i.e., available. However, what makes this different in the case of bandwagon is that there is no willingness to share information. We do not buy a CD because we want to share our choice with other customers in the store; the willingness is lacking. Docility is based on the active sharing of data, opinions, suggestions, etc. You listen to a comment because you want to, and you give an opinion because you want to. The perspective is completely different. One last interesting point is that of cognition. While bandwagon has been defined as a logical fallacy (see Chapter 5), docility is more a strategy to avoid fallacies or use them in a more meaningful way. The use of social channels does not necessarily imply fallacy avoidance but extends the possibilities and opportunities for our cognition to operate in a more consistent and successful way or, to state it more clearly, more rationally. Mindfulness becomes socially distributed and hence subject to some sort of organizational check when docility is widespread. At the end of this paragraph, I believe it is apparent that docility is not bandwagon and is not even close to it! However, the relation we should consider at this point is that docility can help organizations limit the emergence of bandwagons (and maybe of other similar fallacies) through mindfulness that becomes distributed among individuals.
Individual Social Responsibilityxxxii Social responsibility is one of the most widely discussed, analyzed, published, and practiced ideas that has recently hit the ground of management. To be precise, the idea is not recent at all, but it received a full consideration in recent years. This doesn’t seem a fad because it is backed by deep social and economic concerns on our uncertain future. The attention has always been over corporations.xxxiii In fact, the acronym that is used more frequently is CSR, or “corporate social responsibility.” Focusing on individual social responsibility can be interesting and unveil important circumstances under which a responsible behavior finds an explanation.
Understanding Docility
127
Managers, not corporations, sit in the company boards, speak at meetings, go to work, take responsibilities, have duties, and ultimately make decisions. Therefore, emphasis on individuals seems to be well-placed, at least. I am not arguing that this focus is new or unusual, but it is not prevalent. While socially responsible behaviors are growing, a cognitive approach to social responsibility is lacking. This is what this section is about: Is there a way to find evidence that an individual’s cognition supports social responsibility? As the reader may guess, docility is my answer. How? Social responsibility could also be defined as the tendency of individuals to establish or “maintain cognitive advantages from the social resources (channels) that they exploit more frequently.”xxxiv Otherwise stated, social responsibility helps individuals keep and reinforce connections to their most exploited (or relevant) social channels. What is a cognitive advantage? It can be defined in a relative way, through a comparison between one’s cognition and that of the others. A cognitive advantage may be also defined in terms of better performance and results. It is a procedural and substantive enhancement that can be achieved if we exploit external resources in the most appropriate way. This means that external resources need to be tied up to one’s cognition: A resource is not good for everybody, and different resources may lead different individuals to similar outcomes. Therefore, a resource is useful when it provides the individual with an advantage; otherwise, people tend not to use the same resource again (unless they are forced to do so). When comparing types of resources, we have to admit that some social resources need a special care if we want to use them more than once.xxxv To explain this better, a brief inquiry on what responsibility is might be helpful. As already stated above, external social resources have been defined as they relate to other individuals’ willingness to share information, while non-social resources do not show this relation. We used a tree before to exemplify the second, and we can use a newspaper article for a resource of the first type. Of course, the idea of a tree or of a river that is held in our mind is significantly influenced by its socially construed representations. However, it is not (to some extent, at least) a product of human intervention, production, shaping, etc. It is not a work of art, it is Mother Nature. In turn, a social resource/channel is something that a tree or a river is not. It is “a mediator of socially-based information, in which the sender actively gives that communication, and the receiver actively takes it.”xxxvi The example of a newspaper article serves the purpose of showing the difference. An external resource becomes “social” only if and when a reader makes use of it. Advice giving and taking offer a better example (see below). The short recap on social and non-social resources helps us answer the question: How could responsibility lead to a cognitive advantage? Responsibility can be seen as a reinforcement mechanism employed in the exploitation of social channels. A short example may define this better. Consider the journal article and think of its author. Imagine that the journalist decides to write false information for the only purpose of gaining visibility. Due to the professional code, this person faces the risk of being fired right away if someone finds out. However, this is not the worst case scenario. In fact, he may face a dramatic exclusion: Readers may label the journalist
128
9 The “Docile” Organization
as unreliable and stop reading his articles. Otherwise put, the social channel ceases to be exploited because of the journalist’s irresponsible behavior. When irresponsible behavior is perceived as repeatable, it makes the use of the social channel less likely. This is how responsibility may be intended as a reinforcement of the use of any given social channels; this, in turn, means that distributed cognitive processes are reinforced, or more stable. Now, and from a more detailed perspective, what is “responsibility”? To “take responsibility for something” is equivalent to the expression to “have a duty in regard to something.” This definition may be sufficient for general purposes; however, responsibility may be internal or external. The common way to think of responsibility is internal. This is the case of somebody taking responsibility for an action, a thought, or something else. External responsibility is a subtler concept as it can be associated with self-deception or bad faith. This happens when “self-deception, or bad faith, creates a situation in which human beings relinquish freedom and externalize responsibility.”xxxvii A person in a condition of bad faith deceives himself “by constructing a limited reality that does not take into account the full range of choices available to him, and this, alas, is a condition in which many people live all their lives. It is from himself that he is hiding the truth; the deceiver and the deceived coalesce into a single consciousness in a way that must be distinguished from true mental illness or malfunction of consciousness.”xxxviii Therefore, the idea of social responsibility can be redefined as internal only or, on a negative angle, as bad faith avoidance. Moreover, this approach suggests that social responsibility enhances and extends the range of choices that individuals have. On the contrary, those who lean on external responsibility are more limited than they think. In supporting the exploitation of social channels, social responsibility seems to be connected to those individual attitudes that we have called docile behaviors. It happens that this appears to be a reinforcement or, better, a byproduct of docility attitudes.
A Theoretical Framework for Advice Giving and Taking The definition of docility includes, but is not limited to, advice taking and giving. This is a trivial point since it is part of the way we defined docility. However, I believe that docility does not explain the why of advice taking and giving; it explains the how. Individuals take advice because this is the way their cognition works: they lean on external resources. First of all, advice is defined in the literature (and in this text in particular) in a very narrow and strict way. It is a recommendation that an individual offers to another individual, and it contains a specific suggestion on a decision. Advice is a suggestion on how to make a certain decision. This makes it a very specific kind of social interaction. Suppose the decision maker (judge) is about to send an e-mail to the boss with a recommendation on quitting a product line that is not selling well. An advice from a coworker is of the type “do it” or “don’t do it”, while another
What Is a Docile Organization?
129
source may be more vague. For example, it may happen that the judge receives an e-mail with data on people losing their jobs because of the cuts in the company’s product lines; or, the e-mail may be from the boss and suggests the judge look at new data on the quality of those products. It is apparent that docility explains advice taking-giving, as well as other information that helps people make decisions. From this angle, the study of docility is a general theory of decision making in social environments. There is no general theory of advice. I suggest that docility could offer a framework where we can put advice giving and taking together with all of the other byproducts of docility.xxxix However, these processes don’t seem to be byproducts over part of what docility is, at least at first sight. The question that needs to be addressed is whether the decision maker has to be docile in order to take and use the advice. Or, is a person that simply exploits others’ expertises, knowledge, or positions for personal and selfish purposes a docile individual? Or, when the purpose of taking the advise is not that of using it but that of pleasing the advisor, does this still fall under a typical docile behavior? The reasons why people take advice can be very complex. For example, when an individual is obsessed with a specific and single goal, let’s say career advancement, he or she could do anything to achieve that objective. Among the strategies that one can exploit, there is the one that includes a fake reaction to advice coming from people in power. To please these people that are in charge of making decisions on the judge’s career, the judge can take the advice even when he or she thinks it is not good. Once again, is this person docile? Docility is about using information from social channels to make decisions; however, there is trace of purpose, motivation, or of other psychological and cognitive variables that distinguish non-docile individuals from dociles.
What Is a Docile Organization? Scholars scarcely employ Simon’s concept of docility. This, as discussed already at the beginning of this chapter, is based on (a) its vague link to bounded rationality, and (b) the fact that it needs a strong cognitive approach to support it. In this chapter I have presented a more accurate and specific definition of docility and linked it to the distributed cognition approach and rationality. One of the most important implications of what we have examined in the previous pages is to consider whether organizations take docility as part of their structure. What I mean here is that if the large majority of people are docile and show docile attitudes, then the organization may include facilitating mechanisms that support and foster the emergence of these behaviors. This is what I mentioned above and falls under the name of docility effect. Organizations are complex social environments that function on the basis of, among many other things, individuals exchanging relevant information and making decisions on the basis of these pieces of data. It is likely that organizations provide individuals with appropriate
130
9 The “Docile” Organization
social channels where they can be docile. These “structured” mechanisms may fall under rules and/or under informal or behavioral norms. A few examples could help understand what I mean by this. The airway companies, supported by the International Civil Aviation Organization (ICAO), have specific rules for their pilot-copilot communicationsxl during emergency operations. The general rule is that the pilot must listen to what the copilot has to say, and act accordingly. Pilots get trained on this. Of course, experienced pilots with the largest number of flights could come out to also be those that are right most of the time. However, studiesxli on the outcome of situations where this explicitly docile behavior was not taking place have been a disaster. I am not an expert on risky surgeries, but I imagine there must be a procedure that helps surgeons get support and advice from other individuals in the room. This is to limit the risk of doing the wrong thing. The team is involved in making a decision that could save a life. The final decision is at the expense of the one who makes it, though opinions have to be expressed. Information and communication technologies (ICT) help companies develop docility. Many companies have a sort of in-house chat line where employees can exchange messages on their jobs or on particular projects they are working on. This facilitates the number of interactions between people and induces the exchange of information, some of them useful to decision making. The corporate “dedicated” chat line is an example of a modern social channel where individuals easily get their chances to take other points of view under consideration. Although it is not very easy to get sophisticated advice with the chat line, it is possible to exchange relevant and sensitive information that could help the decision-making processes. It could also foster a team or group decision making when the final decision is shared by the majority or all of the members of a team. On a different take, but close to ICT, are corporate cell phones and advanced walkie-talkies. Here the interchange varies depending on the intensity with which people need to communicate with each other. However, these are channels through which individuals get information, comments, and suggestions that help them to make decisions. What I mention here are communication tools that become docility enhancers or facilitators to the extent that the individual use them to make decisions. Communication does not necessarily lead to a decision-making activity. Therefore, the point here is not that the company using these communication tools is docile, but that these tools favor and push the probability that docility emerges among organizational members. Docility can also be related to a specific behavior that is supported by organizational culture. It is the case of cooperation. Depending on national cultural variables and on the way these are integrated in the organization, individuals may have the tendency to work in isolation, to cooperate with all employees, or to cooperate with coworkers only. This tendency may also depend on industry standards and individual personality traits. However, all other variables being equal, the organization may push people toward more or less cooperation. The last case, that of cooperation that is not affected by the power structure of the organization, is very important because
Notes
131
it depends on behavior. You cannot force people to cooperate if they are not willing to do so. The result is a cooperation that results in limited outcomes that do not exploit its full potential. In other words, those with the higher managerial positions have to show themselves willing to cooperate with everybody else in their team. If the perception of their attitudes is not cooperative, then people shall not cooperate in any meaningful way. They face the probability to be let down by their boss, for example. When individuals show their attitude to cooperate, it means that they are being docile and inviting other people to do the same. And this is another aspect of how a docility effect can be achieved. The organization can integrate docile attitudes in its structure and enhance people’s docility tendencies. This, in turn, makes the whole organization docile.
