Evaluative Morphology from a Cross-Linguistic Perspective

Evaluative Morphology from a Cross-Linguistic Perspective Evaluative Morphology from a Cross-Linguistic Perspective By...

Author: Lívia Körtvélyessy

227 downloads 1559 Views 3MB Size Report

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

DOWNLOAD PDF

Evaluative Morphology from a Cross-Linguistic Perspective

Evaluative Morphology from a Cross-Linguistic Perspective By

Lívia Körtvélyessy

Evaluative Morphology from a Cross-Linguistic Perspective By Lívia Körtvélyessy This book first published 2015 Cambridge Scholars Publishing Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2015 by Lívia Körtvélyessy This monograph has been written with the financial support from the VEGA grant project 1/0094/12. All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-7160-5 ISBN (13): 978-1-4438-7160-0

To my parents

CONTENTS

List of Figures, Maps, Tables and Charts ................................................... ix Acknowledgments .................................................................................... xiv Abbreviations ......................................................................................... xviii Introduction ................................................................................................. 1 Chapter One ................................................................................................. 5 A Theory of Evaluative Morphology 1.1 General ............................................................................................. 5 1.2 Standard Average European............................................................. 6 1.3 Evaluative Morphology ................................................................. 20 Chapter Two .............................................................................................. 48 Research 2.1 Objectives ...................................................................................... 48 2.2 Method of research ........................................................................ 48 2.3 Evaluative Morphology Saturation ................................................ 55 2.4 Results............................................................................................ 55 2.5 Conclusions .................................................................................. 137 Chapter Three .......................................................................................... 144 Phonetic Iconicity in Evaluative Morphology 3.1 General ......................................................................................... 144 3.2 Iconicity ....................................................................................... 144 3.3 Phonetic iconicity, sound symbolism, phonetic symbolism......... 147 3.4 Previous research into phonetic iconicity .................................... 150 3.5 Research ....................................................................................... 153 Conclusions ............................................................................................. 220

viii

Contents

Appendix A ............................................................................................. 223 Appendix B.............................................................................................. 228 Bibliography ............................................................................................ 231 Index ........................................................................................................ 251

LIST OF FIGURES, MAPS, TABLES AND CHARTS

Figures Figure 1-1 The place of Evaluative Morphology according to Scalise (1984) ................................................................................................... 27 Figure 1-2 An overview of categories according to Bauer (2004)............. 29 Figure 1-3 Jurafsky’s synchronic-diachronic model of the semantics of diminutives according to Jurafsky ................................................... 33 Figure 1-4 Prieto’s radial model for diminutives....................................... 37 Figure 1-5 Prieto’s radial model for augmentatives .................................. 38 Figure 1-6 Grandi’s semantic primitives for EM ....................................... 38 Figure 1-7 Model of evaluative word formation........................................ 45 Figure 1-8 Radial model of EM semantics ................................................ 47 Figure 3-1 Drawing for Task 1 ................................................................ 155 Figure 3-2 Drawing for Task 2 ................................................................ 157 Figure 3-3 Drawings for Task 3............................................................... 158 Figure 3-4 Drawing for Task 4 ................................................................ 160 Figure 3-5 Drawing for Task 4 ................................................................ 160

Maps Map 1-1 Map of Europe .............................................................................. 9 Map 1-2 Languages spoken in Europe ...................................................... 10 Map 2-1 Languages of Africa .................................................................... 56 Map 2-2 Languages of Australia-New Guinea .......................................... 63 Map 2-3 Languages of North America ...................................................... 68 Map 2-4 Languages of South America ...................................................... 75 Map 2-5 Languages of SE Asia + Oceania ................................................ 80 Map 2-6 Languages of Eurasia .................................................................. 88 Map 2-7 Languages of SAE ......................................................... Centrefold Map 2-8 SAE language saturation ................................................ Centrefold

x

List of Figures, Maps, Tables and Charts

Tables Table 1-1 SAE according to Haspelmath (1998) ....................................... 13 Table 1-2 SAE according to Décsy (2000) ................................................ 13 Table 1-3 SAE according to Ramat (2000) ............................................... 14 Table 1-4 SAE according to van der Auwera (1998b) .............................. 14 Table 1-5 Various approaches to central SAE languages .......................... 15 Table 1-6 Different perspectives of evaluative morphology ..................... 39 Table 2-1 Languages of Africa–EM saturation ......................................... 56 Table 2-2 Languages of Australia-New Guinea–EM saturation ................ 64 Table 2-3 Languages of North America - saturation ................................. 68 Table 2-4 Languages of South America–saturation................................... 76 Table 2-5 Languages of SE Asia + Oceania–saturation ............................ 81 Table 2-6 Languages of Eurasia–saturation............................................... 89 Table 2-7 Overview of evaluative morphology by areas ........................... 94 Table 2-8 Ranking of the areas by the data of Table 2-7 ........................... 95 Table 2-9 EM, DIM and AUG saturation values by areas ......................... 95 Table 2-10 Ranking of the areas by the data of Table 2-9 ......................... 96 Table 2-11 Ranking of the sample languages by EM saturation ............... 97 Table 2-12 Ranking of the sample languages by DIM saturation.............. 98 Table 2-13 Ranking of the sample languages by AUG saturation ............. 99 Table 2-14 Word-formation processes in diminutive formation ............. 100 Table 2-15 Word-formation processes in augmentative formation ......... 101 Table 2-16 Cognitive categories expressed by diminutives..................... 102 Table 2-17 Cognitive categories expressed by augmentatives ................ 102 Table 2-18 Word classes by areas ........................................................... 103 Table 2-19 Languages of SAE................................................................. 104 Table 2-20 EM saturation values by languages ....................................... 105 Table 2-21 Evaluative morphology in individual language families ....... 120 Table 2-22 The ranking of the sample languages according to their total EM saturation............................................................................. 121 Table 2-23 The ranking of the sample languages according to their DIM saturation ................................................................................... 122 Table 2-24 The ranking of the sample languages according to their AUG saturation .................................................................................. 123 Table 2-25 Evaluative morphology: SAE vs. World ............................... 127 Table 2-26 Ranking of the areas by the data of Table 2-10 ..................... 128 Table 2-27 EM, DIM and AUG saturation values by macro-areas.......... 129 Table 2-28 Ranking of the areas by the data of Table 2-27 ..................... 130 Table 2-29 Word-formation processes in diminutive formation ............. 131 Table 2-30 Word-formation processes in augmentative formation ......... 132


xi

Table 2-31 Cognitive categories by areas - diminutives .......................... 136 Table 2-32 Cognitive categories by areas - augmentatives...................... 136 Table 2-33 Classification of languages by EM saturation ....................... 138 Table 2-34 Classification of languages by DIM saturation ..................... 138 Table 2-35 Classification of languages by AUG saturation .................... 139 Table 3-1 Spanish, Age group 6-11, Task 4, proposed words ................. 165 Table 3-2 Spanish, Age group 12-15, Task 4, proposed words ............... 169 Table 3-3 Spanish, Age group 16-19, Task 4, proposed words ............... 171 Table 3-4 Spanish, Age group above 19, Task 4, proposed words .......... 174 Table 3-5 Task 2 A tiny mon, Diminutives, overview by age groups ...... 175 Table 3-6 Task 2 A huge mon, Augmentatives, overview by age groups ... 175 Table 3-7 Task 3 Poem, overview by age groups .................................... 176 Table 3-8 Hungarian, Age group 4-5, Task 4, proposed words ............... 179 Table 3-9 Hungarian, Age group 6-11, Task 4, proposed words ............. 182 Table 3-10 Hungarian, Age group 12-15, Task 4, proposed words ......... 185 Table 3-11 Hungarian, Age group 16-19, Task 4, proposed words ......... 188 Table 3-12 Hungarian, Age group above 19, Task 4, proposed words .... 190 Table 3-13 Task 2 A tiny mon, Diminutives, overview by age groups .... 191 Table 3-14 Task 2 A huge mon, Augmentatives, overview by age groups...................................................................................................... 191 Table 3-15 Task 3, overview by age groups ............................................ 192 Table 3-16 German, Age group 4-5, Task 4, proposed words ................. 195 Table 3-17 German, Age group 6-11, Task 4, proposed words ............... 198 Table 3-18 German, Age group 12-15, Task 4, proposed words ............. 201 Table 3-19 German, Age group 16-19, Task 4, proposed words ............. 204 Table 3-20 German, Age group above 19, Task 4, proposed words........ 206 Table 3-21 Task 2 A tiny mon, Diminutives, overview by age groups .... 207 Table 3-22 Task 2 A huge mon, Augmentatives, overview by age groups ................................................................................................ 207 Table 3-23 Task 3, overview by age groups ............................................ 208 Table 3-24 Summary by languages, Task 1, Age group 4-5.................... 209 Table 3-25 Summary by languages, Task 1, Age group 6-11.................. 209 Table 3-26 Summary by languages, Task 1, Age group 12-15................ 209 Table 3-27 Summary by languages, Task 1, Age group 16-19................ 210 Table 3-28 Summary by languages, Task 1, Age group above19 ........... 210 Table 3-29 Summary by languages, Task 2, Diminutives, all age groups ................................................................................................ 211 Table 3-30 Summary by languages, Task 2, Augmentatives, all age groups ................................................................................................ 213 Table 3-31 Summary by languages, Task 3, all age groups .................... 215

xii


Charts Chart 2-1 A desired proportion of languages............................................. 53 Chart 2-2 An achieved proportion of languages ........................................ 53 Chart 3-1 Spanish, Age group 4-5, Task 2, Diminutives ......................... 161 Chart 3-2 Spanish, Age group 4-5, Task 2, Augmentatives .................... 162 Chart 3-3 Spanish, Age group 6-11, Task 1 ............................................ 163 Chart 3-4 Spanish, Age group 6-11, Task 2, Diminutives ....................... 163 Chart 3-5 Spanish, Age group 6-11, Task 2, Augmentatives .................. 164 Chart 3-6 Spanish, Age group 6-11, Task 3 ............................................ 164 Chart 3-7 Spanish, Age group 12-15, Task 1 .......................................... 166 Chart 3-8 Spanish, Age group 12-15, Task 2, Diminutives ..................... 167 Chart 3-9 Spanish, Age group 12-15, Task 2, Augmentatives ................ 167 Chart 3-10 Spanish, Age group 12-15, Task 3 ........................................ 168 Chart 3-11 Spanish, Age group 16-19, Task 1 ........................................ 169 Chart 3-12 Spanish, Age group 16-19, Task 2, Diminutives ................... 170 Chart 3-13 Spanish, Age group 16-19, Task 2, Augmentatives .............. 170 Chart 3-14 Spanish, Age group 16-19, Task 3 ........................................ 171 Chart 3-15 Spanish, Age group above19, Task 2, Diminutives ............... 172 Chart 3-16 Spanish, Age group above 19, Task 2, Augmentatives ......... 173 Chart 3-17 Spanish, Age group above 19, Task 3 ................................... 173 Chart 3-18 Hungarian, Age group 4-5, Task 1 ........................................ 177 Chart 3-19 Hungarian, Age group 4-5, Task 2, Diminutives ................... 177 Chart 3-20 Hungarian, Age group 4-5, Task 2, Augmentatives .............. 178 Chart 3-21 Hungarian, Age group 4-5, Task 3 ........................................ 178 Chart 3-22 Hungarian, Age group 6-11, Task 1 ...................................... 180 Chart 3-23 Hungarian, Age group 6-11, Task 2, Diminutives ................. 180 Chart 3-24 Hungarian, Age group 6-11, Task 2, Augmentatives ............ 181 Chart 3-25 Hungarian, Age group 6-11, Task 3 ...................................... 181 Chart 3-26 Hungarian, Age group 12-15, Task 1 .................................... 183 Chart 3-27 Hungarian, Age group 12-15, Task 2, Diminutives ............... 183 Chart 3-28 Hungarian, Age group 12-15, Task 2, Augmentatives .......... 184 Chart 3-29 Hungarian, Age group 12-15, Task 3 .................................... 184 Chart 3-30 Hungarian, Age group 16-19, Task 1 .................................... 186 Chart 3-31 Hungarian, Age group 16-19, Task 2, Diminutives ............... 186 Chart 3-32 Hungarian, Age group 16-19, Task 2, Augmentatives .......... 187 Chart 3-33 Hungarian, Age group 16-19, Task 3 .................................... 187 Chart 3-34 Hungarian, Age group above 19, Task 1 ............................... 188 Chart 3-35 Hungarian, Age group above 19, Task 2, Diminutives ......... 189 Chart 3-36 Hungarian, Age group above 19, Task 2, Augmentattives .... 189 Chart 3-37 Hungarian, Age group above 19, Task 3 ............................... 190


xiii

Chart 3-38 German, Age group 4-5, Task 1 ............................................ 193 Chart 3-39 German, Age group 4-5, Task 2, Diminutives....................... 194 Chart 3-40 German, Age group 4-5, Task 2, Augmentatives .................. 194 Chart 3-41 German, Age group 4-5, Task 3 ............................................ 195 Chart 3-42 German, Age group 6-11, Task 1 .......................................... 196 Chart 3-43 German, Age group 6-11, Task 2, Diminutives ..................... 197 Chart 3-44 German, Age group 6-11, Task 2, Augmentatives ................ 197 Chart 3-45 German, Age group 6-11, Task 3 .......................................... 198 Chart 3-46 German, Age group 12-15, Task 1 ........................................ 199 Chart 3-47 German, Age group 12-15, Task 2, Diminutives ................... 199 Chart 3-48 German, Age group 12-15, Task 2, Augmentatives .............. 200 Chart 3-49 German, Age group 12-15, Task 3 ........................................ 200 Chart 3-50 German, Age group 16-19, Task 1 ........................................ 202 Chart 3-51 German, Age group 16-19, Task 2, Diminutives ................... 202 Chart 3-52 German, Age group 16-19, Task 2, Augmentatives .............. 203 Chart 3-53 German, Age group 16-19, Task 3 ........................................ 203 Chart 3-54 German, Age group above 19, Task 1 ................................... 204 Chart 3-55 German, Age group above 19, Task 2, Diminutives ............. 205 Chart 3-56 German, Age group above 19, Task 2, Augmentatives ......... 205 Chart 3-57 German, Age group above 19, Task 3 ................................... 206 Chart 3-58 Morphological forms vs. age groups ..................................... 213

ACKNOWLEDGMENTS

I wish to express my gratitude to all the informants who kindly supplied the necessary data for my research and answered my questions with endless patience. Without their support the research would never have come true. The informants are in alphabetical order by language: Akan Amele Aghul Anejom̃ Archi Bafut Bahasa Indonesian Bardi Belarussian Breton Bulgarian Catalan Cirecire (Khoekhoe) Croatian Czech Danish Dargwa Datooga Diola-Fogny Dutch Embera Endo (Nandi) English Estonian Faroese Finnish French Gã Gagauz German

Clement K. I. Appah John R. Roberts Timur Maisak John Lynch Marina Chumakina Pius N. Tamanji Franz Mueller Claire Bowern Sviatlana Rudaya, Alena Rudenka Greg Stump Boyan Nikolaev Max Wheeler Andy Chebanne Ivo Mihajlović Marcela Grygerková Laurie Bauer Dmitrij Ganenkov Roland Kießling Kirsten Fudeman Jan Don Charles A. Mortensen Joost Zwarts Rochelle Lieber Annika Kilgi Hjalmar Petersen Vesa Koivisto, Johanna Laakso Eva Fričová M.E. Kropp Dakubu Astrid Menz Roswitha Fischer, Mia Körtvélyessy


Greek Hausa Hungarian Ilocano Inguish Irish Italian Itelmen Japanese Kabardian Karao Karelian Karo Batak Kashubian Ket Khwarshi Kinubi/Luganda Konni Krongo Kwaza Ladino Lakota Latvian Lavukaleve Lithuanian Lower Sorbian Luxembourgish Macedonian Maltese Maori Montana Salish Nêlêmwa Nenets Nivkh Norwegian Occitan Ossetic Polish Portuguese

Marios Mavrogiorgos Paul Newman Anna Fenyvesi Carl Rubino Johanna Nichols Paul Geraghty Nicola Grandi Jonathan Bobaljik Mark Volpe John Colarusso Sherri Brainard Pekka Zaikov Geoff Woollams Hania Makuraté, Dusan-Vladislav Pazdjerski Edward Vajda Zaira Khalilova Xavier Luffin Michael Cahill Mechthild Reh Hein van der Voort David Monson Bunis Regina Pustet Agnė Navickaitė-Klišauskienė Angela Terrill Bonifacas Stundžia, William R. Schmalstieg Viktor Zakar Peter Gilles Viktor Zakar Martin R. Zammit Winifred Bauer Sarah Thomason Isabelle Bril Beáta Wagner-Nagy Ekaterina Gruzdeva Eli Anne Eiesland Patric Sauzet David Erschler Gregorz A. Kleparski, Marcin Kudła Bernhard Pöll

xv


xvi

Provençal Roman Romansch Russian Scottish Gaelic Serbian Siar-Lak Siwi Slavey Slovak Slovene Spanish Swahili Tabassaran Tatar Tibetan Turkic Turkish Ubykh Udi Udmurt Ukrainian Upper Sorbian Vitu Votic Welsh Xhosa Yoruba Yurakaré Zazaki

Philippe Blanchet Ilincan Vasile Barbla Etter Ekaterina Barancheeva William Lamb Stanimir Rakić Friedel Martin Frowein Lameen Souag Keren Rice Miloslava Sokolová Andreja Žele Crisitina Suarez, Fernandez Rubiera Ellen Contini-Morava Dmitrij Ganenkov Ilmira Miftachova Scott DeLancey Martine Robbeets Nihan Ketrez John Colarusso Dmitrij Ganenkov Elena Rodionova Anatol Shevel Viktor Zakar Rene van den Berg Tatiana Agranat Gwenllian Awbery D.J. Lloyd Erin Shay Rik van Gijn Mesut Keskin

I also wish to thank all those who helped me with data for the experiment in phonetic iconicity. Without them, I could not have completed this project. The experimentors are ordered in alphabetical order by language: German Hungarian Spanish

Joachim Grzega, Maria Magee, Marja Meinl, Klaus Schneider, Pawel Sickinger, Verena Schneeweis Dóra Pődör Jesús Fernández-Domínguez, Nieves Pascaul Soler


xvii

This manuscript would not have been possible without the unfailing support of my family and colleagues.

ABBREVIATIONS

3 33 3p A:DCL ABS ADJ ADV AG AGT AI AMP AN ART ATT AUG C CF CL CNV1 DEC DEF DEM DIM EM EMP EMS ERG EXCLM F FREQ HAB IE IPF INT LNK

Third person Third person non-singular Third person plural inclusive Affirmative declarative (sentence type) Absolutive Adjective Adverb Agent Agentive case Animate intransitive verb stem Amplification Animate Article Attenuation Augmentative Consonant Circumfix Classifier Converb type 1 Declarative mood clitic Definitive noun gender suffix Demonstrative Diminutive Evaluative Morphology Emphatic Evaluative Morphology Saturation Ergative Exclamation Feminine Frequentative Habitual Indo-European Imperfective aspect Intensive Linking element


LOC M MEA N NOM NRL P PAT PFV PJR PL PN PRS PTCL R RDP REF REP RL RP/P SAE SE (Asia + Oceania) SEQ SG SP POSS TOP V VSC VWC VWF WALS WFR

xix

Locative Masculine Measure Neuter Nominative Non-relational prefix Possessive Predictive Areality Theory Perfective Pejorative Plural Personal name Present Particle Realis Reduplication Reflexive/reciprocal Reported speech particle, i.e. hearsay evidential Relational prefix Realis past/present Standard Average European South-east Asia and Oceania Sequential Converb Singular Speed form, indicating action undertaken before departing Possessive Topic particle Vowel Semantic category value Word class value Word-formation value World Atlas of Language Structures Word-Formation Rule

INTRODUCTION

The focus of this book results from the intersection of evaluative morphology (EM) and areal typology. First, it deals with the field of evaluative morphology, in particular, with the morphologically expressed concepts of diminutiveness and augmentativeness at the level of the system of language. By implication, it does not take into consideration analytically expressed ‘evaluatives’ (discussed, for instance, by Haas 1972) common in English (little boy), Chinese (小犬 xiǎo quǎn ‘doggy’), Turkish (yavru kopek ‘doggy’) and other (predominantly) isolating languages. For the same reason, it disregards context-dependent examination of evaluativeness (cf., for example, Seuren 1978, Bierwisch 1989, Neeleman et al. 2004, Rett 2008). It should be noted, however, that the difference between the langue and the parole is not necessarily clear. Supriyanto (pers. com.) demonstrates both the relativity of these notions as well as that of diminutiveness and augmentativeness with an example from Javanese where a diminutive adjective may also be interpreted as augmentative. Thus, gedhi ‘big’ is not really big, but may be big for a mouse, whereas a mouse as big as a calf would be guedhé ‘big by all standards’. This relative nature of diminutiveness and augmentativeness will be stressed on several occasions in this book. While the concept of evaluative morphology has attracted the attention of morphologists as a result of Scalise’s thought-provoking idea of the existence of a third morphology (1984), various aspects of morphological diminutives and augmentatives were examined both before and after Scalise. These aspects include the place of evaluative morphology (Stump 1993, Bauer 2004, Carstairs-McCarthy 1992); evaluative morphology in individual languages (Stankiewicz 1968 – Russian; Zelinkova 1993, Böhmerová 2011 – Slovak; Bolozky 1994 – Hebrew; Schneider 2003 – English; Grandi 2005 – Sardinian; Mattes 2006 – Bikol; Watson 2006 – Arabic); evaluative morphology as an inflectional category (Brown and Dryer, ms., Derzhanski 2005); semantics of evaluative morphology (Wierzbicka 1984, Jurafsky 1993, 1996, Grandi and Scalise 2000); etymology of evaluative affixes (Matisoff 1991); evaluative morphology acquisition (Richard et al. 1996, Savickiené and Dressler 2007); evaluative morphology and pragmatics (Dressler and Barbaresi 1994, 2001); morphonology (Gregová 2011), etc.

2

Introduction

This book examines evaluative morphology from the perspective of areal typology. The main focus is on the use of diminutives and augmentatives in Standard Average European (SAE). While Whorf’s idea of a European linguistic territory sharing a number of substantial features (1956) does not seem to have caught much attention in the decades following its birth, nearly falling into oblivion, the late 20th and early 21st centuries witnessed the resurrection of this idea, especially thanks to van der Auwera and Baoill (1998), Décsy (2000), Ramat (2000), Haspelmath (2001), Heine and Kuteva (2006) and others. Numerous important publications were engendered within an extensive EU-supported project called EUROTYP (typology of languages in Europe). As noted by Heine and Kuteva (2006), the rich research activity within the EUROTYP project has revealed that “there are multiple networks of typological similarities both uniting and dividing the languages of Europe” (2006, 36). Furthermore, they stress that “(i) there are clusterings of linguistic properties in Europe that are unlikely to occur in other parts of the world; (ii) at least some of these clusterings are unlikely to be due to genetic relationship or coincidence” (2006, 35). Despite the effort invested in research on SAE, there are certain fundamental problems with the concept of SAE itself. First, the (eastern) borders of the SAE territory are vague and remain undefined so far. For this reason, different authors specify varying numbers of SAE languages. Second, the number of linguistic parameters that have been examined is limited. The twelve SAE features identified by Haspelmath (2001, 1494-1501) are exclusively syntactic and inflectional categories. Phonological, derivational, lexical-semantic and a number of other features are almost completely ignored (with few exceptions like van der Auwera’s (1998b) discussion of lexical and phraseological similarities and Ramat’s (2000) analysis of selected wordformation issues). Consequently, SAE identification and specification are far from being complete. Evaluative morphology is an area that has not been studied in relation to Standard Average European, and therefore, it is still an untilled area. In accordance with Haspelmath’s requirement that SAE features only be defined against the situation in the “rest” of the world, my research aims to identify the status of diminutives and augmentatives on the SAE territory by comparing two samples of languages, the SAE sample and the world sample. This objective predetermines the cross-linguistic and typological character of this book, and the method of research, including language sampling and data collection and evaluation. Finally, as its second major objective, the present book examines phonetic iconicity as a theoretically challenging area of evaluative


3

morphology. Its challenging nature was long connected to the putative language universal (Universal #1926) in the Konstanz Archive (Plank and Filimonova). Thomas Payne (1997, 110)1 formulated the idea of the universal nature of phonetic iconicity, suggesting that the concept of diminutiveness tends to be represented by front high vowels and the concept of augmentativeness by high back vowels. This vowel-based iconicity was later extended and completed by a consonant-based iconicity, postulating palatalized consonants as iconic symbols of diminutiveness (Ultan 1978, Nieuwenhuis 1985, Bauer 1996). Strangely enough, some twenty years before Payne it was observed by Ultan and Nieuwenhuis that this kind of symbolism is geographically restricted (Ultan 1978, 553-554, Nieuwenhuis 1985, 107-110). The postulated phonetic symbolism has, therefore, been called into question by a number of linguists. Bauer (1996, 201), for illustration, analysed a sample of 50 languages which enabled him to draw a conclusion that “[t]here does not appear to be any universal principle of sound symbolism operating in markers of the diminutive and augmentative such that palatal articulation correlates with diminutives and not with augmentatives”. Furthermore, Gregová, Körtvélyessy and Zimmermann (2010), based on a sample of 60 languages, give additional support to the above-mentioned view of the areal nature of phonetic iconicity. It is for this reason that the present contrastive research avoids verifyng the validity of the aforementioned Universal #1926. Instead, the primary objective consists in the examination of the interrelation between phonetic iconicity and the sociolinguistic factor of age in the field of diminutives and augmentatives in three different languages. The structure of the book is as follows: Chapter 1 deals with evaluative morphology from a cross-linguistic perspective, relying on data from 203 languages (72 SAE languages and 131 world languages). Since one of the central questions is the relevance of evaluative morphology to Standard Average European (SAE), section 1.2 emphasizes the fundamental problems associated with Standard Average European, including the very notion of SAE, the languages belonging in SAE, the delineation of its borders and, consequently, the languages that fall within the SAE territory, the internal structure of SAE (the core languages and periphery), and the catalogue of the defining SAE features. These issues establish the background that makes it possible to compare two samples of data: a SAE sample and a world sample. No 1 The beginnings of research into phonetic iconicity within evaluative morphology date back to the early 20th century (e.g., Jespersen 1922 and Sapir 1929).

4

Introduction

cross-linguistic research is feasible without defining the basic terms. Section 1.3 deals with the second major topic, i.e., evaluative morphology. The central issues that determine any cross-linguistic research in this field include the place of evaluative morphology within the overall system of morphology as well as the semantic, formal and word-formation aspects of evaluative morphology. A new cognitive model projected onto a new radial model of evaluative morphology is proposed. The model is founded on the idea of evaluative morphology as a continuum in which prototypical cases express the meaning of quantity under or above the default value. The relation between augmentatives and diminutives is viewed as that of a scale. The evaluation process is implemented within four basic cognitive categories (SUBSTANCE, ACTION, QUALITY and CIRCUMSTANCE). The theoretical considerations are supported with numerous examples. With the scene being prepared for empirical research, Chapter 2 then outlines the objectives and method of research and presents and discusses the results of data analysis, first by the individual criteria, including the semantic, word-formation, and word-class criteria, then by comparing the data obtained for the SAE and world samples. The world sample is analysed and commented on with regard to five geographical territories. The samples are evaluated in terms of a newly introduced parameter of Evaluative Morphology Saturation. Since the main purpose of this chapter is to determine the relevance of evaluative morphology as a feature of SAE, the two samples are compared and typical features of evaluative morphology within the SAE are identified in the form of a proposal of EM-related Euroversals. The analysis of SAE languages by means of the Evaluative Morphology Saturation criterion results in the identification of the SAE borders and the internal classification of the SAE languages in terms of the core and periphery. This classification is represented on a SAE map. Chapter 3 deals with phonetic iconicity in evaluative morphology, with the primary objective being verification of the validity of Universal #1926. Section 2.2 clarifies the terms phonetic iconicity and sound symbolism with regard to three great figures in the field of semiotics, de Saussure, Peirce and Sapir. Section 2.3 concentrates on phonetic iconicity in its diverse manifestations and various features and former research in this area. Section 2.4 presents the outcomes of my psycholinguisticallyoriented experimental research. The research covers three typologically different languages, in particular, Hungarian, German and Spanish. After introducing the experiment, the analysis of the data obtained is organized by the individual languages and, within each of them, by age group and task. Finally, the data are compared with the aim of identifying meaningful correlations and drawing relevant conclusions.

CHAPTER ONE A THEORY OF EVALUATIVE MORPHOLOGY

1.1 General The first chapter concentrates on two central topics whose intersection underlies both experimental and empirical research presented in chapter 2. In particular, the concept of Standard Average European is discussed from the perspective of evaluative morphology. More specifically, the book is aimed at a comparison of the categorical nature, structural features, evaluative-formation strategies, semantics and productivity of evaluative morphology in the Standard Average European languages and the languages of the world. It should be noted that evaluative morphology is discussed throughout this book from the system level without reflecting its pragmatic facet. Interestingly, while research into theoretical foundations of evaluative morphology has yielded a few models (cf. Jurafsky 1993, 1996; Grandi 2005; Mutz forthcoming; Prieto forthcoming) and a relatively vivid discussion of the place of evaluative morphology in the system of morphology, little has been done in the description of the evaluative morphology systems of individual languages,1 and even less written on the pragmatic aspects of this subpart of morphology. Dressler and Merlini Barbaresi (1994, 2001) and Schneider (2003) rather represent an exception to this fact. The pragmatic aspects of evaluative morphology is a topic which, for obvious reasons, requires its specific research methods, pursues its specific research objectives and, by implication, lies beyond the scope of this book. The same is true of research into analytic forms and/or morphologically simple words (monemes) used to express evaluation. The complexity of the problems raised during decades of discussion on evaluative morphology makes it, similar to derivational and inflectional morphology, a research field of its own, including its specific topics, problems and research methods. It is for this reason that the title of this book is Evaluative Morphology from a cross-linguistic perspective 1

An important step forward in this respect is Grandi and Körtvélyessy (forthcoming), a volume mapping the state-of-the-art in the field and providing description of evaluative morphology in 50 languages of the world.

6

Chapter One

and that inclusion of any of the above-mentioned issues would go far beyond the scope of this research. A comparison of an SAE sample with a world sample of languages will give support to my hypothesis that evaluative morphology in SAE languages carries certain features that make it distinct from the rest of the world. Therefore, it should be included in the list of the features characteristic of the SAE linguistic area. Furthermore, this chapter deals with the distribution of evaluative morphology on the SAE territory by introducing the parameter of Evaluative Morphology Saturation (EMS). This parameter specifies the capacity of evaluative morphology in individual languages and makes it possible to identify the core and periphery of SAE and the differences between SAE and the rest of the world with reference to evaluative morphology. Theoretical questions of Standard Average European are discussed in section 1.2, and those of evaluative morphology in section 1.3.

1.2 Standard Average European 1.2.1 The notion of Standard Average European Whorf (1956) proposed the term Standard Average European when studying American Indian languages. To facilitate their comprehension Whorf contrasted them with familiar languages and came to the conclusion that there are substantial differences in the grammatical system of Hopi on one hand, and European languages on the other. In his view, these differences reflect the differences between the Hopi culture and their comprehension of the world and the European (western) culture and “European” comprehension of the world (including different comprehensions of the concepts of time, space and substance): “Since with respect to the traits compared there is littler difference between English, French, German, or other European languages with the possible (but doubtful) exception of Balto-Slavic and non-Indo-European, I have lumped these languages into one group called SAE, or ‘Standard Average European’” (ibid., 138). The label Standard Average European did not catch much attention and may have fallen into oblivion were it not rediscovered in recent decades, presumably in connection with the idea of a united Europe. While various projects were implemented, the most significant, extensive and prolific has been the EUROTYP project. Its main objective was to prove that besides numerous common genetic features, European languages bear

A Theory of Evaluative Morphology

7

resemblance to each other thanks to certain shared features resulting from language contact. While its achievements are remarkable, the EUROTYP project has indicated many problems that can be summarized in the following three questions: 1) What should be analysed – languages or dialects? 2) What are the borders of Standard Average European? What languages are at its nucleus and which languages are on its periphery? 3) What should be the catalogue of Standard Average European features?

1.2.2 What should be analysed – languages or dialects? The question is rather tricky when one realizes how difficult it is to put a clear borderline between standard language and dialect. Interestingly, while Haspelmath (1998, 2001) and van der Auwera (1998a, 1998b), in principle, base their studies on standard languages, Haspelmath (1998) also includes in the SAE nucleus some northern Italian dialects. Heine and Kuteva (2006) stress the importance of, inter alia, dialect variation as a way to account for “the dynamics underlying grammatical categorization in European languages” (ibid. 33). Both language-centred and dialectcentred approaches have their advantages and disadvantages. The number of dialects is greater than the number of languages, and they are more open to influences than the standardized variety of a language. For illustration, the Šariš dialect spoken in the eastern part of Slovakia shares a part of its vocabulary with the Hungarian language as a result of long and extensive language contact throughout their common history in the AustrianHungarian Empire. For example, kalap ‘hat’ exists both in the Šariš dialect and the genetically unrelated Hungarian, but not in Slovak where the equivalent word is klobúk. To take one more example, this time from the more resistant category of grammar, the analytic future tense construction of the Zemplín dialect, combining future form of the auxiliary and the past tense of the finite verb such as nebudzem chodzela, ‘I will not go.PAST.F.’ (literally, ‘I will not went’), does not exist in the standard Slovak but does in the neighbouring standard Polish. Thus, a large number of dialects can change the search for common features into an unfeasible enterprise: in the small territory of Slovakia there are no less than 29 dialects (Palkovič 1981). At the same time, the data on standard languages spoken in Europe is easily accessible because these languages have been studied for centuries. This can hardly be said of dialects. To collect dialectal data of

8

Chapter One

the same quality and extent is a large project in itself and is well beyond the scope of the current research. Comparing dialects instead of languages could bring new and interesting results, but is currently an unattainable goal. In my research I will rely on languages that usually appear on maps of European languages.

1.2.3 The borders of Standard Average European To begin, there is no agreement concerning the delineation of SAE borders. Heine and Kuteva (2006, 27) even ask if it is “really possible to argue that there is a European Sprachbund or linguistic area with clearly defined boundaries?” As the following discussion shows this question is fully justified. The first problem concerns the criteria for the delineation of the SAE borders. What one can take for granted is, logically, the fact that SAE languages are spoken in Europe. But what is the linguistic Europe? What criteria should be chosen for its delineation? There are several possibilities – linguistic, geographical, political, cultural, historical, and religious – each providing slightly different results. Décsy (2000), for example, starts his book with geographical delimitation of Europe. His picture of Europe is different from that found in Whorf (1956), who applies the criterion of culture. Others like Kortmann (1998a, 1998b), van der Auwera (1998b), Haspelmath and Buchholz (1998) and Haspelmath (1998, 2001) rely on morphosyntactic criteria. For the purpose of my research I use Price’s (1998) delimitation of Europe that primarily (not exclusively, though) relies on geography: • Iceland in the north is included. • While from the geographical point of view inclusion of some western islands belonging to Portugal is questionable (Portuguese Atlantic islands, e.g. Madeira), the only language spoken in their territory is Portuguese. • Malta is considered a European country because from the geographical point of view it is closer to Sicily than to the North African coast (Price 1998, xi). • While from the geographical point of view Cyprus ought to be considered part of Asia, from the linguistic point of view it is the same case as that of Malta – the languages spoken on Cyprus are Greek and Turkish, both of which are European languages.

A Thheory of Evaluaative Morphologgy

9

Map 1-1 Mapp of Europe

• In thhe south-east the obvious dividing linee between Eu urope and Asia is the Caucassus. A • Mounntains, which leave Geeorgia, Armeenia, and Azerbaijan geogrraphically ouutside Europe. As Price ppoints out (19 998: xii), thesee countries wish to be considered Europeean countries and their languuages should be b included in n Europe. Connsequently, th his means incluuding about 40 4 other lang guages of thee Caucasian language famillies. The sam me approach was adoptedd by van derr Auwera (20111).The easternn border of Europe is esstablished by the Ural Mounntains. In thee questionablee area betweeen the southerrn end of the U Urals and thee Caspian Seea, it is hard to identify a natural bordeer; Price opts for the politiical criterion – the border is i created by thhe frontier of the t Republic of o Kazakhstann. • This geographical delimitation corresponds c w with the follow wing map of lannguages spokeen in Europe:

10

Chapter One

Map 1-2 Languages spoken in Europe

The following list comprises 145 languages. This list is built on the basis of the World Atlas of Language Structures (WALS), in particular the region section. The minimal and maximal latitude and longitude values of Europe were entered in the WALS geographical section, generating the language families listed below. The list was compared to that of van der Auwera (1998a). Languages that displayed discrepancies were checked in WALS again, this time in the language section. The area of their usage was verified. If the area fell within the geographical delimitation of Europe, it remained on the list. The number of Romani and Yiddish varieties seems high: WALS lists 8 Romani and 4 Yiddish languages. In contrast to other languages of Europe, neither Romani nor Yiddish have a fixed territory of usage. Due to the strong influence of the languages with which they live, they are often divided into regional varieties, e.g. Yiddish


11

(Bessarabian), Yiddish (Lithuanian), Yiddish (Lodz). Even the detailed list offered by WALS is not complete, as evident by the omission of Carpathian Romani.1 For this reason, neither Romani nor Yiddish is in the following list. FAMILY: ALTAIC Genus: Mongolic (1) Moghol Genus: Turkic (17 languages) Krymchak, Azerbaijani, Karaim, Karachay-Balkar, Tatar, Noghay, Bashkir, Crimean Tatar, Urum, Tatar-Noghay (Alabugat), Chuvash, Tatar (Mishar), Karakalpak, Gagauz, Noghay (Karagash), Kumyk, Turkish. FAMILY: AFRO-ASIATIC Genus: Semitic (4 languages) Arabic (Kormakiti), Maltese, Neo-Aramaic (Persian Azerbaijan), NeoAramaic (Amadiya). FAMILY: BASQUE Genus: Basque (1 language) Basque. FAMILY: INDO-EUROPEAN Genus: Albanian (1 language) Albanian. Genus: Armenian (2 languages) Armenian (Eastern), Armenian (Western). Genus: Baltic (2 languages) Lithuanian, Latvian. Genus: Celtic (7 languages) Cornish, Irish (Munster), Manx, Gaelic (Scots), Breton, Irish, Welsh. Genus: Germanic (12 languages) German, English, Dutch, Frisian, Luxemburgeois, Swedish, Icelandic, Yiddish, Alsatian, Norwegian, Danish, Faroese. Genus: Greek (2 languages) Greek (Modern), Greek (Cypriot). Genus: Indic (1 language) Romani. Genus: Iranian (8 languages)

1

Cf. Rácová and Horecký (2007).

12

Chapter One

Ossetic, Zazaki, Tat (Muslim), Kurmanji, Gilaki, Talysh (Southern), Hawrami, Mazanderani. Genus: Romance (15 languages) Catalan, Sardinian, Romanian, Galician, Provençal, Italian, Occitan, Ladin, French, Portuguese, Romansch, Spanish, Ladino, Asturian, Moldavian. Genus: Slavic (17 languages) Polish, Slovene, Bosnian, Polabian, Serbian-Croatian, Sorbian (Upper), Russian, Kashubian, Slovincian, Sorbian, Czech, Slovak, Macedonian, Belorussian, Ukrainian, Sorbian (Lower), Bulgarian. FAMILY: KARTVELIAN Genus: Kartvelian (4 languages) Mingrelian, Svan, Georgian, Laz. FAMILY: NAKH-DAGHESTANIAN SUBFAMILY: DAGHESTANIAN Genus: Avar-Andic-Tsezic (13 languages) Hunzib, Chamalal, Hinukh, Bagvalal, Karata, Bezhta, Tindi, Akhvakh, Godoberi, Khvarshi, Tsez, Avar, Andi. Genus: Lak-Dargwa (2 languages) Lak, Dargwa. Genus: Lezgic (10 languages) Archi, Kryz, Aghul, Lezgian, Tsakhur, Rutul, Tabassaran, Udi, Khinalug, Budukh. Genus: Nakh (3 languages) Tsova-Tush, Chechen, Ingush. FAMILY: NORTHWEST CAUCASIAN Genus: Northwest Caucasian (5 languages) Abaza, Adyghe, Ubykh, Kabardian, Abkhaz. FAMILY: URALIC; SUBFAMILY: FINNO-UGRIC Genus: Finnic (15 languages) Karelian, Estonian, Liv, Veps, Izhor, Yazva, Finnish, Votic, Mari (Hill), Mari (Meadow), Mordvin, Komi-Zyrian, Komi-Permyak, Udmurt, Saami. Genus: Ugric (3 languages) Khanty, Mansi, Hungarian. The presented list of languages and map 1-2 refer to languages spoken in Europe. Obviously, two language families prevail – the Indo-European and the Uralic. In the following paragraphs various approaches to the scope of SAE are compared, including the delimitation of the SAE core


13

and periphery. In particular, Tables 1-1, 2, 3, 4 represent the views of Haspelmath, Décsy, Ramat and van der Auwera, respectively: Table 1-1 SAE according to Haspelmath (1998) Zone Nucleus Core Periphery clearly outside

Languages Dutch, German, French, northern Italian dialects Other Germanic and Romance languages, West and South Slavic and Balkan languages East Slavic, Baltic, Balto-Finnic, Hungarian, perhaps Basque, Maltese, Armenian, Georgian Celtic languages, Turkic, Abkhaz-Adyghean, NakhDaghestan, eastern Uralic

Table 1-2 SAE according to Décsy (2000) Zone SAE The Viking The Littoral The Peipus The Rokytno The Danube The Balkan The Kama Isolates Diaspora Languages

Languages German, French English, Italian, Russian Danish, Norwegian, Icelandic, Faeroese, Irish. ScotchGallic. Welsh, Breton, Swedish, Lapp, Finnish. Vepian, Frisian, Dutch, Basque, Spanish. Portuguese, Maltese Estonian, Votic, Livonian, Latvian Polish, Lithuanian, Belorussian, Ukrainian, Kashubian Czech, Slovak, Hungarian, Slovenian, Serbo-Croatian (Serbian, Croatian, Bosnian) Rumanian, Moldovan, Bulgarian, Makedonian, Albanian, Greek, Turkish Chuvash, Cheremis (Mari), Votyak, Mordvin, Ziryaun (Komi), Yurak (nenets), Kalmyk Luxemburgish, Romansch, Sorbian, Gagauz Yiddish, Ladino, Karaim. Romany (Gypsy), Armenian

14

Chapter One

Table 1-3 SAE according to Ramat (2000) Zone More or less central members More or less peripheral members Very marginal even outside the SAE

Languages All Romance, West-Germanic, Balkan and BaltoSlavic North-Germanic, Balto-Finnic, Hungarian, Georgian and Armenian. Celtic languages, Basque, Maltese and the Eastern European languages (Turkic, Abkhaz-Adugheab, Nakh-Daghestanian, Eastern Uralic languages

Table 1-4 SAE according to van der Auwera (1998b) Zone Charlemagne Sprachbund Balkan Sprachbund Northern around Danish

Languages French, Italian, German, Dutch, Polish Rumanian, Albanian, Bulgarian, Greek Danish, English, Norwegian, Faroese

The most comprehensive list of European languages is proposed by Décsy. His criteria are based on a combination of geographical and cultural criteria. Whereas, van der Auwera and Ramat identify three zones of SAE languages. Haspelmath (1998) distinguishes between the nucleus and the core SAE languages. In his later work (2001) he, however, narrows his nucleus to two languages only, French and German, while Dutch and Italian are moved to other Germanic and Romance languages. In spite of these differences, the nucleus (Haspelmath), Standard Average European (proper) (Décsy), and Charlemagne Sprachbund (van der Auwera) play roughly the same role within the SAE system as they resulted from the same effort to identify the heart of SAE. As Ramat’s differentiation is not based on particular languages but on language phyla, it is excluded for the purpose of the comparison. Table 1-5 provides a summary of three different views:


15

Table 1-5 Various approaches to central SAE languages Linguist Haspelmath

Name of the zone Nucleus

Décsy

SAE

van der Auwera

Charlemagne Sprachbund

Languages Dutch, French, German, northern Italian dialects English, French , German, Italian, Russian Dutch, French, German, Italian, Polish

As it follows from the previous overview, each of the three linguists chose a different label for the centre of SAE: Haspelath distinguishes between core and nucleus, Décsy recognizes only five languages as SAE languages and divides Europe into a number of smaller linguistic areas, and van der Auwera speaks of the so-called Charlemagne area. Irrespective of the labels used the areas coincide in considering French and German the central SAE languages. While Haspelmath also lists Italian he confines himself to the northern dialects of Italian and later (2001) eliminates Italian from the nuclear SAE languages. Dutch appears to have a similar status – Décsy does not mention it and it is also excluded by Haspelmath (2001). Both Décsy and van der Auwera try to include at least one Slavonic language, either Russian or Polish. Since Décsy justifies his selection by the number of speakers of a given language, the inclusion of Russian is a logical consequence of his approach. In addition to the identification of the central SAE languages, Ramat and Haspelmath agree on langauges outside the SAE zone. They eliminate Celtic languages, Turkic, Abkhaz-Adugheab, Nakh-Daghestanian and Eastern Uralic languages. Agreement between them is also found in the peripheral, or marginal, zone, including languages like Basque and Maltese. One of the most recent discussions on this topic is Heine and Kuteva (2006). They, essentially, rely on Haspelmath’s division (2001) based on examination of 39 languages for nine features. There are languages having five or more of the nine features and then languages having as few as two of the evaluated features. On the basis of Haspelmath’s research that represents a “systematic analysis”, offering “a way of defining SAE languages based on fairly unambiguous quantitative evidence” (Heine and Kuteva 2006, 25), the following can be referred to as SAE languages (ibid.): • All Romance languages • All Germanic languages • All Slavic languages

Chapter One

16

• • • •

The Baltic languages, Latvian and Lithuanian Albanian Greek Hungarian

Finnish, Estonian, Maltese, Basque, Turkish, the Celtic languages, Georgian, Armenian, etc. are excluded from SAE languages. On the other hand, Heine and Kuteva are not concerned with the question of the nuclear area of Europe. The question of SAE delimitation is fairly controversial. Although Heine and Kuteva apply Haspelmath’s results, they also point out that Haspelmath did not avoid bias in favor of Indo-European languages and the western half of Europe. Interestingly, he returned to the original, Westoriented idea of Whorf, who assumes, as already mentioned that “…the grammar of Hopi bore a relation to Hopi culture and the grammar of European tongues to our own ‘Western’ or ‘European’ culture” (Whorf 1956, 138). The previous overview of different approaches to the range and the internal structure of the SAE languages suggests the following: first, if there is anything like a SAE core/nucleus, French and German are its crucial parts, at least from the point of view of the parameters examined and evaluated so far; second, the number and the structure of the remaining SAE languages depends on various factors, including the approach of the linguist, typological features studied, data interpretation and evaluation, and criteria taken into consideration. The lack of agreement on the delimitation of the SAE languages has had its consequences on the methodology of the present research. I preferred not to identify the list of the SAE languages before the research itself. Instead, the SAE territory, and by implication, the languages examined were tentatively delimited geographically, by the borders of Europe, and subsequently delimited on the basis of the research results.

1.2.4 What is the catalogue of Standard Average European features? Haspelmath (2001, 1492) defines a linguistic area as an area where “a number of geographically contiguous languages share structural features which cannot be due to retention from a common protolanguage and which give these languages a profile that makes them stand out among the surrounding languages.” It follows from this definition that the catalogue of structural features is crucial to the notion of Sprachbund. This raises the


17

question of what features are relevant to the discussion of SAE and, therefore, should be examined. Haspelmath (ibid.) employs the following twelve morphosyntactic features to determine the degree of typicality of the individual SAE languages: • • • • • • • • • • • •

Definite and indefinite articles Relative clauses with relative pronouns ‘Have’ perfect Nominative experiencers Participial passive Anticausative prominence Dative external possessors Negative pronouns and lack of verbal negation Particles in comparative constructions Relative-based equative constructions Subject person affixes as strict agreement marker Intensifier-reflexive differentiation

Ramat (2000) provides a list of (prevailingly) morpho-lexical SAE characteristics: • Building derivational compounds of the -ation, -ing, -ness type (e.g. investigation, learning, fitness). • Word-formation rules with suffixes and prefixes – the possibility of building new words, mainly abstract nouns, thus keeping pace with the general development of sciences and techniques via parallel and translatable lexical system. • A stock of prefixes and suffixes of cultivated origin that are used in many European standard languages: auto-, anti-, -ismo, -logia. • Diachronic drift toward grammatical transparency (“diagrammacity”). Van der Auwera’s (1998b, 815) features range over inflectional morphology, syntax and lexicon: • • • • • •

Analytic expression formats Simplified case paradigms Presence of definite and indefinite articles Use of ‘have’ and ‘be’ as auxiliaries Non-pro-drop character Relatively fixed word order at the level of sentence–SVO order

Chapter One

18

• • • •

Phrasal word order: prepositions and postposed genitives Accusativity: agreement of the finite verb with the subject Passive allows the agent Lexical and phraseological similarities

The three lists differ from each other. One of the major problems facing each of these lists of features is the vague selection criteria. Although Haspelmath (2001) states that the features have not been selected randomly, he rather weakly argues that the features were known to show a distribution that supports the SAE hypothesis. This issue is, however, crucial. If we want to identify a language area, the criteria for the selection of the features must be stated unambiguously. It goes without saying that these features should range over the individual structural levels of language, including phonology, inflectional and derivational morphology, syntax, and lexicon. Any conclusions concerning the number of SAE languages and the core languages drawn from a limited, incomplete and/or biased selection of features are premature. Omitting any of the fundamental language features from this kind of areal typology research may distort the results, or may result in “false belief” of linguistics (such as that of the universal nature of phonetic iconicity referred to in the Introduction). As will be demonstrated by the results of my research, the range of languages as well as the core of SAE languages heavily depend on the feature(s) examined. The core SAE languages in terms of evaluative morphology differ substantially from those identified by morphosyntactic criteria in Haspelmath (2001). By implication, it may be assumed that the actual number of SAE languages and the degree of SAE typicality only result from the data obtained from the evaluation of all relevant typological features. Terminologically, there is an apparent effort to use an umbrella term for features common to (the majority of) European languages. Décsy (1973, 196) uses the term Europeme for linguistic properties found in European languages when he asks: Gibt es Eigenarten in der inneren Struktur (Phonetik, Grammatik, Syntax, Wortschatz), die nur in den Sprachen unseres Kontinents nich aber in der Idiomen anderer Erdteile anzutreffen sind? Kann man einen Katalog dieser Europeme zusammenteilen?

Van der Auwera (1998b, 812) introduces the notion of Euroversal. In this respect, Heine and Kuteva (2006, 13) point out that there are two possible definitions of EuroversalError! Bookmark not defined.. They can be understood as “linguistic properties found in most or all European


19

languages but which may also be found in non-European languages” (ibid.), or, as maintained by van der Auwera (1998b, 812), as “claims that are held for all or most of the languages of Europe”. The term, coined as a blend of Europe and universal, suggests that it is not clear whether the features are typical of the languages of Europe only, or whether they can be identified as language universals. This hesitation is expressed by van der Auwera when he states that “in this modest and very tentative way, some of the Euroversals are possible Universals. It seems reasonable to assume that the more the Euroversals depend on semantic and/or functional factors, the higher the chance that the Euroversals are Universals” (ibid.). This hesitation may well be a consequence of the fact that the analysis of European languages is not contrasted with languages spoken outside Europe. The only exception seems to be Haspelmath (2001). The above discussion of the SAE languages makes it possible to pinpoint the fundamental problems of research into SAE that still need addressing: • In spite of a number of projects, experiments, and studies, the views on the range of SAE languages differ. • In spite of some agreement on the existence of a nucleus, typologists do not agree on those languages that comprise the nucleus. The catalogue of features that have been examined for the sake of identifying both SAE languages and SAE core languages appears to be incomplete and shows bias in favour of morphosyntactic features. There are two reasons for this situation. First, the starting point of each typologist is different as there are many language features to be compared. The methods of and approaches to their identification/selection are not unified. Consequently, the results necessarily differ; sometimes they are even contradictory. Second, the attention of the previous studies has been almost exclusively focused on European languages; since they are not put into contrast with languages spoken outside Europe, it is difficult to say whether the evaluated features are really the features of (the majority of) SAE languages (genuine Euroversals), or are also shared by other languages of the world. Hence, one can hardly draw conclusions relevant to SAE unless the SAE data is complete and compared to that of the global sample of languages. A logical consequence of these considerations is the question pertaining to whether anything like the SAE Sprachbund actually exists. Is

20

Chapter One

it not just a desideratum? An answer to this question will require much effort from typologists in the direction outlined above. To avoid the indicated shortcomings of previous investigations, I predefine neither the borders of SAE nor its core. The languages chosen for my research have been taken from the list of languages spoken within the territory of Europe as defined in 1.2.3 above. None of them has been eliminated before conducting my research. The data collected is analysed and the outcomes are compared to languages spoken outside the European territory. The comparison will show whether the SAE languages share some common features in the field of evaluative morphology that are not present significantly in other languages of the world, and if so, then the data will make it possible to determine the SAE borders by checking the proposed languages for presence or absence of the features examined. Section 1.2 was devoted to the introduction of the main challenges for research into SAE. The purpose of section 1.3 is to do the same for the field of evaluative morphology.

1.3 Evaluative Morphology “…despite the enormous literature in the subject, mystery still enshrouds these forms [expressive derivations]” —Beard (1995, 165)

1.3.1 General Evaluative morphology came to prominence with Scalise (1984) and his thought-provoking idea of the existence of “third morphology”. If one checks the subject index of various books on morphology or lexical semantics, or browses through voluminous reference books, e.g. An Encyclopedia of Language (Routledge 1990) or The Handbook of Linguistics (Blackwell 2003), evaluation, diminutives or augmentatives are hardly ever mentioned. On the other hand, linguists are aware of their existence and discuss them within the framework of expressive morphology and semantics (Zwicky and Pullum 1987, Fortin 2011), the Jackendoffean lexical sematic framework (Gràcia and Turon 2000), expressive word-formation (Szymanek 1988), extragrammatical morphology (Dressler 2005), morphopragmatics (Dressler and Merlini Barbaresi 1994, 2001), the theory of gradation (Bierwisch 1989), etc. Within the time span of almost 30 years various languages were analysed, studies into evaluative morphology were conducted, and numerous papers were written. In spite of this fact, a


21

common platform is still missing – a consensus on the nature and the scope of evaluative morphology. Grandi (2011, 6) states that evaluation, roughly speaking, implies two different perspectives: 1) objective, descriptive or quantitative, represented by the semantic primitives SMALL and BIG; 2) qualitative or subjective represented by the semantic primitives GOOD and BAD. While the former perspective of evaluation is based on denotation, the starting point of the latter is connotation, or association. Referring to this classification, Grandi and Körtvélyessy (forthcoming) maintain that qualitative evaluation is a judgment of how useful, good, successful, lovely, tender, etc. something is. It can be threefold (in particular, neutral, positive, or negative) with subtle differences in the degree of “positivity” or “negativity” of the (perceived) quality. If the term “evaluation” is applied to the field of linguistics, it refers to various linguistic constructions at different levels of language expressing the concept of evaluation by different linguistic means. Linguistic constructions of this nature can be referred to as evaluative constructions. Haas (1972, 148) identifies six methods of expressing diminutiveness: consonant and/or vowel symbolism, reduplication, affixation, syntactic modification, suppletion, and grammatical displacement. The evaluative marker can be expressed by tonal variation, analytic construction, synthetic construction, etc. From this it follows that evaluative constructions range over the whole language system. If morphological means are used to express evaluation, we speak about evaluative morphology: (1) EVALUATION+MORPHOLOGY → EVALUATIVE MORPHOLOGY => expression of evaluation by means of morphology (evaluative morphological constructions). By implication, sound symbolism, syntactic modification or grammatical displacement, all included in Haas’ list, do not fall within the scope of evaluative morphology. In their prototypical realization, evaluative morphological constructions consist of the evaluated base, which is an existing (actual) word, and the evaluative marker. The base prototypically expresses the neutral value, and the evaluative marker assigns a new value to the standard within the limits set by the already mentioned semantic primitives GOOD-BAD and SMALL-BIG. The base

Chapter One

22

is represented by a word stem and the evaluative marker can be realized by various word-formation processes, most commonly by suffixation (2), prefixation (3) and reduplication (4), less frequently by compounding (5) and some other devices discussed later. (2) Armenian

hast-lik thick/heavy-DIM ‘plump’ (Dum-Tragut 2009, 665)

(3) Breton

gour-lano AUG-tide ‘high tide’ (Stump, pers. com.)

(4) Omaha

gdhejā~ja spotted~RDP ‘spotted in many places’ (Hale 2001, 41)

(5) Malayalam

aaʈʈin-kuʈʈi aatə-kuʈʈi goat-young ‘kid’ (Asher and Kumari 1997, 398)

The morphological nature of the evaluative marker cannot be universally identified with derivational morphology, even though derivational realization of morphological evaluatives may be perceived as prototypical. It will be, demonstrated, however, that inflectional realization of evaluation is not infrequent in some languages. Cross-linguistically, we may speak of a scale between inflection and derivation, with individual languages assuming different positions on this scale. According to Dressler and Barbaressi (1994, 53), the Italian term valuativi covers diminutives, augmentatives and pejoratives. What strikes one is the bias of this delineation of evaluative morphology in favor of negative connotations. What this term fails to cover are ameliorative, affectionate and other positive connotations. Irrespective of this prejudice, their approach aptly identifies two central areas at the level of langue, in particular, diminutiveness and augumentativeness. According to Bauer (1997, 537), diminutivisation and augmentativisation are the core areas of evaluative morphology. However, diminutives and augmentatives are merely two aspects of evaluation. It follows from the previous discussion that evaluative morphology deals with morphologically complex forms. The main objectives of


23

research into evaluative morphology include the identification of its place in the system of morphology in individual languages as well as from the cross-linguistic perspective; delimitation of its scope bearing in mind fuzzy boundaries and overlapping with various areas of derivational and inflectional morphology; and analysis and description of evaluative constructions formed by morphological means. As such, it deals with “typical” morphological issues, including strategies of complex word formation, questions of word-classes, grammatical categories, semantic (cognitive) categories and morphophonological effects of evaluative formation. The following section discusses both form and semantic facets of evaluative morphology as well as the question of its scope.

1.3.2 The place of evaluative morphology in the system of morphology One central concern of cross-linguistic research into evaluative morphology is its place within the system of morphology. The significance of this question stems from the widely held belief that evaluative morphology is derivational in nature. Interestingly, if one were to go through various criteria proposed for the distinction between inflection and derivation, the picture that would emerge for evaluative morphology seems very conclusive:2 • EM is derivational in its function of extending the lexicon (even though there are many cases of extending the lexicon by way of inflectional processes). • EM is derivational in its non-obligatoriness in syntactic constructions. • EM is derivational owing to the competition between various evaluative formations. • EM is derivational because it rests on a multiplicity of rules rather than on uniformity. • EM is derivational because its meanings are much more specific than those of inflectional morphology. • EM is derivational because it can be class-changing. • EM is derivational because it is recursive. 2

The list is far from being complete, but the criteria adduced here are sufficient to support the argumentation.

Chapter One

24

• EM is derivational because its paradignmatic organization is weaker than that of inflectional morphology. • EM is derivational because its affixes typically precede the inflectional ones. • EM is inflectional because it typically involves smaller meaning changes than derivational morphology. Out of these ten criteria, nine indicate the derivational nature of EM. Recent research, however, provides ample evidence that the situation is much more fuzzy from a cross-linguistic perspective, or even from the perspective of individual languages some of which exclusively employ derivational techniques (e.g. Slavic languages), some of them both derivational and inflectional devices (e.g. Occitan, Italian and Supyire), and some of which completely rely on inflectional techniques (e.g. Bantu languages). The fuzziness of the problem is also manifested by the fact that each of the above-mentioned nine criteria favouring the derivational nature admits (not infrequently) a number of counter-examples. Thus, for example, it will be shown throughout this book that there are numerous cases of inflection-based evaluative formations; there are cases of applying evaluative-formation rules after rather than before inflectional rules; there are numerous languages without recursive capacity of evaluative suffixes; and that the paradigmatic organization of evaluative morphology may be very systematic such as that of Slovak adjectives diminutivized by the suffix -učk (peknučký ‘nice-DIM’, milučký ‘lovely-DIM’, choručký ‘ill-DIM’, dobručký ‘good-DIM’, etc.). Strictly speaking, there are four possibilities: 1) 2) 3) 4)

Evaluative morphology is a subcategory of inflection. Evaluative morphology is a subcategory of derivation. Evaluative morphology has its own specific position. Evaluative morphology is a part of the inflection-derivation continuum.

As indicated above, there is ample evidence that options 1 and 2 are unjustified from a cross-linguistic perspective despite the traditional belief that evaluative morphology is a part of derivational morphology, even though not its prototypical part (cf., for example, Dressler and MerliniBarbaresi 1994 and Grandi 2011, forthcoming). Scalise (1984), Beard (1995) and Fortin (2011) are proponents of option 3. While Scalise’s observations are based on Italian, Fortin takes a broader view by examining specific instances from various languages in order to conclude


25

that “the traditional paradigm of inflectional and derivational morphology breaks down when it comes to expressive affixes” (2011, 74). By implication, Fortin stresses, we ought to consider EM on its own terms (ibid.).3 Booij (2000, 361) assumes that evaluative morphology “is, like inflection, transparent for the syntactic category and the gender of the base, but that is felt as derivation as far as the semantic change involved is concerned”, thus adhering to option 4.5 Inflectionwise, Brown and Dryer demonstrate that the diminutive category in Walman, a language spoken in Papua New Guinea, “is clearly an inflectional category that is realized not on the noun itself, but on various words that agree with nouns or with noun phrases…. The diminutive affix -l- is suffixed, prefixed, or infixed to words that stand in agreement with the nouns” (1): (6)

pelen l-aykiri. dog 3SG.DIM-bark ‘The puppy is barking’

The next example illustrates multiple agreement with the diminutive affix in different positions: (7)

Pelen pa ten l-o dog that 3SG.DIM-be ‘That puppy is large’

lapo-l. large-DIM

Jóola Banjal reserves particular inflectional classes for marking diminutives and augmentatives, respectively: (8) Jóola Banjal (DIM) (AUG)

eḀjoba ‘dog’ ju-joba ‘small dog’ gaḀjoba ‘big dog’

PL. suḀjoba PL. muḀjoba PL.ba-joba

(Creissels, forthcoming) In Shona “evaluatives are from a set of N-class prefixes that mark grammatical class” (Déchaine, Girard, Mudzingwa, Wiltschko, forthcoming):

3

It should be stressed, however, that Fortin’s conclusions are drawn on the basis of several selected cases and are based on just one part of evaluative-formation devices, i.e., affixes. Other evaluative formation processes are not reflected in his work.

Chapter One

26

(9) Shona mù-kómáná CL1-boy ‘boy’ kà-kómáná CL12-boy ‘small (puny) boy’

rù-kómáná CL11-boy

‘thin skinny boy’ chì-kómáná1 CL7-boy ‘small (sturdy) boy’

To address the problem of the place of evaluative morphology within the system of Italian morphology, Scalise (1984, 132-133) proposes six criteria distinguishing evaluative morphology from derivational and inflectional morphology: a) They change the semantics of the base (e.g. lume–lumino; ‘lamp– little lamp’). b) They allow the consecutive application of more than one rule of the same type, and at every application the result is an existent word (cf. fuoco–fuocherello–fucherellino ‘fire–little fire–nice little fire’). c) They are always external with respect to other derivational suffixes and internal with respect to inflectional morphemes (e.g. contrabbanderieculo “small time smuggler” = Word contrabbando “contraband”)+Derivational suffix (-iere “agentive”)+Evaluative suffix (-ucolo “pejorative“)+Inflectional morpheme (-i “masc., pl.”). d) [T]hey allow, although to a limited extent, repeated application of the same rule in adjacent cycles (e.g. carinino “nice+dim. +dim.”). e) They do not change the syntactic category of the base they are attached to. f) They do not change the syntactic features or the subcategorization frame of the base. Scalise (1984, 133), regarding the six criteria, concludes: While the properties in (a) and (b) would lead us to group the evaluative suffixes with derivational affixes, the properties in (e) and (f) would lead us to group them with inflectional affixes. The properties in (c) and (d), however, are different from those of both derivational and inflectional affixes. On the basis of these last two properties we can conclude that the evaluative suffixes cannot, in fact, be grouped with either of the other types of affixes.

Within the framework of level-ordering morphology he places a separate block of evaluative rules between derivational and inflectional


27

rules, implicitly supporting option 3 of the four possibilities mentioned above: Figure 1-1 The place of Evaluative Morphology according to Scalise (1984)

While this idea raised considerable attention, it also aroused criticism by indicating the weaknesses of Scalise’s proposal. Stump (1993) points out several difficulties: • The model presumes that inflectional rules always apply after all word-formation rules. • The properties of evaluative rules are not so distinctive so as to unquestionably ‘deserve’ this special position. • The model does not explain why evaluative rules display precisely the cluster of six criteria. • Criterion c) in Scalise’s list is plainly falsified in many languages. • Criterion f) is sometimes falsified. Stump (1993, 6) proposes an alternative conception of morphology according to which “the special properties of evaluative rules follow directly from their status as Category-Preserving Rules” and suggests renaming Scalise’s properties of evaluative affixes as evaluative rules. Within his theory of Partial Lexical Rules (1991) Stump gives the following classification of morpholexical rules:4 4

Root and words are understood as follows: “The ROOT of a lexeme is the base form on which the (non-suppletive) forms in that lexeme’s inflectional paradigm are constructed. Ordinarily, a lexeme will have a single root, though in cases of heteroclisis, two or more roots may be associated with a single lexeme. WORDS,

28

Chapter One

(a) Inflectional rules: (i) root-to-word (ii) root-to-stem; stem-to-stem; stem-to-word (iii) word-to-word (b) Category-changing rules of derivation and compounding: (i) root-to-root (ii) word-to-root (c) Category-preserving rules of derivation and compounding: (i) headroot -to-root (ii) headword -to-word The difference between (b) and (c) is that while category-changing rules inflict all new specifications (including specifications on syntactic category) on their output, category-preserving rules “allow their output to inherit at least (a) its syntactic category and (b) one of its morphosyntactic feature specifications from its derivational base (or from its head, in the case of category preserving compounding); in at least some instances, they allow their output to inherit all of its morphosyntactic or purely morphological feature specifications from its base” (1993, 19). Stump concludes that rules of evaluative morphology in general belong to the category-preserving type, and he identifies two types of categorypreserving rules: 1. headroot-to-root; 2. headword-to-word. The first rule produces roots from roots. The second rule produces both roots from roots and fully inflected words from other fully inflected words. As soon as rules of evaluative morphology are covered by category-preserving rules of derivation and compounding, their properties can be predicted, and there is no need to list them as Scalise did, nor is there any need to assign the evaluative rules a peculiar position of third morphology. Evaluative rules share their properties with the category-preserving rules of derivation and compounding. Consequently, Stump includes evaluative morphology in the field of derivation. Bauer (1997, 541), too, comments on Scalise’s six properties with the aim of identifying the place of evaluative morphology on the inflectionderivation axis. He maintains that Scalise’s points (b) and (d) make EM unlike either derivation or inflection; (e) and (f) make it unlike derivation; (a) could be like either; and (c) makes EM a marginal member of both categories. Later, in his (2004) paper, Bauer returns to the discussion of the scope of inflection and derivation by reflecting on various approaches in contrast, are the fully inflected forms occupying the different ‘cells’ in a lexeme’s paradigm” (1993, 18).