Summary This chapter has introduced and explored the concept of docility. First we have defined the conditions that allow docility to emerge: (a) community, (b) communication standards, and (c) public availability of information. Second, we have seen that docility can be used to define different roles and tendencies of individuals in organizations. Third, docility has been related to social responsibility, bandwagon, and advice giving and taking. Fourth, I examined how the docility effect may define a “docile organization.” We are now ready for some concluding remarks.
Notes i. ii. iii. iv. v. vi.
vii. viii. xi. x.
See Humphrey (1976). Dunbar (1998) and Dunbar and Shultz (2007). Richerson and Boyd (1998, 2005). Simon (1993b, p. 156). Simon (1993b, p. 156). See Maturana and Varela (1987), for an example of this. Recently, the theory of nicheconstruction builds on some inconsistencies of a strict Darwinism (Odling-Smee et al., 2003). It is A Theory of Docile Society, Secchi (2007). See for example Odling-Smee et al. (2003). This was a first version of the paper on super-docility; Secchi and Bardone (2009b). The original model by Simon (1993b) presents a set of three equations that define fitness. In particular, they define the fitness of the selfish (fS), that of the unintelligent altruist (fU)—individuals that do not discriminate with whom to be altruistic—, and of the intelligent altruist (fI). The equations are fS = fn + faI x qI x cI + faUx qU x cU fI = fn + fd x dI + fal xqI x cI + faU x qU x cU−c x cI fU = fn + fd x dU + faI x qI x cI + faU x qU x qU x cU x cU − c x cU Parameters are defined as follows: “fn is normal fitness; fd x dI and fd x dU are increments of fitness for docility; faI and faU are increments in fitness from others’ altruism; cI and cU represent the extent to which I and U are altruistic; dI and dU denote the
132
9 The “Docile” Organization
xi.
xii. xiii. xiv. xv. xvi. xvii. xviii. xix. xx. xxi. xxii. xxiii. xxiv.
xxv. xxvi.
xxvii. xxviii.
xxix.
xxx. xxxi.
abilities of I and U to benefit from docility; qI and qU are percentages of I and U in the population. The cost of altruism is c” (Simon 1993b, pp. 157–158). If we attribute the following values to parameters, “fn = 1.01; fd = 0.02; c = 0.005; faI = 0.01; faU = 0.005; qI = 1/3; qS = 1/3; pU = 1; dI = 2; cI = 0.8; cU = 1; dU = 1.” (Simon, 1993b, p. 158), in 30 generations, altruists reach 72% of the population while selfish decrease to 18%. In 721 generations, society has only intelligent altruists (or highly docile individuals; Secchi, 2007). The most interesting steps are those of behavioral economics (see notes on previous chapters). However, there are other parts of the field that move in equally interesting directions; see Frank (2004). Details of the model summarized here are in Secchi (2007). Data here stays for comments, suggestions, recommendations, and information. I believe that techniques developed by Social Network Analysis (e.g., Knoke and Yang, 2008) are particularly helpful to structure a study on docility in organizations. Bardone and Secchi (2006, 2009), Secchi (2007), and Secchi and Bardone (2009b). See Humphrey (1976), Dunbar (1998), and Dunbar and Shultz (2007). Secchi and Bardone (2009b, pp. 347–348). In Hutchins (1995), there is a chapter on learning that explicitly connects the DC approach to this cognitive activity. What I present here is an elaboration of findings published in Secchi and Bardone (2009b). See also Bardone and Secchi (2009). This is what is formalized in Secchi and Bardone (2009b). This example comes from a real case and I witnessed it happening while working as consultant for a bank. The result comes from the model analyzed in Secchi and Bardone (2009a). Granovetter (1978). A bandwagon effect emerges when “the demand for a commodity is increased due to the fact that others are also consuming the same commodity” (Leibenstein, 1950, p. 189). This happens when consumers “purchase a commodity in order to get into ‘the swim of things’; in order to conform with the people they wish to be associated with; in order to be fashionable or stylish; or, in order to appear to be ‘one of the boys’” (Leibenstein, 1950, p. 189). Fiol and O’Connor (2003). In a working paper with Bardone and Secchi (2009), we elaborate on bandwagon and docility and present a formal model and a simulation to study the conditions that support these phenomena in organizations. A second paper is forthcoming and takes one step further in mathematical modeling. See for example, Schein (1990, 1996). The stability of a corporate/organizational culture is also supported by a positive social and economic performance. When everything goes extremely well or even fine it is difficult to think out of the box and to challenge established and common behavior patterns. On the contrary, when a crisis emerges and the organization is challenged by financial shortcomings or the management is threatened by external stakeholders, it is easier that criticism and “deviant” (from the ordinary) behavior emerges. Organizational behavior studies focus on the so-called macro aspect of the phenomenon so that interest is in the processes of imitation between companies (e.g., Abrahamson and Rosenkopf, 1993, 1997). A mix of micro and macro analyses can be found in Fiol and O’Connor (2003). Sociology offer insights on micro-organizational behavior (see, for example, Chiang 2007). I am trying to describe tendencies, in any case I believe that this is what always happens. The word distributed is very important. Docility is an individual trait but also a community property (that comes from its preconditions). Therefore it is the network of relations
Notes
xxxii. xxxiii. xxxiv. xxxv.
xxxvi. xxxvii. xxxviii. xxxix.
xl. xli.
133 that supports or rejects docile behaviors. There is no mindfulness—i.e., conscious rational behavior—when the organization (other human beings, the community) does not support this cognitive activity. It is intuitive that one individual cannot be continuously mindful for a long period of time. That individual needs social support to be as mindful as possible. This is when organizational docility comes into play (see below in the text) and this is also the difference between what is in this book and what Fiol and O’Connor (2003) write in their work. This paragraph summarizes my paper “The Cognitive Side of Social Responsibility,” Journal of Business Ethics (2009). This is apparent from literature reviews: Garriga and Melé (2004) and Secchi (2007). Secchi (2009, p. 575). This makes also the difference between non-docile and docile. Non-docile individuals don’t care about exploiting the same social channel more than once since they don’t have a socially oriented (or pro-social) mind. Their attitude can be thought of as a response to immediate needs, to a focus on the short term. Preservation of social relations is not their objective. This provides a micro-explanation of the reason, according to orthodox neoclassical economists, social responsibility finds no place in their theoretical system. Secchi (2007, p. 575). Magnani (2007, p. 129). Magnani (2007, p. 131). It is not in the economy of this book to analyze the byproducts of docility. However, a short list may include cooperation, altruism, social responsibility, advice taking-giving (Secchi, 2009). I took this from Gladwell (2002). It is, again, Gladwell (2002) that offers a vivid and effective reconstruction of this point.
Chapter 10
Conclusions
Here we go! This is the final section of this book on rationality and decision making. After this tour of my thoughts on some of the most important aspects of individual and organizational life, the final question that every author has to address at the end of his/her work is, “What have we accomplished?” This is not an easy question, especially for a book that attempts to merge a significant amount of otherwise scarcely connected models, approaches, and theories. Since what I presented here is a theory of human behavior in organizations, the result is based on implications of considering the distributed cognition approach and docility as key components of decision-making processes in organizations. There are different sets of arguments that we can address in this concluding section. Although these remarks provide insight on the material in this work, none of them is a true conclusion strictu sensu. They serve as guidance for future research to me and to anybody who is willing to join this exciting thought adventure. The first part of this chapter is dedicated to the exploration of past studies in the light of the distributed cognition approach and docility. The second section questions the appropriateness of the old distinction between group and individual decision making. Third comes a methodological discourse, and finally an overview of the meanings of the proposed redefinition of rationality.
The Point on Rationality Among the achievements of this book is the idea of “extendable” rationality. This concept is based on the fact that the two limits defined by Herbert Simon do not seem to hold if we switch from a “separatist” or “isolationist” to a “distributed” paradigm. Rationality is bounded because of limited access to information and limited computational capabilities. The former is the external, and the latter is the internal limit. If this divide falls apart we can think of rationality as based on the communication (interplay) between internal and external limits. From here, I have argued that the results we get present a completely different picture of individual rationality. What emerges from the break of this old inside-outside barrier is a new rational being, capable of modifying the limits of rationality depending on how these two bounds D. Secchi, Extendable Rationality, Organizational Change and Innovation, C Springer Science+Business Media, LLC 2011 DOI 10.1007/978-1-4419-7542-3_10,
135
136
10
Conclusions
are coupled. The external is not external anymore, it is part of cognition and defines how rational human beings can be in their decision-making processes. Rationality is therefore not stable, it is not defined once and for all for each one of us. It depends on how cognition changes itself depending on this internal-external interplay. Rationality changes as time passes by, as availability of new artifacts or external resources emerges, as fitness to the social environment grows. Our rationality is time sensitive. This is easy to understand if we think of ourselves or if we observe how people change during their lifespans. Cognitive processes of older people are not the same as those of a child; we can argue that rationality changes as well when it is confronted with age. However, if we compare two individuals, one can always argue that their limits are different because of their “internal” build-up rationality. In fact, for this thought experiment to be of some value, we can look at ourselves. How has your cognition changed since you were a child? Consider an individual who is a PhD student right now: How do you think his cognition changed since he was a child? What we do all our lives is learn. Some learning mechanisms are more accurate and work better than others; however, it is this process that modifies the way we develop our rationality. Moreover, it changes the way we make decisions. This is not only true when we consider the evolution of our cognition over time, it is particularly significant for specific decision-making processes. Individuals don’t make decisions at a definite moment in time. A decision may take hours, days or even months (sometimes years) to be made. During this period new resources may become available, new insights on the problem may appear, variables can acquire different meanings, etc. We become acquainted with external resources, how to use them, how they affect our cognition, and it is this experience that lets us change the way we make decisions. Sometimes everything happens in a limited period of time where the decision process changes through manipulation of external resources, or “through doing.” And this makes our rationality “extendable” over time. Extendable means that it has potential to move, that bounds are not fixed or stable; whether it does so at its higher potential or not depends on the integration between internal and external resources. This leads to the second variable. The process through which resources become available is worth analyzing as far as cognition is concerned. The process through which individuals learn how to exploit these resources is ongoing. A decision process made in isolation then means that the individual has no contact with anything (e.g., visual, touch, hearing) which makes it perfect from a neoclassical economist perspective, although impossible. Isolated decision making is something that communicates directly and internally with the brain only. I am not aware of any common practice that allows anything like that. In a real decision-making setting, rationality could change in relation to resources that become available at different times. It may be a piece of information, an advice, a new thought on a previous comment (re-projecting), an activity that you perform (e.g., write, read, speak, behave in a specific way). These activities move the bounds of individual rationality so that a solution becomes more apparent. The third variable that helps explain the extendibility of rationality is the fitness. A decision and its process are rational if they fit a specific social environment. This
What Are We Mapping?