29

and delimitations, e.g. that of Plank (1984) and Anderson (1982). Furthermore, taking no preliminary stance as to whether morphology is derivational or inflectional, transpositional, valency-changing, evaluative and Lexicon-expanding, Bauer identifies and compares categories from five points of view. The overview is provided in Figure 1-2 (2004, 284): Figure 1-2 An overview of categories according to Bauer (2004) Contextual

Inherent

ValencyEvaluative Transpositional Changing Agreement No agreement Lexeme-maintaining Creating new lexemes Class-maintaining Class-changing Grammatical Lexical Paradigmatic

Lexiconexpanding

Nonparadigmatic

The chart breaks the prototypical understanding of inflection and derivation and gives evaluative morphology a specific position. Although it is not a position between inflectional and derivational morphology (in contrast to Scalise’s view), it confirms the existence of evaluative morphology and assigns it a unique position. In particular, Bauer speaks of the “outlandish” nature of evaluative morphology when he notes that “evaluative morphology is placed awkwardly here, since it is typically class-maintaining, though it can be class-changing” (2004, 286).5 Similarly to Bauer, although from a different perspective, Beard (1995) also speaks of a unique position of evaluative morphology. He distinguishes three types of derivational rules: feature value switches, expressive derivations and functional L-derivations.6 Expressive derivations “reflect the attitude of the speaker toward the size and merit of the referent” (ibid.). According to Beard (1995, 163), they reflect at least five functions universally: diminutive, augmentative, Pejorative, Affectionate and Honorific. More particularly, they

5

Apparently, Bauer is not right in saying that evaluative morphology does not have agreement. The opposite is illustrated in Walman examples (6) and (7) above. As stressed by one of the reviewers of the first version of the manuscript, there are many languages in which evaluative morphology triggers agreement. 6 “L-derivation rules might be called Category Feature Rules since they operate on the grammatical level” (Beard 1995, 52). They are the morphological operations of the Lexicon.

30

Chapter One

• do not change the meaning or lexical class of the lexemes over which they operate; • generate nouns from nouns, verb from verbs and adjectives from adjectives; • express prejudices of the speaker as to whether the referent is smaller, larger, more likeable, or more threatening than other members of its semantic category; • are always optional and subjective; • appear recursively; • are not binary but gradient, capable of marking degrees. In Beard’s understanding, the core of evaluative morphology, diminutiveness and augmentativeness, is included in a specific type of rules labelled “expressive derivation” which is projected onto its unique position, and he concludes that “[m]uch more remains to be said about this class of derivations, but at this point it remains a mysterious type” (1995, 175). Szymanek (1988) adheres to the more traditional view of evaluative morphology by postulating its derivational nature, but is more cautious in this respect and places evaluative morphology to the expressive periphery of derivation which does not correlate to any cognitive category (1988, 106-109). Even more radically, Carstairs-McCarthy (1992, 107) assumes that expressive affixes behave “differently in some respects from other kinds of derivational morphology”. The discussion of the place of evaluative morphology greatly benefited from the impact of the work of a non-morphologist. The idea of Rosch (1976, 434) that there is “considerable evidence that some natural categories are continuous rather than definitively bounded entities” and that “some natural, continuous categories seem to be structured cognitively into items which differ in their degree of prototypicality” has had considerable influence not only upon psychology, the main field of her research activities, but also upon the whole range of natural and human sciences, including linguistics. The idea of fuzziness turns up, for example, in Lakoff (1973, 459), who exemplifies it by the “birdiness” hierarchy: robins are more bird-like than eagles, eagles more so than fowl (like chickens, ducks and geese), chickens more so than penguins and pelicans, and penguins more so than bats. Plank (1994), Dressler (1989), Bauer (1988) and others introduce the idea of prototypical nature of inflection and derivation. Dressler (1989, 5) assumes that the classification of morphological categories into inflectional and derivational cannot be done in a discrete way. Similarly, Stump (1993, 22) maintains that “inflectional


31

rules, category-changing rules and category-preserving rules are not assigned to separate components, but are instead free to interact with one another to the extent allowed by the restrictions specifying the root, stem, or word status of their input and output”. In principle, a number of morphologists, making use of the invaluable contribution of cross-linguistic typological research, have come to the conclusion that “the inflection/derivation distinction is not absolute but allows for gradience and fuzzy boundaries” and that “we are dealing with a continuum from clear inflection to clear derivation with ambiguous cases in between” (Haspelmath 1996, 47). This view is also supported by Štekauer (forthcoming) who – based on a cross-linguistic analysis of various reasons for the fuzziness of the inflection/derivation border and of various criteria proposed for the distinction of the two morphological areas – concludes that “it is not possible to draw a clear-cut borderline between inflection and derivation and that the relation between these two areas of morphology is best treated ‘as a cline rather than a dichotomy’ …, with prototypical cases at both ends of the cline”. It goes without saying that the awareness of the cline-like nature of the derivation–inflection relation could not but be reflected in the discussion of the place of evaluative morphology. From this point of view, evaluative morphology as a part of the morphological system, sharing (from a crosslinguistic point of view) the properties of both derivational and inflectional parts of the cline, should not be treated as a unique morphological level, as the third morphology. This position gets further support from the varying situation in individual languages. Consequently, it may be assumed that EM lies within the limits of the inflection-derivation continuum. This is most obvious when one examines this area of linguistics from a crosslinguistic perspective. But evaluative morphology is not the only example of fuzzy boundaries between two categories which were treated in the past as disjunctive categories. Let us, for example, recall Beard (1982) who gives a number of arguments suggesting that the category of number, traditionally considered as an inflectional category, behaves in many respects as a derivational category. However, even more important in our context is Beard’s idea of the common foundations of inflection and derivation in IE languages. Beard (1995) argues for their parallel typological and historical development, maintaining that L(exical)derivations and inflection may express the same categories, but, on the other hand, that the same morpheme may mark derivational and inflectional categories sharing identical functions (no doubt a factor that has contributed to the inflection–derivation cline in general, and in evaluative morphology in particular).

32

Chapter One

1.3.3 Semantics of evaluative morphology Evaluation implies expression of a judgment about the evaluated object. There is an evaluator who evaluates and an evaluated object. The object is assessed according to the standards and traditions (default value) established in the particular speech community traditions; according to the evaluator’s education, profession, generic knowledge and experiences; and, certainly, based on a specific situational and/or linguistic context. From the pragmatic perspective, diminutives and augmentatives express an evaluation “according to the evaluator’s intentions, perspectives and standards of evaluation” (Dressler and Merlini Barbaresi 1994, 153). All these factors contribute to the relativism of evaluation by means of diminutives, augmentatives and other evaluative categories. However, it should be noted that this relativism, which is well illustrated in the Supriyanto’s “mouse” example (mentioned in the Introduction), is not only a matter of the pragmatic facet of evaluative morphology. Lexical semantics in general offers numerous instances and extensive discussion of this sort of relativism, which usually takes the “predicate and context dependent standard of comparison (Kennedy 2007, 3) as a reference point.7 As such, any evaluation can be either descriptive/objective or qualitative/subjective. The process of evaluation, appreciation, and assessment is a mental operation in which the value of the evaluated object is stated. Traditionally, evaluation is connected to diminutives and augmentatives. While they are core evaluative categories, the scope of evaluative morphology is not restricted to them. This was observed many decades ago. For example, Dokulil (1962) considered evaluative formations a modificational category. Modification means that the original meaning is semantically modified, usually enriched. Modifications can be of various sorts and Dokulil (1962, 46) distinguishes the following modificational onomasiological categories: • Diminutive onomasiological category – the concept is modified by the diminutive marker.

7

For illustration, Kennedy (2007), which provides a number of references to the former research, is an analysis of the relativity of the meaning of the positive form of gradable adjectives in sentential context such as “expensive” in the sentence “The coffee in Rome is expensive”. The truth condition of such a statement is, obviously, relativized by the context. On this topic, see also Bierwisch (1989), Rett (2008), and Neeleman et al. (2004).


33

• Augmentative onomasiological category – understood as counterpart to the diminutive onomasiological category: the concept is modified by an augmentative marker. • Derivation of female names from male names and vice versa – the concept of animate being is modified by a gender marker. • The young of animate beings – the concept is modified by marker with the meaning “not grown-up”. In any case, an impetus for a widespread and serious discussion was given much later by Jurafsky (1996) – one of the most elaborated models dealing with the semantics of diminutives (but not augmentatives). Jurafsky proposes a model of synchronic and diachronic semantics of the diminutive category and applies Lakoff’s radial category model of semantics. His proposal is a type of structured polysemy “that explicitly models the different senses of the diminutive and the metaphorical and inferential relations which bind them” (Jurafsky 1996, 533). From the synchronic point of view, the model illustrates numerous and, sometimes, contradictory senses of diminutiveness. From the diachronic point of view, the model expresses the generalizations of the classical mechanisms of semantic change (metaphor, abstraction and inference). Jurafsky also adds a new mechanism of semantic change, lambda abstraction, which “accounts for the rise of qualification meaning and second-order predicates in the diminutive” (ibid.). As such, the model acts as a kind of archaeology of meaning (1996, 536). Figure 1-3 Jurafsky’s synchronic-diachronic model of the semantics of diminutives according to Jurafsky

Chapter One

34

Jurafsky (1993, 542) claims that the central category of diminutiveness, i.e., “child”, is historically prior to the other senses and therefore motivates them metaphorically and inferentially. In the presented model, Jurafsky distinguishes four mechanisms of semantic change: Metaphors. There are two types of semantic change expressed by metaphors – metaphors for gender and metaphors for centrality and marginality. Metaphors for gender. Jurafsky points out two paradoxes. The first is the linking of female gender with both the diminutive and the augmentative cross-linguistically. The second is the asymmetric use of diminutives and female augmentatives for body parts. The link between female gender and diminutives/augmentatives arises from various sources. In Romance languages, for example, the link results from conflation of the Latin collective suffix with the feminine suffix, based on similar realization. In the languages of Southeast Asia, a morpheme originally meaning “mother” has grammaticalized to the augmentative throughout the region (Matisoff 1991). In Indo-European and Afro-Asiatic languages the same morpheme is used for female markers and diminutive markers. At the same time, from the pragmatic point of view, there is a strong tendency typical of women to use diminutives. Fundamental metaphors for gender are, e.g. ORIGINS ARE MOTHERS; IMPORTANT THINGS ARE MOTHERS (augmentatives); WOMEN ARE CHILDREN/SMALL THINGS; SMALL THINGS ARE WOMEN (diminutives). Metaphors of centrality and marginality. They explain the second paradox mentioned in A. Diminutive markers can be used to express intensification (10a) and approximation (10b): (10) a. Mex. Spanish b. Dom. Spanish

ahor-ita now-DIM ‘immediately’ ahor-ita now-DIM ‘soon’

Diminutive markers can also mark the centre, or prototype, of a social category (Japanese edo ‘Tokyo’, edokko ‘Tokyoite’ and the social marginal Fuzhou Chinese huaŋ-ŋiaŋ ‘foreigner’). Jurafsky (ibid.) conveys three metaphors responsible for these semantic changes: SOCIAL GROUPS ARE FAMILIES (the group member in the source domain corresponds to the child in the target domain), CATEGORY


35

CENTRALITY IS SIZE (links central, or prototypical, members of a category to large size, and peripheral or marginal members of a category to small size), MARGINAL IS SMALL. Conventialization of inference. The construction with a diminutive marker is naturally associated with certain inferences such as affection for children and prototype exemplars of small objects. Such a construction can be lexicalized. It is the case of classificatory diminutives – an object denoted with a diminutive construction is a small object belonging to the same semantic field as the larger object, for example, booklet. Generalization or bleaching. An example of this semantic change is the English suffix -ish (boy-ish). Boyish is less specific than boy: its meaning is more abstract. Lambda-Abstraction-Specification refers to the sense of approximation. The direction in which the diminutive modifies the predicate depends on the direction of the relevant scale. If there is a scale y with a point x and this point is diminutivized, the resulting meaning is lower than x on y. In the scale of colours there is green. Greenish is less: it is lower than green. Lambda-abstraction takes one predicate in a form and replaces it with a variable. The resulting expression is now a second-order predicate, since its domain includes a variable which ranges over predicates. For the diminutive, this process takes the original concept ‘small(x)’, which has the meaning ‘smaller than the prototypical exemplar x on the scale of size’, and lambda-abstracting it to ‘lambda(y) (smaller than the prototypical exemplar x on the scale y)’ (Jurafsky 1996, 555).

Jurafsky assumes that the morphological category of diminutiveness originates in the sense of “child”. He considers both semantic and pragmatic origins of diminutive morphemes in order to show that in “almost every case in which a historical origin can be determined for a diminutive morpheme, the source was either semantically related to ‘child’ (e.g. a word meaning ‘child’ or ‘son’), or pragmatically related to ‘child’ (e.g. a hypocoristic suffix on names)” (1996: 562).8 This idea has been supported by many others (e.g. Matisoff, 1991, Bakema and Geeraerts 2000). Matisoff (1991, 303) maintains that while “child” is the source of diminutive markers, “mother” is the source of augmentative morphemes in 8

Cf. Dressler and Merlini Barbaresi (2001) for a criticism of Jurafsky’s childbased radial model based on cross-linguistic data, principles of language acquisition as well as available diachronic data. Fortin (2009) also provides some counter-examples in the field of Spanish expressive adjectives.

Chapter One

36

SE Asian languages. He gives, among others, the following examples from Thai: (11)

hŭa-mɛ̂ɛ-tiin head-mother-foot ‘big toe’

(12)

lûuk-mɛɛw DIM-cat ‘kitten’

This has also been confirmed by my sample data. Examples (13), (14), and (15) show that this phenomenon is not restricted to SE Asia: (13) a. Bafut b. Bafut

mu-nili’i child-eye ‘little eye’ ma-nili’i mother-eye ‘big eye’ (Tamanji, pers.comm.)

(14) Kwaza

dara-to’hoi green-CL.child ‘baby’ (van der Voort 2004, 172)

(15) Nelemwa

axoomo naabuc mother-mosquito ‘big mosquito’ (Brill, pers.com.)

While not a universal feature of the semantics of evaluative formations, Mutz (forthcoming) seems to be right when saying that “[c]rosslinguistically, the semantic source of the diminutive and augmentative meaning tends to be an anthropocentric one”. Matisoff’s observation of the “mother” source for augmentatives in many languages obtains an interesting sort of support from diachronic cognitive semantics. The available analyses indicate frequent pejoration of female-specific words, especially those in the sematics of which an element ADULT is discernible. This, then, may account for the frequent pejoration of augmentatives whose source is one particular type of female adult being–mother (Kleparski 1986, 1996, 2000, 2004, Kochman-Haladyj 2008, Wlodarczyk-Stachurska 2010).


37

Jurafsky’s seminal paper with his radial model (1996) meant an impulse to more intensive research into the semantics of evaluative formations. One of the more attractive contributions in this field comes from Prieto (forthcoming) who evaluates diminutives and augmentatives within the framework of cognitive linguistics, and provides physiological, psychological, biological (ontogenetic) and neurological evidence for the claim that the core concept of evaluative morphology is that of size. He maintains that owing to the biological and psychological importance of the concept of size, the cognitive foundation of the evaluative meaning is the semantic primitive “big” for augmentatives and “small” for diminutives. Furthermore, the evaluative meaning is physiologically grounded via our embodied experience. Prieto points out in this connection that evaluative meanings correspond with the general principles of phylogenetic and ontogenetic development from concrete thinking to abstract thinking and from concrete concepts to abstract concepts. In evaluative formation, this process is manifested in the direction from a more quantity-based primitive notion towards a more quality-based abstract concept. Based on these principles and on the Lakoffean idea of metaphorical and metonymical chaining, Prieto proposes two radial models. While Jurafsky’s model is confined to diminutives, Prieto is more comprehensive and provides separate models for diminutives and augmentatives. The core semantic category of the former is “littleness” and that of the latter is “bigness”: Figure 1-4 Prieto’s radial model for diminutives

38

Chapter One

Figure 1-5 Prieto’s radial model for augmentatives

Mutz (forthcoming), who, like Jurafsky, takes the diachronic perspective as a starting point for the discussion of the semantics of evaluatives, points out that Jurafsky’s model does not work universally, and that, for example, many diminutive suffixes in Indo-European languages have developed from derivational affixes meaning “related to” or “similar to” as a result of refunctionalization (ibid.). She therefore provides a modified model which shows “the network-like panchronic category of the diminutive, whereby the synchronic links between the senses only partially correspond to unidirectional diachronic paths”. Grandi (2005) maintains that the meaning of evaluation can be represented by four semantic primitives: small, big, bad, good. In the process of evaluation, the evaluative morphological marker changes/modifies the basic meaning: it adds a semantic component by which it differentiates the evaluated object from the standard: Figure 1-6 Grandi’s semantic primitives for EM

STANDARD VALUE +

GOOD BAD SMALL BIG

Grandi (2005) assumes that the evaluated base and the new evaluating construction usually have the same referent. Between the evaluated base and the new evaluating construction there is the relationship of hyponymy. Moreover, a construction is evaluative when it meets two criteria, one related to the semantic level and one related to the formal level. The first criterion indicates that the evaluative construction “has the function of assigning to a concept a value different from that of the ‘standard’ (within the semantic scale to which it is a part of), without resorting to no parameters of reference external to the concept itself” (2005, 188). The


39

second criterion suggests that an evaluative construction should include explicit expression of the standard value and an evaluative marker. Grandi (2011a) combines the descriptive and the qualitative perspectives and semantic operations assigned to them with a semantic scale that has a positive and a negative pole. From a descriptive perspective, a shift towards the positive end indicates growth, intensification of the actual feature (e.g. physical dimension: gatto ‘cat’ > gattone ‘big cat’). A shift towards the negative end results in decrease of the actual feature (gatto ‘cat’ > gattino ‘small cat’). The combination of the qualitative perspective and a shift towards the positive or negative end indicates the feelings of the speaker. It should be stressed, however, that the positive/negative evaluation without involvement of the feelings of the speaker is not infrequent. An interesting example comes from two major dialects of Portuguese. As noted by Pöll (forthcoming) whereas mulherão means ‘big and fat woman’ (with the expected pejorative nuance) in European Portuguese, the same word has a positive meaning ‘particularly pretty and attractive woman’ in colloquial Brazilian Portuguese, this being a clear system-level difference. Grandi adopts the idea of Wierzbicka (1984, 108) who assumes that “the set of universal semantic primitives must be included in the set of concepts which have been lexicalized in all languages,” and summarizes his ideas in Table 1-6 below where small and big are semantic primitives representing the descriptive side of evaluation, and good and bad represent the “qualitative” (discourse) side (Grandi: 2011a): Table 1-6 Different perspectives of evaluative morphology

Shift towards the ‘+’ end Shift towards the ‘-’ end

Descriptive perspective Prototypical augmentatives Prototypical diminutives

Qualitative perspective Prototypical amelioratives Prototypical pejoratives

There is no doubt that various shades of meaning have been identified in traditional morphology and lexicology and labeled with diverse semantic category names such as hypocoristics (endearments), pejoratives, depreciatives, intensives, attentuatives, etc. Szymanek (1988, 108) is, however, rather sceptical about these labels and assumes that “[t]hese putative categories are not accountable for in cognitive terms. Besides, certain recurrent clusters of emotive, evaluative and/or attitudinal content

40

Chapter One

are simply hardly nameable, and cannot be referred to by means of single, general labels” (ibid.). A diminutive, for example, may refer to an object smaller than the default quantity for its category, e.g. Slovak dom ‘house’ > domček ‘house-DIM’. However, it can also express the speaker’s attitude to the object of extralinguistic reality. This can be either affectionate or pejorative. The diminutive domček in the sentence Bývame v žltom domčeku na kopci (‘We live in a yellow house-DIM on a hill’) can refer to the small size of the house and also to the endearing attitude of the speaker toward the house while the house itself can exceed the default size. The sentence To bývate v takom domčeku? (‘You really live in such a houseDIM?’) expresses the speaker’s disgust and negative attitude irrespective of the size of the house. If it is the size of the house that makes the speaker use a diminutive, its meaning is purely denotative, lacking any personal assessment. In the second case, it is not only the referent that influences the speaker’s utterance – it is his/her attitude to and/or expectations concerning the house, and the meaning refers to the quality “other than standard” in the value system of the speaker. The semantic interpretation is influenced by the speaker’s attitude. This example shows that while Grandi’s approach aims to capture the meaning of evaluative constructions comprehensively, i.e., by combining the quantitative and the qualitative criteria, it raises the question of what belongs to objective (quantitative) and what to subjective (qualitative) evaluation. The borderline is usually very difficult to draw due to the fuzzy borders and the interwovenness of the denotative and the connotative meanings. From the communicative point of view, it may be assumed that the vast majority of uses of evaluative formations represent a combination of these two aspects of evaluation, and that they depend on the specific speaker and the context. It should be stressed, however, that a purely denotative (not lexicalized) use of morphological evaluatives (“uncontaminated” by subjective/qualitative evaluation) is possible. To illustrate, Slovak diminutive loptička ‘little ball’ refers to any (soccer) ball which is smaller than the default ball. Here the meaning of loptička is not lexicalized – unlike its next meaning – ‘tennis ball, table-tennis ball’ where it is lexicalized. In addition, evaluative formation is not only the matter of nouns. There are many actions of diminutive/augmentative nature (less/more than the default quantity of action) which seem to be purely quantity-based such as Slovak skackať ‘to perform many very small jumps’. The previous discussion showed close interrelation between semantics and pragmatics. While the majority of morphologists and semanticists derive pragmatic meaning from the semantic one (e.g. Jurafsky 1996,


41

Mutz, forthcoming), Dressler and Barbaresi (2001, 43) maintain that pragmatics is superordinate to semantics and propose the pragmatic feature [fictive], which, in their view, is more basic than the semantic features [small] and [big].

1.3.4 Semantics of evaluative morphology: a new proposal This section proposes a new approach to the semantics of evaluative morphology in which evaluative morphology is treated as a continuum such that prototypical cases express the meaning of quantity under or above the default value. In this connection, let us refer, for example, to Apresjan’s discussion (1992: 373ff.) where, in reference to antonymous adjectives, he speaks of the deviations from the norm such as “greater than the norm” (Russian boľšoj ‘big’) and “less than the norm” (Russian malen’kij ‘small’). This principle is applied by Apresjan to other wordclasses, too. In my approach, however, evaluative morphology is founded in the “supercategory” of Quantity that includes not only the core areas of evaluation (diminutiveness and augmentativeness) but also other semantic categories. These categories include various manifestations of Aktionsart and pluractionality whose concepts of multiplicity, iterativity, frequentativity, distributiveness, etc. are of quantitative nature (cf. Štekauer et al. 2012). This approach also encompasses the semantic categories of attenuation (deintensification) and intensification. As indicated above, the idea of the deviation from the norm, from the default value, is not new. What is new in my approach is that the reference point, i.e., the standard or default value, is anchored in the fundamental cognitive categories SUBSTANCE (including human beings, material objects), ACTION (including processes, actions proper and states), QUALITY (including properties, features, characteristics), and CIRCUMSTANCE (location, time, manner of action, cause of action, etc.). This approach is in accordance with the cognitive approach to semantics, suggesting that meaning is constructed at the conceptual level. In Prieto’s (forthcoming) words, conceptualization is “a linguistic process in which linguistic forms such as EVALs [evaluative affixes] are the ‘opening door’ for ‘a room’ full of conceptual operations (physical diminutivisation, endearment, pejoration, and so on in the case of DIMs [diminutives])…” In other words, Prieto understands evaluative meanings as a conceptual structure, that is, the conceptual structure of size evaluation. I assume that the key feature of evaluative morphology is the capacity of a language to express morphologically the semantics of “less than / more than the default quantity”, with the concept of default quantity being

Chapter One

42

a relative one: the default (system-level) value is language specific, influenced by many factors such as culture, traditions of a speech community, and one’s knowledge and experience. The meaning of the other-than-default quantity can pertain to any of the aforementioned cognitive categories of SUBSTANCE, ACTION, QUALITY, AND CIRCUMSTANCE. By implication, the specific value of standard quantity and any deviations from it may bear on the quantity of both physical and abstract objects, the quantity of actions, processes and events, the quantity of quality and features, and the quantity of particular circumstance. This establishes four basic categories of evaluative morphology, in particular, the Quantity of Substance (16), the Quantity of Action (17), the Quantity of Quality (18) and the Quantity of Circumstance (19). These cognitive categories may be expressed by nouns, verbs, adjectives, adverbs, and also pronouns. (16) a. English b. Czech c. Welsh d.Wanano e. Imonda f.Tapiete

(17) a. Saami

star-let star-DIM ‘little star’ (my example) hvězd-ičk-a hvězd-DIM-NOM.SG.F ‘little star’(my example) llyfr-yn book-DIM s‘mall book’ (Awbery, pers.com) tʉ--̴ka stick-DIM ‘little stick’ (Stenzel 2004, 167) pfeia-tõ stone-DIM ‘small stone’ (Seiler 1985, 57) ɨwɨra-mi stick-DIM ‘little stick’ (González 2005, 193) vourdit vourdit-DIM wait-DIM ‘wait for a while’ (Sammallahti 1998, 116)


b. Belorussian

43

папа-гуляць papa-guľac FREQ-walk ‘to walk a lot’ (Rudenka and Rudaya, pers. com.) c. Provencal tre-muda AUG-change ‘change a lot’ (Blanchet, pers.com.) mehci-ne-hs-ə̆-w-à d. Passamaquoddy end-die-DIM-AI-(3)-3ABS ‘the unfortunate one has died’ (LeSourd 1995, 108) e. Montana Salish esl*n’an’aq’wi9 ‘he’s stealing just a little’ (Thomson, pers. com.) f. Arabana nhukathirnda yanhi-mda-pa-rnda excessively .speak-SP-INT-PRS ‘He talks too much altogether’ (Hercus 1994, 203) (18) a. Basque b. Tatar c. Udi d. Tzutujil e. Djaru

9

handi-handia handi-RDP ‘very big’ (Hualde and de Urbina 2003, 462) küg-elcem blue-ATT ‘blueish’ (Miftachova, pers. com.) maˁjin-ǯä black-ATT ‘blackish’ (Ganenkov, pers. com.) kaqk-oj white-ATT ‘whitish’ ( Dayley 1985, 221) jambi-jambi big-RDP ‘very small’ (Tsunoda 1981, 234)

From esna*q’wi ‘he’s stealing’.

Chapter One

44

(19) a. Hungarian

b. Udmurt c. Anamuxra d. Agta e. Italian f. Slovak

lassacskán lassan-cskán slowly-DIM ‘very slowly’ (my example) êuog-ak fast-AUG ‘very fast’ (Rodionova, pers. com.) sarwe~sarwe quickly~quickly ‘very quickly’ (Ingram 2001, 105) fu~fúráb RDP-afternoon ‘late afternoon’ (Healey 1960, 9) ben-in-o well-DIM-NOM.SG ‘quite well’ (Grandi, pers. com.) tepl-učk-o teplý-učk-o warm-DIM-ADV ‘pleasantly warm’ (my example)

This conception of evaluative morphology can be represented by the following dynamic model of evaluative formation: As illustrated in Figure 1-7, the process of evaluation starts in extralinguistic reality. The point of departure is the need of (a member of) a speech community to evaluate an object of extra-linguistic reality. This need is reflected at the cognitive level. The process of evaluation starts with quantification in terms of the basic cognitive categories (Quantity of Substance, Quantity of Action, Quantity of Quality, and Quantity of Circumstance). If there is a need for qualitative evaluation, based on the metaphorical shifts SMALL IS CUTE and BIG IS NASTY, the evaluation takes the qualification path. This account is given support from Mutz (forthcoming), a diachronic and cross-linguistic analysis of evaluative categories. To illustrate, the Slovak diminutive equivalent of ‘cat’ is mačička. It can refer not only to size ‘little cat’ and age ‘kitten’, but also to qualities like tenderness, beauty, or cuteness. At the level of the language system, cognitive categories are expressed by semantic categories like diminutive, augmentative, pejorative, ameliorative, pluractionality, attenuation, intensification, Aktionsart, etc. Concrete realization of these semantic categories comes into existence by means of the markers of evaluative


45

morphology. If needed, the final evaluative construction undergoes phonological changes. The output leaves the level of langue and enters the level of parole where it can obtain various additional shades of emotive colouring, depending on the specific context, e.g. admiration, contempt, etc. Figure 1-7 Model of evaluative word formation

46

Chapter One

How can this model be projected onto a radial model of EM semantics? In his radial model, Prieto (forthcoming) aptly replaces Jurafsky’s central category child with the broader category LITTLE. It is proposed here that the central category little be substituted by an even much broader category of QUANTITY. The reasons are twofold. First, while Jurafsky’s model is confined to diminutives, Prieto provides two separate radial models for diminutives and augmentatives. Thus, none of these authors treats diminutives and augmentatives as two central categories of evaluative morphology in a unified fashion. This comes as a bit of a surprise because diminutives and augmentatives can be viewed as two poles of a quantitatively defined EM cline, related via a common reference point, in particular, the default value. This default value may be viewed as a prototypical exemplar in regard to quantitative evaluation. Second, while Prieto concentrates on the size of objects, quantitative evaluation with regard to the prototypical default value in the form of various EM meanings (frequentativeness, intensity, duration, distribution, attenuation, exaggeration, approximation, size, social position, etc.) is performed for both diminutives and augmentatives within the scope of the above-mentioned cognitive categories of SUBSTANCE, ACTION, QUALITY and CIRCUMSTANCE. In other words, various EM meanings radiate from each of the four cognitive categories. Obviously, individual languages differ in concrete implementation of these cognitively founded options. What is proposed here is to take the prototypical, quantitatively defined default value as the central reference point of a radial model of EM semantics. Diminutives and augmentatives are then viewed as deviations from the prototypical value in any of the cognitive categories (Figure 1-8). It may happen that a particular EM meaning is implemented within two different cognitive categories such as attenuation, which can take the form of a reduced QUALITY (smallish, reddish) as well as reduced ACTION (Slovak skackať ‘to perform very small jumps’ – smaller than the prototypical default value). An interesting sort of support to the unified model of EM semantics comes from cases like the above-mentioned Slovak example skackať ‘to perform many very small jumps’ which indicates not only diminutiveness of ACTION (jumps smaller than the referential default value), but also, simultaneously, augmentativeness of action (iterativity–more than the default value of one jump).


47

Figure 1-8 Radial model of EM semantics10

A few examples will illustrate this point: The adjective blackish represents the semantic category of Attenuation within the cognitive category of QUALITY and deviates from the default value by indicating a lower level of blackness. Slovak pomaličky ‘slowly-DIM’ is an adverb expressing the semantic category of Manner of Action within the cognitive category of CIRCUMSTANCE and indicates the level under the default value, i.e., ‘less than slowly’. The reduplicated pluractional verb tsO-mO~tsO-mOi ‘turn over and over, many things in many places’ (Kropp-Dakubu, pers. com.) from the Ga language realizes the semantic categories of Iterativity and Distributiveness within the cognitive category of ACTION and indicates the higher quantity of action than the default one (one action aimed at one thing at one place). The Khwarshi reduplicated adjective ungo-ungoyab ‘really real’ (Khalilova, forthcoming) realizes the semantic category of Intensification within the cognitive category of Quality by indicating a higher quantity of a quality. Taking the Qualification path, Slovak spinkať ‘sleep-DIM’ realizes the semantic category of hypocorism within the cognitive category of Action. 10

Unlabelled lines indicate other possible meanings.

CHAPTER TWO RESEARCH

2.1 Objectives Given the absence of any large-scale cross-linguistic research into Evaluative Morphology, the purpose of my empirical research is to identify universal tendencies and examine the nature of evaluative morphology, its semantic and formal characteristics from the perspective of the theoretical principles outlined in the preceding section and assess the relevance of Evaluative Morphology to the concept of Standard Average European by analysing and comparing the SAE sample and the world sample.