137
is to say that individuals need to use their docility to be rational. Implications of this are socially responsible behavior, cooperation, altruism, advice taking and giving, and the tendency to avoid bandwagons (whenever possible). All of these processes help individuals make decisions that work in a given social environment. It must be clear by now that I am not presenting any theory of unbounded or full rationality. The whole point can be reduced to the fact that rationality is not limited in the way Simon (and many others after him) thought it was. It is not limited by internal and external variables; it is limited by a composition of these two sets of variables, where what is internal and what is external is not exactly definite ex ante. Decision makers overcome limitations when they exploit external tools so that their cognition adapts to them, and they can make a decision. This process highlights the fact that what makes human beings so peculiar is not the fact that they are bounded. Quite the opposite. We are exceptional—compared to other animals—because we can modify these bounds through their manipulation; we can extend the potential of our rationality and cognition depending on this interplay. Our rationality is not stable, our limits are relative to specific conditions and situations. And even when we can define these limits, they may change very fast due to interaction with external resources. This instability is what makes our rationality extendable, adaptable, plastic, and modifiable.
What Are We Mapping? When we consider the experiments that I presented in the first part of this book (Chapters 4 and 5) then, what do they look like? If bounds are not the most important variable that define human beings, then what do these experiments really measure? Is it accurate to say that they present maps of our bounded rationality? Or, are they mapping something else? A typical outcome of studies that map bounded rationality is to show how old paradigms don’t work. These are very important results when we think of what it takes to abandon well-established scientific paradigms. Studies of bounded rationality have shown us that the idea of a fully rational individual has very limited explanatory (descriptive) and normative (prescriptive) power. However, these studies are oriented towards the past; anomalies of the past paradigm are what they are concerned with. Moreover, students of bounded rationality are not interested in defining rationality. It is rare to find inquiries on what rationality means, on the meaning of bounds, or on the assumptions that hold when embracing BR.i When I presented the maps of bounded rationality, I have intentionally hidden these points since scholars are mapping something they forgot to ask about. The concept remained was that of Simon as he defined it more than 50 years ago. Anyway, what if rationality is not bounded the way students of BR think it is? What are they mapping then? I believe that it is possible to have a new reading of these experiments and of their findings. Of course, for some this process is easier than for others since they
138
10
Conclusions
were not set up to measure change in rationality bounds. Take the San AntonioSan Diego experiment.ii In order to show how heuristics work, researchers asked European and American students which city is larger between San Diego and San Antonio. Answers vary depending on personal knowledge, experience, and the fact that you have heard about one city at least. The point is that if you don’t know the answer, you go with the guess that is close to the only clue you may have (e.g., heard of San Antonio/San Diego through movies, friends). This simple experiment shows that individuals are not stuck in their position, but they adjust their perspectives depending on the level of interaction they have with the external resource. Results show how people move their bounds when they have no or limited clues on what the right answer is. Of course, in a nonexperimental setting this aspect falls down in a second as Web search engines can be exploited. When these are not available then advice can easily come up as a resource.iii When nothing like that is available, then we are defining a situation that is very difficult to become real. What is that study really analyzing? What rationality is that of not asking for further information when needed? It is a rationality that shows its adaptability to a particular (extreme) situation; researchers were showing how rational individuals cope with uncertainty remodeling their rationality to fit the needs of that moment with available information. Bounds were moving. The proponents of this experiment discuss about bounded rationality as an “adaptive toolbox,” and this is probably a link we may find between their findings, distributed cognition, and the idea of extendable rationality. We can easily reinterpret many other experiments illustrated in the previous pages of this book through an extendable rationality perspective. And I believe that this rereading makes more sense than the original view of BR simply because (a) it includes it and (b) it forces focus on the plasticity and malleability of human rationality. Researchers were mapping exactly this aspect of human beings: the adaptability to a different set of constraints in the use of external resources. They just didn’t know it.
The Individual and the Group One of the ideas that remained latent in this work is that of groups. I explicitly assumed that individuals live in social environments, where organizations are the most diffused and effective places where their lives are spent. However, many of us work in environments that are not composed of thousand people that participate in large organizations. Every one of us has his or her own social niche composed by a limited number of other individuals that help to make decisions. A significant amount of the literature on decision making is thus dedicated to team and group decision making. These studies usually stress how different the individual is from the group-related decision process.iv They also highlight how organizational decision making can be analyzed more properly if groups and teams are considered. Members of organizations, whether they are workers, middle
The Individual and the Group
139
managers, shop managers, top managers, or members of the board of directors, do not make decisions in isolation. The more complex the decision, the less these people tend to act as if they are alone. As I have specified in the previous pages, there are mechanisms that prevent people from making decisions without involving other people. I think these studies are particularly relevant for the study of decision making, and I would like to be provocative in pushing the argument forward. Is the groupindividual distinction still valid? How individualistic can an individual decision maker be in organizational settings? The use of external social resources is but a marginal part of the way individuals make decisions. Our cognition is social, our rationality takes advantage of this intertwined relation between social channels and decisions. Decision making in organizations is a social process. There are two major reasons. First, even though we have stated (more than once at this point) that individuals lean on external social resources and that they exploit social channels to make decisions, we haven’t clearly specified what makes the difference between an individual doing that as an organizational member or as an isolated individual. Is rational thinking within organizations different? Why? One of the most significant variables that makes the organizational setting different is the pertinence of the social element that affects each of the following:
– Social roles. Every member of the organization plays a role that depends on position, authority, and power.v This role is usually defined by social norms and values that are specific to the organization. There are two aspects that help to define individual social roles: leadership and hierarchy.vi The latter is the structure through which power is formally established in the organization, while the former relates to individual capabilities to be respected, authoritative, find support, affect other people’s thinking and behavior, share values, and lead. Sometimes the two stay together, sometimes they do not. – Culture. Organizations develop specific values and rules that can be partly written in codes and lean partly on behavior.vii This culture is related to the national culture, but it is also original and dependent on its members. Depending on organization size, subculturesviii may develop and define behavior and thinking for a limited group of people that, for example, work together or meet with some sort of continuity, or that establish specific and long-term social relations. – Goals. Organizations have goals. The usual explanation for the existence of organizations is that they exist to accomplish goals that cannot be achieved by a single individual’s effort. Another explanation is that organizations are more efficient than markets when it comes to acquiring specific resources (e.g., labor) and letting them work. Goals vary depending on their importance so that we may have a hierarchy (primary, secondary, tertiary, etc., or goals, sub-goals, sub-sub-goals, etc.,). They define everyday activities in organizations, and also help to understand the organization of labor and responsibilities in organizations. The same goal can be achieved through a multiple array of different strategies; how formal
140
10
Conclusions
these strategies are vary as well depending on organizational culture, social roles, and docility. Students of organizations add more than these factors. When they define organizations, they also add the environment, production process, and participants.ix However, my purpose is not to define an organization but to understand what makes it different for making decisions in that setting. Therefore, I am not disregarding the importance of organizational elements as they emerge from decades of studies, I am only trying to state their relative importance in relation to my arguments. This probably results in a clear-cut distinction or dismissal of some of these organizational elements. A short explanation of why I disregard these three factors may be helpful to get the point. I dismiss the first just because “having an environment” is not typical of an organization, since everything has an environment. Everything we can think of exists in a given environment. Moreover, according to the approach we have introduced here, a clear-cut distinction between the organization and its environment is not that relevant: The organization is what it is because of its environment, or because of processes of interaction with the environment.x The production process may well be something that happens at the individual level also, e.g., writing a report is a typical “individual” production process (of course, I should include resources—e.g., computer—that the individual uses), and therefore it does not define organizations only. The third element, participants, is somehow implicit in my arguments and cannot be regarded as typical of an organization: Unorganized groups of people have their participants too. I believe that the difference is in the social role (formal and informal) that each member has in the organization. If all of these conditions hold and if we integrate them with docility then, social relations become the basis for individual decision making in organizations. A group is defined by individuals that establish social relations on the basis of a formal and/or informal structure. Organizational groups include the three factors listed above and specific channels to facilitate the emergence of docility. Otherwise stated, decision making is a social phenomenon. The second reason why groups are fundamental for the study of decision making in organizations is that individuals think of consequences of their actions. A decision may be successful in terms of goals, but sometimes the decision maker prefers not to pursue it to its end simply because it can be harmful for the group. In other words, a decision may be particularly good for the single individual but harmful for the relations established in the group. This also includes communication issues, i.e., when a group member thinks she has to communicate to the upper levels but prefers not to do that in the interest of the group. I am not arguing that this is what always happens, I am just emphasizing that these are social phenomena that constitute part of usual organizational life. There are multiple loyalties in an organization, and these play a significant role in how individuals make decisions. Is it the R&D department that counts the most or the people you work with? Is it the organization as a whole or the department where you work? The closer the social relations, the less likely is an individual dissonance in the decision-making process.
A Methodological Note
141
As it may be apparent by now, this is another aspect of docility that we have called social responsibility in one of the previous chapters. More than on a logic of consequences, the individual behaves on the basis of his or her responsibility to the social environment (group) he or she partakes. As we know, this is a logic of appropriateness, together with one of adaptiveness. This is a short explanation of the reasons why I suggest that the study of individual decision making in organization should be intertwined with that of group decision making. We cannot understand the former without the latter, and vice versa.