2.2 Method of research My research into evaluative morphology of both the world languages and SAE languages consisted of several stages: • Preparing a data sheet (Appendix 1). • Language sampling: basically, the principle of convenience (e.g., Bauer, 1997) and the principle of random sampling (e.g., Bakker 2013) were followed. The proportionality criterion was applied in compiling the world sample. • Data acquisition: the data was obtained from informants (experts in morphology and/or experts in a particular language) or data sheets were completed on the basis of reference books. Whenever possible both methods were combined. • First analysis of the data sheets: reliability criterion. If needed, informants were contacted again for further explanation. If the questionable data was not verified, the data sheet was excluded from further analysis. If possible, a new source of information for the particular language was identified. • Second analysis of data: calculation of saturation values. • Evaluation and comparison of the data: world sample vs. SAE sample.

Research

49

2.2.1 Preparing a data sheet The data sheet proposed for this research consists of four parts (see Appendix 1): 1) General 2) Word-formation 3) Cognitive categories 4) Word classes General data The aim of the introductory part of the questionnaire was to specify: • the source of the data – either a printed source, a website, or the name and email of the informant); • language specification – language family, genus and its geographical area Word-formation data The word-formation part consists of two sections: a section devoted to diminutives and a section devoted to augmentatives. The structure of both of these sections is the same. The informants were asked to provide data on word-formation processes used in the evaluative morphology of the specific language. The following word-formation processes were included: prefixation, suffixation, infixation, root-and-pattern derivation (transfixation), compounding, incorporation, reduplication, and conversion. In addition to a box on the questionnaire corresponding to each process, an empty box was including to capture any additional data. The informants were also asked to provide an example, the literal translation of constituent morphemes and its English translation. The basic word-formation processes are subdivided into parameters seeking more specific data. For example, the prefixation and the suffixation processes include the following questions: • Does your language have more than one DIMINUTIVE/AUGMENTATIVE prefix/suffix? • Is more than one DIMINUTIVE/AUGMENTATIVE prefix/suffix possible in a word? • Can one and the same DIMINUTIVE/AUGMENTATIVE prefix/suffix be used more than once in the same word? • Are there DIMINUTIVE/AUGMENTATIVE prefixes/suffixes that can have a meaning other than DIMINUTIVE/AUGMENTATIVE?

Chapter Two

50

In the case of reduplication, the data sheet asks about the type of reduplication used in evaluative formation of a particular language, notably, the complete (20), partial preposing (21), partial postposing (22) and partial infixing (23) types of reduplication. (20) East Dangla

(21) a. Arabana

sooɛ ~sooɛ small~small ‘very small’ (Shay, pers. com.) kalti~kaltiRi RDP~hill

b. Yurakare

‘small hill’ < ‘hill’ (Hercus 1994, 98) maj~matata RDP~big ‘really big’ (van Gijn 2006, 136)

(22) Anamuxra

kuxbai~xbai kuxbai~ RDP ‘very slowly’ (Ingram 2001, 105)

(23) Alawa

rukukalařa ru~ku~kalařa ‘very long’ (Sharpe 1972, 68)

In some of the completed data sheets, especially for SAE languages, markers like mini-, maxi-, super- were included by informants. They were assigned either to the prefixation or to the compounding process. Their inclusion in the data sheets raises certain doubts. These elements also occur in languages that by other criteria appear to be poor at evaluative morphology, e.g. in Danish supermarket or Dutch giga problem. These internationalisms of Latin and/or Greek origin (sometimes also labelled as combining forms of neo-classical compounds – cf., for example, Bauer 1983) may be assumed to be productive in each language, especially in Europe. However, their inclusion in the data sheets was strongly subjective, and heavily depended on the particular informant. To avoid this kind of distortion of facts they were excluded from further research with the exception of Romance languages.

Research

51

Cognitive categories The semantic part of the data sheet draws on the principles of Dokulil’s (1962) onomasiological approach and his use of four fundamental cognitive categories of SUBSTANCE, ACTION, QUALITY and 1 CIRCUMSTANCE. By implication, the data sheet asks for the productive formation of evaluatives within the categories of QUANTITY OF SUBSTANCE, QUANTITY OF ACTION, QUANTITY OF QUALITY and QUANTITY OF CIRCUMSTANCE. Word class data The section on word classes is based on the traditional classification of parts of speech. The basic question is which of the following main word classes benefit from evaluative formation in a given language: nouns (24), verbs (25), adjectives (26), adverbs (27). The possibility of class change due to diminutive/augmentative formation was also examined (28). (24) Tutelo

pha:xti: bpha-xti head-AUG ‘nose’ (Oliverio 1996, 143)

(25) Polish

płakusiać płak-uś-ać cry-DIM-INF ‘cry (affectionate)’ (Kleparski, pers. com.)

(26) English

green-ish (my example)

(27) Welsh

lluny-et far-DIM ‘at some distance’ (Wheeler, pers. com.)

(28) Ladin

stra-vent-é AUG-wind-INF

‘to blow hard (in the context of wind)’ (Siller-Runggaldier, forthcoming)

1

For the meaning of these cognitive categories see Point 1.3.4.

52

Chapter Two

2.2.2 Language sampling One of the basic aims of my research has been to compare two language samples in order to find out whether evaluative morphology in the SAE languages features certain characteristics distinct from those in the languages outside the European territory. A positive reply to this question means that Evaluative Morphology may be considered as one of the defining characteristics of the Standard Average European. World sample The basic reference source for the sampling of languages was the World Atlas of Language Systems. It enables one to combine the data on areal typology with that concerning the genetic classification of languages. Besides the traditional genetic classification into language families, genera etc., the WALS languages are also grouped into macro-areas. WALS identifies six macro areas counting 2488 languages. 590 languages are spoken in Africa, accounting for 24% of the total number of the WALS languages; Eurasia has 368 languages (15%); SE Asia + Oceania 507 (20%); Australia-New Guinea 389 (16%); North America 381 (15%) and South America 253 (10%). In the sampling of world languages the proportionality principle was applied. It means that my sample maps the percentages of the individual macro-areas. For example, the languages spoken in the macro-area of Australia-New Guinea represent 16% of all the WALS languages. My world sample counts 132 languages; 16% of 132 is approximately 22 languages. Therefore, my world sample counts 22 languages spoken in the area of Australia-New Guinea. In some cases the unavailability of (reliable) data has precluded me from keeping to the desired proportionality principle. However, the deviation is never greater than 2-3 languages. The pie charts below illustrate the desired proportion of languages and that achieved:

Research

53

Chart 2-1 A desired proportion of languages

Chart 2-2 An achieved proportion of languages

As the pie charts suggest, the highest discrepancy is in the macro areas of Africa and SE Asia + Oceania. The proportionality principle requires at least 31 African languages. I managed to collect the data on 29 languages. Similarly, while 26 languages are required for SE Asia + Oceania, my sample counts 24 languages. On the other hand, the required number of languages for North America was exceeded by 2 languages.

54

Chapter Two

It should be pointed out that the actual number of completed data sheets was higher (156); however, following the initial analysis some of them were excluded because the data were incomplete and/or were deemed unreliable. In these cases, the criterion of completeness/ reliability overpowered the proportionality criterion. SAE sample The sampling of the SAE languages was based on a much simpler principle: to cover as many languages as possible. A potential source of information for each language from the list of languages mentioned in 1.2.3 was identified. Out of 145 languages on the list, data on 71 languages was collected.

2.2.3 Analysis of the data sheets The analysis of data sheets was based on the following algorithm: IS THERE ANY EVALUATIVE MORPHOLOGY? 1. If no, the saturation of the evaluative morphology is 0. 2. If yes: A. Are there only diminutives? a. If no, proceed to C. b. If yes: What word-formation processes are employed? How are they further specified? What semantic categories are expressed? What word classes are represented? What is the saturation of EM? B. Are there only augmentatives? a. If no, proceed to C. b. If yes: What word-formation processes are employed? How are they further specified? What semantic categories are expressed? What word classes are represented? What is the saturation of EM? C. Are there both diminutives and augmentatives? If yes analyse diminutives as in Ab and augmentatives as in Bb. Count the final saturation by adding up the saturations in Ab and Bb.

Research

55

2.3 Evaluative Morphology Saturation It follows from the algorithm that for the sake of data evaluation the parameter of Evaluative Morphology Saturation (EM saturation) is introduced. EM saturation is a mean of three values: word-formation value (VWF), cognitive category value (VSC) and word class value (VWC). They are numerical representations of productive use of word-formation processes, cognitive categories and word classes in evaluative morphology in a language: (29) SEM = (VWF+VSC+VWC): 3 The word-formation value (VWF) reflects the number of word formation processes and their specification. If the number of wordformation processes used in evaluative formation is 1 or 2, it is evaluated by 1 point; the use of 3 or 4 processes is valued by 2 points etc. Extra points are added for subparameters such as the possibility of repetitive use of evaluative affixes. The cognitive category value (VSC) is based on the presence of cognitive categories in the evaluative morphology of a language. Each cognitive category is assigned one point. If a language, for example, expresses the categories of Quantity of Substance and Quantity of Quality, the presence of 2 cognitive categories is evaluated by 2 points. If 3 categories are expressed, the score is 3, etc. Word classes and the corresponding word class value (VWC) are evaluated analogically. EM saturation is counted separately for diminutives and augmentatives where relevant. The total EM saturation is a sum total of the diminutive (DIM) and the augmentative (AUG) saturations. In the following sections the data acquired for each macro area will be presented. Furthermore, the results will be discussed and the world language sample will be compared to the SAE language sample.

2.4 Results 2.4.1 Languages of Africa The African area is one of the most numerous from the point of view of languages – it counts 590 languages representing 64 genera and 7 language families. With the exception of Hadza all language families and 21 language genera are represented by 29 languages in my sample.

Chapter Two

56 Map 2-1 Languages of Africa

Out of these 29 languages 17 have evaluative morphology (58%). Ten languages productively form diminutives but no augmentatives. Hausa, an Afro-Asiatic language, has only augmentatives. Both diminutives and augmentatives are productively formed in 6 languages. The highest total EM saturation is in two languages spoken in western Africa, Bafut (3.67) and Konni (3.34). On the other hand, 5 of 12 languages without EM are concentrated in eastern Africa. Bafut and Konni also feature the highest DIM and AUG saturation values. Table 2-1 Languages of Africa–EM saturation Family

Genus

Language

NigerCongo

Bantoid

Kwa

Aghem

EM satur. 1.00

DIM satur. 1.00

AUG satur. 0

Bafut

3.67

2.67

1.00

Kinubi/ Luganda Swahili

0

0

0

2.33

1.33

1.00

Akan

1.00

1.00

0

Gã

0.00

0.00

0.00

Research

Gur

Konni

3.34

1.67

1.67

Supyire

1.67

1.67

0.00

Diola-Fogny

2.00

1.00

1.00

Jamsay/ Dogon Yoruba

0.00

0.00

0.00

0.00

0.00

0.00

Central Cushitic Semitic

Agaw

1.00

1.00

0.00

Amharic

0.00

0.00

0.00

East Chadic West Chadic Omotic

Dangaleat

1.00

1.00

0.00

Hausa

1.00

0.00

1.00

Northern Atlantic Dogon Defoid AfroAsiatic

Maale

2.33

1.33

1.00

Mina

0.00

0.00

0.00

Rendille

0.00

0.00

0.00

Siwi

0.00

0.00

0.00

Tamashek

1.00

1.00

0.00

Datooga

0.00

0.00

0.00

Endo (Nandi)

0.00

0.00

0.00

Fur

Fur

0.00

0.00

0.00

West Saharan Songhay

Kanuri

1.00

1.00

0.00

Koyra Chiini

1.00

1.00

0.00

MoruMa’di Central Khoisan Southern Atlantic Kadugli

Ma’di

2.33

1.00

1.33

Cirecire (Khoekhoe) Kisi

1.33

1.33

0.00

0.00

0.00

0.00

Krongo

1.33

1.33

0.00

0.98

0.68

0.28

BiuMandara East. Cushitic Berber NiloSaharan

Khoisan Atlantic Kadugli

57

Nilotic

TOTAL SATURATION

Chapter Two

58

Diminutives Word-Formation2 10 out of 17 languages are suffixing. Bafut (30) and Krongo (31) make use of prefixation: (30) mu-nili’i DIM-eye ‘little eye’ (Tamanji, pers. com.) (31) tìn-kàfà dim-axe ‘little axe’ (Reh 1985, 158) Reduplication was attested in Dangaléat (32): (32)

sooɛ~sooɛ small~small ‘very small’ (Shay, pers. com.)

and compounding in Koyra Chiini (33): (33) fufu-tondi-ije grinding.stones-child ‘small grindstone’ (Heat 1999, 78) With the exception of Bafut, Koyra Chiini and Supyire, all sample languages make use of one word-formation process only. Although Bafut, Koyra Chiini, and Supyire make use of two word-formation processes, one is always suffixation. Additionally, change of paradigm was attested in 5 languages, 4 of which are from the Niger-Congo family. For example, in Aghem (Hyman 1979, 24) some nouns can be transferred from gender class 7/8 to gender class 11/12 to acquire a diminutive meaning. (34) fƚ́-fú ‘small rat’ < kƚ́-fú ‘rat’ Another Niger-Congo language, Diola-Fogny, forms singular and plural diminutives by means of class 10 and 11 prefixes which are attached

2

In the following sections the heading Word formation is used in its broader sense as an umbrella term for diminutive and augmentative formation.

Research

59

to words of different inflectional class, such as -ko ‘head’ (5/6) and -ɲil ‘child’ (1/2) (Aronoff and Fudeman 2005, 62): (35) jibεkεl/mubεkεl ‘palm-oil tree’ jikit/mukit ‘type of small antelope’ In Supyire, gender 3 is the gender of small things. According to Carlson (1994, 5), moving a root into gender 3 may have the effect of allowing it to denote a smaller object than in another gender. In Swahili (Contini-Morava, pers. com.), the noun class prefix ki- (class 7) replaces the plural prefix m- of class 8: (36) kitoto ‘small child, infant’ < mtoto ‘child’ Change of paradigm was attested also in an Afro-Asiatic language, Tamashek (Heat 2005, 161): (37) t-e-h æ̏nni-t-t (feminine) ‘mediocre dwelling’ < é-hæn (masculine) ‘dwelling’ In Afro-Asiatic languages diminutivisation and augmentativisation very often result from gender shift (Mous, pers. com.). Masculine is usually used for augmentatives and feminine for smallness, but it can also be the other way around. In Agaw, for example, the feminine gender of nouns refers to female beings, or is a diminutive or pejorative (Meinhof 1912, 22-25). While masculine nouns have no particular ending, the feminine singular is consistently marked with -a: (38) a. b.

séna ‘sister’ < sén ‘brother’ gséŋa ‘bitch’ < gséŋ ‘dog’

African languages give strong support to Jurafsky’s theory of the semantic origin of diminutives. In Bafut (Tamanji, pers. com.) mu- is the regular word for ‘child’, and in Akan (Appah, pers. com.) the only diminutive morpheme known and documented is the morpheme -ba. This morpheme is understood to have been derived from the word ɔ̀bá ‘child’. In Khokhoe ‘child’ can be used as a suffix (Chebanne, pers. com.):

Chapter Two

60

(39) a. b.

duu |ua eland.(antelope)-child ‘little antelope’ |ao |ua snake-child ‘little snake’

A similar case can be identified in Koyra Chiini (Heat 1999, 78), where X-ije literally means ‘child of X’ in the sense of ‘a smaller entity associated with X’. If X denotes a physical object, X-ije denotes a smaller object physically associated with it, or a small X. If X denotes collectivity, mass, location, or abstraction, X-ije denotes an individual: (40) fufu-tondi-ije ‘small grindstone’ < fufu-tondi ‘grinding stones’ Cognitive categories 14 languages of the sample can express Quantity of Substance (41), 4 Quantity of Quality (42) and 2 Quantity of Action (Konni and Bafut (43), i.e., the languages with highest saturation). Quantity of Circumstance was not attested. 12 languages express only one cognitive category, mainly Quantity of Substance. Supyire and Konni can express 2 and Bafut 3. (41) Cirecire

|ao-|ua snake-child ‘small snake’ (Chebanne, pers. com.)

(42) Kanuri

sə́lə́mà-ràm black-ATT ‘blackish’ (Hutchinson 1981, 67)

(43) Bafut

bwiin-tə sleep-DIM ‘sleep a little’ (Tamanji, pers. com.)

Word classes The most productive word class is nouns (14 languages) followed by adjectives (4 languages). Verbs were attested in Bafut and Konni. No adverbs were identified. Maale, an Omotic language which also expresses diminutives by the feminine marker -éll, can diminutivize the pronoun meaning ‘this’ (Amha 2001, 71):

Research

61

(44) Yénnó mís’-éll-ó dóngo? as-á that.F:ABS tree-F-ABS five person.PL-NOM bukínt-I wolla túg-á-ne gather-CNV1 together up.root-IPF-A.DCL ‘Five people gather together and uproot that (small) tree.’ According to Amha (ibid.), the masculine demonstrative form hayí ‘this’ (which when referring to things and non-humans denotes something big) can be combined with the diminutive marker -má to express relative smallness: (45) hayi-mmá ‘this.NOM’ this.M-DIM In addition, when ha ‘this F/M’ (which is the base for both masculine and feminine demonstratives) is combined with the diminutive, it refers to an even smaller object (Amha, ibid.): (46) ha-mmá ‘this.NOM’ this-DIM Correlations The correlations found do not show any deviations from the expectations. In particular, suffixation highly correlates with both Quantity of Substance and nouns. This appears to be one of the most typical characteristics of evaluative morphology in African languages. Another typical feature is the change of paradigm that correlates, basically, with nouns and Quantity of Substance. This was attested in 5 languages. Besides Tamashek, all of these are from the Niger-Congo family. Another typical feature is shift of gender. Augmentatives Word formation None of the 7 languages with augmentatives make use of more than one word-formation process. There are 3 suffixing languages (Hausa, Konni and Maale); Bafut is a prefixing language. Bafut forms augmentatives by means of the prefix ma-; which in the form maa is the regular term for ‘mother’: (47) ma-nili’i AUG-eye ‘big eye’ (Tamanji, pers. com.)

Chapter Two

62

Xhosa (not included in my sample), a Bantu language, also uses the feminine affix (inkosi > inkosikazi, ‘Lord, Lady’). In case of ambiguity, the augmentative suffix -kazi is not used unless the noun is feminine; in that case the suffix becomes a little pejorative: (48) umfazikazi omkhulukazi (‘a great big fat woman’) (Lloyd, pers.com.) Two languages, Diola-Fogny and Swahili, make use of the change of paradigm. In Diola-Fogny (Aronoff and Fudeman 2005, 63) the noun class is changed from class 5 or 9 to class 12: (49) buko < -ko ‘head’ (50) fuɲil < -ɲil ‘child’ Swahili (Contini-Morava, pers. com) replaces a noun class prefix with zero (class 5) or with the prefix ma- (plural, class 6): (51) toto ‘large, ungainly child’ < mtoto ‘child’ An interesting situation occurs in Maale, an Osmotic language. While diminutives can be coined (in addition to other possibilities) by a feminine gender marker -éll, the masculine gender marker -ats is used for augmentatives. (52) máár-átsí maƷƷ-ínt-é-ne house-M.NOM build-PAS-PF-A.DCL ‘The (big) house is built’ (Amha 2001, 71) Cognitive categories and word classes With the exception of Hausa, all languages can express Quantity of Substance. Hausa, however, does express Quantity of Quality. None of the languages expresses Quantity of Circumstance. Only one language, Konni, makes use of two cognitive categories: Quantity of Substance and Quantity of Action. This situation is reflected in word classes, too. The two cognitive categories produce augmentatives falling within two word classes–nouns and verbs. All languages except for Hausa produce nouns. Hausa, a language with no diminutives, changes the word class (Newman, pers.com.):

Research

63

(53) gabjeejeè ‘bulky’ < gabzaà ‘heap up a lot of something’ As the number of languages with augmentatives is very low, it is not possible to identify any correlations. The most common correlation of Suffixation-Quantity of Substance-nouns can only be identified for 2 languages (Konni and Maale).

2.4.2 Languages of Australia and New Guinea According to WALS, there are 389 languages divided into 125 language genera and 54 language families. My sample covers, as illustrated in Map 2-2 and Table 2-2, 22 languages from 15 genera and 8 families. Map 2-2 Languages of Australia-New Guinea

Out of 22 languages, 13 have no evaluative morphology, 2 of them (Bāgandji and Imonda) have diminutives only, and 7 produce both diminutives and augmentatives. The highest total EM saturation value is 4.00 (Arabana, Alawa and Sentani, each representing a different language family). The highest DIM saturation (2.33) is found in Anamuxra, and the highest AUG saturation occurs in Djaru (2.67). In Djaru, the EM saturation of augmentatives is higher than the saturation of diminutives.

Chapter Two

64

Table 2-2 Languages of Australia-New Guinea–EM saturation Language Alyawarra

EM satur. 0

DIM satur. 0

AUG satur. 0

Arabana

4.67

2.00

2.67

Bāgandji

1.00

1.00

0.00

Diyari

0

0

0

Djaru

1.33

1.33

2.67

Yidin

0

0

0

Bachamal

3.33

2.00

1.33

Nyulnyulan

Bardi

0

0

0

Maran

Alawa

4.00

2.00

2.00

Gaagudju

Gaagudju

0

0

0

Yangmanic

Wardaman

0

0

0

Wororan

Worrorra

0

0

0

Madang

Amele

0

0

0

Anamuxra

3.67

2.33

1.33

Kobon

0

0

0

Koiarian

Koiari

2.67

1.33

1.33

West Papuan Border

Hatam

Hatam

0

0

0

Border

Imonda

1.00

1.00

0.00

Sko

Krisa

I’saka

0

0

0

Solomons East Papuan Lower Sepik-Ramu Sentani

Solomons East Papuan Lower Sepik

Lavukaleve

0

0

0

Yimas

0

0

0

Sentani

Sentani

4.00

2.00

2.00

1.16

0.68

0.60

Family Australian

Genus Pama-Nyungan

Anson Bay

Trans-New Guinea

TOTAL SATURATION

Out of 22 sample languages, 13 have no evaluative morphology, 2 of them (Bāgandji and Imonda) have diminutives only, and 7 productively form both diminutives and augmentatives. The highest total EM saturation value is 4.00 (Arabana, Alawa and Sentani, each representing a different

Research

65

language family). The highest DIM saturation value (2.33) is found in Anamuxra, the highest AUG saturation occurs in Djaru (2.67). In Djaru, the EM saturation of augmentatives is higher than the saturation of diminutives. Diminutives Word-Formation The evaluative morphology of languages spoken in the Australia-New Guinea area makes use of two word-formation processes–reduplication and suffixation. Reduplication is usually full reduplication: (54) Bachamal

Arabana

pøccal̯ ak~pøccal̯ ak small-small ‘tiny’(Ford 1990, 94) karla~ (k)arla creek-creek ‘very small creek’ (Hercus 1994, 98)

But there are also languages with partial reduplication (Arabana, Alawa, Anamuxra): (55) Alawa

paʃa~paʃaʃa RDP~little

‘very little’ (Sharpe 1972, 54) For an example of partial postposing reduplication, see the Anamuxra example (19) above. Reduplication is only used in 7 languages (4 Australian, 2 Trans-New Guinean and 1 Sentani). The languages with diminutives use suffixation exclusively. Both of these word formation processes are used in Anamuxra, a language of the Trans-New Guinea family and the Madang genus. In Bāgandji, the only diminutive suffix is -ulu and its basic meaning is ‘singular’ (Hercus 1982, 81): (56) ηidja-ulu mūrba-ulu one child ‘one single small child’ Cognitive categories 5 languages express only one cognitive category. In the suffixing languages Bāgandji and Imonda it is Quantity of Substance, in Djaru,

66

Chapter Two

Alawa and Kioari it is Quantity of Quality. Anamuxra, a language with the highest DIM saturation, can express Quantity of Circumstance only. Both Quantity of Substance and Quantity of Quality are present in Bachamal and Senatni, and two cognitive categories, Quantity of Substance and Quantity of Action can be found in Arabana. Quantity of Substance can be expressed in 5 languages (Arabana, Bāgandji, Bachamal, Imonda and Sentani); the same number of languages express Quantity of Quality (Djaru, Bachamal, Alawa, Koiari, Senatni); only one language, ArabanaWankangurru, has the cognitive category of Quantity of Action. Quantity of Circumstance is only expressed by Anamuxra. Word classes The situation of cognitive categories correlates with the situation of word classes. The two suffixing languages Bāgandji and Imonda (expressing only Quantity of Substance) make exclusive use of nouns. Only adjectives are productively formed in Djaru and Koiari. Arabana, the only language with the cognitive category of Quantity of Action, uses both verbs and nouns. Two word classes are also used in Bachamal and Anmuxra, nouns and verbs. Sentani is the only language with adverbs forming diminutives. In addition, the diminutivisation process in Sentani can change the word class (Cowan and Jan 1965, 12): (57) duka~duka ‘massive’ < duka ‘stone’. Correlations The correlations attested between word-formation processes and cognitive categories are those between reduplication and Quantity of Substance, and between reduplication and Quantity of Quality. Subsequently, there is a correlation between reduplication and nouns, and between reduplication and adjectives. The cognitive category of Quantity of Substance correlates with nouns, Quantity of Quality with adjectives. The analysis of correlations allows us to draw the following conclusion: The evaluative morphology of diminutives in the languages of the Australia-New Guinea area preferably use reduplication. It is predominantly nouns and adjectives that are reduplicated. These express the cognitive categories of Quantity of Substance and Quantity of Quality. Augmentatives Only one word-formation process coining augmentatives was attested in my sample of Australian-New Guinean languages–reduplication. It can be complete

Research

(58) Arabana

67

murla~murla skink~skink ‘large skink’ (Hercus 1994, 96)

or partial (cf. an example of partial infixing reduplication in (23) above). Arabana has three cognitive categories (Quantity of Substance, Quantity of Quality and Quantity of Action). Sentani expresses (just) two cognitive categories (Quantity of Quality and Quantity of Action). The remaining languages express only one cognitive category. Bachamal, Alawa and Koiari express Quantity of Quality and Djaru Quantity of Circumstance. The strongest correlation is among reduplication, Quantity of Quality and adjectives.

2.4.3 Languages of North America According to WALS, 381 languages spoken in the area of North America are divided into 93 genera and 49 language families. My sample includes 23 languages representing 16 genera and 14 language families. Map 2-3 illustrates the position of languages; Table 2-3 summarizes their classification and EM saturation. Out of 23 languages, 8 have no evaluative morphology. Out of 15 languages with evaluative morphology, 7 have only diminutives. The highest EM saturation occurs in Kwakw’ala, a Northern Wakashan language, followed by Micmac (Algonquian), Lakhota (Siouan) and West Greenlandic (Eskimo-Aleaut). More specifically, Kwakw’ala also has the highest DIM and AUG saturation values. The group of languages spoken in the southern part of North America have either no evaluative morphology (Totonac, Sochiappan, Jakaltec), or the value of EM saturation is as low as 2 (Tzutzujil, Pipil, Yukatec). Similarly, languages spoken on the eastern coast (Tutelo, Mississaga, Passamaquoddy, Wintu) display no evaluative morphology or the saturation value is low. In Micmac and West Greenlandic the AUG saturation is higher than the DIM saturation. In the languages with both diminutives and augmentatives, the values of DIM and AUG saturation are balanced.

Chapter Two

68

Map 2-3 Languages of North America

Table 2-3 Languages of North America – saturation Family

Genus

Language

EM satur.

DIM satur.

AUG satur.

Mayan

Mayan

Cakchiquel

0

0

0

Jacaltec

0

0

0

Tzutujil Yucatec Maya

2

1

1

2

2

0

Carrier

0

0

0

Slavey

1.33

1.33

2.67

Numic

Comanche

1.33

1.33

0

Aztecan

Pipil Passamaquod dy

2

2

0

1.33

1.33

0

Micmac

3.67

1.67

2

Nipmuck

1

1

0

Mississaga

1.33

1.33

0

Na-Dene UtoAztecan

Algic

Athapaskan

Algonquian

Research

69

Yurok

Yurok

0

0

0

Diegueno

2.67

1.33

1.33

Wakashan

Yuman Northern Wakashan

Kwakw’ala

6.33

3.33

3

Siouan

Siouan

3.00

2.00

1.00

Totonacan

Totonacan Central Salish

0

0

0

1.33

1.33

0

Chinanyecan

Lakota Misantla Totonac Montana Salish Sochiapan Chinantec

0

0

0

Siouan EskimoAleut

Siouan EskimoAleaut

Tutelo West Greenlandic

2.67

1

1.67

3

1

2

Caddoan

Caddoan

Wichita

0

0

0

Penutian

Wintuan

Wintu

0

0

0

1.52

1.00

0.64

Hokan

Salishan OtoManguean

TOTAL SATURATION

Diminutives Word-formation Four word-formation processes are used for the coining of diminutives–compounding, reduplication, suffixation and stem-consonant alternation. The majority of the languages use only one word-formation process. Compounding was attested in Tutelo (59) and Tzutzujil (60): (59) Tutelo īstab-la-ksīk lip-INSTR-DIM ‘small lip’ (Oliverio 1996, 207) (60) Tzutujil saq-perepoj white-white.and.flat ‘whitish’ (Dayley 1985, 221) Reduplication is employed in Yucatec Maya and Montana Salish. In Montana Salish, reduplication is combined with the use of the voiceless lateral fricative (Thomson, pers. com.):

Chapter Two

70

(61) a. b.

l*c~ci*txw ‘little house’ < ci*txw ‘house’ l*c~cq’e*l*p ‘a little Douglas fir tree’ < cq’e*l*p ‘a Douglas fir tree’

The most frequent word-formation process is suffixation, usually employed in languages spoken in the central part of North America (Comanche, Lakhota, Nipmuck). In Pipil, the suffix -tsin has either a diminutive or reverential meaning: (62) mistun-tsín cat-DIM ‘kitten’ (Campbell 1985, 49) The diminutive plural is formed by reduplication of the suffix -tsin (Campbell 1985, 49). Another interesting case of suffixation can be found in Passamaquoddy. According to LeSourd (1995, 104), “the diminutive stems of one type of transitive verbs are productively derived from a partly inflected form of the basic stem”. In such cases, the inflectional suffix is repeated after the diminutive suffix whenever inflectional morphology calls for its use. Recursive diminutivisation then includes three identical inflectional morphemes. Thus, for example, diminutive and plural affixes can interact and be repeated in a single stem in several possible orders (ibid.). The diminutive stems are derived from plural stems and plurals are formed from diminutives– inflection precedes diminutivisation, which contradicts the universal order of word-formation and inflection (LeSourd 1995, 126): (63) hkihka-n-hǝtǝ-ss-ò-kk all-die-PL-DIM-AI-(3)-33AN.ABS ‘all of the poor little ones are dead’ In addition to suffixation, a special case of diminutivisation, i.e., diminutive consonant symbolism, can be found in Nipmuck (and in Algonquian languages in general). The process involves replacement of certain consonants in a word by other consonants, effecting a diminutive meaning. According to Gustafson (2000, 30), “this commonly involves the palatalization of *t to phonetic *[c], a process which may or may not be accompanied by a diminutive suffix.” An example (Pentland 1975, 241) from the Cree language in (64) is followed by an example (65) from Nipmuck (Gustafson 2000, 31): (64) acihk or acihkošiš ‘little caribou’ < atihk ‘caribou’

Research

71

(65) cekwal ‘little frog’ < *tekwala ‘frog’ While (65) raises no doubts Gustafson’s (2000: 30) examples in (66) may be called into question in terms of evaluative morphology: (66) a. macahkamikesiw ‘he behaves badly’ < *rnataxkamikesiwa b. macehtākwat ‘it is heard to be bad, it is bad news’ < *matehta:kwatwi As Gustafson (2000, 30) points out, *t to *[c] palatalization often connotes pejorative meaning and that is why it is an example of diminutive consonant symbolism rather than a generalization from those forms in which the palatalization of *mat- ‘bad’ is a phonologically regular process. No doubt, diminutiveness and pejorativeness do not exclude each other, which can be illustrated with an example from Slovak: (67) učit-el-ík teach-AG.M-DIM ‘teacher-DIM.PJR’ Diegueno, a Yuman language from the Hokan family, makes use of stem-consonant alternation, too. Voiced laterals alternate with corresponding voiceless lateral spirants. These alternations underlie the distinction of size: “forms with voiced laterals imply smallness, those with voiceless laterals bigness or intensity“ (Langdon 1970, 101): (68) nyily ‘to be black’ vs. nyiƚy ‘to be very black’ (69) ɔsaƚy ‘hand, arm’ vs. ɔsaly ‘little hand’ Cognitive categories The cognitive category Quantity of Substance is expressed by 14 languages (out of 15 languages with evaluative morphology). The only language that does not express this category is Tzutujil. Quantity of Quality was identified in 5 languages, Quantity of Action in 6, and Quantity of Circumstance only in Kwakw’ala, the language with the highest EM saturation in my sample of North American languages. Tzutujil, Tutelo and West Greenlandic express only one cognitive category; the rest of the languages usually express two (Quantity of Substance and either Quantity of Quality or Quantity of Action). Only in Kwakw’ala were all cognitive categories attested.