A Methodological Note The study of organizational behavior has made dramatic improvements in the last few decades. One of the points that always troubled me was the almost null advancement on the idea of bounded rationality. The recent “cognitive revolution”xi shed a new light on the field and added insights that otherwise would have remained only marginal (e.g., shared cognition processes, sense-making approaches).xii When this happened, I started to think that the attribute “behavior” was close to its end in this field, and that it was about to be replaced by the term “cognition.” This didn’t happen. What happened was that the “old” OB division of the American Academy of Management had a spin-off and generated the Managerial and Organizational Cognition (or MOC) division. The old paradigm didn’t give up and generated a new (and smaller) group of people interested in cognition that probably believes this is the future of organizational studies. I am not interested in recriminations or in making a proposal for name changes (in fact, I am a member of both divisions!) but it is apparent that mainstream management failed to embrace the “new” that was coming. Why? Thomas Kuhn explains this better than me. With the advent of cognitive science in the organizational behavior field, we can ask if “behaviorism” is definitely dead or if it is still alive. While I believe it is in very good shape, I also hope that the field starts moving forward. The question is one of motivations that bring people to behave the way they do. It is not only a psychological approach that we need, it is a cognitive one: Do we still need/want to base our analysis on behavior and overlook the inner core of human mental processes? Isn’t thinking coming back of age? What is behavior without an explanation of how our cognitive processes work? More is needed. In the recent years a new and very promising field was born: neuroeconomics. Students in this field explore individual choices mapping their brain activities; it is an extension of what is the field of behavioral economics. It is, once again, opposed to mainstream economics and its “should” approach to science; that really looks like a product of the nineteenth century now. The term “behavioral” in economics stays for “new” and “evidence-based” as opposed to “old” and
142
10
Conclusions
“theory-based.” I suggest that this may be the same that should happen in the OB field with cognition and behavior. However, both approaches are related to management and organizational practices, and they complete and integrate each other more than they offer contrasting views of the individual (as the old and new economics). I hope this work helps in this shift toward the neurobehavioral approach to the field of organizational behavior and management. There is a second methodological point that I would like to mention, although very shortly. Scientific fields, as many of human activities, are not exempted from fads. I do not write this because I hope that the use of cognition is not such a thing but because I believe that the use of systems theory in organizational studies has been a fad. Systems theory emerged as a field soon after World War II and left two branches of studies to the scientific community: (a) systems theory and (b) cybernetics. The first is the study of interrelations between components that make every system work and of relations between the system and other systems (known as system-environment relations). The second is the study of control systems in animals and machines. While the second did not abandon the common scientific approach, systems theory presented a completely different and, at that time, new approach to science. The scientific approach and its methodology is reductionist. It consists in decomposition of a given phenomenon in its parts and analyzes these in isolation. The whole is a simple sum of component parts. For example, the study of a molecular structure is defined through the atoms that compose it, as well as a study of the emergence of crime in society can be the result of a study on individual behaviors, attitudes, education, and so on. Systems theory thinkers had another approach to science. They believed that the whole is not a result of the simple sum of its component; there is something that cannot be explained by this simple sum function. Individual parts are different from when they are together with the other parts of the system. This is apparent in organizations where, for example, culture emerges only in relation to its members or, for example, when the head of the finance department is defined through her team, coworkers, leadership, authority, and everything that is organization related. This is the reason why organizational scholars adopted a systems theory approach. To have an idea of how significant this approach was (maybe still is) for organizational studies, one can read Richard Scott’s classification.xiii My question, then, is if we agree that systems approach is relevant when it comes to studies of organizations, why does reductionism prevail? With no exceptions, studies on heuristics, biases, emotions, ethics, decision making, and cognition in organizations follow a perfect reductionist schema. This is a major weaknesses of experiments and surveys that provide the basis on which our theories of organizations are based. I believe that the distributed cognition approach is a contemporary attempt to let the system effect re-emerge in our studies. Docility is also a perspective that starts and ends in a system of cognitive resources; the docility effect is a simple way of taking systematic effects into consideration when studying decision making.
Extendable Rationality
143
Extendable Rationality What is really new with extendable rationality? Why am I suggesting moving from the bounded rationality approach? Hasn’t it been successful? This work followed a simple model to present extendable rationality. The first part of the book has been dedicated to the definition of bounded rationality and to the attempts to map it. The second part introduced the distributed cognition approach showing that bounds are not stable; humans can “stretch” (or extend) them to fit needs for rational decisions. This stretching has been defined through the logic of adaptiveness that explains “through doing,” change processes, and the social dimension of decisions. Love is a wonderful sentiment, although it is not appropriate for scientists. I am not in love with the idea of bounded rationality nor with that of extendable rationality. What I suggest is that the latter is more useful for a number of reasons. This entire book is about these reasons. Before getting into them once again, and for the last time in this book, I would like to offer a definition of this extendable rationality: Rationality is extendable when it adapts to the changing conditions that define available cognitive resources, providing workable solutions to a given problem.
The idea of rationality as extendable suits organization scholars’ needs because it extends and integrates the original idea of bounded rationality more than being opposed to it. I like to think of this as an extension of that original idea, a study on rational bounds and on how they work. At the very end, this theory of an extendable rationality continues the legacy of Herbert Simon and all other students of bounded rationality simply because it answers the same question: How do people actually make decisions? The answer that this book provides to the question is particularly related to the distributed cognition approach and to the docility hypothesis: 1. Instead of focusing on limitations, this theory puts emphasis on potential. It is not particularly important how bounded our rationality is than it is how we overcome these limitations to make decisions. Isn’t this what we do all the time when uncertainty grows high? Rationality is exactly about how we are capable of exploiting our limitations at their best, moving the boundaries of what was a limit for us. 2. The continuity with the bounded rationality theory stays in that we do have limits, but they are not stable and they cannot be defined as internal or external. Limits are both internal and external; a problem-solving activity is defined also on the basis of what resources the decision maker can exploit. I have pushed these bounds to a higher and systemic level, I should say: It is the system of internal and external resources that defines our rationality. And it is the interplay that moves these bounds. This is a theory that takes into consideration the fact that resources available together with individual efforts change the perspective of rationality and decision making. This aspect was missing in theories of bounded rationality.
144
10
Conclusions
3. Another significant difference from the theory of bounded rationality is that the decision maker does not get suboptimal results. The decision maker gets viable, feasible, or workable solutions to a given problem. This changes the perspective of the analysis. I am not interested in making a difference with the neoclassical theory; Simon had that problem. Neoclassical theory is not the benchmark for the extendable rationality theory because it is not for real decision makers! My purpose is to get an idea of rationality that is more related to a real decisionmaking process; there individuals get solutions (alternatives, options) that work. The questions I always had were, suboptimal to what? What is the optimum? If we have constraints, we can only get sub-something results. However, having a benchmark that we cannot define is like not having a benchmark at all. Why leave this benchmark then? I suggest that rational decision makers have the problem to make decisions that fit a particular situation, and that they don’t care about what is an optimum or the best. If they care, these are always relative concepts that are defined depending on the decision maker, the problem, the organization, and available resources. 4. This theory extends rationality to external resources and in particular to social resources. Docility emerges not because individuals are boundedly rational but because of its opposite: Individuals move their bounds and can be rational through the exploitation of social channels. All in all, this theory of extendable rationality considers the problem of bounds from the opposite side from what a theory of bounded rationality does. I do not state that human beings have no bounds, what I argue is that this is not the most important characteristics of their rationality. Individuals are rational because they can move their cognitive bounds and adapt through external resources from their environment. This enhances their fitness, especially when levels of docility grow at the individual and at the organizational levels. This is an approach that looks at the “bright” side of rationality. We do have limitations, but we also overcome these limitations. This distinct aspect makes us social human beings, not its opposite.
Notes i. What I mean here is that there are no inquiries on the nature of bounds. These studies show how these bounds affect individual behavior but the focus is on the outcome, rarely on the cognitive mechanisms underlying this process. There is a generalized lack of philosophical and epistemological background in studies of BR. At least, few are those that discuss or take into consideration what is the starting point of the theory. And, since scholars do not take a stand, one may assume that they share whatever was originally in the theory of BR. This means that everybody that writes on or “maps” bounded rationality is a logical positivist, as Simon clearly states in his book (1947, chapter 3, p. 55). Do they know? ii. Gigerenzer, Todd, and ABC Research Group (1999). The authors here present their idea of an ecological rationality. However, they are too shy and continue to use the old paradigm of bounded rationality that expands to the environment and becomes situational.
Notes
145
iii. Of course, one may argue that this was not the goal of the experiment as it was conducted. However, it is not the point of the book to falsify important results of previous experiment but to suggest that there might be implications overlooked by experimenters. It is quite the opposite then, that of highlighting that these experiments are far richer than one may expect. iv. For a review, see Kerr and Tindale (2004). v. Pfeffer (1992). vi. In this respect, see Martin (2007) and Clawson (2006). vii. This is explained in the first chapter of Scott (2003). viii. Schein (1990, 1996). ix. This is, once again, from Scott (2003). x. An application of recent developments on niche-construction (Odling-Smee et al., 2003) to organization theory may be helpful to support this argument. For an exploratory study of its application, see Hench and Secchi (2009). xi. Ilgen et al. (1994) and Hodgkinson and Healey (2008). xii. Weick and Roberts (1993), Weick (1995), Langan-Fox et al. (2001), Langfield-Smith and Wirth (1992), Laroche (1995), and Cannon-Bowers and Salas (2001). xiii. This is found in Scott (2003), Organizations. Rational, Natural, and Open Systems.
Afterword
The journey always arrives at destination. In this case, the destination is a starting point for a new journey. I hope you’ll enjoy it. It is unusual for an author, but I believe it is fundamental in this case. While I was writing this book I was thinking all the time of how alone I am in this travel. Besides Emanuele, I do not know of anybody else that is interested in studying a “new” approach to rationality. This is probably because we have spent our PhD years in European countries, or probably because there are better theories out there. In any case, write me (
[email protected]) if you want to do research with me, if you want just to exchange opinions, or if you simply found something interesting or something that you hated in this book. Thank you! Davide Secchi P.S. Am I being docile?
D. Secchi, Extendable Rationality, Organizational Change and Innovation, C Springer Science+Business Media, LLC 2011 DOI 10.1007/978-1-4419-7542-3,
147
References
Abelson, R.P. 1963. Computer simulation of ‘hot cognitions.’ In Computer simulation and personality: Frontier of psychological theory, eds. S. Tomkins, and S. Messick. New York: Wiley. Abrahamson, E., and L. Rosenkopf. 1990. When do bandwagon diffusions roll? How far do they go? And when do they roll backwards? A computer simulation. Academy of Management Best Paper Proceedings, 50:155–159. Abrahamson, E., and L. Rosenkopf. 1993. Institutional and competitive bandwagons: Using mathematical modeling as a tool to explore innovation diffusion. Academy of Management Review 18:487–517. Abrahamson, E., and L. Rosenkopf. 1997. Social network effects on the extent of innovation diffusion: A computer simulation. Organization Science 8(3):289–304. Adams, F., and K. Aizawa. 2008. The bounds of cognition. Malden, MA: Blackwell. Albrecht, K. 2006. Social intelligence. The new science of success. San Francisco, CA: JosseyBass. Allen, I.E., and J. Seaman. 2008. Staying the Course. Online Education in the U.S. The Sloan Consortium. http://sloanconsortium.org/sites/default/files/staying_the_course_2.pdf, Retrieved 3 October, 2010 10:48 AM. Anderson, M.H. 2006. How can we know what we think until we see what we said? A citation and citation context analysis of Karl Weick’s ‘the social psychology of organizing.’ Organization Studies 27(11):1675–1692. Andrade, E.B., and D. Ariely. 2009. The enduring impact of transient emotions on decision making. Organizational Behavior and Human Decision Processes 109(1):1–8. Argyris, C. 1986. Reinforcing organizational defensive routines: An unintended human resources activity. Human Resource Management 25(4):541–555. Ariely, D. 2008. Predictably irrational. The hidden forces that shape our decisions. New York, NY: Harper Collins. Ariely, D., and G. Loewenstein. 2006. The heat of the moment: The effect of sexual arousal on sexual decision making. Journal of Behavioral Decision Making 19(2):87–98. Ashkanasy, N.M., W.J. Zerbe, and C.E. Härtel. E. 2005. A bounded emotionality perspective on interpersonal behavior in organizations. In Emotions in organizational behavior, eds. C.E. Härtel, W.J. Zerbe, and N.M. Ashkanasy, 113–115. Mahwah, NJ: Lawrence Erlbaum Associates. Ashkanasy, N.M., and C.S. Daus. 2005. Rumors of the death of emotional intelligence in organizational behavior are vastly exaggerated. Journal of Organizational Behavior 26:441–452. Augier, M. 2000. Models of Herbert A. Simon. Perspectives on Science 8(4):407–443. Augier, M. 2004. March’ing towards ‘a behavioral theory of the firm’ James G. March and the early evolution of behavioral organization theory. Management Decision 42(10):1257–1268. Aumann, R.J. 1997. Rationality and bounded rationality. Games and Economic Behavior 21: 2–14.