Chapter Two

72

Word classes Discussion on word classes in languages of North America is impeded by the predominantly polysynthetic nature of the languages. My approach to word classes is a traditional (Latin-based) one. It is a well-known fact that the Latin-based classification is very often inapplicable to polysynthetic languages. For example, in Lakhota (Pustet, pers. com.) verbs can be combined with the suffix -la, but semantically -la does not modify the verb but an argument of the verb. Consequently, although Lakhota is identified as a language with diminutivized verbs in my analysis, it does not exhibit the cognitive category Quantity of Action that usually correlates with verbs. However, for the sake of further comparison with other language areas, I also applied the traditional classification to North American languages. Diminutivized nouns are used in 14 languages (besides Tzutzujil). 7 languages out of 15 use verbs, 6 languages use adjectives, and 1 (Kwakw’ala) uses adverbs. The class change was attested in Pipil (Campbell 1985, 50): (70) tsapa-tsin short-DIM ‘shorty’ Tzutzujil, Nipmuck, Mississaga, Tutelo and West Greenlandic make use of only one word class (noun, except for Tzutzujil that produces diminutivized adjectives); the rest of languages use two word classes (usually nouns and either verbs or adjectives). In Pipil and Lakhota three word classes were identified. In Lakhota these are nouns, verbs and adjectives; Kwakw’ala uses nouns, verbs, adjectives and adverbs; in Pipil nouns, adjectives. In addition, Pipil employs a rather rare word-class change mechanism. Correlations 10 languages exhibit correlation between suffixation and the cognitive category of Quantity of Substance. 7 languages exhibit correlation between suffixation and nouns. High correlation (14 languages) was also attested between Quantity of Substance and nouns (as well). In the languages of the North American area (10 languages), the strongest correlation among word-formation process, cognitive category and word class is that of suffixation, Quantity of Substance and nouns, respectively. On the other hand, only one language features the co-occurrence of reduplication, Quantity of Substance and Quantity of Action (Yucatec

Research

73

Maya). Quantity of Circumstance co-occurs with suffixation only in one language (Kwakwa’la), too. Augmentatives Augmentatives are present in 8 languages. Word-formation processes used for augmentivisation are compounding (Tzutujil), suffixation (Slavey, Kwakw’ala, Lakhota, Tutelo, West Greenlandic), prefixation (Micmac), reduplication (Tutelo) and stem-consonant alternation (Diegueno). The standard situation is just one word-formation process. As noted by Dayley (1985, 221), compounding in Tzutujil is restricted to a set of adjective compounds that are built on a monosyllabic colour term: saq ‘white, clear’, q’eq ‘black’, rax ~ rex ‘green’, and kaq ‘red’. Compounded with these colour terms are bound stems formed with a monosyllabic root plus the suffix -V1C1oj. Each of the bound stems only occurs with one, two, or three colour terms. The compounds formed with the colour terms plus bound stems usually indicate either a colour very much like the one denoted by the colour term (i.e., an example par excellence), or one sort of like it: (71) a. b. c.

rex-mulumuj green-piled.up ‘very green’ saq-tilitoj white-knock.fruit.of. trees ‘very white’ saq-muqumuj white-cloudy ‘very cloudy’ (Dayley, ibid.)

Stem-consonant alternation in Diegueno has been mentioned above. However, in augmentatives the alternation is exhibited not only in root consonants, but also in a prefix, as illustrated in the following examples (Langdon 1970, 102): (72) ləxup ‘large hole’ < əxu’ k wəlxup ‘nostril (i.e. nose hole)’ tələwik ‘crooked, bent (large)’ < tələwik ‘crooked, bent’ The case of Micmac, a prefixing language, is interesting from two points of view. First, it is the only language in my sample of NorthAmerican languages that uses different affixation processes for diminutives (suffixation) and augmentatives (prefixation). Second, it is the

Chapter Two

74

only prefixing language in my sample. Three prefixes are used: kji- (with nouns), kesi- or ksi- (with adjectives and verbs, respectively) (Hewson and Bernard 1990, 36), e.g.: (73) kji-sipu AUG-river ‘great river’ An interesting example of augmentative use of reduplication of adjectives comes from Tutelo: (74) a.

sā~sabe RDP~black

b.

‘black here and there’ gdhejā ~ja spotted~RDP ‘sptted in many places’

A similar meaning, i.e., ‘of different kinds’, can also be attached to reduplicated nouns: (75) nūjin͂ ~jīn͂ -ga < nūjin͂ ga ‘boy’ boy~RDP ‘boys of different sizes and ages’ (Hale 2001, 41) The most frequently expressed cognitive category is Quantity of Substance (all languages besides Lakhota and Tzutzujil), followed by Quantity of Quality in 5 languages (Tzutzujil, Slavey, Kwakwa’la, Lakhota and West Greenlandic). This situation is mirrored in word classes–the most productive is nouns (all languages besides Lakhota and Tzutzujil), followed by verbs. Thus, the strongest correlation is that between suffixation, Quantity of Substance and nouns. On the other hand, no correlation can be identified for reduplication, stem alternation or prefixation.

2.4.4 Languages of South America WALS includes 253 South American languages grouped into 92 genera and 71 language families. My sample consists of 15 languages from 15 genera and 15 language families–each language thus represents a different genus and family. This phenomenon is typical of South America–

Research

75

while the proportion of languages to genera in South America is 2.75, the same proportion for the languages spoken in Eurasia is 9.9. Map 2-4 Languages of South America

Two South American languages without evaluative morphology (Movima and Trumai) are spoken in the central part of South America. 13 languages have evaluative morphology. While 9 of them can produce both diminutives and augmentatives, 3 languages (Embera, Maipure, Tarna Quecha) have only diminutives, and Wari can only produce augmentatives. The highest EM saturation was identified for Yurakare (5.00), followed by Jaqaru (3.33). Both languages are spoken in the central part of South America. Yurakare also has the highest DIM saturation. Jaqaru is the language with highest AUG saturation. Interestingly, the AUG saturation in Jaqaru is higher than its DIM saturation. In the remaining languages, the DIM and the AUG saturation values (if both are present) are in balance. Movima, as indicated above, was analysed as a language with no evaluative morphology. This should be, however, taken with certain reservations. While Haude (pers. com.) assumes that there are no diminutive or augmentative morphemes in Movima, she points out the use of classifiers that have a highly restricted capacity to form diminutives or augmentatives. Let us mention, for example, the classifier -di whose main meaning is ‘seed, grain’. In (76) it seems to function as a diminutive:

Chapter Two

76

(76) awaro-di ‘little parrot’< awaro ‘parrot’ Similarly, -La, a combination of voiceless lateral fricative + a, is used to denote smallish, pointed objects like spoons or teeth (77): (77) bilaw-La ‘little fish’< bilaw ‘fish’ As stressed by Haude, while more examples can be found, none of the classifiers is productive or used by all speakers. Table 2-4 Languages of South America – saturation

Embera

EM satur. 1.33

DIM satur. 1.33

AUG satur. 0

Aymaran

Jaqaru

3.33

1.00

2.33

Kwaza

Kwaza

Kwaza

2.00

1.00

1.00

Arawakan

Arawakan

Maipure

1

1

0

Panoan

Panoan

Matses

2.67

1.33

1.33

Mosetenan

Mosetenan

Moseten

2

1

1

Movima

Movima

Movima

0

0

0

Nambikuaran

Nambikuaran

2.00

1.00

1.00

Tupian

Tupi-Guarani

Sabanê/ Nambikuara Tapiete

2.67

1.33

1.33

Quechuan

Quechuan

1.00

1.00

0

Trumai

Trumai

Tarna Quecha Trumai

0

0

0

Tucanocan

Tucanocan

2.00

1.00

1.00

Warao

Waroa

Wanano/Gu anano Warao

2.00

1.00

1.00

ChapacuraWanham Yuracare

ChapacuraWanham Yuracare

Wari

1.00

0

1.00

Yurakare

5.00

3.33

1.67

2.00

0.87

0.84

Family

Genus

Language

Choco

Choco

Aymaran

TOTAL SATURATION

Research

77

Diminutives Word-formation Word-formation processes which are used for the coining of diminutives in the South American languages include suffixation, reduplication, compounding and change in vowel/nasality of a noun. Suffixation is present in 8 languages (out of 12) and it is the most productive word-formation process. In Jaqaru, the most productive EM suffix is -cha (Hardman 2000). It can express a range of meanings, including reduction in importance, emotional limitation regarding the referent as well as a sort of lament, or apology: (78)

Jaqaru

Na.ch.kas.wa. ‘It’s just me alone now.’ (Hardman 2000, 22)

In Tapiete (González 2005), the suffix -mi can also modify the lexical meaning of the noun through the formation of a new kinship term: (79) a. b.

shé-sɨ-mi 1SG.POSS-mother-DIM ‘my maternal aunt’ shé-ru-mi 1SG.POSS-father-DIM ‘my maternal uncle’

Reduplication (both complete and partial) is present in Embera (80) and Matses (81): (80) pure~purea RDP~red ‘reddish’ (Mortensen, pers. com.) (81) ?senad piu~piu isombi deer red~red see.PAST.1A ‘I saw the reddish deer.’ (Fleck 2003, 496) Compounding in Sabanê: (82) olumata maysunon bull-male.youngling ‘male bull calf’ (de Araujo 2004, 105)

Chapter Two

78

Clitics are used in Matses: (83) chido-n-uid-tsëc-bi aid daëdca-quid woman-ERG-only-DIM-EMP that.one weave-HAB ‘Only women weave those [baby carrying straps]’ (Fleck 2003, 284) and change in vowel and/or nasality of a noun in Moseten (Sakel 2004, 100): (84) ïwï ‘little child’ < äwä ‘child’ The maximum number of word-formation processes per language is two. In particular, two processes are used in Matses (reduplication and clitics) and Nambikuara (suffixation and compounding). Cognitive categories All four cognitive categories can be found in South American languages. However, Quantity of Circumstance was only present in Yurakare, and Quantity of Action only in Yurakare and Tarna Quecha. On the other hand, Quantity of Quality can be expressed in four languages and Quantity of Substance in nine of them. Yurakare, the language with the highest DIM saturation, can express all four cognitive categories. Two cognitive categories, Quantity of Substance and Quantity of Quality, can be expressed in Matses. The rest of the languages only make use of one cognitive category, usually Quantity of Substance. Word classes The most frequent word class is nouns, followed by adjectives. Nouns are present in 8 languages (basically the same as those expressing the cognitive category of Quantity of Substance). Only two languages make use of more than one word class–Matses and Yurakare. Matses uses nouns and adjectives; Yurakare nouns, adjectives, verbs and adverbs. In Yurakare, the same suffix -nñu is used with nouns, verbs and adjectives: (85) a. ëshshë-nñu ka-tütü-ø matata ëshshë a-dojo=y stone-DIM 3SG-sit;be-3 big stone 3SG.P-body=LOC ‘The little stone is on top of the big stone’ b. ana-ja-lë baja-nñu-ø DEM-MEA-AMP subside-DIM-3 ‘The water has subsided a little’

Research

c.

79

ñuñujulë-nñu-ø ta-ëtëjñu small-DIM-3 1PL-son ‘Our son is very small’ (van Gijn 2006, 120)

No example of class change was attested. Correlations The highest correlation is among suffixation, Quantity of Substance and nouns (7 languages). Augmentatives Augmentatives are produced in 10 languages. Wari, a ChapacuraWanham language, has no diminutives but it can produce augmentatives. As stated by Everett and Kern (1997, 346), most degrees of quality are expressed by modification of the verb root via compounding: (86) mixem na wom black 3S.RP/P cotton ‘The clothes are dirty’ vs. Mixem tamana na wom black much 3S.RP/P cotton ‘The clothes are very dirty’ The most frequent word-formation process is suffixation (7 languages). In Kwaza, the morpheme -te(te) with ‘augmentative effect’ on verbs (van der Voort 2004, 261-262) can also be attached to nouns, but then its meaning is ‘pure, real’ (2004, 207). As noted by van der Voort, it is probably a form borrowed from the Tupi languages. Compounding was attested in Wari, clitics in Matses and prefixation in Moseten. Augmentation in Moseten is limited to a group of nouns such as plants or body parts (Sakel 2004, 101): (87) chhi-yiij-si’ AUG-leg-LNK.F ‘big-legged (woman)’ With the exception of Jaqaru and Yurakare all languages use only one word-formation process. Jaqaru and Yurakare use reduplication and suffixation. No language expresses the cognitive category Quantity of Circumstance. On the other hand, Quantity of Substance was attested in 6

80

Chapter Two

languages; Quantity of Qualitny and Quantity of Action in 3 languages. Yurakare can express two cognitive categories (Quantity of Quality and Quantity of Action) and Jaqaru three (Quantity of Substance, Quantity of Quality and Quantity of Action). The situation in word classes corresponds with cognitive categories. Mostly nouns are used (6 languages). Two word classes (verbs and adjectives) are used in Yurakare and three (nouns, verbs, adjectives) in Jaqaru. The highest correlation (7 languages) was attested between suffixation, Quantity of Substance and nouns.

2.4.5 Languages of SE Asia + Oceania In the area of South-East Asia + Oceania there are 507 languages grouped in 43 genera and 6 language families. My sample covers 24 languages from 10 genera and 3 families. Map 2-5 Languages of SE Asia + Oceania

11 languages out of 24 are without evaluative morphology. 10 of them are from the Austronesian language family. 13 languages have evaluative morphology. 7 of them have both diminutives and augmentatives; 3 of them can only produce diminutives (Anejom, an Austronesian, Oceanic language; Dolakha Newar and Dumi, both Sino-Tibetan, Bodic) and 3 languages only augmentatives (Nelemwa, Siar Lak, and Austronesian, Oceanic language and Thai, a Tai-Kadai, Kam-Tai language). The highest total EM saturation is characteristic of Agta (5.67) followed by Karao and

Research

81

Muna (both 4.33). These languages are from the Austronesian language family. The difference in EM saturation value between Muna and the language with the next highest EM value, Tibetan (2.67), is fairly high at 1.66. In fact, the saturation of the remaining languages falls sharply: the value of the total EM saturation in 6 languages is less than 1.00. With the exception of 3 languages that can only form augmentatives, the AUG saturation in Agta and Meithei is higher than the DIM saturation. Table 2-5 Languages of SE Asia + Oceania – saturation

Austronesian

Family

Anejom

EM satur. 1

DIM satur. 1

AUG satur. 0

Boumaa Fijian

0

0

0

Kaliai-Kove

0

0

0

Maori

0

0

0

Nalik

0

0

0

Nelemwa

1.00

0

1.00

Siar Lak

1.00

0

1.00

Vitu

0

0

0

Agta

5.67

2.67

3.00

Ilocano

0

0

0

Karao

4.33

3.33

1.00

Bahasa Indonesia

0

0

0

Karo Batak

0

0

0

Central MalayoPolynesian Sulawesi

Kambera

0

0

0

Muna

4.33

3.33

1.00

South HalmeraWest New Guinea

Taba

0

0

0

Genus

Language

Oceanic

Northern Philippines

Sundic

Chapter Two

82

Sino-Tibetan

Bodic

Jkuki-ChinNaga TibetoBurman Kam-Tai

Dolakha Newar

1.00

1.00

0

Dumi

1.67

1.67

0

Kham

2.00

1.00

1.00

Manange/Gurung

0

0

0

Meithei

2.33

1.00

1.33

Mongsen Ao

2.00

1.00

1.00

Tibetan

2.67

1.67

1.00

1.00

0

1.00

1.26

0.79

0.51

Thai TaiKadai TOTAL SATURATION

Diminutives Word-formation The most frequent word-formation process is suffixation. With the exception of Karao, all these languages (Dumi, Tibetan, Dolakha, Kham, Mongsen Ao) are spoken in South-East Asia and belong to the SinoTibetic language family. The second most frequent word-formation process is reduplication, present in the Austronesian languages Karao, Agta and Anejom̃ , and one Sino-Tibetan language Jkuki-Chin-Naga language (Meithei). Examples are given in (88) and (89): (88) Agta

at~átu RDP~dog ‘puppy’ (Healey 1960, 7)

(89) Meithei piklək~piklək sémmu pik-lək~pik-lək sém-u small-type~small-type make-IMP ‘Make it of a smaller size’ (Chelliah 1997, 270) In Karao, reduplication can be both complete and partial (Brainard, pers. com): (90) a.

trak~trak CCVC~trak ‘toy truck’

Research

b.

83

ba~badiy CV~baliy RDP~house ‘toy house’

In Muna, it can be combined with prefixation. The prefix ka- in combination with reduplication is used with a simple noun base van der Berg (1989, 295): (91) ka-kontu~kontu ka-stone~stone ‘small stone’ with a reduplicated base: (92) ka-kuhu~kuhuti ka-owl ‘little owl’ as well as with a bound root: (93) ka-bua~bua ‘girl (aged 12-15)’ The same prefix can also have other meanings. If attached in combination with another prefix and reduplication to a verb, it expresses simultaneous action. Ka- itself, without reduplication, attached to a verbal base can produce abstract nouns, result of action or instrument. Muna makes use of circumfixation, too. Nominal circumfix sa- -ha can be attached to verbal bases (with possessive inflection). Its meaning is ‘hardly, barely, just enough, only just’: (94) sa-wanu-ha-no CF-get.up-CF-his ‘he can barely get up’ (van der Berg 1989, 314) The majority of languages in this geographical area make use of one word-formation process. Only the above mentioned Muna uses two. The same is true of Agta and Anejom̃ which combine reduplication and prefixation. Karao is the only language with three attested word-formation processes (suffixation, reduplication and reduplication with infixation).

84

Chapter Two

Cognitive categories Quantity of Substance can be expressed in 8 languages. Quantity of Action and Quantity of Quality can be expressed in 4 languages. Quantity of Circumstance was attested in one language, Dumi (van Driem 1993, 79): (95) Yak’ka hoː-pəts-ɨ-ki dzaˈkha…-tsu?u Over.there come-all-23S-no.sooner.had slow…DIM aːn-piː-si ?e. return-come-REF REP ‘No sooner had he got a way off [from the house] than he slipped back [to the house] ever so slowly and stealthily’ 5 languages express only one cognitive category. Anejom̃ , Dolakha Newari and Ao express Quantity of Substance, Kham of Quantity of Quality, and Meithei Quantity of Action. Karao, Dumi and Tibetan express two cognitive categories, and Muna and Agta, both Austronesian languages, can express 3 cognitive categories – Quantity of Substance, Qunatity of Quality and Quantity of Action. Word classes By implication, the most frequently used word class is nouns (8 languages), followed by adjectives (6 languages) and verbs (4 languages). Adverbs were attested only in Dumi. Corresponding with the situation in cognitive categories, 5 languages make use of one word class; Karao, Dumi and Tibetan of two word classes; and Muna with Agta exploit three word classes. Correlations The correlations are not surprising–the strongest one is between the cognitive category of Quantity of Substance and nouns, attested in 8 languages. A correlation between suffixation and Quantity of Substance was attested in 5 languages, the same number of languages featuring an interrelation between suffixation and nouns. A correlation between suffixation, Quantity of Substance and nouns was present in 5 languages. 3 of them (Dumi, Dolakha Newari and Ao) are from the Sino-Tibetan language family and all of them are spoken in South-East Asia, except for Karao.

Research

85

Augmentatives There are 3 languages in this geographical area that have augmentatives but not diminutives (Nelemwa, Siar Lak, Thai). Siar Lak and Nelemwa are from the Austronesian language family, Oceanic genus. Thai is classified as a Tai-Kadai language from Kam-Tai genus. However, their AUG saturation is quite low at 1.00. The highest AUG saturation is found in Agta (3.00). The next highest saturation value, found in Meithei, is much lower at 1.33. The most frequent word-formation process is reduplication, attested in 5 languages. 3 of them are spoken in South-Asia (Thai, Tibetan, Meithei) and 2 of them are languages of the Philippines (Agta, Karao): (96) Thai

wáan~wăan sweet~sweet ‘very sweet’ (Iwasaki and Ingkaphirom 2005, 35)

(97) Tibetan sal-le-ba be.bright-RDP-ADJ.SUFFIX ‘verybright /very clear’ (Hill, pers. com.) (98) Meithei ŋəwrok~ŋəwrok pót too.white~too.white thing ‘excessively white item’ (Chelliah 1997, 270) (99) Karao manpanga~panga man-CVCV-panga IMPF-PL-branch ‘to grow lots of branches’ (Brainard, pers. com.) (100) Agta

da~dagkal RDP~big ‘very big’ (Healey 1960, 8)

Muna, however, combines reduplication with prefixation. In example (101) the prefix mansi- means ‘more items than expected’ (van der Berg 1989, 299-300): (101)

ta-do-mansi-tandu-tandu-mo only-3pR-quite.a.few-horn-horn.EMP ‘there were quite a few horns (buffaloes)’

Chapter Two

86

The next most frequent word-formation process is compounding. While this word-formation process is not present in diminutives, it was attested in 3 augmentivizing languages, an example comes from Nelemwa: (102) Nelemwa

axoom naabuc mother-mosquito ‘a big mosquito’ hulaxa mwa old.person-house ‘big house’ (Bril, pers. com.)

(103) Siar Lak

ép tan wang the-mother-canoe ‘big canoe’ (Frowein, pers. com.)

(104) Mongsen Ao tǝɹ nuksǝnsaŋpaɁ nǝ “u áhŋáɁ tǝji la lǝɹaɹùɁ ɹǝ.” tǝ̀-ǝɹ nuksǝnsaŋ-pàɁ nǝ u a-hŋáɁ tǝ-ji 3 thus-SEQ PN-M AGT EXCLM NRL-fish RL-mother la lǝɹa-ǝ̀ɹ-ùɁ ɹǝ TOP descend-come.PRS-DEC PTCL ‘And, Noksensangba [said] “Ah! A huge fish is coming down [the river]” (Coupe 2007, 460) Surprisingly, suffixation was attested only in one (Kham) out of 10 augmentivizing languages: (105)

chya-ldyu a.little.bit-AUG ‘enormous’ (Watters 2002, 140)

The criterion of cognitive category displays an unexpected result, too– the category of Quantity of Substance was attested as many times as Quantity of Quality (5 languages). Quantity of Action was present in 2 languages (Agta and Meithei) and Quantity of Circumstance in Agta: (106)

fu-fúráb ‘late afternoon’ < fúráb ‘afternoon’

Research

87

The majority of languages express only one cognitive category–either Quantity of Substance or Quantity of Quality. Meithei can express two cognitive categories, Quantity of Quality and Action–surprisingly none of them is Quantity of Substance. Similarly, Agta can express all cognitive categories except for Quantity of Substance. The criterion of word class maps this situation. Nouns were attested in 6 languages, and adjectives in 4 languages. Agta makes use of 4 word classes, one of them being pronouns: (107)

ad-áddu ‘very many’ < addu ‘many’

Correlations The correlations, too, deviate from the situation in the other language areas. They do not follow the more or less common pattern of suffixation – Quantity of Substance – nouns. The strongest correlation was attested between Quantity of Substance and nouns (5 languages, except for Ao belonging to the Austronesian family) and reduplication and Quantity of Quality (4 languages, except for Agta all of them spoken in South-Asia). Only three languages feature a correlation between compounding, Quantity of Substance and nouns (Nelemwa, Siar Lak, Ao). Reduplication, Quantity of Quality and adjectives co-occur in Agta, Tibetan, and Thai. From the semantic point of view, my data gives support to Matisoff (1991) who maintains that in the SE Asian languages a lexical source for an augmentative morpheme is mother; child is the source for a diminutive morpheme. According to Coupe (2007, 273), for example, the diminutive suffix -za in Ao has grammaticalized from the relational noun tə-za (RLchild) ‘child’, which is no longer used independently as a lexical noun. Similarly, there is a morpheme that appears to be in the process of grammaticalizing from the lexical noun tə-ji ‘mother’ (RL -mother). In Siar, augmentivisation is achieved by forming a nominal compound involving the noun _ta-_ ‘mother’, e.g. (Frowein, pers. com.): (108)

_tan liwan_ mother knife ‘big knife / bush knife’

And, as exemplified in (15), Nelemwa also employs the ‘mother’ morpheme for augmentatives.

Chapter Two

88

2.4.6 Languages of Eurasia WALS identifies 386 languages representing 39 genera and 19 language families in the area of Eurasia. When eliminating the languages spoken outside the postulated SAE area that belong to this area, my sample covers 19 languages. To make the picture of evaluative morphology in the world languages complete, these 19 languages were grouped into one sample labelled Eurasia. The 19 languages in this sample represent 10 language families and 12 genera. Map 2-6 Languages of Eurasia

The three languages without evaluative morphology are from different language families: Nepali is an IE language, Kamssaian is a Uralic language and Mangghuer an Altaic language. 16 languages have evaluative morphology. 5 of them have only diminutives and 1 language, Southern Dravidian Kannada, forms only augmentatives. In 10 languages, both diminutives and augmentatives were attested. The highest total EM saturation was identified for the Tungusic language Udihe (4.33), followed by Nenets (4.00) from the Samoyedic genus, and Japanese (3.33). The lowest EM saturation (1.00) was calculated for Yeniseian Ket and Southern Dravidian Malayalam. The highest DIM saturation was calculated for Nenets and Udihe (both 2.33). In addition, Udihe is the language with the highest AUG saturation.

Research

89

Table 2-6 Languages of Eurasia – saturation

Gujarati

EM satur. 2.00

DIM satur. 1.00

AUG satur. 1.00

Hindi

1.33

1.33

0

Nepali

0

0

0

1.33

1.33

0

Iranian

Trinidad Bhojpuri Kurmanji

1.33

1.33

0

Samoyedic

Nganasan

2.67

1.67

1.00

Family

Genus

Language

IE

Indic

Uralic

Altaic

Tungusic

Nenets

4.00

2.33

1.67

Kamassian

0

0

0

Selkup

1.67

1.67

Udihe

4.33

2.33

2.00

Mongholic Dravidian

Southern Dravidian

0

0

0

Mangghuer Kannada

1.67

0

1.67

Malayalam

1.00

1.00

0

Ainu

Ainu

Ainu

2.33

1.33

1.00

ChukotkoKamchatkan

Itelmen

2.00

1.00

1.00

Japanese

Southern ChukotkoKamchatkan Japenese

Japanese

3.33

2.00

1.33

Yeniseian

Yeniseian

Ket

1.00

1.00

0

Nivkh

Nivkh

Nivkh

2.33

1.00

1.33

Yukaghir

Yukagir

Kolyma

2.00

1.00

1.00

1.81

1.12

0.68

TOTAL SATURATION

Diminutives Word-formation The most frequent word formation process exploited in the evaluative morphology of Eurasian languages is suffixation. Out of 15 languages with diminutives it was attested in 14 of them. As an example of high productivity of diminutive suffixes in this cohort of languages, Nikolaeva

Chapter Two

90

and Tolskaya (2001, 149) maintain that the Udihe diminutive suffix -zig’a is “the most productive suffix of all nominal derivational suffixes and attached to practically any noun with concrete semantics.” (109)

b’ata-zig’a boy-DIM ‘little boy’ zolo-zig’a stone-DIM ‘small stone, pebble’

The only exception to the suffixation tendency is Malayalam which relies on compounding. Compounds are usually coined to express the meaning ‘young one’. The determinatum in these compounds is kuʈʈi ‘child’ (Asher and Kumari 1997, 398): (110)

paʈʈikkuʈʈi dog-young ‘puppy’ puuccakkuʈʈi cat-young ‘kitten’

Another word formation process is prefixation attested in Japanese (111) and Ainu (112): (111)

(112)

ko-gawa DIM-river ‘small river’ (Volpe, pers. com.) ru-sesek ATT-be.warm

‘lukewarm’ (Refsing 1986, 178) Both Japanese and Ainu make use of not only prefixation, but also employ suffixation for evaluative formation. Similar to Malayalam (kuʈʈi), the Japanese (ko-) and Ainu (-po) diminutive markers originate from the same word meaning ‘child’.

Research

(113) Japanese

ko-inu DIM-dog ‘puppy’ (Volpe, pers. com.)

(114) Ainu

cep-po fish-DIM ‘small fish’(Refisng 1986, 159)

91

In comparison to the other geographical areas, the internal structure of individual word-formation processes is fairly rich. The use of more than one diminutive suffix is common (e.g., Bhojpuri, Udihe, and Nenets). Diminutive suffixes can have other meanings, too. In Bhojpuri, for example, diminutive terms of kinship also express affection, and when referring to older relatives, they also express respect. This function is observed only in vocative forms (Mohan 1978, 51): (115)

ē chaũriyā, kā karat bā? ‘Hey, little girl, what are you doing?

When the Japanese diminutive suffix -ko ‘child’ is attached to a place name, it refers to a person who comes from this place and has grown up there all his/her life (Jurafsky 1996, 547): (116)

edo-kko edo (old Tokyo)-child ‘a person who comes from Tokyo (i.e., the child of Tokyo)’

Cognitive categories 14 out of 15 languages can express Quantity of Substance. The exception is Nivkh which expresses Quantity of Action: (117)

xavu-jowarm.sth-DIM ‘warm sth. a little’ (Gruzdeva 1998, 30)

Quantity of Action can be expressed in 5 languages and Quantity of Quality in 3 languages. With the exception of Nivkh, the category of Substance is either the only category expressed, or it is combined with Quantity of Action/Quality. On the other hand, Quantity of Action and Quantity of Quality never occur as the only cognitive categories. Quantity of Circumstance was not attested in the sample.

Chapter Two

92

Word classes By implication, the word class of nouns is the most frequently attested–it was attested in 14 languages with the exception of verbs in Nivkh. All languages make use of nouns (except for Nivkh). Verbs are used in 5 languages and adjectives in 4 of them. Udihe is the only language with 3 word classes–nouns, verbs, adjectives. Adverbs were not identified in the sample. Correlations Suffixation correlates with the cognitive category Quantity of Substance in 13 languages. The same number of languages manifests the correlation between suffixation and nouns. The cognitive category Quantity of Substance correlates with nouns in 14 languages. By implication, the correlation between suffixation, Quantity of Substance and nouns was identified in 13 languages. Augmentatives Augmentatives can be morphologically formed in 10 out of 19 Eurasia sample languages. Only in 1 language, Kannada, they exist without diminutives. The evaluation is very similar to that of diminutives. The most frequent word-formation process is suffixation, attested in 7 languages; prefixation is used in 4 languages: (118) Ainu

ar-askay INT-able ‘extremely able’ (Refsing 1986, 178)

(119) Japanese

chō-kyokka-shugi AUG-national-ism ‘ultranationalism’ (Volpe, pers. com.)

(120) Gujarati

əti-krur AUG-cruel ‘very cruel’ (Cardona 1965, 162)

(121) Kannada

hemmara AUG-tree ‘giant tree’ (Aronoff and Sridhar 1988, 182)

Research

93

No other word-formation process is used for the purpose of augmentative formation. No combinations of word-formation processes were observed. The cognitive category Quantity of Substance was identified in 7 languages; in 5 languages it is the only cognitive category expressed. Quantity of Action was present in 5 languages. In Nivkh and Ainu it is the only cognitive category used. Quantity of Quality was found in Udihe and Nenets but it never occurs as a single category. Both Udihe and Nenets use augmentatives to express at least two cognitive categories: Quantity of Substance and Quantity of Quality (Nenets)’, Quantity of Quality and Quantity of Action (Udihe). No examples of Quantity of Circumstance were observed. Nouns were identified in 6 languages, Verbs only in 3 (Udihe, Nivkh and Ainu) and adjectives in 3 (Gujarati, Nenets and Udihe). Only two word classes are used in Nenets (nouns and adjectives) and Udihe (verbs and adjectives). No significant correlations were found. Suffixation, Quantity of Substance and Nouns were attested in 4 languages. Suffixation, Quantity of Action and verbs co-occur in 3 languages spoken in the eastern part of Eurasia (Udihe, Nivkh, Ainu).

2.4.7 World sample The world sample consists of 132 languages. 83 (62.88%) of them have evaluative morphology; 49 (37.12%) have no evaluative morphology. 29 languages (21.97%) can productively form only diminutives and 6 languages (4.55) only augmentatives. Both diminutives and augmentatives are productively formed in 48 languages (36.36%). The highest percentage of the occurrence of evaluative morphology is demonstrated in South America (86.67%), followed by Eurasia (84.21%). There is nearly a 20% gap between Eurasia and North America (65.22%). Australia and New Guinea (40.09%) have the lowest number of EM languages. A similar ranking of geographical areas is in terms of DIM/AUG presence in languages. 60% of languages in South America productively form both diminutives and augmentatives. South America is followed by Eurasia (52.63%) and North America (34.78%). The lowest percentage (24.14) was identified for Africa. The percentage values for SE Asia + Oceania are roughly the same (29.67%). The highest proportion of languages that produce only diminutives is in North America (30.43%) and Africa (30.03%). The region South Asia + Oceania has the highest proportion of languages forming only augmentatives (12.5%). This area is followed by South America (6.67%). An overview of this sort of data is given in Table 2-7:

Chapter Two

94

Table 2-7 Overview of evaluative morphology by areas

10

7

31.82

47

52.63

29.67

20.69

34.78

60

7

35.60

6

6.67

0

0

1

3.45

3

6

12.5

1

8

5.26

3 0

0

9

0.00

83

1

4.55

2

%

20

9

%

30.43

5

%

34.03

16

DIM+ AUG

12.5

3

only AUG

26.32

86.67

49

65.22

132

58.62

13

54.67

Australia+ New Guin.

22

13

84.21

3

1 0

40.09

Eurasia

19

17

62.88

11

13.33

SE Asia + Oceania

24

34.78

12

7

41.38

Africa

29

15

45.83

8

3

15.79

South America

23

13

59.09

2

%

37.12

North America

15

Total

%

only DIM

9.09

EM

22.73

no EM

Area

No. of lang.