149
150
References
Axelrod, R. 1976. The structure of decision: The cognitive maps of political elites. Princeton, NJ: Princeton University Press. Barclay, L.J., D.P. Skarlicki, and S.D. Pugh. 2005. Exploring the role of emotions in injustice perceptions and retaliation. Journal of Applied Psychology 90(4):629–643. Bardone, E. 2008. Human cognition as chance-seeking system. PhD dissertation submitted to the Department of Philosophy, University of Pavia. Unpublished. Bardone, E., and D. Secchi. 2006. Challenging bounded rationality. The cognitive role of external resources. Presented at the Academy of Management, Atlanta (USA), August 11–16. Bardone, E., and D. Secchi. 2009. Distributed cognition: A research agenda for management. In Current topics in management. Organizational behavior, performance, and effectiveness ed. M.A. Rahim, vol. 14, 183–207. New Brunswick, NJ: Transaction. Bazerman, M.H. 1994. Judgement in managerial decision making (3rd edition). New York, NY: Wiley. Beach, L.R. (ed.) 1998. Image theory. Theoretical and empirical foundations. Mahwah, NJ: Lawrence Erlbaum Associates. Bechara, A., and A. Damasio. 2005. The somatic marker hypothesis: A neural theory of economic decision. Games and Economic Behavior 52:336–372. Benartzi, S., and R.H. Thaler. 1999. Risk aversion or Myopia? Choices in repeated gambles and retirement investments. Management Science 45(3):364–381. Betsch, T. and S. Haberstroh. (eds.) 2005. The routines of decision making. Mahwah, NJ and London: Lawrence Erlbawm. Bilboa, I. 2010. Rational choice. Cambridge, MA: MIT Press. Block, W., N. Snow, and E. Stringham. 2008. Banks, insurance companies and discrimination. Business and Society Review 113(3):403–419. Bonaccio, S., and R.S. Dalal. 2006. Advice taking and decision-making: An integrative literature review, and implications for the organizational sciences. Organizational Behavior & Human Decision Processes 101(2):127–151. Bono, J.E., and M.A. Vey. 2005. Toward Understanding Emotional Management at Work: A Quantitative Review of Emotional Labor Research. In Emotions in organizational behavior, eds. C.E. Härtel, W.J. Zerbe, and N.M. Ashkanasy., 213–234. Mahwah, NJ: Lawrence Erlbaum Associates. Bornmann, L., and H.-D. Daniel. 2006. What do citation counts measure? Counts measure? A review of studies on citing behavior. Journal of Documentation 64(1):45–80. Camerer, C. 1998. Bounded rationality in individual decision making. Experimental Economics 1:163–183. Camerer, C. 2007. Neuroeconomics: Using neuroscience to make economic predictions. Economic Journal 117(March):C26–C42. Camerer, C., G. Loewenstein, and M. Weber. 1989. The curse of knowledge in economic settings: An experimental analysis. Journal of Political Economy 97:1232–1254. Cannon-Bowers, J.A., and E. Salas. 2001. Reflections on shared cognition. Journal of Organizational Behavior 22:195–202. Carnall, C.A. 1990. Managing change in organizations. New York, NY: Prentice-Hall. Case, D.O., and G.M. Higgins. 2000. How can we investigate citation behavior? A study of reasons for citing literature in communication. Journal of the American Society for Information Science 51(7):635–645. Chao, G.T., and H.W. Willaby. 2007. International employment discrimination and implicit social cognition: New directions for theory and research. Applied Psychology: An International Review 56(4):678–688. Chiang, Y.-S. 2007. Birds of moderately different feathers: Bandwagon dynamics and the threshold heterogeneity of network neighbors. Journal of Mathematical Sociology 31:47–69. Clark, A. 2003. Natural-born cyborgs. Minds, technologies, and the future of human intelligence. New York, NY: Oxford University Press. Clark, A. 2008. Supersizing the mind. Embodiment, action, and cognitive extension. New York, NY: Oxford University Press.
References
151
Clark, A., and D.J. Chalmers. 1998. The extended mind. Analysis 58(1):7–19. Clawson, J.G. 2006. Level three leadership. Getting below the surface (3rd edition). Upple Saddle River, NJ: Prentice Hall. Conlisk, J. 1996. Why bounded rationality? Journal of Economic Literature 34(2):669–700. Daft, R.L., R.H. Lengel, and L.K Trevino. 1987. Message equivocality, media selection, and manager performance: Implications for information systems. MIS Quarterly September:355–366. Damasio, A. 1994. Descartes’ error: Emotion, reason, and the human brain. New York, NY: Avon. Deephouse, D.L. 1996. Does isomorphism legitimate? Academy of Management Journal 39: 1024–1039. De George, R.T. 1999. Business ethics (5th edition). Upper Saddle River, NJ: Prentice Hall. Dennis, A.R., and S.T. Kinney. 1998. Testing media richness theory in the new media: The effects of cues, feedback, and task equivocality. Information Systems Research 9(3):256–274. Descartes, R. 1984/1641. The philosophical writings of Descartes (trans: Cottingham, J., Stoothoff, R., Murdoch, D.). Cambridge: Cambridge University Press. Dickson, P.R., and A.G. Sawyer. 1990. The price knowledge and search of supermarket shoppers. Journal of Marketing 54:42–53. Dreyer, T.P. 1993. Modeling with ordinary differential equations. Boca Raton, FL: CRC Press. Drolet, A.L., and M.W. Morris. 1995. Communication media and interpersonal trust in conflicts: The role of rapport and synchrony of nonverbal behavior. Paper presented and the Academy of Management meeting, Canada: Vancouver. Dunbar, R.I.M. 1998. The Social Brain Hypothesis. Evolutionary Anthropology 6(5):153–190. Dunbar, R.I.M., and S. Shultz. 2007. Evolution in the Social Brain. Science 317(5843):1344–1347. Earley, P.C. 2002. Redefining interactions across cultures and organizations: Moving forward with cultural intelligence. Research in Organizational Behavior 24:271–299. Earley, P.C., and E. Mosakowski. 2004. Toward culture intelligence: Turning cultural differences into a workplace advantage. Academy of Management Executive 18(3):151–157. El-Shinnawy, M., and M.L. Markus. 1997. The poverty of media richness theory: explaining people’s choice of electronic mail vs. voice mail. International Journal of Human-Computer Studies 46:443–467. Emmerling, R.J., and D. Goleman. 2005. Leading with Emotions. Leadership Excellence 22(7): 9–10. Eysenk, H.J., and L. Kamin. 1981. Intelligence: The battle for the mind. London: MacMillan. Festinger, L. 1957. A theory of cognitive dissonance. Stanford, CA: Stanford University Press. Fiol, C.M., and E.J. O’Connor. 2003. Waking up! Mindfulness in the face of bandwagon. Academy of Management Review 28(1):54–70. Foss, N.J. 2003a. Bounded rationality and tacit knowledge in the organizational capabilities approach: an assessment and a re-evaluation. Industrial and Corporate Change 12:185–201. Foss, N.J. 2003b. Bounded rationality in the economics of organizations: ‘much cited and little used’. Journal of Economic Psychology 24:245–264. Frank, R.H. 2004. What price the moral high ground? Ethical dilemmas in competitive environments. Princeton and Oxford: Princeton University Press Freeman, E.R., A.C. Wicks, and B. Parmar. 2004. Stakeholder theory and ‘the corporate objective revisited.’ Organization Science 15(3):364–369. Friedman, M. 1953. Essays in positive economics. Chicago, IL: University of Chicago Press. Gabbay, D.M., and J. Woods. 2007. Seductions and Shortcuts. Fallacies in the cognitive economy. Book in progress, http://www.dcs.kcl.ac.uk/staff/dg/SeductionsandShortcutscomplete.pdf, Retrieved January 12, 2009 9:36 AM. Gardner, P., and D. Berry. 1995. The effect of different forms of advice on the control of a simulated complex system. Applied Cognitive Psychology 9:S55–S79. Garriga, E., and D. Melé. 2004. Corporate social responsibility theories: Mapping the territory. Journal of Business Ethics 53:51–71. Gazzaniga, M.S. 2008. Human. The science behind what makes us unique. New York, NY: Harper Collins.
152
References
Gibbs, R.W. Jr. 2005. Embodiment and cognitive science. Cambridge: Cambridge University Press. Gigerenzer, G., H. Brighton. 2009. Homo heuristicus: Why biased minds make better inferences. Topics in Cognitive Science 1:107–143. Gigerenzer, G., and R. Selten. (eds.) 2001. Bounded rationality. The adaptive toolbox. Cambridge, MA: MIT Press. Gigerenzer, G., P.M Todd, and ABC Research Group 1999. Simple heuristics that make us smart. New York, NY: Oxford University Press. Gilboa, I. (2010). Rational choice. Cambridge, MA: MIT Press. Gladwell, M. 2002. The tipping point: How little things can make a big difference. New York, NY: Back Bay Books. Gladwell, M. 2008. Outliers. The story of success. New York: Little, Brown, and Co. Goff, P.A., J.L. Eberhardt, M.J. Williams, and M.C. Jackson. 2008. Not yet human: Implicit knowledge, historical dehumanization, and contemporary consequences. Journal of Personality and Social Psychology 94(2):292–306. Goldman, A., and K. Mason. 2007. Simulation. In Philosophy of psychology and cognitive science, ed. P. Thagard, 267–293. Amsterdam: Elsevier. Goldstein, D.G., G. Gigerenzer. 1996. Reasoning the fast and frugal way: Models of bounded rationality. Psychological Review 103(4):650–669. Goleman, D. 1995. Emotional intelligence. New York, NY: Bantam Books. Golinelli, G.M. 2005. L‘approccio sistemico al governo dell‘impresa, vol. I – L‘impresa sistema vitale. Padova: CEDAM. Granovetter, M. 1973. The strength of weak ties. American Journal of Sociology 78(6): 1360–1380. Granovetter, M. 1978. Threshold models of collective behavior. American Journal of Sociology 83(6):1420–1443. Granovetter, M. 1985. Economic action and social structure: The problem of embeddedness. American Journal of Sociology 91:481–510. Grove, A. 1999. Only the paranoid survive: How to exploit the crisis points that challenge every company. New York, NY: Broadway Books. Groves, R.M., J.F. Fowler, M.P. Couper, J.M. Lepkowski, E. Singer, and R. Tourangeau. 2009. Survey methodology (2nd edition). Hoboken, NJ: Wiley. Grüne-Yanoff, T. 2007. Bounded Rationality. Philosophy Compass 2/3:534–563. Guerrera, F. 2009. Welch denounces corporate obsessions. Financial Times, March 13. http://www.ft.com/cms/s/0/3ca8ec2e-0f70-11de-ba10-0000779fd2ac.html?nclick_check=1, Retrieved 8 May 2009 6:56 PM. Hanoch, Y. 2002. ‘Neither an angel nor an ant:’ Emotion as an aid to bounded rationality. Journal of Economic Psychology 23:1–25. Hanoch, Y., S. Wood, and T. Rice. 2007. Bounded rationality, emotions and older adult decision making: Not so fast and yet so frugal. Human Development 50:333–358. Hansen, H.H. 2002. The straw thing of fallacy theory: the standard definition of ‘fallacy.’ Argumentation 16(2):133–155. Harris, R. 2004. Integrationism, language, mind and world. Language Sciences 26(6): 727–739. Härtel, C.E., W.J. Zerbe, and N.M. Ashkanasy. 2005. Emotions in organizational behavior. Mahwah, NJ: Lawrence Erlbaum Associates. Harvey, N., and I. Fisher. 1997. Taking advice: Accepting help, improving judgment, and sharing responsibility. Organizational Behavior and Human Decision Processes 70:117–133. Hedlund, J., D.R. Ilgen, and J.R. Hollenbeck. 1998. Decision accuracy in computer-mediated versus face-to-face decision-making teams. Organizational Behavior and Human Decision Processes 76:30–47. Heine, S.J., and D.R. Lehman. 1997. Culture, dissonance, and self-affirmation. Personality and Social Psychology Bulletin 23:389–400. Hench, T., and D. Secchi. 2009. Organizational niche-construction and stakeholder analysis: Concepts and implications. Philosophy of Management 8(3):47–64.
References
153
Hodgkinson, G.P., and M.P. Healey. 2008. Cognition in organizations. Annual Review of Sociology 59:387–417. Holyoak, K.J., and P. Thagard. 1995. Mental leaps. Analogy in creative thought. Cambridge, MA: MIT Press. Hoshino-Browne, E., A.S. Zanna, S.J. Spencer, M.P. Zanna, S. Kitayama, and S. Lackenbauer, 2005. On the cultural guises of cognitive dissonance: The case of Easterners and Westerners. Journal of Personality and Social Psychology 89(3):294–310. Humphrey, N.K. 1976. The social function of intellect. In Growing points in ethology, eds. P.P.G. Bateson, and R.A. Hinde, 303–317. Cambridge, MA: Cambridge University Press. Hutchins, E. 2000. Distributed cognition. Working paper. Available online at: http://eclectic.ss.uci.edu/∼drwhite/Anthro179a/DistributedCognition.pdf. Accessed 12 Jan 2009 11:00 AM. Hutchins, E. 1995. Cognition in the wild. Cambridge, MA: MIT Press. Ilgen, D.R., D.A. Major, and L.T Spencer. (1994). The cognitive revolution in organizational behavior. In Organizational behavior: The state of the science, ed. J. Greenberg, 1–22. Mahwah, NJ: Lawrence Erlbaum Associates. Jones, G.R., and J.M. George. 2009. Contemporary management (5th edition). New York, NY: McGraw-Hill. Kahneman, D. 2003. A perspective of judgement and choice. Mapping bounded rationality. American Psychologist 58(9):697–721. Kahneman, D., and A. Tversky. 1979. Prospect theory: An analysis of decision under risk. Econometrica 47(2):263–292. Kahneman, D., J.L. Knetsch, and R.H. Thaler. 1990. Experimental tests of the endowment effect and the Coase theorem. Journal of Political Economy 98:1325–1348. Kahneman, D., J.L. Knetsch, and R.H. Thaler. 1991. The endowment effect, loss aversion, and status quo bias. Journal of Economic Perspectives 5(1):193–206. Katz, M.L., and C. Shapiro. 1985. Network externalities, competition, and compatibility. American Economic Review 75:424–440. Kavanaugh, A., D. Reese, J. Carroll, and M. Rosson. 2005. Weak ties in networked communities. Information Society 21(2):119–131. Kerr, N.L., and R.S. Tindale. 2004. Group performance and decision making. Annual Review of Psychology 55:623–655. Khalil, E.L. 2004. What is altruism? Journal of Economic Psychology 25:97–123. Klaes, M., and E.-M. Sent. 2005. A conceptual history of the emergence of bounded rationality. History of Political Economy 37(1):27–59. Knoke, D., and S. Yang. 2008. Social network analysis (2nd edition). Thousand Oaks, CA: Sage. Knudsen, T. 2003. Simon’s selection theory: Why docility evolves to breed successful altruism. Journal of Economic Psychology 24:229–244. Knudsen, T., and D.A. Levinthal. 2007. Two faces of search: Alternative generation and alternative evaluation. Organization Science 18(1):39–54. Kock, N. 2005. Media richness or media naturalness? The evolution of our biological communication apparatus and its influence on our behavior toward E-communication Tools. IEEE Transactions on Professional Communication 48(2):117–130. Krackhardt, D., and M. Kilduff. 2002. Structure, culture and Simmelian ties in entrepreneurial firms. Social Networks 24(3):279–290. Kunda, Z. 1999. Social cognition. Making sense of people. Cambridge, MA: MIT Press. Laland, K.N. 2001. Imitation, social learning and preparedness as mechanisms of bounded rationality. In Bounded rationality. The adaptive toolbox, eds. G. Gigerenzer, and R. Selten, 233–247. Cambridge, MA: MIT Press. Langan-Fox, J., A. Wirth, S. Codea, K. Langfield-Smith, and A. Wirth. 2001. Analyzing shared mental models. International Journal of Industrial Ergonomics 28:99–112. Langer, E.J. 1989. Minding matters: The consequences of mindlessness-mindfulness. In Advances in experimental social psychology, ed. L. Berkowitz, vol. 22: 137–173. San Diego, CA: Academic Press.
154
References
Langfield-Smith, K., and A. Wirth. (1992). Measuring differences between cognitive maps. Journal of the Operational Research Society 43(12):1135–1150. Langley, A., H. Mintzberg, P. Pitcher, E. Posada, and J. Saint-Macary. 1995. Opening up decision making: The view from the black stool. Organization Science 6(3):260–279. Laroche, H. 1995. From decision to action in organizations: Decision-making as a social representation. Organization Science 6(1):62–75. Lazarus, R.S. 1984. On the primacy of cognition. American Psychologist 39(2):124–129. Lazarus, R.S. 1991. Emotion and adaptation. New York, NY: Oxford University Press. Leibenstein, H. 1950. Bandwagon, snob, and Veblen effects in the theory of consumers’ demand. Quarterly Journal of Economics 64(2):183–207. Levinthal, D., and C. Rerup. 2006. Crossing an apparent chasm: Bridging mindful and less-mindful perspectives on organizational learning. Organization Science 17(4):502–513. Lewin, K. 1951. Field theory in social science. New York, NY: Harper & Row. Liu, Y., and P.L. Perrewé. 2005. Another look at the role of emotion in the organizational change: A process model. Human Resource Management Review 15(4):263–280. Locke, E.A. 2005. Why emotional intelligence is an invalid concept. Journal of Organizational Behavior 26: 425–431. Love, N. 2004. Cognition and the language myth. Language Sciences 26(6):525–544. Luce, R.D., and H. Raiffa. 1989. Games and decisions. Introduction and critical survey. Mineola, NY: Dover. Magnani, L. 2001. Abduction, reason, and science. Processes of discovery and explanation. New York, NY: Kluwer. Magnani, L. (ed.) 2006. Model-based reasoning in science and engineering. Cognitive science, epistemology, logic. London: College. Magnani, L. 2006. Mimetic Minds. Meaning Formation Through Epistemic Mediators and External Representations. In Artificial cognition systems, eds. A. Loula, R. Gudwin, and J. Queiroz. 327–357. Hershey, PA: Idea Group. Magnani, L. 2007. Morality in a technological world. Knowledge as a duty. Cambridge: Cambridge University Press. Magnani, L., and Nersessian, N.J. (eds.). 2002. Model-based reasoning. Scientific discovery, technological innovation, values. New York, NY: Kluwer/Plenum. Magnani, L., N.J. Nersessian, and P. Thagard. (eds.). 1999. Model-based reasoning in scientific discovery. New York, NY: Kluwer/Plenum. Magnani, L., D. Secchi, and E. Bardone. 2006. Docility and ‘through doing’ morality: An alternative approach to ethics. Working paper, 8, Department of Economics. Varese: University of Insubria. Maramatsu, R., and Y. Hanoch. 2005. Emotions as a mechanism for boundedly rational agents. The fast and frugal way. Journal of Economic Psychology 26:201–221. March, J.G. 1994. A primer on decision making. New York, NY: Free Press. March, J.G., and Z. Shapira. 1987. Managerial Perspectives on Risk and Risk Taking. Management Science 33(11):1404–1418. March, J.G., and H.A. Simon. 1958. Organizations. London: Wiley. Martin, R.L. 2007. The opposable mind: How successful leaders win through integrative thinking. Cambridge, MA: Harvard Business School Press. Mas-Colell, A., M.D. Whinston, and J.R. Green. 1995. Microeconomic theory. New York, NY: Oxford University Press. Mathis, R.L., and J.H. Jackson. 2008. Human resource management (12th edition). Mason, OH: Thomson South-Western. Maturana, H., and F.J. Varela. 1987. The three of knowledge: The biological roots of human understanding. Boston, MA: New Science Library. Meerschaert, M.M. 1999. Mathematical modeling (2nd edition). London: Academic. Merton, R.K. 1996. The Matthew effect, ii. In On social structure and science, ed. P. Sztompka. 318–336. Chicago, IL: Chicago University Press.
References
155
Meyer, J., and W.R. Scott (eds.). 1983. Organizational environments: Ritual and rationality. Beverly Hills, CA: Sage. Michel, A.A. 2007. A distributed cognition perspective on newcomers’ change processes: The management of cognitive uncertainty in two investment banks. Administrative Science Quarterly 52(4):507–557. Milkovich, G.T., and J.M. Newman. 2008. Compensation (9th edition). New York, NY: McGrawHill. Mithen, S. 1999. Handaxes and ice age carvings: hard evidence for the evolution of consciousness. In Toward a science of consciousness iii. The third Tucson discussions and debates, eds. A.R. Hameroff, A.W. Kaszniak, and D.J. Chalmers. 281–296. Cambridge, MA: MIT Press Moore, D., T.R. Kurtzberg, L.L. Thompson, and M.W. Morris. 1999. Long and short routes to success in electronically mediated negotiations: Group affiliations and good vibrations. Organizational Behavior & Human Decision Processes 77(1):22–43. Mousavi, S., and J. Garrison. 1992. Toward a transactional theory of decision making: creative rationality as functional coordination in context. Journal of Economic Methodology 10(2): 131–156. Mumby, D.K., and L.L. Putnam. 1992. The politics of emotion: A feminist reading of bounded rationality. Academy of Management Review 17(3):465–486. Munier, B., R. Selten, D. Bouyssou, P. Bourgine, R. Day, N. Harvey, D. Hilton, M.J. Machina, P. Parker, J. Sterman, E. Weber, B. Wernerfelt, and R. Wensley. 1999. Bounded rationality modelind. Marketing Letters 10(3):233–248. Nagel, E., and J. Newman. 1958. Gödel’s proof. New York, NY: NYU Press. Neale, M.A., and M.H. Bazerman. 1991. Cognition and rationality in negotiation. New York, NY: Free Press. Nelson, R.R., and S.G. Winter. 1973. Toward an evolutionary theory of economic capabilities. American Economic Review 63:440–449. Nelson, R. and S. Winter. 1982. An evolutionary theory of economic change. Cambridge, MA: Belknap Press. Newell, A. and H.A. Simon. 1972. Human problem solving. Englewood Cliffs, NJ: Prentice-Hall. Newton, K. 1997. Social capital and democracy. American Behavioral Scientist 40(5): 575–586. Nickerson, R.S. 1999. How we know—and sometimes misjudge—what others know: Imputing one’s own knowledge to others. Psychological Bulletin 125:737–759. Norman, D. 1993. Things that make us smart. New York, NY: Addison Wesley. Odling-Smee, F.J., K.N. Laland, and M.W. Feldman. 2003. Niche construction: The neglected process in evolution. Cambridge, MA: Princeton University Press. Oh, H., G. Labianca, and M.-H. Chung. 2006. A multilevel model of group social capital. Academy of Management Review 31(3):569–582. Olive, D. 2001. The man who taught an elephant to dance: General Electric’s Jack Welch was one of the first celebrity CEOs. National Post. Don Mills, Ont.: Sep 7, p. C16. Pahor, M., M. Škerlavaj, V. Dimovski. 2008. Evidence for the network perspective on organizational learning. Journal of the American Society for Information Science & Technology 59(12):1985–1994. Panksepp, J. 1982. Toward a general psychobiological theory of emotions. The Behavioral and Brain Sciences 5:407–467. Panksepp, J. 1992. A critical role for ‘affective neuroscience’ in resolving what is basic about basic emotions. Psychological Review 99:554–560. Panskepp, J. 2007. Criteria for basic emotions: Is DISGUST a primary ‘‘emotion’’? Cognition and Emotion 21(8):1819–1828. Patokorpi, E. 2008. Simon’s paradox: Bounded rationality and the computer metaphor of the mind. Human Systems Management 27:285–294. Payne, J.W., J.R. Bettman, and E.J. Johnson. 1993. The adaptive decision maker. New York, NY: Cambridge University Press.