Table 2-7 shows the ranking of language areas according to five criteria: no evaluative morphology, occurrence of evaluative morphology, only diminutives, only augmentatives, both diminutives and augmentatives. It follows from Table 2-8 that South America occupies the top position in the

Research

95

columns EM and DIM+AUG. Australia and New Guinea ranks at the bottom. Table 2-8 Ranking of the areas by the data of Table 2-7 EM

DIM only

South America

North America

Eurasia

Africa

AUG only SE Asia+Oceania South America

DIM+AUG

North America

Eurasia

Eurasia

Africa

South America

Africa

SE Asia+Oceania Australia+New Guinea


Australia+New Guinea

North America Australia+New Guinea SE Asia+Oceania

North America

Africa

South America Eurasia

These results are further supported by the criterion of EM saturation. Table 2-9 gives the total EM saturation, the DIM saturation and the AUG saturation for the individual areas: Table 2-9 EM, DIM and AUG saturation values by areas EM saturation 0.98

DIM saturation 0.68

AUG saturation 0.28

1.16

0.68

0.60

1.52

1.00

0.64

South America

2.00

0.87

0.84

SE Asia+Oceania

1.26

0.79

0.51

Eurasia

1.81

1.12

0.68

Africa Australia+New Guinea North America

Table 2-10 ranks the language areas according to the EM, DIM and AUG saturation values. Clearly, South America, Eurasia and North America have the highest levels of saturation. On the other hand, the values for SE Asia + Oceania, Australia + New Guinea and Africa are rather low.

Chapter Two

96

Table 2-10 Ranking of the areas by the data of Table 2-9 EM saturation

DIM saturation

AUG saturation

South America

North America

South America

Eurasia

Eurasia

Eurasia

North America

South America


SE Asia+Oceania

SE Asia+Oceania

North America

Australia+New Guinea Africa

Africa Australia+New Guinea

SE Asia+Oceania Africa

The data for South America and Eurasia are very similar. South America, represented by 15 languages, has 2 languages without EM. Eurasia with 19 languages has 3 languages without EM. In both of these areas, the saturation values are analogical: minimum number of languages with saturation 1, the saturation values range from 1.33 to 3.33. An exception to this tendency is the South American language Yurakare (5.00) and the Eurasian languages Udihe (4.33) and Nenets (4.00). A similar situation can also be found in the languages of North America. The values range from 1.33 to 3. The only exception is Kwakw’ala (6.33). The data for Africa is different. 8 languages out of 17 have saturation as low as 1.00. Only 2 languages have a saturation above 3. All EM saturation values for this area fall below 4.00, with the highest being that of Bafut (3.67). The saturation values in the remaining two areas are scattered. Australia + New Guinea is the area with the lowest percentage for EM languages (40.09%). Languages spoken in this area have either no EM, or the saturation value is high; 3 languages have the EM saturation of 4.00, and two of them above 3.00. Similarly, SE Asia + Oceania, the area with the highest “no EM” percentage (59.09), has 2 languages with EM saturation above 4.33. One language, Agta, has an EM saturation value 5.67. On the other hand, 4 languages (out of 13 with EM) have a saturation of 1.00. Tables 2-11, 12, 13 show the top languages by EM saturation, DIM saturation and AUG saturation, respectively.

Research

97

Table 2-11 Ranking of the sample languages by EM saturation Family

Kwakw’ala

Wakashan

Agta

Austronesian

South America

Yurakare

Yurakare

Australia+New Guinea SE Asia+Oceania SE Asia+Oceania Eurasia Australia+ New Guinea Australia+ New Guinea Eurasia

Arabana-

Australian

Karao

Austronesian

Muna

Austronesian

Sulawesi

4.33

Udihe

Altaic

Tungusic

4.33

Alawa

Australian

Maran

4.00

Sentani

Sentani

Sentani

4.00

Nenets

Uralic

Samoyedic

4.00

Africa North America Australia+ New Guinea

Bafut

Niger-Congo

Bantoid

3.67

Micmac

Algic

Algonquian

3.67

Madang

3.67

Gur

3.34

North America SE Asia+Oceania

Genus

EM satur.

Language

Area

Northern Wakashan Northern Philippines Yurakare PamaNyungan Northern Philippines

6.33 5.67 5.00 4.67 4.33

Africa

Konni

Trans-New Guinea Niger-Congo

South America

Jaqaru

Aymaran

aymaran

3.33

Eurasia

Japanese

Japenese

Japanese

3.33

Anamuxra

Chapter Two

98

Table 2-12 Ranking of the sample languages by DIM saturation Area SE Asia+Oceania SE Asia+Oceania North America South America Africa SE Asia+Oceania Australia+ New Guinea Eurasia

Language

Family

Genus

DIM satur.

Karao

Austronesian

Northern Philippines

3.33

Muna

Austronesian

Sulawesi

3.33

Kwakw’ala

Wakashan

Northern Wakashan

3.33

Yurakare

Yurakare

Yurakare

3.33

Bafut

Niger-Congo

2.67

Agta

Austronesian

Bantoid Northern Philippines

2.67

Trans-New Guinea Uralic

Madang

2.33

Samoyedic

2.33

Altaic

Tungusic

2.33

Mayan

Mayan

2.00

Lakota

Siouan

Siouan

2.00

Anamuxra Nenets Udihe Yucatec Maya

Eurasia North America North America North America Australia+ New Guinea Australia+ New Guinea Australia+ New Guinea Australia+ New Guinea Africa

Tutelo

Siouan

Siouan

2.00

Bachamal

Australian

Anson Bay

2.00

Arabana

Australian

PamaNyungan

2.00

Sentani

Sentani

Sentani

2.00

Alawa

Australian

Maran

2.00

Supyire

Niger-Congo

Gur

1.67

Eurasia

Nganasan

Uralic

Samoyedic

1.67

Research

99

Table 2-13 Ranking of the sample languages by AUG saturation Area

Family

Agta

Austronesian

Kwakw’ala

Wakashan

Djaru

Australian

Arabana

Australian

Jaqaru

Aymaran

Aymaran

2.33

Micmac

Algic

Algonquian

2.00

Udihe

Altaic

Tungusic

2.00

Alawa

Australian

Maran

2.00

Sentani

Sentani

Sentani

2.00

Yurakare

Yurakare

Yurakare

1.67

Konni

Niger-Congo

1.67

Eurasia

Kannada

Dravidian

Eurasia

Nenets

Uralic

Gur Southern Dravidian Samoyedic

SE Asia+Oceania North America Australia+ New Guinea Australia+ New Guinea South America North America Eurasia Australia+ New Guinea Australia+ New Guinea South America Africa

Genus

AUG satur.

Language

Northern Philippines Northern Wakashan PamaNyungan PamaNyungan

3.00 3.00 2.67 2.67

1.67 1.67

The tables confirm the above comments. Africa, with its generally low values, appears rarely in the three tables above. Eurasia is represented by 3 languages in each of the tables, and South America by 2 in both Table 211 and Table 2-13 and by 1 language in Table 2-12. On the other hand, the Australian + New Guinea area has 5 languages in Table 2-12, 4 languages in Table 2-11 and 3 languages in Table 2-13. Similarly, SE Asia + Oceania have 3 languages in Table 2-11 and Table 2-12, but only one in Table 213. The tendencies observed can be formulated as follows: The regions with very high (Eurasia, South America) and very low (Africa) total EM saturation values are characterized by remarkable homogeneity of EM saturation values of individual languages. On the other hand, the areas

Chapter Two

100

with EM values between the two extremes (SE Asia + Oceania and Australia + New Guinea) feature considerable scattering of EM saturation values for individual languages. The most frequent word-formation processes used for the formation of diminutives are summarized in Table 2-14. The number next to the area name gives the number of languages with EM. The table makes it possible to conclude that the most frequent word-formation process is suffixation. It is typical especially true of the languages of Africa and Eurasia. In Australia + New Guinea and SE Asia + Oceania, a competing wordformation process is reduplication. In SE Asia + Oceania, reduplication is sometimes combined with infixation and prefixation. Other processes not included in Table 2-14 are stem alternation in North America, the change of paradigm in South America, and circumfixation in SE Asia + Oceania. The general tendency is the use of one word-formation process, mainly suffixation. A strong tendency towards change of paradigm is observable in the languages of Africa, and reduplication is preferred in languages of Australia & New Guinea.

Prefixation

Reduplication

Compounding

Change of paradigm

Africa (17) Australia+ New Guinea (9) North America (15)

10

2

1

1

5

3

0

8

0

0

9

0

2

2

0

South America (13)

8

0

1

1

0

SE Asia+Oceania (13)

6

2

4

0

0

Eurasia (16)

14

2

0

1

0

Area (No. of ang.)

Wordformation process

Suffixation

Table 2-14 Word-formation processes in diminutive formation

Table 2-15 gives the same kind of information for augmentatives:

Research

101

Prefixation

Reduplication

Compounding

Change of paradigm

Africa (17) Australia+ New Guinea (9) North America (15)

3

1

1

0

2

0

0

7

0

0

6

1

1

1

0

South America (13)

7

1

2

1

0


1

0

5

3

0

Eurasia (16)

7

3

0

0

0

Area (No. of lang.)

Wordformation process

Suffixation

Table 2-15 Word-formation processes in augmentative formation

The data indicates that the suffixation tendency for augmentatives is not as strong as that for diminutives. Languages of SE Asia + Oceania and Australia and New Guinea prefer reduplication. The word-formation processes not covered by Table 2-15 include stem alternation in 2 languages spoken in South America (Moseten) and North America (Diegueno) and reduplication with prefixation in SE Asia + Oceania (Muna). Cognitive categories and word classes Table 2-16 summarizes the number of diminutive cognitive categories expressed in the individual geographical areas. It demonstrates the preference for Quantity of Substance, followed by Quantity of Quality. The occurrence of the cognitive category Quantity of Circumstance is rare. The general tendency observed is the expression of a single cognitive category, mainly Quantity of Substance.

Chapter Two

102

Circumstance

Action

Quality

Area (No. of lang.)

Substance

Cognitive categries (Quantity of …)

Table 2-16 Cognitive categories expressed by diminutives

Africa (17)

14

4

2

0

Australia+New Guinea (9)

5

5

1

1

North America (15)

14

5

6

1

South America (13)

9

4

2

0


8

4

4

1

Eurasia (16)

14

3

5

0

The same comment applies to Table 2-17 presenting the augmentative cognitive categories.

Circumstance

Action

Quality

Area (No. of lang.)

Substance

Cognitive categries (Quantity of …)

Table 2-17 Cognitive categories expressed by augmentatives

Africa (17)

5

1

1

0


1

5

2

0

North America (15)

6

5

3

1

South America (13)

6

3

3

0


5

5

2

1

Eurasia (16)

7

2

4

0

Research

103

The word class table provides support to the afore mentioned cognitive category tables. Nouns have the highest frequency of occurrence, followed by adjectives.

AUG

DIM

AUG

DIM

AUG

DIM

AUG

DIM

AUG

Pronouns

Change of Class

DIM

Adverbs

Adjectives

AUG

Verbs

DIM

Word class

Nouns

Table 2-18 Word classes by areas

Africa (17)

14

5

2

1

4

0

0

0

0

1

1

0

Australia+New Guinea (9) North America (15) South America (13) SE Asia+ Oceania (13) Eurasia (16)

5

1

1

1

6

5

1

3

1

0

0

0

12

6

7

4

6

5

1

1

0

0

0

0

8

6

3

3

4

4

1

0

0

0

0

0

8

6

4

2

6

4

1

0

0

0

0

1

14

6

5

3

3

3

0

0

0

0

0

0

Area (No. of lang.)

Based on the preceding analysis, some general tendencies and observations can be formulated. The areas with the highest total EM saturation values – South America, Eurasia, and North America – mainly use suffixation for the coining of diminutives and augmentatives. Suffixation usually correlates with the cognitive category of Quantity of Substance and with nouns. In addition, the languages of both Americas display some peculiarities such as clitics, change of vowel and/or nasality in nouns, and stem alternation. Phonetic iconicity is fairly common. The change of paradigm is typical of Africa, the area with homogeneous but low saturation values. Two significant correlations have been identified: suffixation – Quantity of Substance – nouns and change of paradigm – Quantity of Substance – nouns, especially in Niger-Congo languages. Gender shift is also characteristic of African languages. In the languages of SE Asia + Oceania, Quantity of Substance and nouns correlate mainly

Chapter Two

104

with reduplication. Similarly, in the languages of Australia & New Guinea, reduplication correlates with Quantity of Substance and Quantity of Quality, (and) nouns and adjectives.

2.4.8 SAE sample As explained in 1.2.3, the starting point for the delimitation of SAE languages was the geographical criterion. It was combined with the regional section of WALS. In this way, 145 languages were chosen from 7 language families. My SAE sample counts 71 languages representing each family with the exception of Indic (see centrefold, map 2.7). Table 2-19 illustrates the number of genera and languages for individual language families. The first column gives the number of genera/languages in WALS, the second column is the number of genera/languages on the SAE territory and the last column gives the number of genera/languages in my sample. Table 2-19 Languages of SAE Family Altaic Afro-Asiatic Basque Indo-European Kartvelian NakhDaghestan Northwest Causasian Uralic

Number of genera WALS SAE Sample 3 2 1 12 1 1 1 1 1 10 10 9 1 1 0

Number of languages WALS SAE Sample 65 18 3 141 4 1 12 1 1 176 68 43 4 4 0

4

4

4

28

28

8

1

1

1

7

5

3

3

2

2

27

18

13

While the number of languages covered in my sample is high compared to the majority of the preceding research projects in this area, the relatively large gap between the total number of the SAE languages covered and the number of languages given by WALS, on one hand, and the number of languages on my list, on the other, is motivated by practical reasons. First, WALS lists, unlike my sample, all varieties of one language, e.g. the number of ‘Germans’ listed in WALS is 16. Second, the only genus not represented in my sample is Indic. In the SAE territory it is represented by Romani. It was not included for two reasons: (i) There are

Research

105

at least 8 varieties of Romani and (ii) the varieties are strongly influenced by the contact with surrounding languages. As a result, Romani spoken in Slovakia, for example, is very different from Romani spoken in Hungary. Table 2-20 summarizes the EM saturation values. The highest EM saturation value is 7.67, calculated for some Romance and Slavic languages (Italian, Occitan, Slovak, and Slovene). A little lower value (7.33) characterizes Portuguese, Belorussian and Czech; Provencal and Polish have 7.00. The highest EM saturation outside the Romance and Slavic genera is that of Greek (6.66), followed by Turkish (5.00) and Basque (4.34). Table 2-20 EM saturation values by languages Family

Genus

Indo-European

Romance

Slavic

Catalan

EM satur. 4.00

DIM satur. 2.33

AUG satur. 1.67

French

5.34

2.67

2.67

Italian

7.67

4.67

3.00

Ladino

4.67

3.00

1.67

Occitan

7.67

3.67

4.00

Portuguese

7.33

4.33

3.00

Language

Provencal

7.00

3.67

3.33

Romansch

5.34

2.67

2.67

Roman

4.67

2.67

2.00

Spanish

6.00

3.33

2.67

Belorussian

7.33

3.67

3.67

Bulgarian

5.33

2.67

2.67

Croatian

3.00

2.00

1.00

Czech

7.33

4.00

3.33

Kashubian

5.33

3.33

2.00

Macedonian

4.66

3.33

1.33

Lower Sorbian

5.67

3.67

2.00

Polish

7.00

3.67

3.33

Russian

5.67

3.67

2.00

Slovak

7.67

3.67

4.00

Chapter Two

106

Germanic

7.67

4.00

3.67

Ukrainian

5.33

3.00

2.33

Upper Sorbian

6.00

4.00

2.00

Serbian

7.67

4.66

3.33

Danish

2.00

1.00

1.00

Dutch

4.00

1.67

2.33

English

2.33

2.33

0

Faroese

1.33

1.33

0

German

4.34

2.67

1.67

Icelandic

2.33

2.33

2.33

Luxemburgish

1.33

1.33

0

Norwegian

4.00

1.33

2.67

Breton

3.67

2.67

1

Irish

1.33

1.33

0

Scottish Gaelic

3.00

1.33

1.67

Welsh

2.00

2.00

0

Latvian

3.00

2.00

1.00

Lithuanian

4.66

3.33

1.33

Ossetic

3.33

2.33

1

Zazaki

5.33

3.33

2.00

Albanian

Albanian

1.33

1.33

0

Armenian

Armenian

2.00

2.00

0

Greek

Greece

6.66

3.33

3.33

Avar-AndicTsezic

Khvarsi

0

0

0

Lak-Dargwa

Dargwa

0

0

0

Aghul

0

0

0

Archi

0

0

0

Lezgian

0

0

0

Tabassaran

0

0

0

Celtic

Baltic

Iranian

Nakh-Daghestanian

Slovene

Lezgic

Research

Nakh

1.33

0

Ingush

1.00

1.00

0

Karelian

2.00

2.00

0

Estonian

2.00

2.00

0

Finnish

2.00

1.00

1.00

Votic

2.33

2.33

0

Mari meadow

1.33

1.33

0

Mordvin

1.67

1.67

0

Komi-Permyak

0

0

0

Udmurt

1.67

1.67

0

Saami

1.67

1.67

0

Mansi

3.33

3.33

0

Khanty

0

0

0

Hungarian

3.33

3.33

0

Adyghe

1.00

1.00

0

Kabardian

1.00

1.00

0

Ubykh

0

0

0

Gagauz

1.66

1.66

0

Altaic

1.33

Turkic

Tatar

4.00

2.67

1.33

Turkish

5.00

3.00

2.00

Basque

Udi

Basque

Basque

4.34

2.67

1.67

Afro-Asiatic

Uralic

Finnic

107

Semitic

Maltese

3.00

North west Caucas.

Ugric

Northwest Caucasian

AVERAGE SATURATION VALUE

3.4

2.00

1.00

2.18

1.26

The following sections discuss each language family in the SAE area from the point of view of evaluative morphology.

Chapter Two

108

Afro-Asiatic There are four Afro-Asiatic languages in the SAE area. In my sample, they are represented by Maltese. Maltese forms both diminutives and augmentatives. The word-formation process used for their coining is suffixation, and also root-and-pattern in the case of diminutives. According to Zammit (pers. com.), the morphology of the Maltese diminutive includes insertion of the infix -eyye/a- and -ayye/a- between the 2nd and 3rd radicals of nouns of Arabic origin: (122)

tfayyel ‘little boy’ < tifel ‘boy’

The cognitive categories expressed are Quantity of Substance and Quantity of Quality. The word classes used are nouns and adjectives. Altaic The Altaic language family is represented by three languages in my sample. All languages are from the Turkic genus. Two languages, Turkish and Tatar, have both diminutives and augmentatives, whereas Gagauz has only diminutives. The languages use only one word-formation process for the coining of diminutives–suffixation. Tatar and Turkish have more than one diminutive suffix. In addition, Turkish diminutive formation makes use of recursiveness of diminutive suffixes: (123)

yavaş-ça-cık slow-DIM-DIM ‘very slow’ (Ketrez, pers. com.)

Furthermore, the same diminutive suffix can be used more than once in a word: (124)

küçü-(k)çü-(k)çük small-DIM-DIM ‘very very small’ (ibid.)

As communicated by my informant, Ilmira Miftachova, in Tatar the term diminutive is only used for nouns, but the Tatar verb has also some morphemes that can have a diminutive function such as -štır in example (125). They are usually described as ‘Aktionsarten’ of the verb:

Research

(125)

109

qara-štır look.after-DIM ‘to look after once in a while’

While Gagauz expresses two diminutive cognitive categories, Quantity of Substance and Quantity of Quality, Tatar and Turkish can express three of them–Quantity of Substance, Quantity of Quality, and Quantity of Action and Circumstance, respectively. Word classes correspond with cognitive categories–Gagauz uses nouns and adjectives, Tatar and Turkish nouns, adjectives and verbs/adverbs. Both Tatar and Turkish use reduplication for augmentivisation, both partial and complete. The cognitive category Quantity of Quality dominates. Turkish also expresses Quantity of Circumstance. While Miftachova (pers. com.) points out that the term ‘augmentative‘ is not used in Tatar grammars, examples can be found like that in (126): (126)

zur~zur big~big ‘very big’

Basque Even though there are twelve languages listed in WALS for the Basque family, all of them are dialects of the Basque language. This is also the reason why these languages are represented by only one in my sample. Although Basque is a separate language family, it does not manifest any substantial deviations from the other SAE languages in terms of evaluative morphology. Consequently, while Haspelmath (2001) eliminates Basque from SAE area it is a part of SAE based on the EM parameter – an important piece of evidence that the definite borders of the SAE territory can only be delimited after complex evaluation of all the relevant parameters. Basque produces both diminutives and augmentatives. The main wordformation process used is suffixation and, in the case of augmentatives, also reduplication. There are multiple diminutive suffixes and (typically) complete reduplication in Basque: (127)

handi-handi big-big ‘very big’ (Hualde and de Urbina 2003, 462)

Chapter Two

110

Quantity of Substance, Quantity of Quality and, in the case of diminutives, also Quantity of Circumstance are expressed by corresponding word classes. Nakh Daghestan The Nakh-Daghestan language family is represented by eight languages in my sample. None of them have augmentatives and only two of them have diminutives. They primarily make use of suffixation; but one particular language, Udi, also employs reduplication. Ingush expresses Quantity of Substance and Udi Quantity of Quality. As Ganenkov (pers. com.) points out, Udi borrowed the means to express a high degree of quality from Azerbaijani: (128)

map~macːi RDP~white ‘very white’

Northwest Caucasian There are 3 languages from the Northwest Caucasian family in my sample. One of them, Ubykh, has no evaluative morphology. Adyghe and Kabardian have no augmentatives and the DIM saturation is very low. Diminutives are coined by means of suffixation, and only the sematic category Quantity of Substance is expressed by means of nouns. Uralic My sample encompasses twelve languages representing two genera. Each of the languages has some evaluative morphology, but, with the exception of Finnish, only diminutivisation was attested. Furthermore, there is only one word-formation process used–suffixation. Hungarian, Karelian, Mansi, and Votic each have multiple diminutive suffixes. Estonian allows the use of a diminutive suffix to be used more than once in the same word as illustrated in (129b.) in contrast to (129a.): (129) a.

b.

kassi-kene cat-DIM ‘little cat’ kassi-kese-kese-kene cat-DIM-DIM-DIM ‘very little cat’ (Kilgi, pers. com.)

Research

111

Recursiveness is also possible in Mari (130): (130)

tol-al-al-aš come-DIM-DIM-INF ‘come (affectionate)’ (Alhoniemi 1993, 152)

An interesting fact is that the highest EM saturation is observable in Hungarian and Mansi (3.33), the languages surrounded by Slavic languages with a very high saturation (Slovak 7.67, Ukrainian 5.33, Russian 5.67). The high saturation values of Hungarian and Mansi especially reflect the fact that both can express all four cognitive categories, while the rest of the Uralic languages usually express only Quantity of Substance and Quantity of Quality. The following example illustrates the otherwise rarely used category Quantity of Circumstance: (131)

lassa-cskán slowly-DIM ‘very slowly’ (my example)

The only Uralic language with (very few) augmentatives, Finnish, coins them by means of compounding (132) and reduplication (133): (132)

(133)

emä-munaus dam-blunder ‘a really big blunder’ (Laakso, pers. com.) suuren~suuri RDP~big

‘very big’ (Nenonen, pers. com.) Indo-European The Indo-European language family is the most numerous; my sample counts 42 languages from 9 genera. To get a better picture of the situation, the next discussion focuses on individual genera rather than the language family as a whole. Romance There are 10 Romance languages in my sample, and each language has both diminutives and augmentatives. Each of these languages makes use of suffixation in the process of diminutivisation. Suffixation is rich in terms of its productivity in evaluative formation as well as in terms of the

Chapter Two

112

number of evaluative suffixes per languge. In addition, a particular diminutive suffix can in some cases be used more than once in a word: (134) Spanish

poqu-it-it-o little-DIM-DIM-NOM.SG ‘vey very little’ (Suarez, pers. com.)

Some of the Romance languages also allow more than one diminutive suffix in a word: (135) Italian

bimb(o)-ett(o)-in-o child-DIM-DIM-M.SG ‘small little child’ (Grandi, pers. com.)

In addition to this rich suffixation, French, Portuguese, Provencal and Romansch use reduplication: (136) Provencal

pichoun-pichoun small-RDP ‘very small’ (Blanchet, pers. com)

Prefixation, represented especially by the prefixes of Latin origin (e.g. mini), is used by Catalan (137) and Italian: (137) Catalan minibàsquet DIM-basketball ‘junior basketball’ (Wheeler, pers. com.)

Occitan shows a rare case of diminutivisation by gender shift from feminine to masculine: (138) dalh-0 (masculine) ‘little scythe’ < dalh-a (feminine) ‘scythe’ (Sauzet, pers. com.) Each sample language can express the cognitive category Quantity of Substance, and the majority of them also Quantity of Quality and Quantity of Action. Even the relatively rare Quantity of Circumstance is expressed in as many as 7 languages. 5 languages can express all four cognitive categories, 4 languages express three and Catalan only two. As the morphology of Romance languages is based on Latin, the cognitive categories usually correspond with the word classes. Verbs are fairly

Research

113

common. The categories of Aktionsart are commonly represented in evaluative formation within the cognitive category of Action, and the same is true of the cognitive category Attenuation: (139) chuviscar ‘to drizzle’ < chover ‘to rain’ (Pöll, pers. com.) The strongest correlation is between suffixation, Quantity of Substance and nouns. Very strong also is the correlation between suffixation, Quantity of Quality and adjectives, attested in eight languages. Reduplication usually correlates with Quantity of Substance and nouns. The situation of augmentatives is very similar. The most frequently attested word-formation process is suffixation with rich specifications. Reduplication was identified in Provencal and Romansch. Prefixation, based on Latinate prefixes, was found in French, Italian, Occitan and Provencal. Gender shift was only attested in Italian: (140)

buc-a hole-F ‘big hole’

buc-o hole-M ‘hole’ (Grandi, pers. com.)

Each of the sample languages expresses at least two cognitive categories, usually Quantity of Substance and Quantity of Quality/Action. Occitan can express all of them. Slavic There are 14 Slavic languages in my sample, and each of them productively coins both diminutives and augmentatives. Each language uses suffixation for evaluative formation, expresses the cognitive category Quantity of Substance, and forms evaluative nouns. Suffixation in Slavic languages is rich, too. Only two languages, Macedonian and Bulgarian, use one word-formation process (suffixation), the rest of the languages use at least 2 word-formation processes. The highest number of processes is used by Slovak and upper Sorbian; both languages make use of suffixation, prefixation, infixation, and prefixal-suffixal derivation. The word-formation process that was not identified in other language families within SAE is the prefixal-suffixal derivation that usually correlates with Quantity of Action and verbs: (141) Belorussian

па-свiст-ва-ць ‘to whistle a little bit’ < свiстаць ‘to whistle’ (Rudenka and Rudaya, pers. com.)

Chapter Two

114

(142) Kashubian

(po)baw-k-ac sã ‘to play a little’ < bawic sã ‘to play’ (Pazdjerski and Makuraté, pers. com.)

Infixation as a process used for the coining of diminutives occurs in three languages (Slovak, Slovene, Upper Sorbien). The next example comes from Slovene: (143) stop-ic-lj-ati ‘to step lightly’ < stop-ic-a-ti < stopiti ‘to step’ (Žele, pers.com.) Each of the languages expresses more than one cognitive category. Two categories are expressed by Croatian and three by Bulgarian and Ukrainian. The rest of the languages express all four cognitive categories. Quantity of Substance and Quantity of Action are present in 14 languages; Quantity of Quality in 13 and Quantity of Circumstance in 11. Word class representation, in principle, corresponds with cognitive categories. The situation in augmentatives is very similar to that in diminutives. All languages make use of suffixation, 10 of them use prefixation, one language uses prefixal-suffixal derivation. 9 languages make use of two and Belorussian of three word-formation processes. Each language can express Quantity of Substance and Quantity of Action. Quantity of Quality is expressed in 13 languages of the sample and Quantity of Circumstance in 11 languages. Germanic 4 out of 8 Germanic languages have both diminutives and augmentatives. The remaining four have only diminutives. Danish and German make use of two word-formation processes, namely suffixation and compounding. The rest of them rely on a single process–suffixation; in Norwegian it is prefixation: (144)

bitte-liten DIM-little.M/N ‘very little’ (Eiesland, pers. com.)

Prefixation in Norwegian is recursive. While suffixation is not so rich as in Slavic or Romance languages, the majority of Germanic languages have more than one diminutive suffix. On the other hand, in Icelandic there is no special diminutive suffix. There are suffixes that are, among

Research

115

other things, preferably used for the coining of diminutives (Grönke 1955, 79). Each language expresses more than one cognitive category. Quantity of Substance is always expressed and is usually accompanied by Quantity of Quality. German and Icelandic can express three cognitive categories– the Qunatity of Substance, Quantity of Quality and Quantity of Action. The latter is exemplified in (145): (145) German

lächeln ‘smile a little’ < lachen ‘smile’ (M. Körtvélyessy, pers. com.)

Each language has evaluative nouns, three of them have adjectives, two have verbs, and only Dutch has evaluative adverbs: (146) Dutch

lief-jes sweet-DIM ‘in a sweet.DIM way’ (Don, pers. com.)

English also provides an example of class change: (147)

brownie < brown (my example)

Even though the German grammatical gender changes to neuter as a result of diminutivisation (e.g. der Tisch ‘table’ – das Tischlein ‘little table’), this process is not typical of all Germanic languages. For example, according to Krummes (pers. com.), the gender of the base word in Luxemburgish normally remains unchanged in the diminutive and does not switch to neuter. Luxemburgish exhibits another interesting feature: singular diminutive forms are coined from plural bases, which implies a violation of the level-ordering hypothesis: (148)

Kätzercher ‘kitten’ < Kätzchen (pl.) < Katz ‘cat’ (Krummes, pers.com.)

Augmentatives are poorer in terms of the number of word-formation processes–there are only two of them, namely, prefixation and compounding. Prefixation is only used in Dutch. Three languages from the sample make use of compounding: Danish, German, and Norwegian. As pointed out by the informants, augmentatives are rare. Thus, for example, Bauer (pers. com.) assumes that there are no aumentatives in Danish except for those with the element stor- ‘big’ as a compound first element,

Chapter Two

116

e.g. Storkoebenhavn ‘Greater Copenhagen’. Each language can express the cognitive category Quantity of Substance. The cognitive category Quantity of Circumstance is present in Dutch and Norwegian: (149) Dutch

kei-hard stone-hard ‘very hard.adv’ (Don, pers. com)

Celtic Each Celtic language in my sample has both diminutives and augmentatives, however, the EM saturation is very low. Except for Breton, the Celtic languages only make use of suffixation. Breton uses both prefixation (150a) and suffixation (150b) (Stump, pers. com.): (150) a.

gou(r)-vab DIM-son

b.

‘grandson’ bag-ig boat-DIM ‘little boat’

An interesting feature of diminutives is their plural formation, where the evaluative rule applies after the specific inflectional rule–another case of violation of the level-ordering hypothesis: (151)

bag-où-ig-où boat-PL-DIM-PL ‘little boats’ (Stump 2005, 57)

Another intriguing feature of Breton is the homonymy of the diminutive and augmentative prefix gou(r)-, which historically goes back to two different sources. Interestingly, this homonymy is not accompanied by equal productivity. While the augmentative prefix is productive (152), the productivity of its diminutive counterpart is very low (Stump, pers. com.). (152) gour-lano AUG-flow ‘high tide’ An interesting parallel to the relationship of plurality and diminutives in Breton is found in Welsh (Awbery, pers. com.):

Research

(153)

117

plant-os children-DIM ‘little children’

The evaluative morphology of Breton is the richest among the Celtic languages in the sample: it can express three cognitive categories, Welsh can express two and the rest of the sample languages can only express one. None of the languages can express the cognitive category Quantity of Action. Interestingly, there is no suffixation in augmentatives, similar to Germanic languages. Two languages with augmentatives, Breton and Scottish, use prefixation for expressing Quantity of Substance: (154) Scottish

mòr-chuairt AUG-trip ‘big-trip’ (Lamb, pers. com.)