156
References
Peng, W.-S. 1992. A critique on H.A. Simon’s administrative behavior theory. Public Administration Quarterly 16(2):254–264. Pennington, D.C. 2000. Social cognition. London: Routledge. Perry, M. 2003. Distributed cognition. In Hci, models, theories, and frameworks: Toward a Multidisciplinary Science ed. J.M. Carroll. 193–224. London: Morgan Kaufman. Pierce, C.S. 1955. Abduction and induction. In Philosophical writings of pierce, eds. C.S. Peirce J. Buchler. 150–156. New York, NY: Dover. Pfeffer, J. 1992. Managing with power. Politics and influence in organizations. Cambridge, MA: Harvard Business School Press. Plato, 2005. Phaedrus (trans: Rowe, C.). London: Penguin Books. Polanyi, M. (1966). The tacit dimension. London: Routledge. Politser, P. 2008. Neuroeconomics. A guide to the new science of making choices. New York, NY: Oxford University Press. Popper, K. 1935/2002. The logic of scientific discovery. New York, NY: Routledge. Post, T., M.J. van den Assem, G. Baltussen, and R.H. Thaler. 2008. Deal or No Deal? Decision Making under Risk in a Large-Payoff Game Show. American Economic Review 98(1): 38–71. Prigogine, I., and I. Stengers. 1986. La nouvelle alliance. Paris: Gallimard. Purdy, J. M., P. Nye, and P.V. Balakrishnan. 2000. The impact of communication media of negotiation outcomes. International Journal of Conflict Management 11(2):162–187. Rabin, M., and R.H. Thaler. 2001. Risk Aversion. Journal of Economic Perspectives 15(1): 219–232. Reid, T.R. 2004. The United States of Europe. The new superpower and the end of American supremacy. New York, NY: Penguin Books. Richardson, K. 2000. The making of intelligence. New York, NY: Columbia University Press. Richerson, P.J., and R. Boyd. 1998. The evolution of human ultra-sociality. In Indoctrinability, ideology, and warfare: Evolutionary perspective, eds. I. Eibl-Eibesfeldt, and F.K. Salter, 71–95. New York: Berhahn Books. Richerson, P.J., and R. Boyd. 2005. Not by genes alone: How culture transformed human evolution. Chicago, IL: Chicago University Press. Rothschild, K.W. 2001. The reluctant rebel or glamor and poverty of the homo oeconomicus. Kyklos 54(2/3):445–452. Rubinstein, A. 1998. Modeling bounded rationality. Cambridge, MA: MIT Press. Schein, E.H. 1990. Organizational culture. American Psychologist 45(2):109–119. Schein, E.H. 1996. Culture: The missing concept in organization studies. Administrative Science Quarterly 41:229–240. Schein, E.H. 1999. Kurt Lewin’s change theory in the field and in the classroom: Notes toward a model of managed learning. Reflections 1(1):59–72. Schein, E.H. 2002. Models and tools for stability and change in human systems. Reflections 4(2):34–46. Scott, W.R. 1995. Institutions and organizations. Thousand Oaks, CA: Sage. Scott, W.R. 2003. Organizations. Rational, natural, and open systems (5th edition). Englewood Cliffs, NJ: Prentice-Hall. Secchi, D. 2007. A theory of docile society: The role of altruism in human behavior. Journal of the Academy of Business and Economics 7(2):146–160. Secchi, D. 2007. Utilitarian, managerial, and relational theories of corporate social responsibility. International Journal of Management Reviews 9(4):347–373. Secchi, D. 2010. Banks and Discrimination: How Far Can We Go with Competition? A Reply to Block, Snow, and Stringham. Submitted to Business and Society Review, March 2009. Secchi, D. 2009. The Cognitive Side of Social Responsibility. Journal of Business Ethics 88(3):565–581. Secchi, D., and E. Bardone. 2009a. An Organizational Model of Bandwagon. Working Paper No. 1, College of Business Administration, University of Wisconsin–La Crosse.
References
157
Secchi, D., and E. Bardone. 2009b. Super-docility in organizations. International Journal of Organization Theory and Behavior 12(3):339–379. Selten, R. 1998. Features of experimentally observed bounded rationality. European Economic Review 42:413–436. Semler, R. 1994. Why my former employees still work for me. Harvard Business Review January– February:64–74. Sent, E.-M. 1997. Sargent versus Simon: Bounded rationality unbound. Cambridge Journal of Economics 21:323–338. Sent, E.-M. 2005. Simplifying Herbert Simon. History of Political Economy 37(2):227–332. Shakun, M.F. 2001. Unbounded Rationality. Group Decision and Negotiation 10:97–118. Shampanier, K., N. Mazar, and D. Ariely. 2007. Zero as a special price: The true value of free products. Marketing Science 26(6):742–757. Shier, D.R., and K.T. Wallenius. 2000. Applied mathematical modeling. A multidisciplinary approach. Boca Raton, FL: Chapman & Hall/CRC. Silver, W.S., and T.R. Mitchell. 1990. The status quo tendency in decision making. Organizational Dynamics 18(4):34–46. Simon, H. 1955. A behavioral model of rational choice. Quarterly Journal of Economics LXIX: 99–118. Simon, H.A. 1999. Interview with Herbert A. Simon. Working paper, CISTEMA, Copenhagen Business School. Simon, H.A. 1947/1997. Administrative behavior (4th edition). New York, NY: The Free Press. Simon, H.A. 1957. Models of man. New York, NY: Wiley. Simon, H.A. 1959. Theories of decision-making in Economics and behavioral science. American Economic Review 49(3):253–283. Simon, H.A. 1960. The new science of management decision. New York, NY: Harper & Row. Simon, H.A. 1978. Rationality as process and a product of thought. American Economic Review 68:1–14. Simon, H.A. 1990. A mechanism for social selection and successful altruism. Science 250(4988):1665–1668. Simon, H.A. 1993b. Altruism and economics. American Economic Review 83(2):156–161. Simon, H.A. 1993a. The human mind: The symbolic level. Proceedings of the American Philosophic Society 137(4):638–647. Simon, H.A. 1979. Rational decision making in business organizations. American Economic Review 69:493–513. Slaugher, J.E., and S. Highhouse. 2003. Does matching up features mess up job choice? Boundary conditions on attribute-salience effects. Journal of Behavioral Decision Making 16: 1–15. Small, H. 2004. On the shoulders of Robert Merton: Towards a normative theory of citation. Scientometrics 60(1):71–79. Sniezek, J.A., G.E. Schrah, and R.S. Dalal. 2004. Improving judgment with prepaid expert advice. Journal of Behavioral Decision Making 17:173–190. Starbuck, W.H. 1983. Organizations as action generators. American Sociological Review 48(1): 91–102. Staw, B.M., and L.D. Epstein. 2000. What bandwagons bring: Effects of popular management techniques on corporate performance, reputation, and CEO pay. Administrative Science Quarterly 45:523–556. Sternberg, R.J. (ed.). 1999. Handbook of creativity. Cambridge: Cambridge University Press. Sundaram, A.K., and A.C. Inkpen. 2004. The corporate objective revisited. Organization Science 15(3):350–363. Sunstein, C.R. 2005. Conformity and dissent. Public Law and Legal Theory Working Paper No. 34, University of Chicago. Sutton, J. 2004. Representation, levels and context in integrational linguistics and distributed cognition. Language Sciences 26(6):503–524.
158
References
Thagard, P. 2007. Introduction to the Philosophy of Psychology and Cognitive Science. In Philosophy of psychology and cognitive science, ed. P. Thagard. ix–xvii. Amsterdam: Elsevier. Thagard, P. (ed.) 2007. Philosophy of psychology and cognitive science. Amsterdam: Elsevier. Thaler, R.H., and C.R. Sunstein. 2008. Nudge. Improving decisions about health, wealth, and happiness. New Haven & London: Yale University Press. Thomas, D.C., and K. Inkson. 2003. Cultural intelligence. San Francisco, CA: Berrett-Kohler. Todd, P.M. 2001. Fast and frugal heuristics for environmentally bounded minds. In Bounded rationality. The adaptive toolbox, eds. G. Gigerenzer, and R. Selten, 51–70. Cambridge, MA: MIT Press. Todd, P.M., and G. Gigerenzer. 2003. Bounding rationality to the world. Journal of Economic Psychology 24:143–165. Tversky, A., and D. Kahneman. 1974. Judgment under Uncertainty: Heuristics and Biases. Science 185:1124–1130. Van der Veer Martens, B., and A.A. Goodrum 2006. The diffusion of theories: A functional approach. Journal of the American Society for Information Science and Technology 57(3): 330–341. Verplanken, B., V. Myrbakk, and E. Rudi. 2005. The measurement of habit. In The routines of decision making, eds. T. Betsch and S. Haberstroh, 231–247. Mahwah, NJ & London: Lawrence Erlbawm. Vicari, S. 1991. L’impresa vivente. Itinerario in una diversa concezione. Milano: Etas. Von Foerster, H. 2002. Understanding systems. Conversations on epistemology and ethics. New York, NY: Kluwer/Plenum. Von Neumann, J., and O. Morgenstern. 1944. Theory of games and economic behavior. Cambridge, MA: Princeton University Press. Wallister, B. 2008. Cognitive economics. Berlin & Heidelberg: Springer. Walsh, J.P. 1995. Managerial and organizational cognition: Notes from a trip down memory lane. Organization Science 6(3):280–321. Weber, E.U., and E.J. Johnson. 2009. Mindful judgment and decision making. Annual Review of Psychology 60:53–85. Weick, K. 1995. Sensemaking in organizations. Thousand Oaks, CA: Sage. Weick, K., and K.H. Roberts. 1993. Collective mind and organizational reliability: the case of flight operations on an aircraft carrier deck. Administrative Science Quarterly 38:357–381. Weiss, H.M., K. Suckow, and R. Cropanzano. 1999. Effects of justice conditions on discrete emotions. Journal of Applied Psychology 84(5):786–794. Werhane, P.H. 1999. Moral imagination and management decision making. New York, NY: Oxford University Press. Wilson, R.A. 1994. Wide computationalism. Mind 103:351–372. Wilson, R.A. 2004. Boundaries of the mind. Cambridge: Cambridge University Press. Wilson, R.A. 2005. Collective memory, group minds, and the extended mind thesis. Cognitive Processing 6(4):227–236. Woods, J. 2004. The death of argument. Dordrecht: Kluwer. Yaniv, I., and E. Kleinberger. 2000. Advice taking in decision making: egocentric discounting and reputation formation. Organizational Behavior and Human Decision Processes 83:260–281. Zafirovski, M. 1999. The ‘unbearable lightness’ of economic approach to economic behavior in the social setting: Rational action and the sociology of the economy. Journal for the Theory of Social Behavior 29(3):301–334. Zajonc, R.B. 1984. On the Primacy of Affect. American Psychologist 39(2):117–123. Zanna, M.P., and J. Cooper. 1974. Dissonance and the pill: An attributional approach to studying the arousal properties of dissonance. Journal of Personality and Social Psychology 29:703–709. Zelený, M. 2001. Obituary. Herbert A. Simon (1916–2001). Human Systems Management 20:3–4. Zhang, J. 1997. The nature of external representations in problem-solving. Cognitive Science 21(2):179–217.