Baltic Lithuanian diminutive formation is extremely rich, making use of a very long list of diminutive suffixes that can express all cognitive categories. On the other hand, only Quantity of Substance was attested with augmentatives. The other Baltic language, Latvian, uses suffixation for diminutivisation, which can express two cognitive categories, notably, Quantity of Substance and Quantity of Quality. An intriguing feature of Lastvian is that augmentatives are formally identical to diminutives. The actual meaning, either diminutive or augmentative, follows from the context (Navickaitė-Klišauskienė, pers. com.). It should be noted that none of these two classes of words are coined by morphological means. An exception is the prefix pār-: (155)

pār-cilvēks AUG-man ‘superman’

Iranian The two Iranian languages in my sample, Ossetic and Zazaki, use suffixation and reduplication for the coining of diminutives. The example in (156) illustrates the latter possibility in Zazaki (Keskin, pers. com.): (156)

kıtab~mıtab book~RDP ‘little book’

Chapter Two

118

Ossetic also makes use of compounding with byn/bun ‘bottom, root’. In combination with colour terms, it yields attenuative meaning (Erschler, forthcoming) (157)

šyrχ-byn / surχ-bun ‘reddish’

Ossetic and Zazaki express the cognitive categories of Quantity of Substannce and Quantity of Quality; in addition, Zazaki also expresses Quantity of Circumstance, for instance, hebıkê ‘a little bit’ (Keskin, pers. com.). In both languages augmentatives were also attested. While Zazaki makes use of the cognitive category of Quantity of Substance, Ossetic uses Quantity of Quality. The word-formation processes used include suffixation and compounding in Ossetic and suffixation and reduplication in Zazaki. Greek The EM saturation value of Greek is very high at 6.66. Both diminutives and augmentatives can express all four sematic categories representing four word classes: nouns, verbs, adjectives and adverbs. For diminutivisation, the word-formation processes of suffixation and prefixation are used; augmentatives also make use of neoclassical compounding. The combining forms like super- can also be used with verbs (Mavrogiorgos, pers. com.): (158)

super-apolamvano super-enjoy ‘I enjoy very much’

Armenian Armenian has only diminutives. They are coined by suffixation. Two cognitive categories were attested: Quantity of Substance and Quantity of Quality. The word classes used are nouns and adjectives as exemplified in (159) and (160): (159)

tn-ak ‘small house’ < tun ‘house’ tun-ak house-DIM ‘small house’

Research

(160)

119

p’ok’r-ik pok’r-ik small-DIM ‘very small’ (Dum-Tragut 2009: 665)

Albanian Similarly to Armenian, Albanian has only diminutives coined by suffixation. Only one cognitive category is expressed, i.e., Quantity of Substance. The use of suffixes for the diminutivisation process depends on the gender of the particular noun: the suffix -th, for example, is used to derive masculine diminutives (161), while -zë/-ëz is used to derive feminine diminutives (162): (161)

zog-th bird-DIM ‘little bird’

(162)

bark-ëz boat-DIM ‘little boat’ (Camaj 1984: 80)

Summary of SAE Table 2-21 summarizes the presence of evaluative morphology in individual language families. Out of 71 SAE languages, 60 have evaluative morphology (86%); 10 languages, especially from the NakhDaghestan language family, have no EM (14.3%). 39 languages (55.71%) have both diminutives and augmentatives, and 21 languages (30%) have only diminutives. There is no language with augmentatives only. Tables 2-22, 23, 24 provide, respectively, the ranking of the sample languages according to EM saturation, diminutive saturation, and augmentative saturation values.

Chapter Two

120

Table 2-21 Evaluative morphology in individual language families Family

Indo-European

Afro-Asiatic Altaic Basque Romance Slavic Germanic Celtic Baltic Iranian Albanian Armenian Greek NakhDaghestan Northwest Caucasian Uralic TOTAL

Number of languages 1 3 1 10 14 8 4 2 2 1 1 1 8

DIM only 0 1 0 0 0 4 2 0 0 1 1 0 2

With EM AUG DIM only &AUG 0 1 0 2 0 1 0 10 0 14 0 4 0 2 0 2 0 2 0 0 0 0 0 1 0 0

3

2

0

0

1

12 71

9 21

0 0

1 40

2 10

Without EM 0 0 0 0 0 0 0 0 0 0 0 0 6

Research

121

Table 2-22 The ranking of the sample languages according to their total EM saturation Ranking

Language

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

Italian, Occitan, Slovak, Slovene, Serbian Belorussian, Czech, Portuguese Polish, Provencal Greek Spanish, Upper Sorbian Russian, Lower Sorbian French, Romansch Bulgarian, Kashubian, Ukrainian, Zazaki Turkish Ladino, Roman Lithuanian, Macedonian German, Basque Catalan, Dutch, Norwegian, Tatar Breton Ossetic, Hungarian, Mansi Croatian, Scottish Gaelic, Latvian, Maltese English, Icelandic, Votic Armenian, Danish, Estonian, Finnish, Karelian, Welsh Mordvin, Saami, Udmurt, Gagauz Albanian, Faroese, Irish, Luxemburgish, Udi, Mari Meadow Ingush, Adyghe, Kabardian Khvarsi, Dargwa, Aghul, Archi, Lezgian, Tabassaran, Komi-Permyak, Khanty, Ubykh

16. 17. 18. 19. 20. 21. 22.

EM saturation 7.67 7.33 7.00 6.66 6.00 5.67 5.34 5.33 5.00 4.67 4.66 4.34 4.00 3.67 3.33 3.00 2.33 2.00 1.67 1.33 1.00 0

Chapter Two

122

Table 2-23 The ranking of the sample languages according to their DIM saturation Ranking

Language

1. 2. 3.

Italian, Serbian Portuguese Czech, Slovene, Upper Sorbian Belorussian, Occitan, Provencal, Lower Sorbian, Polish, Russian, Slovak Greek, Kashubian, Macedonian, Spanish, Lithuanian, Zazaki, Hungarian, Mansi Ladino, Ukrainian, Turkish Basque, Bulgarian, Breton, French, German, Romansch, Roman, Tatar Catalan, English, Icelandic, Ossetic, Votic Latvian, Croatian, Welsh, Armenian, Karelian, Estonian, Maltese Dutch, Mordvin, Udmurt, Saami, Gagauz Faroese, Luxemburgish, Norwegian, Irish, Scottish Gaelic, Albanian, Udi, Mari Meadow Danish, Ingush, Finnish, Adyghe, Kabardian, Khvarsi, Dargwa, Aghul, Archi, Lezgian, Tabassaran, Komi-Permyak, Khanty, Ubykh

4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

DIM saturation 4.67 4.33 4.00 3.67 3.33 3.00 2.67 2.33 2.00 1.67 1.33 1.00 0

Research

123

Table 2-24 The ranking of the sample languages according to their AUG saturation Ranking

Language

1. 2. 3.

Occitan, Slovak Belorussian, Slovene Provencal, Czech, Polish, Greece, Serbian Italian, Portuguese French, Romansch, Spanish, Bulgarian, Norwegian Ukrainian, Dutch, Icelandic Roman, Kashubian, Lower Sorbian, Russian, Upper Sorbian, Zazaki, Turkish Catalan, Ladino, German, Scottish Gaelic, Basque Macedonian, Lithuanian, Tatar Croatian, Danish, Breton, Ossetic, Latvian, Finnish, Maltese English, Faroese, Luxemburgish, Irish, Welsh, Albanian, Armenian, Khvarsi, Dargwa, Aghul, Archi, Lezgian, Tabassaran, Udi, Ingush, Karelian, Estonian, Votic, Mari meadow, Mordvin, Komi-Permyak, Udmurt, Saami, Khanty, Mansi, Hungarian, Adyghe, Kabardian, Ubykh, Gagauz

4. 5. 6. 7. 8. 9. 10. 11.

AUG saturation 4.00 3.67 3.33 3.00 2.67 2.33 2.00 1.67 1.33 1 0

The data above suggests that evaluative morphology is especially heavily saturated in the Romance and Slavic languages, and at the opposite end are the Nakh-Daghestan, Northwest Caucasian, and Uralic languages. The data in Tables 2-22, 23, 24 give additional support to the significance of introducing EM evaluation as a complex parameter, such as EM saturation. To take one example, the ranking of Dutch, as noted by one of the reviewers of the first draft of the manuscript, “is highly unexpected, given the extensive literature which shows that diminutives (but not augmentatives) are very productive in Dutch”. The reviewer refers to Shetter (1959, 1) who observes that “everyone with some knowledge of the Dutch language has been struck by the frequency of its use of diminutives”. In addition, as stated in the review, “both recent grammars

124

Chapter Two

(e.g., Donaldson 1997, 57) and longitudinal studies on acquisition (e.g., Gillis 1997; Souman and Gillis 2007) have confirmed this assumption”. These observations clearly contradict the low ranking of Dutch in Tables 2-22 and 2-23. The explanation of these discrepancies is, however, simple. The relatively low total EM and diminutive EM saturation values follow from the complexity of the EM saturation parameter: its value results from three fundamental variables: the number of cognitive categories represented by diminutives/augmentatives, the number of word-formation processes employed for the formation of diminutives/augmentatives (including subcategories), and the number of word classes represented by EM. In Dutch, diminutives are represented by only two word classes and two cognitive categories, and are formed by only one word-formation process (only one suffix). This is what constitutes the difference between Dutch and the languages with a much higher diminutive saturation (i.e. the languages which cover four cognitive categories, four word classes and employ various word-formation processes. This also explains the unexpected result for Dutch augmentatives because the number of cognitive categories and word classes in augmentatives is higher than in diminutives. Finally, the discrepancies between the general expectations for Dutch and the results based on the EM saturation parameter also result from different perspectives of investigation. Shetter himself speaks of the “frequency of use”, but the parameter of EM saturation defines the richness of evaluative morphology at the level of the system. The latter does not have to correspond with the frequency of use and/or focus on the richness of evaluative morphology in one particular word class. Word formation Diminutives The number of languages in my sample is 71, and 62 languages have evaluative morphology. Each language uses suffixation for coining diminutives. Prefixation was attested in 17 languages of the sample; all of them are from the Indo-European family. Compounding is used in 6 languages: four Germanic languages and two Iranian languages, Ossetic and Zazaki. A similar combination can be found in six reduplicating languages; four languages are Romance, one is Iranian, and one, Udi, is from the Nakh-Dargwa language family. Two less widespread wordformation processes are gender shift (Occitan) and root-and-pattern (Maltese). Suffixation is the only word formation process used in 27 languages, nearly half of the languages with evaluative morphology. 19 languages make use of two word-formation processes, and 5 languages make use of

Research

125

three processes. Two Slavic languages, Slovak and Upper Sorbian, employ four processes. Augmentatives Augmentatives were attested in 40 of the 71 languages from the sample. The most frequent word-formation process is suffixation, attested in 28 languages, followed by prefixation (20 languages), compounding (7 languages) and reduplication (6 languages). Gender shift is employed in Italian only. 19 languages make use of only one word-formation process; the most frequent is suffixation (9 languages). Over half of the 40 languages with augmentatives exhibit multiple word-formation processes, of which 17 languages use two and 4 (Italian, Occitan, Provencal and Belorussian) rely on three word formation processes. Cognitive categories While Quantity of Substance and Quantity of Quality can function as an exclusive cognitive category in evaluative formation of a particular language, Quantity of Quality and Quantity of Circumstance have to be accompanied by either of them. This enables us to formulate the following implicational universal: (163)

IF QQ or QC THEN QS and/or QQ

Diminutives Quantity of Substance is expressed in 59 languages; two exceptions in the 61 language sample are Norwegian and Udi. 45 languages express Quantity of Quality, 30 languages Quantity of Action, and 26 languages Quantity of Circumstance. Only 11 languages express one cognitive category (Quantity of Substance); Norwegian and Udi express Quantity of Quality. Only two cognitive categories are expressed in 16 of the languages, while 13 express three categories, and 18 express all four categories, all 47 of which are Indo-European languages. Augmentatives Only three languages do not express Quantity of Substance – Basque, Turkish and Ossetic (the only category they express is Quantity of Quality). Quantity of Action is expressed by 25 languages, and both Quantity of Quality and Quantity of Circumstance are expressed by (different subsets of) 18 languages. Only 10 languages have one cognitive category, while 5 can express all cognitive categories.

126

Chapter Two

Word classes Word classes mostly map the cognitive categories. For both diminutives and augmentatives, the most frequently used word class is nouns, followed by adjectives and adverbs. 15 languages make use of 3 word classes, and 15 languages use as many as four classes in the formation of diminutives. In augmentivisation, 6 languages make use of 4 word classes and 10 of them three. Correlations Diminutives The correlation suffixation – Quality of Substance – nouns was attested in 56 SAE languages, suffixation – Quantity of Quality – adjectives in 39 and suffixation – Quantity of Action – verb in 23. A weaker correlation of prefixation – Quantity of Substance – nouns was identified for 14 languages, and prefixation – Quantity of Action – verb for 13 languages. Augmentatives Even though the correlation of suffixation – Quality of Substance – nouns is strong, it is weaker than the same correlation for diminutives. It was attested in 29 out of 39 languages. The correlation of prefixation – Quantity of Substance – nouns was identified in 17 languages.

2.4.9 SAE and the world EM saturation The aim of this section is to compare the SAE area with the rest of the world from the point of view of evaluative morphology in order to identify any differences between these languages that might indicate the uniqueness of SAE in terms of this typological feature. A positive reply to this research question will mean that evaluative morphology can be treated as one of the defining characteristics of Standard Average European and, by implication, may be used (in combination with other distinctive features) for the delimitation of the SAE borders as well as for the identification of the core and the periphery of SAE. For the sake of comparison, Tables 2-11, 12, 13, 15 are reiterated here as Table 2-25, 26 27, 28 and completed by the corresponding SAE data.

Research

127

Table 2-25 Evaluative morphology: SAE vs. World

7

1 0

7

31.82

4 7

35.60

52.63

29.67

20.69

34.78

60

6.67 0 3.45

6

6

12.5

3 0

0

8

5.26

8 3

1

9

0.00

4 9

3

DIM+ AUG %

4.55

132

2

20

World sample

9

30.43

1 3

34.03

22

5

1

12.5

Australia + New Guinea

1 6

0

26.32

3

3

1

9.09

19

1 0

only AUG %

22.73

Eurasia

1 3

86.67

1 1

65.22

24

58.62

SE Asia + Oceania

1 7

54.67

1 2

7

84.21

29

3

40.09

Africa

1 5

only DIM %

62.88

8

13.33

23

34.78

North America

1 3

41.38

2

%

45.83

15

EM

15.79

No EM %

59.09

South America

No. of lang.

37.12

Area

2 2

0

4 0

56.33

6 2

0

9

30.98

71

87.32

SAE

12.6

Chapter Two

128

The proportion of languages without EM is much lower in the SAE languages – only 12.6% compared to 37.12% in the world sample. By implication, the proportion of languages with evaluative morphology clearly favours the SAE sample – 85.91% compared to 62.88%. While the percentage of languages with diminutives only is nearly the same in both samples, there are no SAE languages with augmentatives only, which can be captured in the following implicational Euroversal: (164) Euroversal 1 IF Augmentatives THEN Diminutives. The world sample, however, provides 6 counterexamples (Wari, Hausa, Nelemwa, Siar Lak, Thai, Kannada), justifying the limitation of this universal to SAE languages. Furthermore, the percentage of languages with both diminutives and augmentatives is much higher in the SAE sample than in the world sample: 56.33% and 35.60%, respectively. If the SAE area is included in Table 2-10, then the situation is as follows: Table 2-26 Ranking of the areas by the data of Table 2-10 EM

DIM only

DIM+AUG

Africa

AUG only SE Asia+Oceania South America

SAE

North America

South America Eurasia

SAE

Eurasia

Eurasia

North America

Eurasia

Africa

South America

Africa Australia+New Guinea



North America Australia+New Guinea SE Asia+Oceania

North America SAE

South America SAE

Africa

Research

129

From the point of view of presence vs. absence of evaluative morphology, SAE competes with languages of South America. Although the percentage of languages with EM is lower in the SAE sample, the difference is insignificant in terms of the number of analyzed languages. The criterion of EM saturation provides much more conclusive evidence of the unique position of evaluative morphology in SAE: Table 2-27 EM, DIM and AUG saturation values by macro-areas

Africa Australia+New Guinea North America

EM saturation 0.98

DIM saturation 0.68

AUG saturation 0.24

1.16

0.68

0.60

1.52

1.00

0.64

South America

2.00

0.87

0.84

SE Asia+Oceania

1.26

0.79

0.51

Eurasia

1.81

1.12

0.68

SAE

3.4

2.18

1.26

The EM saturation in the SAE sample is significantly higher than in any other macro-area. It is followed by South America, where the EM saturation is lower by 1.4, which is a significant gap. The SAE languages have the highest DIM saturation, too. The difference between the SAE DIM saturation and the next highest DIM saturation (North America) is 0.78. The AUG saturation follows the same tendency – the SAE AUG saturation (1.26) is followed by South America (0.84). Importantly, Table 2-27 proves the advantage of and thus justifies the introduction of the EM saturation parameter (as a complex parameter reflecting three defining characteristics of EM – word-formation, semantics, word class) into the assessment of the status of evaluative morphology in various macro-areas of the world. In addition, it demonstrates that the mere presence vs. absence of a feature is not a sufficient criterion for this kind of research. This observation obtains strong support from the comparison of the data for SAE vs South America and North America. While the comparison of the percentages on the basis of the presence/absence of diminutives, augmentatives, and EM as a whole in these macro-areas is very tight, the individual saturation values provide

Chapter Two

130

indisputable and conclusive evidence for evaluative morphology as a characteristic feature of the SAE languages. This is evident from Table 2-28: Table 2-28 Ranking of the areas by the data of Table 2-27 EM saturation

DIM saturation

AUG saturation

SAE

SAE

SAE

South America

North America

South America

Eurasia

Eurasia

Eurasia

North America

South America


SE Asia+Oceania

SE Asia+Oceania

North America


Africa

SE Asia+Oceania

Africa


Africa

If we move from the level of macro-areas to individual languages, the highest EM saturation in the world sample was identified for Kwakw’ala (6.33), a North American language from the Wakashan family. It is comparable to Greek (6.66) or Spanish and Upper Sorbian (6.00). However, in Table 2-20 there are 10 SAE languages with EM saturation higher than 6.66. Similarly, Karao (3.33), an Austronesian language, has the highest DIM saturation in the world sample. The same DIM saturation value was calculated for Greek, Kashubian, Macedonian, Spanish, Lithuanian, Zazaki and Hungarian. There are 12 SAE languages with higher saturation. The comparison of AUG saturation is similar: Austronesian Agta (3.00) can be compared to Italian and Portuguese. 8 SAE languages have an AUG saturation above 3.00. Based on this data the following Euroversal can be formulated: (165) Euroversal 2 Evaluative morphology in the SAE languages differs from that in the world’s languages by much higher EM, DIM and AUG saturation values. Word formation It follows from Table 2-29 that SAE mirrors the tendencies in the world languages – the most frequently used word-formation process for diminutiviziation is suffixation, followed by prefixation, compounding, and reduplication. However, out of 62 languages with diminutives, only 28 make use of one word-formation process, the rest employ at least two processes.

Research

131

At the same time, the processes of either suffixation or prefixation are richly structured in the SAE languages. Nearly each language has more than one diminutive suffix, and, frequently, a diminutive suffix can be used more than once in a word. Furthermore, recursiveness of diminutive suffixes is a typical feature of SAE languages in general. This structural richness and the number of diminutivizing word-formation processes make the SAE languages unique when compared to the world sample.

Area

Suffixation

Prefixation

Reduplication

Compounding

Change of paradigm

Table 2-29 Word-formation processes in diminutive formation

Africa (17)

10

2

1

1

5


3

0

8

0

0

North America (15)

9

0

2

2

0

South America (13)

8

0

1

1

0


6

2

4

0

0

Eurasia (16)

14

2

0

1

0

SAE (71)

62

17

6

6

0

Similar observations apply to augmentative formation. Both SAE and world languages show a tendency towards suffixation, even though the difference in the percentages between the suffixing and the prefixing languages are smaller than in diminutives. However, the diversification of the SAE word-formation processes is richer, with 44% of the SAE languages using at least two word-formation processes. The previous discussion makes it possible to formulate the following Euroversal: (166)

Euroversal 3 Contrary to the world’s languages, there is a very strong tendency for SAE languages to make use of more than one word-formation process, more than one affix, and to employ evaluatiuve affixes (sometimes the same suffix) in a recursive way.

Chapter Two

132

Area

Suffixation

Prefixation

Reduplication

Compounding

Change of paradigm

Table 2-30 Word-formation processes in augmentative formation

Africa (17)

3

1

0

0

2


0

0

7

0

0

North America (15)

6

1

1

1

0

South America (13)

7

1

2

1

0


1

0

5

3

0

Eurasia (16)

7

3

0

0

0

SAE (71)

28

20

6

7

0

Motivation of evaluative markers One of the most intriguing questions of evaluative morphology is the nature of markers representing the concepts of diminutiveness and augmentativeness. A cross-linguistic analysis mapping the situation in the languages of the world cannot but have recourse to Jurafsky’s model of diminutive meanings presented in Figure 1-3 above. My data has confirmed the central claim of his model, i.e., that a frequent source (not universal, though) of diminutiveness is the concept of CHILD. Diminutive markers with the meaning ‘child’ are used in some African languages (Bafut, Akan, Khoekhoe, Koyra Chiini, Cirecire), in Japanese, and in Ao, a language of SE Asia and Oceania. Malayalam, a SE+Oceania language, employs a generalized variant of the concept of CHILD, in particular, the concept of YOUNG. Interestingly, while the concept of CHILD has usually developed into a bound morpheme used in suffixation, the YOUNG-based diminutivsation in Malayalam rests on compounding, which implies a lexical independence of this concept. Another step in the conceptual extension of the diminutive meaning motivation is found in Movima, an Amerindian language spoken in Bolivia. To diminutivize a word, it uses a classifier whose meaning represents the concept of SEED or GRAIN. Both the concept of CHILD and the concept of YOUNG can be associated with tenderness, nicety, vulnerability, fragility, and loveliness. It comes as a surprise, then, that another possible extension in this direction of

Research

133

associations, the concept of WOMAN, is sometimes used for augmentivisation rather than for diminutivisation. This development can be found in Siar, Ao, Bafut and Xhosa. Our data thus gives support to Matisoff (1991) who maintains that in the SE Asian languages a lexical source for an augmentative morpheme is mother. Both the MOTHER and the CHILD motives can be found, for example, in Bafut. It forms dimunitves and augmentatives by using the prefixes mu- and ma-, with mu elsewhere being the regular word for child and maa the regular word for mother (Tamanji, pers. com.). While the Quantity-related motive for the CHILD–MOTHER opposition is evidently related to the ‘axis of life’, with the concepts of CHILD, YOUNG, GRAIN and SEED representing the smaller-than-default quantity of size, age, maturity, etc., it may also be hypothesized that the MOTHER-type of augmentative motivation is culturally determined and relates to the position of women in the respective communities and/or religion. The CHILD-motivated diminutivisation in many African and Asian languages is a phenomenon unknown in the SAE languages. Contrary to this, as pointed out by Mutz (forthcoming), in many Indo-European languages one of the main sources of diminutivisation is the concept of smallness in the form of an adjective (167): (167) Modern Greek DIM micro-< Anc. Greek micros ‘small’ The next important source is the semantic shift of suffixes with the meaning ‘related-to’ or ‘similar-to’ or ‘origin-of’ (Hasselrot 1957; Magni 1999; Mutz 2000, Grandi 2011a), for example: (168) Sanskrit -(i)ka > DIM -ká, e.g.vamra-ká ‘little ant’ (Lühr 2008, 229) The non-existence of CHILD-motivated diminutives in IE languages is another piece of evidence of the unique position of this territory. This observation is enhanced by the situation in Uralic languages where “diminutive formation seems to be an areal phenomenon: in Estonian diminutives are formed largely according to the German pattern, Livonian follows the Latvian pattern and the Eastern Finnic the Russian pattern” (Kiefer and Németh, forthcoming). A different kind of technique which also appears to rest on the male/female-motivated associations is the change of paradigmatic classification (change of inflectional class) of a particular word resulting in, respectively, diminutivized and augmentivized meanings.

Chapter Two

134

Diminutivisation and augmentivisation expressed by the change of gender is typical of Afro-Asiatic languages. So, for example, the feminine gender in Agaw and Maale is used to express diminutiveness, the masculine gender expresses augmentativeness. The Maale example illustrates the opposition between the masculine gender marker -ats (169a) and the feminine gender marker -éll (169b): (169) a.

máár-átsí maƷƷ-ínt-é-ne Together up.root-IPF-A.DCL ‘The (big) house is built’ b. yénnó mís’-éll-ó dóngo?as-á that.F.ABS tree-F-ABS five. person.PL.NOM bukínt-i wolla túg-á-ne gather-CNV1 together up.root-IPF-A.DCL ‘Five people gather together and uproot that (small) tree’ (Amha 2001, 71)

One of the most striking features of the grammatical systems of this group of languages is the reservation of one inflectional class for the category of diminutives. As an example, in Supyire, gender 3 is the gender of small things: nouns with the diminutive suffix -rV are automatically gender 3 singular (Carlson 1994, 105): (170)

ba-rá-ni house-DIM-DEF ‘small house’

In Swahili, Class 5 (replacement of a specific noun class prefix with zero) (171a.) can express augmentativeness, and Class 7 can express diminutiveness (replacement of a specific noun class prefix with ki-) (171b.) (Contini-Morava, pers. com.): (171) a. b.

toto ‘large, ungainly child’ < mtoto ‘child’ kitoto ‘small child, infant’ < mtoto ‘child’

A typologically opposite situation can be found in Occitan, a SAE language: a word with the feminine marker -a refers to normal size. If this feminine marker is replaced with the masculine gender zero morpheme, the word is diminutivized (cf. example (138)). Conversely, the augmentative suffix -kazi in Xhosa is the same as the feminine suffix:

Research

(172)

135

inkosikazi ‘lady’ < inkosi ‘lord’

As noted by Lloyd (pers. com.), where this leads to ambiguity, the augmentative suffix -kazi is not used unless the noun is feminine when the suffix becomes a little pejorative as in: (173)

umfazikazi omkhulukazi ‘a great big fat woman’

The category of Quantity also underlies a different type of conceptual motivation in evaluative morphology, which is related to the category of number, notably, the singular-plural opposition. For illustration, the augmentivizing prefix mansi-, used in the Austronesian language Muna, means ‘more items than expected’, i.e., more than the deafault value. Contrary to this, the basic meaning of the diminutive suffix -ulu in Bāgandji is ‘singular’. Moving to another part of the world, the Celtic languages Welsh and Breton form diminutives by attaching a suffix to a plural form, thus, relating the concept of diminutiveness to the category of Quantity. The same is true of the Germanic language Luxemburgish, where the singular diminutive forms have plural bases (cf. example (148)). And finally, the same phenomenon that contradicts the universal order of derivational and inflectional markers also occurs in the Algic language Passamaquoddy (cf. example (63)). Cognitive categories and word classes The following tables provide the summarized data for cognitive the categories of diminutives and augmentatives. The data suggests that the SAE area reflects some general tendencies. The most frequently expressed cognitive category is Quantity of Substance, followed by Quantity of Quality and Quantity of Action. However, there are some differences between the languages of the world sample and SAE: First, while the languages of the world sample prefer to express only one category, the SAE languages usually express more than one cognitive category. Quantity of Circumstance was rarely attested in the world languages; in the SAE sample, it is much more frequent. Quantity of Action, too, is much more frequent, especially in the Slavonic genus.

Chapter Two

136

Table 2-31 Cognitive categories by areas – diminutives

Africa (17)

QS

QQ

QA

QC

14

4

2

0


5

5

1

1

North America (15)

14

5

6

1

South America (13)

9

4

2

0


8

4

4

1

Eurasia (16)

14

3

5

0

SAE (71)

60

46

27

27

Table 2-32 Cognitive categories by areas – augmentatives QS

QQ

QA

QC

Africa (17)

5

1

1

0


1

5

2

0

North America (15)

6

5

3

1

South America (13)

6

3

3

0


5

5

2

1

Eurasia (16)

7

2

4

0

SAE (71)

37

18

25

18

The data supplies unambigous evidence of the significant differences between the SAE languages and the world’s languages, which range over all cognitive categories. This is especially true of the categories of Quality (only diminutives) and Circumstance, where there is a substantial gap between the percentage data for diminutives: 63.38% (SAE) vs. 30.00% (world) and 38.20% (SAE) vs. 3.60% (world), respectively, and augmentatives (Quantity of Circumstance): 25.35% (SAE) vs.2.4% (world). But the data for the other two cognitive categories also illustrate these differences well. The situation in the category of augmentatives appears to parallel that of diminutives with one impotant exception: it bears on Quantity of Quality where the percentages are roughly identical. The comparison represented in Tables 2-31 and 2-32 leads to the formulation of the following Euroversals:

Research

137

(174) Euroversal 4 The SAE languages on average are characterized by a much higher percentage of occurance of all the cognitive categories within morphological diminutives than the languages of the other areas of the world. (175) Euroversal 5 The SAE languages on average are characterized by a much higher percentage of occurance of all the cognitive categories within morphological augmentatives (with the exception of Quantity of Quality) than the languages of the other areas of the world. All these tendencies are mapped onto the parameter of word classes. Nouns are most frequent, but the number of verbs and adverbs is much higher than in the world sample. While diminutivized verbs were attested in 27% of the world sample languages, their percentage in the SAE sample is higher by 10 %. Even more striking is the comparison of adverbs: in the world sample languages with adverbial diminutives represent only 4.6%, while in SAE it was 34%. In accordance with the data obtained for the cognitive categories, the tendency is to use more than one word class by the SAE languages, which stands in contrast to the world’s languages. Correlations The correlation suffixation – Quantity of Substance – nouns is the strongest in both the world’s languages and SAE languages. In the SAE language area, it is followed by the correlation of suffixation – Quantity of Quality – adjectives.

2.5 Conclusions The previous discussion unambiguously suggests that the SAE language area has some specific EM features that allow it to be distinguished from the rest of the world. The total EM saturation of the individual EM languages ranges from 7.67 to 1.00. Based on this, the languages can be grouped (combined) into the following four groups:

Chapter Two

138

Table 2-33 Classification of languages by EM saturation Total EM saturation 7.67-5.67 5.66-3.67

3.66-1.67 1.66-1.00 0

Languages Italian, Occitan, Slovak, Slovene, Belorussian, Czech, Portuguese, Polish, Provencal, Greek, Spanish, Upper Sorbian, Russian, Lower Sorbian, Serbian French, Romansch, Bulgarian, Kashubian, Ukrainian, Zazaki, Turkish, Ladino, Roman, Lithuanian, Macedonian, German, Basque, Catalan, Dutch, Norwegian, Tatar, Breton Ossetic, Hungarian, Mansi, Croatian, Scottish Gaelic, Latvian, Maltese, English, Icelandic, Votic, Armenian, Danish, Estonian, Finnish, Karelian, Welsh, Mordvin, Saami, Udmurt Albanian, Faroese, Irish, Luxemburgish, Udi, Mari Meadow, Gagauz, Ingush, Adyghe, Kabardian Khvarsi, Dargwa, Aghul, Archi, Lezgian, Tabassaran, Komi-Permyak, Khanty, Ubykh

Similar groups can be identified for DIM saturation: Table 2-34 Classification of languages by DIM saturation DIM saturation 4.67-3.67

3.66-2.67

2.66-1.67 1.66-1.00 0

Languages Italian, Portuguese, Czech, Slovene, Upper Sorbian, Belorussian, Occitan, Provencal, Lower Sorbian, Polish, Russian, Slovak, Serbian Greek, Kashubian, Macedonian, Mansi, Spanish, Lithuanian, Zazaki, Hungarian, Ladino, Ukrainian, Turkish, Basque, Bulgarian, Breton, French, German, Romansch, Roman, Tatar Catalan, English, Icelandic, Ossetic, Votic, Latvian, Croatian, Welsh, Armenian, Karelian, Estonian, Maltese, Dutch, Mordvin, Udmurt, Saami Faroese, Luxemburgish, Norwegian, Irish, Scottish Gaelic, Albanian, Udi, Mari Meadow, Danish, Ingush, Finnish, Adyghe, Kabardian, Gagauz Khvarsi, Dargwa, Aghul, Archi, Lezgian, Tabassaran, Komi-Permyak, Khanty, Ubykh

Research

139

The grouping of languages according to their AUG saturation differs especially in the number of languages with augmentative evaluative morphology. Otherwise, Table 2-35 below is very similar to the previous ones. Table 2-35 Classification of languages by AUG saturation AUG saturation 4.00-3.67 3.66-3.00 2.99-2.00 1.99-1.00

0

Languages Occitan, Slovak, Belorussian, Slovene Provencal, Czech, Polish, Greece, Italian, Portuguese, Serbian French, Romansch, Spanish, Bulgarian, Norwegian, Ukrainian, Dutch, Icelandic, Roman, Kashubian, Lower Sorbian, Russian, Upper Sorbian, Zazaki, Turkish Catalan, Ladino, German, Scottish Gaelic, Basque, Macedonian, Lithuanian, Tatar, Croatian, Danish, Breton, Ossetic, Latvian, Finnish, Maltese English, Faroese, Luxemburgish, Irish, Welsh, Albanian, Armenian, Khvarsi, Dargwa, Aghul, Archi, Lezgian, Tabassaran, Udi, Ingush, Karelian, Estonian, Votic, Mari meadow, Mordvin, Komi-Permyak, Udmurt, Saami, Khanty, Hungarian, Adyghe, Kabardian, Ubykh, Gagauz

These tables can be projected onto maps of saturation. The SAE language saturation map (see centrefold, Map 2.8) is based on a WALS map, with dots for languages and with country boundaries. As it is problematic to produce a map with precisely represented language boundaries, country boundaries were used instead. The biggest problem was posed by the territory of Russia with its many languages. Russian and various other languages, such as Mansi, Khanty, Tatar etc., overlap inside one large territory. The languages are identified with a dot on the WALS map, but the actual territories in which these languages are spoken are much more extensive. At the same time, the eastern boundary of the Russian language is questionable, which is why it is blurred on the map (see centrefold, Mao 2.8). The map (centrefold) and the tables suggest that not only is SAE a unique language area, but it also has its central, core, nuclear languages as well as peripheral languages. In Section 1.2.3, various classifications of the SAE languages by four different typologists arre mentioned, and each

140

Chapter Two

of these identifies the core of the SAE languages (no matter how they label it). Some parallels between their division and my saturation values can be identified. First, languages with highest saturation (7.67–5.67) are basically Italo-Western and Slavic languages. They correspond with Ramat’s “more or less central members” and partially with van der Auwera’s Charlemagne Sprachbund. However, in my division there are no Germanic languages in the core group. At the same time, the SAE core includes Slavic languages, a fact which contradicts Haspelmath’s division. Furthermore, my division includes languages, such as Turkish and Zazaki, which in each of the aforementioned classifications are “clearly outside” (Haspelmath) or “very marginal” (Ramat). Both Haspelmath and Ramat divide the SAE languages into zones. Haspelmath provides a comprehensive list of the languages of the nucleus and is more general in relation to the core and peripheral languages, i.e., he confines himself to the identification of genera. A zone-based division is also used by Ramat who provides a genera-based classification of the individual zones. His zones rely on genetic classification but, geographically, is rather fuzzy, with no clear boundaries between the centre, the periphery, and the marginal zones. His “more or less central members” include all Romance, West-Germanic, Balkan and Balto-Slavic languages. “Very marginal, even outside the SAE” are Eastern Uralic languages. Décsy distinguishes, for instance, the Danube, the Balkan, and the Kama zones, van der Auwera identifies, inter alia, the Charlemagne Sprachbund. Contrary to Haspelmath and Ramat, these zones more or less map onto the geographical criteria. As it follows from this brief overview, use of zones in the SAE territory rests on two complementary principles: the genetic and the geographical principles. Since areal typology is based on shared properties irrespective of the genetic relatedness of languages on a particular territory, I believe that the primary criterion should be geographical, even if the genetic criterion cannot be ignored. In the following description, the individual cohorts of SAE languages are referred to by colours (see centrefold, Map 2.8): • The brown area represents languages with the highest EM saturation. From the geographical point of view, it covers the Mediterranean region, the majority of Central and East Europe (with certain exceptions, e.g. Hungary), and a great part of Southern Europe. From the point of view of genetic classification of languages, it is the territory of Indo-European languages, basically Romance and Slavic genera. • The green area represents languages with the second highest EM saturation. From the geographical point of view, it encompasses

Research

•

•

•

141

the languages spoken in the West Europe, and some languages of Scandinavia and Southern Europe. From the genetic point of view, these are some Romance, Slavic, Germanic, Celtic, and Iranian languages of the Indo-European family, some Altaic languages and Basque. The blue area represents languages with the third highest EM saturation – part of the British Isles, Central and Eastern Europe, North Europe, Scandinavia. Languages spoken on this territory include Germanic, Baltic, Celtic, Slavic, Iranian genera of IndoEuropean languages, Finnic and Ugric genera of the Uralic family, the Altaic family, and the Afro-Asiatic family. The red area represents languages with the lowest EM saturation. Here belongs Albanian, part of the British Isles, Faroe Islands, the territory of the Luxemburgish language, the territory of Gagauz language, Northwest Caucasus, and the eastern part of Russia. The yellow area represents languages without evaluative morphology – two small “islands” in the south-eastern part of Europe, western part of Russia. In the majority of cases these are languages of Nakh-Daghestan language family, but there are also two languages of the Ugric genus. If the parameter of evaluative morphology is included in the catalogue of SAE features, these languages are excluded from the SAE Sprachbund. Their borders delimit the area of the SAE languages. Considering their location on the geographical map, they form the eastern border of the SAE Sprachbund. The western, southern, and northern borders are the Arctic Ocean, the Atlantic Ocean, and the Mediterranean Sea, respectively.