Index
A Abduction, 2, 13, 18, 51, 78 Accessibility, 5, 48–52, 75–76 Advice active, 98, 122, 127 and bandwagon, 82 giving, 2, 82, 100, 102, 104, 107–109, 111, 114, 127–129 and information mediums, 97 and information richness scale, 102 judge-advisor system (JAS), 99–101, 103–104 passive, 108–110 taking, 98–111, 113, 115, 119, 128–129, 133, 137 Altruism, 27, 83, 95, 115–116, 131–133, 137 Ariely, Dan, 17, 38, 79, 94
B Bandwagon and docility, 126, 132 and organizational culture, 124–125 and social relations, 125 Bandwagon effect, 34–35, 39, 50, 119, 123, 132 Bardone, Emanuele, 39, 65, 78, 95, 111, 115, 118, 131–132 Behavioral research, 54 studies, 27, 54, 79 Behaviorism, 141 Bias anchor bias, 33–34 endowment effect, 31–32 and errors, 36–37 and prejudices, 31, 35–36
status quo, 32–33 Biases paradigm, 54–55 Boundedly rationality, 20, 24, 37, 52, 63, 77, 83, 114, 144 Bounded rationality, 1, 3–6, 14, 19–25, 27–39, 41–59, 63–64, 68, 73–74, 76–78, 81–83, 85, 93, 95, 113–114, 129, 137–138, 141, 143–144 Brain, 4–6, 42–43, 46–48, 55, 63–79, 81–95, 97–111, 113–133, 135–145 isolated, 81 C Chalmers, David J., 78 Change, 1–4, 11, 24–25, 28, 32–34, 39, 42, 45, 52–54, 56, 58, 66–69, 71, 76, 82, 84, 86–94, 97, 100, 117–118, 120, 122, 124, 136–138, 141, 143–144 Clark, Andy, 73, 78–79, 90, 95 Cognition cognitive process, 74 distributed (DC), 1, 6, 64–78, 81, 84–89, 91–92, 97–98, 103–105, 110–111, 113, 115, 117–119, 123, 129, 132, 135, 138, 142–143 Cognitive divide, 6, 72, 76 Community, 1, 45, 74, 114, 120–126, 132–133, 142 Computation, 16, 22 Computational, 13–14, 19, 22, 31, 54–55, 57, 64, 76, 135 limits, 31, 54–55 Context, 53, 63, 71, 73, 82, 84, 88–89, 92, 94, 97, 104, 117–118 Cooperation, 27, 130–131, 133, 137 Culture, 45, 71, 79, 83–84, 110, 120, 124–126, 130, 132, 139–140, 142
159
160 D Damasio, Antonio, 57–58, 94–95 Decision making, 1–6, 9–21, 23–24, 29, 31, 35–36, 39, 41–43, 46–47, 49, 51–53, 58, 63–64, 69–70, 72–77, 79, 81–89, 92, 94–95, 97–98, 101, 103, 105–106, 110, 113–117, 119–121, 129–130, 135–136, 138–143 Deduction, 13 Descartes, René, 17, 47, 57 Docile exceptionally docile, 121–123 non-docile, 114–117, 121–123, 129, 133 ordinarily docile, 121–123 organization, 113–133 Docility and altruism, 115–116, 131–133 and bandwagon, 123–126 community, 120 and cooperation, 130–131, 133 degrees of, 117 and environment, 117–118 in organizations, 121–123 prerequisites, 119–121 public availability of information, 121 and social channels, 114, 117, 119–123, 126–130, 133 social responsibility, 126–128 and standards, 120 and time, 118 Docility effect, 118, 129, 131, 142 E Ecological rationality, 55, 144 Emotions and bounded rationality, 46–48 and distributed cognition, 71, 84 and through doing logic, 91 Environment, 13, 24, 35, 39, 44, 52, 55, 64–65, 73–77, 79, 81, 85, 87, 89, 92, 97, 107, 114–120, 124–125, 136–138, 140–142, 144 Error central tendency, 37 extreme tendency, 37 first impression, 37 halo, 36–37 horn, 37 leniency, 37 recency, 37 severity, 37 spillover, 37 Ethics, 27, 85–87, 113, 133, 142
Index Externalization, 70–73, 75, 83–84, 87, 90 External resources artifacts, 4, 81, 136 non-social, 88, 127 social, 87–88, 98, 114, 117–119, 121, 127, 139 social channels, 127, 144 F Fallacy ad hominem, 51 ad populum, 50 ad verecundiam, 50 composition and division, 51 gang of eighteen, 49–50 hasty generalization, 51 Foss, Nicholai J., 1, 18, 25, 77, 81, 93 Framing, 5, 35, 48–51, 75–76, 87, 103 Fully rational, 19–21, 25, 63, 137 G Gabbay, Dov M., 49, 58, 78 Gigerenzer, Gerd, 25, 42, 54–57, 59, 77–78, 144 Granovetter, Mark, 39, 108, 111, 123, 132 H Hanoch, Yaniv, 47, 58, 77, 94–95 Heuristics and biases, 3, 5 and fallacies, 42 fast and frugal (FFH), 42–44, 48, 57–58, 83 High-technology, 89–91, 97 Homo economicus, 25, 64 Hutchins, Edwin, 1, 64, 73, 78–79, 119, 132 I Induction, 12–13, 18, 51 Information discount, 109 limits, 55 and medium, 101–103 richness scale, 102 Innovation, 2, 18, 41, 49, 84, 89–90, 93, 110 technology, 93 Intelligence, 14, 24, 42, 57, 64, 84–85, 94–95 Interplay, 66, 70, 75, 79, 82, 88, 92–93, 104, 135–137, 143 In vacuum, 46, 52, 77, 104 K Kahneman, Daniel, 17, 25, 27, 38–39, 54, 58, 77–79, 111 Knudsen, Thorbjørn, 77, 114
Index L Laland, Kevin, 39, 109, 111 Leibniz, Gottfried W., 46–47 Less-is-more, 55 Logic of adaptiveness and change, 89 and context, 88 and fitness, 93 Logic of appropriateness, 52–53, 56, 72, 87, 141 Logic of consequences, 52, 56, 87, 141 M Magnani, Lorenzo, 18, 78–79, 85–86, 95, 133 Mapping, 45, 56, 83, 137–138, 141 Maps, 27–39, 41–59, 78, 137, 144 March, James G., 18, 24–25, 51–52, 59, 77–78, 94 Medium(s), 33, 67–68, 77, 97–98, 101–107, 110, 117, 119, 123 Mind, 4, 12, 28–29, 34, 37, 39, 43, 46, 54, 57, 64–71, 73, 75, 77–79, 84, 86, 92, 98, 100, 109, 111, 116, 121, 125–127, 133 extended, 73–74, 77 Morality and decision making, 86 and distributed cognition, 85–87 and through-doing, 86–87 Morgenstern, Oscar, 2, 18, 38, 78, 93 N Neoclassical economics, 5, 51, 54, 76, 116 model, 14, 21 Newell, Alan, 46, 57, 95 O Organization decision-making, 135 docile, 113–133 Organizational behavior, 1, 3, 14, 39, 45, 56–58, 76–77, 85, 109–110, 123, 125–126, 132, 141–142 change, 92, 94 culture, 79, 124–125, 130, 132, 140 decision making, 69, 73, 138 routine, 10 P Peirce, Charles S., 13, 18 Plasticity, 67, 75, 78, 81–82, 138 Problem solving, 46, 74–75, 82, 143 Prospect theory certainty effect, 27–31, 38
161 possibility violation, 29, 31, 38 reflection effect, 30, 38 R Rationality adaptive, 86 bounded, 1, 3–4, 19–25, 27–39, 41–59, 63–64, 68, 73–74, 76–78, 81–83, 85, 93, 95, 114, 129, 137–138, 143–144 ecological, 55, 144 full, 20, 25, 63, 77, 93, 97, 137 limited, 23–25 neoclassical model of, 14 procedural, 20–22, 24, 41 substantive, 5, 21, 41 Representation, 4–5, 48–52, 56, 66, 72, 75–77, 79, 103, 127 Re-projecting, 136 Richardson, Ken, 24, 64, 78 Routine, 10, 18, 33, 70–71, 110, 121 S Scott, W. Richard, 57, 142, 145 Simon, Herbert A., 1, 3–4, 6, 13–19, 21, 23–25, 46, 52, 54–55, 57, 59, 63–79, 81, 93, 95, 113–114, 116, 118, 129, 131–132, 135, 137, 143–144 Smart interplay, 66 Social responsibility, 126–128, 131, 133, 141 Subjective expected utility (SEU), 2, 3, 63 Sunstein, Cass R., 38–39, 110–111 T Thagard, Paul, 57, 78–79 Thaler, Richard H., 38–39, 108, 110–111 Through doing logic, 68–70, 87, 91 rationality, 78 Time, 3, 9, 13–15, 21–22, 24–25, 32, 34–35, 42–43, 45–46, 48–51, 55–58, 64, 66–72, 77, 82, 84, 86, 88, 90, 98, 100–102, 107–108, 114–118, 130, 133, 136, 142–143 Todd, Peter M., 56–57, 77–78, 144 Toolbox paradigm, 54–56 Tversky, Amos, 25, 27, 38–39, 54, 77 V Von Neumann, John, 2, 18 W Wilson, Robert A., 73, 78–79 Woods, John, 49, 57–58, 78