The brown area overlaps to some extent with Haspelmath’s and Ramat’s nucleus/central zone. The biggest difference is the absence of Germanic languages in the EM nucleus and the central position of Slavic languages in it. The green area is comparable to Haspelmath’s core. However, in addition to the Germanic, Slavic and Balkan languages, there are also Altaic, Celtic, and Baltic languages and Basque, i.e., the languages that are “marginal even outside” (Ramat), or clearly outside (Haspelmath). On the other hand, in Haspelmath’s “clearly outside” and Ramat’s “marginal even outside”, there are languages of the NakhDaghestan family and Eastern Uralic languages – languages included in the yellow area (languages without EM saturation). Areal classification of languages is based on the fact that languages of a given geographical area share some features, even though they are not

142

Chapter Two

genetically related. The shared features result from language contact. In other words, features of a language depend on its history (genetic classification), which is indispensably “geared” with contacts with other languages. A case in point is the evaluative morphology of two Uralic languages, Hungarian and Mansi. In comparison to other Uralic languages, both of them feature high EM saturation values (3.33). Hungarian is spoken in Central Europe and Mansi in north-western part of Russia. From the geographical point of view, they are surrounded by languages with high EM saturation (Slovak, Ukrainian, Romanian, Serbian, and Slovene in the former case, and dominantly Russian in the latter case). It may be surmised that this unusually high EM saturation value (absent in the other Uralic languages) is due to language contact. This hypothesis is also supported by the phonetic similarity between the Hungarian diminutive suffix -cska and Slovak -čka. Another manifestation of the influence of language contact on the structure of language is, as has been discussed by Laakso (forthcoming) in refence to derivation in Uralic languages: … areal developments and language contacts may have affected the functions of derivation: a good example is the formation of diminutive nouns, which is productive in the Eastern Finnic languages (as in Russian) and in Estonian (as in German), while Finnish patterns with Swedish in not having a regular suffixal diminutive derivation at all.

This phenomenon has also been observed in sociolinguistic studies. Lanstyák (1998), for instance, describes the frequent use of diminutives especially among non-Hungarian dominant speakers of Hungarian language varieties in Slovakia. In summary, the present analysis of evaluative morphology in the SAE languages gives additional support to the idea of the existence of the SAE area. However, it also demonstrates that any conclusions of the external borders and of the internal typicality of structure of the SAE territory are premature. An important step forward yielding tentative conclusions has been achieved by the research of Haspelmath, van der Auwera, Ramat, and others. But there is much more to be done in the future than has so far been achieved. Any relevant conclusions on SAE borders and its internal structure are preconditioned by a much longer and much more comprehensive list of typological parameters that need to be examined and interrelated. The methodology of research and data-evaluation also need to be fine-tuned in order to work with compatible data that can lead to significant and reliable observations. This is important if we realize that, as already stated in 1.2.4, various analyses into SAE features are based on different starting points and different methods, which excludes the

Research

143

possibility of unifying them. Based on my research the following may be suggested: • Inclusion of the parameter of EM saturation in the catalogue of the features of SAE languages. • Application of the method of saturation calculation (adequately adapted to the specificity of the individual freatures examined) to other fields of language analysis. The result will be a unified method which will make it possible to identify the zones of the SAE language territory. • Application of a twofold analysis, i.e., no SAE-oriented studies should be confined to the SAE languages themselves. Relevant results cannot be obtained without comparing the territory of SAE languages with the languages of the world.

CHAPTER THREE PHONETIC ICONICITY IN EVALUATIVE MORPHOLOGY

“The difference between the right word and the almost right word is the difference between lightning and a lightning bug” —Mark Twain

3.1 General When phonetic iconicity grasped my attention for the first time, I was overwhelmed by its intricacies. One of the very first hiccups I had to struggle with was the terminological diversity. Thus, section 3.2 is devoted to the delimitation of the term phonetic iconicity – “an expression that, while seemingly transparent at first sight, conveys many mysteries”. Section 3.3 focuses on the subject matter of my research–synesthesia. Section 3.4 offers an overview of experiments focussed on the area of phonetic iconicity, and the final section discusses the results of the experiments. They are related to the concept of Evaluative Morphology Saturation in order to find out whether or not there is any correlation between iconic expression of evaluatives and the EM saturation value in the selected languages.

3.2 Iconicity Undoubtedly, iconicity has a firm position in our everyday communication. As Fischer (2004) points out, there are at least four different domains where evidence for iconicity can be found: language evolution, language acquisition, language change, and language use in literary texts. My understanding of iconicity leans upon Peirce’s classification of signs. In his conception, icon refers to the relation between sign and object based on similarity, resemblance. Technically, it means that discourse participants share the same experience and

Phonetic Iconicity in Evaluative Morphology

145

knowledge that enable them to employ a particular icon for communication. Iconicity as a semiotic principle is derived from this conception of icon. Anderson (1998, 15) states that “the concept of iconism was ideological disfavor, as it still is among many linguists. It is unfashionable to write about the affective domain in language”, and he goes on to say that “[i]conism presents itself as an unwelcome link between language conceived as an autonomous system versus language as non-autonomous discourse” (1998, 20). In spite of these circumstances, phonetic iconicity has received considerable attention in recent decades. Unfortunately, numerous studies and experiments failed to explain satisfactorily the place and role of iconicity at the level of langue. This might explain why the parts of Wallis’ Gramatica linguae anglicane (1653) dealing with onomatopoeia and phonosymbolism were not translated until 1972 (Anderson 1998, 20), and similarly, why the dissertation thesis on sound symbolism of Finnish suffixes by Rytkonnen (1940) was declined (ibid.). As pointed out by Anderson (1998, 15), iconicity is an affective part of language: the world of emotions and the world of (language) system rarely collocate. As soon as we leave the world of “hard core” linguistics and enter the space of literature, generously employing linguistic creativity, we encounter iconicity in its diverse manifestations, exploited by men of letters (often unconsciously) and later studied by linguists. Examples of iconicity in literature are numerous. The most frequently cited authors are Poe (Raven, the nevermore refrain), Shakespeare (Antony and Cleopatra III.10.1-4 Naught, naught, naught, all naught, I can behold no longer: The Antoniad, the Egyptian admiral, With all their sixty fly, and turn the rudder …) Carroll (Alice in Wonderland, Humpty Dumpty), and, for obvious reasons, Rowling and her Harry Potter series (Slitherin, Draco Malfoy). Given their linguistic background, Rowling and Carroll may have employed the principles of iconism purposefully. It requires a great leap of faith, however, to believe that Shakespeare was aware of its existence. It may be assumed that he used iconic expressions intuitively because of their sound impression. Poe (1845, 163) himself stated that to create feelings of melancholy he used the refrain “Nevermore” because of “o as the most sonorous vowel, in connection with r as the most producible consonant”. However, why these sounds create the feeling of sadness is left unexplained. Supported by Peirce’s conception, it is a relatively simple task to define iconicity. For instance, Anderson (1998, 26) conceives it as “representation and signification based upon partial resemblance, not complete replication.” Merlini Barbaresi (2000, 1) understands it as

146

Chapter Three

“describing a natural link between language and experience, or rather sign and object, based on some type of similarity”. Thus, iconicity is characterized by partial resemblance of an icon to its referent. Extralinguistic reality gives a stimulus, e.g. size, colour, darkness, which is imitated by a specific sound sequence. Merlini Barbaresi (ibid.) points out that iconic procedures are natural because they are efficient, speakers prefer them because they allow easy planning, monitoring, immediacy of perception, and memorization; moreover, they are also equally preferred by children and foreign language learners. Iconicity has great potential for connotation and associative meaning, and it allows the language users to load their utterance with layers of extra meaning (ibid.). Therefore, it is exploited in literature, songs, brand names, and advertisements. The external input is “iconized” (Anderson, 1998, 99) at different levels of the language system – syntax (by means of word order), morphology (word structure) and phonology. A summary of Anderson (1998) and Merlini Barbaresi (2000) provides us with four basic features of iconicity. First, it is a gradable phenomenon, a cline.1 The degree of iconicity of, for example, hmm, shh, tsk is higher than that of screech or rumble because of the onomatopoeic nature of the former. Second, iconicity is a marginal phenomenon in the lexicon, although hmm, shh, tsk are more iconic for native speakers, they have no grammar and violate the rules of English phonotactics. Third, hmm, shh, tsk are based upon partial resemblance and as such they are examples of language economy. On the other hand, Anderson (1998, 34) speaks of the peripheral paradox. According to him, iconic words often use anomalous or peripheral structures, are often playful and “playfulness is a diseconomy in language” (ibid.). Fourth, as a peripheral, marginal language phenomenon it prefers peripheral structures, e.g. reduplication (chit-chat), word extensions (a–an–and–sand), monosyllabic words (mat, rat, fat), consonant repetition (ding-dong), consonant clusters (snort, snarl, snore),

1

The gradability of iconism is discussed from two points of view: that of speaker and that of hearer. Sapir (1929, 227), for example, assumes that the subjects of his research differed in their ability to “feel” sound symbolism. On the other hand, he states that iconicity is in accordance with an unconscious or intuitive logic (not necessarily) based on experience with the stimuli. By implication, sound symbolism depends on individuals and their iconic sensitiveness. This idea was supported not only by linguistic studies (Jespersen, 1933; Oszmianska, 2001, Shrum et al., 2007, etc.) but also psychological/neurological research (Ramachandran and Hubbard 2001, Maurer et al. 2006, etc.).


147

phonotactic deviations (hmm, brr, shh), and onomatopoeia (whoosh, plop).2

3.3 Phonetic iconicity, sound symbolism, phonetic symbolism The terms phonetic iconicity, sound symbolism, phonetic symbolism 3 are used as synonyms in linguistic literature. Jespersen (1933), Ultan (1978), Ohala (1994), Shinohara (forthcoming) prefer sound symbolism; Fisher-Jørgsen (1978), Lapolla (1994), and Sapir (1929) use the term phonetic symbolism, and Gebels (1969) speaks of phonetic iconicity. Thus, these terms may be used interchangeably in reference to the phenomenon in which the relationship between the signifiant and the signifié is not wholly arbitrary. The shape of the signifiant is influenced by associations, feelings, ideas etc. evoked by the signifié itself. Phonetic iconicity is based on binary oppositions, e.g. high vs low; front vs back; rounded vs unrounded; high tone vs low tone.4 These phonetic values correlate with values in semantics such as small vs large; bright vs dark; sharp vs blunt. A discussion of the reason for this correlation was started by Jespersen (1922), Sapir (1929), and Newman (1933), who were followed by many others. Sapir (1929, 235) maintains that the reason can be acoustic or kinaesthetic, or a combination of both. While the acoustic explanation is based on hearing a produced sound, the kinaesthetic explanation reflects the speaker’s view, for example, of the 2

This assumption seems to be too strong, especially in view of the topic of this research. Phonetic iconicity in evaluative morphology is considered to be universal in language (and as such, hardly a peripheral phenomenon in language). And even if it is proved in the course of this book that the universality claim is also too strong of a hypothesis, this phenomenon is a highly productive and inherent part of the word-formation system in a number of languages of the world. 3 The same phenomenon is called by various names in literature: in addition to phonetic symbolism, phonetic iconicity, and sound symbolism, phonosymbolism, iconism, isomorphism, constructional diagrammacity, mimesis, etc. can also be encountered. 4 Binary opposition was proved by experiments. Generally, the experimental methods in phonetic iconicity can be twofold: free-choice and forced choice. In the forced-choice (Sapir 1929), the subjects are directed toward particular dimensions, their task is to choose from two meanings (e.g. small/large or which is smaller). In the free choice experiment the subject is asked to spontaneously state the meaning of a given sound. The results of free-choice experiments are, however, random, non-repeatable and do not prove the existence of phonetic iconicity (e.g. Bentley and Varon 1933).

148

Chapter Three

movement of his articulatory organs. However, Sapir (ibid.) himself points out that the subjects of his experiment differed “somewhat in the psychological basis of the symbolism, some being apparently swayed entirely by the acoustic factor, others by the acoustic factor only or mainly insofar as it was itself supported by the kinaesthetic factor”. He stresses a rather hard-to-define factor – the feeling. Phonetic iconicity originates in the way the hearer feels/perceives the sound or the speaker feels its articulation.5 Thus, phonetic iconicity is influenced by an individual’s psychological functions, by an individual’s experience. The experience with sounds in language is also affected by contrast within the phonemic system of a language. In this context, Anderson (1998, 106) speaks of phonemic relativism conceived as the extent to which inherent acoustic or kinaesthetic features “lend themselves to contrast within the phoneme system of the language”. This means that the correlation between phonetic and semantic values is basically influenced by two factors: cognitive processing and phonemic relativism. The role of feelings is also stressed by Osgood (1960, 168) who assumes that “the phenomena which seem to display generality across human groups regardless of language or culture are essentially connotative – the affective ‘feeling tones’ of meaning which contribute to synesthesia, metaphor and the like”. Osgood characterizes denotation as the multitudinous and arbitrary sets of correlations between perceptual events and linguistic events and connotation as affective feeling tones. The connotative systems display similarities. These similarities are caused by biological factors. Human beings react to certain situations in a similar, automatic, emotional way which creates a shared connotative framework despite different languages and cultures. This can result in synaesthesia. In other words, the shared connotation framework may result in phonetic iconicity. In addition, Osgood finds the basis for synesthesia also in specific relations between human organisms and their generally similar environments. Their stability can be the basis for synesthetic and metaphorical interpretations. He distinguishes two types of these relations: (i) innate (e.g. red colour associated with warmth) and (ii) acquired, developed by learning under similar conditions (e.g. common association of visually large with auditorily loud).6 5

For the idea of feelings see also Tsuru and Fries (1933). In this connection let us mention Morton (1994) and Ohala (1994) who work with the concept of the F0 code. The frequency code as defined by Ohala (1983) is understood as a size code. It explains the striking similarity across languages in the use of rising or high question intonation and falling or low statement intonation, and suggests that sound symbolism is a non-arbitrary connection between certain 6


149

Then, it may be assumed that feelings create connotative meanings and engender synesthesia, metaphor and the like. If the connotative framework is shared by different cultures/languages created by human beings with common biology, then feelings are the same. This fact is reflected in the same type of sounds used for their expression and can be understood as the common ground for the phonetic iconicity. The typology of iconicity (morphological, syntactic, phonological) relates to individual language levels. While morphological and syntactic iconicity are based on a language structure, phonological iconicity grows out of sound. The term phonetic iconicity or sound symbolism is used “when a sound unit such as phoneme, syllable, feature, or tone is said to go beyond its linguistic function as a contrastive, non-meaning bearing unit, to directly express some kind of meaning” (Nuckolls 1999, 228). Generally, five types of phonetic iconicity are distinguished: onomatopoeia, kinaesthesia, synaesthesia, chromaesthesia and phonaesthesia. This classification is common to our European culture. In the Japanese language, for example, the position of phonetic iconicity is much more important. This is also reflected in a special terminology: phonomime (onomatopoeia), phenomime (synaesthesia) and psychomime (phonaesthesia). As stated by Hamano (1998, 1), Japanese has an extensive set of sound-symbolic forms “commonly called giongo/giseigo/gitaigo, ‘mimetic words’, which constitute a highly cohesive system with distinct phonological, semantic and syntactic characteristics”. What interests us in relation to the topic of this chapter is magnitude, or size symbolism/phonetic iconicity. It is based on the small/large opposition. It is believed that fronted high vowels are used to express smallness and back low vowels the opposite. Similarly, at the consonantal level, smallness is supposed to be associated with fronted consonants and/or high tone. Importantly, (phonetic) iconicity is of scalar nature. Degrees of iconicity were proved by experiments–e.g. one language can be more iconic than another (Japanese-French, cf. Oszmianska 2001), and one word can be more iconic than another. This is closely related to the discussion of arbitrariness of linguistic signs.7

classes of speech sounds and the meanings of the words or morphemes. This connection is believed to have a biological evolutionary origin. 7 Like many other linguistic issues, the question of arbitrariness vs nonarbitrariness should be perceived as continuum, a spectrum. As Simone (1990, 121) points out, we can speak about two paradigms, the paradigm of arbitrariness (unmotivated signs) and the paradigm of substance (motivated signs). They represent two ends of one axis, two borders of a continuum.

150

Chapter Three

3.4 Previous research into phonetic iconicity 3.4.1 General Phonetic iconicity has attracted the attention of many linguists. However, not all researchers try to explain the origin of phonetic iconicity. Two pioneers of the research into phonetic iconicity, Jespersen and Sapir, suggested acoustic, kinaesthetic and articulatory explanation, or a combination of them. According to the articulatory explanation (and also kinaesthetic, as it refers to the muscle sense), phonetic iconicity is caused by way of articulation. Thus, for example, small lip aperture is associated with something small. On the other hand, the acoustic explanation focuses on the perception of the sound, e.g. high pitch of the vowel. The articulatory/acoustic explanation is also supported by Diffloth (1994) and Shinohara and Kawahara (forthcoming). French (1972) looks for an explanation in semantics and concludes that the results of studies into phonetic iconicity can be explained by the shared semantic properties of the response words tending to cluster. Koriat (1975) argues that symbolic connotations are stored in the lexicon: certain sounds symbolize certain semantic categories. This is, in his view, an innate capacity of human beings. The innate nature of phonetic iconicity is also postulated by Ohala (1994), a proponent of the theory of frequency code which is believed to be innate, too. Ohala conjectures that the frequency code is the association of high acoustic frequency with smallness and low acoustic frequency with largeness. Fitch (1997) also refers to the frequency code and proposes a new acoustic variable: formant dispersion. Formant dispersion is “the averaged difference between successive formant frequencies, and was found to be closely tied to both vocal tract length and body size” (1997, 1213). Berlin (1994) and Lapolla (1994) provide an explanation close to Sapir’s account. Maurer et al. (2006) draw on biology and stress the role of primary and sensory cortical areas. Although Ranachandran and Hubbard (2001) do not deal with cortical areas, their neurological explanation supports the idea of the sensory nature of synaesthesia. In fact, the idea of sensory nature of sound symbolism appears as early as the ideas of Sapir (1929), who believes that subjects can somewhat feel sound symbolism in sound contrast.

3.4.2 Types of synesthesia As studies and experiments into phonetic iconicity are numerous, in the process of preparation of my own experiment I concentrated on those


151

that focused on size/magnitude symbolism. They include Ultan (1978); Tarte (1974); Shrum and Lowrey (2007); Shinohara and Kawahara (forthcoming); Silverstein (1994); Sapir (1929); Roper et al. (1976); Ohala (1994); Lapolla (1994); Klank et al. (1971), Kong-on (1977); Johnson (1967); Jespersen (1933), Fitch (1997); Diffloth (1994); Berlin (1994); Atzet and Gerard (1965); Aoki (1994). Six publications analyse more than one type of synesteasia: Koriat (1975); Jespersen (1922); Fónagy and Fónagy (1970); Bentley and Varon (1933); Becker and Fisher (1988), Brackbill and Little (1957). Woodworth (1991) and Tanz (1971) deal with proximal and distal forms; Ramachandran and Hubbard (2001) deal with grapheme-colour synaesthesia; Osgood (1960) discusses visual-verbal synaesthesia; Mauerer et al. (2006) deal with shapes. Oszmianska (2001) compares English phonostemes and Japanese mimetics. Specific mention should also be made of Westcott (1971), who discusses iconicity in general; Wertheimer (1958), who tests fittingness and nonfittingness of meaning and sound; Gebels (1969), who deals with words of sensory nature; and French (1972) who analyses the influence of semantics on experiments in phonetic symbolism.

3.4.3 Type of research Phonetic iconicity has mostly been studied from a linguistic perspective. Exceptions to this tendency include Fitch (1997) – biological approach; Maurer et al. (2006) – psychological approach; Ramachandran and Hubbard (2001) – neurological approach; and partially Ohala (1994) – biological.

3.4.4 Languages studied The variety of languages analysed is wide, ranging from IndoEuropean languages (English, Spanish, Czech), Ugric (Hungarian, Finnish), Ameroindian (Navajo) and African (Yoruba) to Japonic (Japanese). In this work, I focus on selected SAE languages.

3.4.5 Research methods In general, the research method chosen is influenced by the research type. Research of biological, neurological, and psychological types was usually carried out by an experiment or observation. The linguistic type of research can be divided into two groups:

Chapter Three

152

• •

Desk research based on the listing and comparing of words from various languages. A typical example is Jespersen (1922, 1933). Experimental, both monolingual and cross-linguistic, making use of various methods.

Furthermore, experimental research methods can be classified into two large groups: forced-choice and free-choice methods. In the forced-choice method, the subject matches the words/sounds given with some meanings, or senses. The free-choice method means that nonsense words vary in their sound and the subject provides the first thing that comes to his/her mind. Another classificatory principle is the nature of stimulus words; from this point of view, the methods can be divided into those based on nonsense words (vowel between two variable consonants) and those relying on existing words. Special attention is paid to the presentation of the stimulus words. For illustration, Atzet and Gerard (1965) who studied the Navajo language stress the fact that oral presentation was used because Navajo does not exist in written form. The procedure variants include an audiovisual method or exclusively visual or exclusively auditory method. In the following overview, the methods are numbered from 1 to 5. Methods 3–5 were described by Brown and Nuttall (1955). • •

•

Nonsense words and arbitrary referents are given; for example, nonsense words and English referents (e.g. Lapolla, 1994). Informants evaluate nonsense words (or words from unknown languages) on a scale. This method, developed by Greenberg and Jenkins (1966), is frequently used. It could also be called the vowel sounds-scales method. In principle, two groups of subjects rate audiotaped vowel sounds on scales. One example is found in Fisher-Jørgsen (1978) in which vowel categories are matched with the members of a selected set of adjective pairs. The English–Foreign-Pairs method. The experimenter gives subjects pairs of contrasting English words (e.g. dark/light, fat/thin). On the other side of a sheet, pairs of foreign language equivalents are written. Words within a pair are randomly arranged. Even though subjects know that a given English word matches one of the two foreign words standing opposite, they cannot tell from the arrangement which of the two it is. Subjects hear the words pronounced (the foreign words by native speakers) in the order in which they appear on the sheet. A good example of the application of this method is Brown, Black and Horowitz (1955).


• •

153

The Foreign–Foreign method. Stimulus words and two response words are given to subjects. Their task is to choose the equivalent of the stimulus word (e.g Maltzman et al. 1956). The Same–Different procedure. As an illustration, Brackbill and Little (1957) listed 50 words constituting a random sample of concepts of high frequency in usage. Instead of using contrasting pairs, Brackbill and Little included such terms as when, first, this, etc. The English forms were translated into Chinese, Japanese, and Hebrew. Subjects were presented with two words at a time (the two being from different languages) and were asked to judge whether they were the same or different in meaning. They were informed that half of the pairs were the same. The four languages were combined in all possible sets of two to yield six different collections of paired words. With this procedure subjects were able to guess with better than chance success for English–Hebrew (53%), Chinese–Japanese (54.8%) and Hebrew–Japanese (52.3%), but with success at or even significantly below chance levels with English–Japanese (50.3%), English–Chinese (49.9%), and Chinese–Hebrew (48.1%).

3.5 Research 3.5.1 General The following section introduces and comments on an experiment in phonetic iconicity. It was motivated by former research into phonetic iconicity and various experiments carried out by, inter alia, Sapir (1929) and Berko-Gleason (1958). Its purpose was to verify the hypothesis that iconicity is primarily bound to the language of small children, in contrast to the language of adult speakers who prefer to express evaluation by way of morphologically complex words. Moreover, the experimental results were related to the EM saturation values identified for the selected languages in the previous chapter, with the aim of finding out whether the age-affected preferences for iconicity vs complex word in evaluative formation correlate with the richness of evaluative morphology. The following sections introduce individual groups of informants and a data sheet used for the experiment. The description is followed by an analysis of the data obtained for individual languages, for individual age groups and for each task; these are compared, correlated, and commented on.

154

Chapter Three

3.5.2 Informants The experiment involved an experimenter, a native speaker of the language under investigation, and a sample of informants, native speakers of the tested language. The sample of informants encompassed five age groups for each language under research: 4-5; 6-11; 12-15; 16-19; and adults above 19. The minimum number of respondents per age group was10 informants.

3.5.3 Language of the experiment The experimenters used their respective mother language for giving instructions, but the words tested by the data sheet were Slovak. The informants completed the data sheets (Appendix 2) on the basis of an audio-recording. The recording was prepared in the Slovak language and recorded by a male native speaker of the Slovak language. The experiment was carried out in different language environments, including Spanish, Hungarian, and German. Each of these languages belongs to the Standard Average European group. Hungarian belongs to the Uralic languages, while Spanish and German belong to two different genera of the IndoEuropean language family. Slovak as a language of another IE genus was chosen as the basic, referential language of the experiment. The motivation underlying the selection of these languages was to include genetically different languages (Finno-Ugric vs two different genera of IE) and structurally different languages (agglutinative vs inflectional). Even more important, their selection resulted from the scope of this monograph, the main focus of which is evaluative morphology in the Standard Average European. It is one of the reasons why languages from other areas of the world, genetically and/or structurally completely different, were not included in the experiment. And finally, Spanish, German, and Hungarian became the experiment languages as a sort of Bauer’s understanding of a convenient sample for the very simple reason of the availability of collaborators willing to perform this kind of research with various groups of informants. Needless to say, it was highly demanding with regard to the low-age group and required a high degree of patience and personal involvement on the part of the collaborators.

3.5.4 Experiment The experimenters explained the tasks in their respective mother languages, with the instructions adapted to the individual age groups. The

Phonetic Iconicity in Ev valuative Morphhology

155

stimuli werre combinatiions of pictures (visual)) and Slovaak words (auditory). T The data sheett included four tasks. Task 1 The firsst task was aimed at phonetic icoonicity in evaluative e morphologyy. It was a foorced-choice task based onn actual Slovaak words. There were three bears inn the picture–aa big one, a ddefault one an nd a small one. The picctures were shhown to the su ubjects and thhe experimenteers, using language appropriate to thhe subjects’ ag ge,8 presentedd the following g task: You can see three beaars. In Slovak, the standard--sized bear is called c medveď. You will heaar two other similar names:: one for a big g bear and onee for a smalll bear. Pleaase match thhe pictures with w a correspoonding name from f the reco ording. (I gott a picture off three bears froom my friend Johnny and he h asked me too give them names. n He cannnot decide whhich name is suitable s for thhe big bear an nd for the smalll bear. A norm mal-size bear is called meddveď. Can you u help me? Johhnny comes frrom a country y far away annd he recordeed the names hee has chosen for me. I will play the nam mes for you an nd you will show w me what to call the bears, ok?) Figure 3-1 Drrawing for Taskk 1

medveď

8 The text in the brackets suuggests how thee task could bee introduced to the lowest age group.

156

Chapter Three

Recording: medveď, medveď (standard size), medvedisko, medvedisko, medvedík, medvedík medveď, medveď (standard size), medvedisko, medvedisko, medvedík, medvedík The experimenter showed the picture below and waited 20 seconds, and then switched on the recording. The order of words in the recording does not correspond with the sequence of pictures. The experimenter played the whole recording, and then played each word separately. The subjects were supposed to match the word they heard with one of the bears in the picture. This was repeated for each of the three words recorded. The choice of the word was based on Universal #1926. There are many evaluative suffixes in the Slovak language and not all of them are purely iconic. However -ík and -isko meet the iconicity requirement at a satisfactory level. If the hypothesis of the association of front vowels with smallness and back vowels with bigness is correct, the informants should match the leftmost bear with medvedík, the second with medveď and the last one with medvedisko. Task 2 The second task is an example of a forced-choice task based on a nonsense word. The experimenter showed the informants the picture below and provided, in his/her native language, the following explanation: This is an animal called mon. What would you call a tiny mon? Listen to the following words and choose a name for it. (My friend Johnny likes animals and he sent me a picture of a mon – it is in the picture. Johnny says that mons can be very tiny. He recorded two names for a miniscule mon and asked me to choose the more proper one. Listen.


157

Figure 3-2 Drawing for Task 2

Recording: moníček, moníček – min, min; moníček, moníček – min, min Mons can also be huge. What would you call a huge mon? Recording: monisko, monisko–mun, mun–man, man monisko, monisko–mun, mun–man, man The task was motivated by Berko’s wug experiment (1958). In this case the factor of age should play an important role. If Berko’s assumption is correct, the age group of 4-5-year-old children should choose min/mun, the age group 6-11 should still prefer an iconic form. Older age groups are expected to prefer morphological forms to the iconic ones. It was assumed that mun should be preferred as a clearly iconic augmentative (because of the u vowel). On the other hand, monisko as a longer word referring to something huge complies with the principle of constructional iconicity. This option is further supported by the back o vowel. Cross-linguistic research carried out by Štekauer, Valera and Körtvélyessy (2012) gives support to the observation that augmentative forms are rare even in those language that form diminutives by derivation. This is why it was hypothesized that speakers of Hungarian, a language without morphological augmentatives, would prefer mun. While the word moníček in Task 2 follows the derivational rules of the Slovak language and is coined by adding the diminutive suffix -íček, the word min is in accordance with phonetic iconicity, and the diminutive form is formed by substituting the front high vowel [i] for the back vowel [o] in the stem of the word. The suffix -íček is a cluster of two front vowels, one front consonant and one back consonant. My expectations were that the informants would refer the iconic min.

Chapter Three

158

Task 3 Task 3 tests what Sapir calls feelings. There are two short poems and two pictures. Nearly each word in the first poem is a morphological diminutive. The second poem consists of augmentative words. The first poem and the corresponding picture are expected to evoke a nice, quiet and playful feeling, and the second poem and the corresponding picture a gloomy and murky one. This should be reflected in the matching task. The experimenter explained: There are two pictures and two poems. What do you think, which poem describes the first picture and which the second? Listen to the following poem and match it with one of the pictures. (My friend Johnny likes writing poems; he is very good at it. He also sent me two new poems and drew pictures with them. I cannot speak his language and I do not know which poem matches the first and the second picture. Can you help me with this?) Figure 3-3 Drawings for Task 3


159

Recording Poem 1 Na lúčke hopkal zajačik Mal čiernučký ňufáčik. Zaňuchal ním mrkvičku, zamával labkou slniečku Poem 2 Zafúkalo vetrisko Zavylo vlčisko Stromisko striaslo listy posledné Temnota zakliala sny nádejné. It was hypothesized that if there is phonetic iconicity in evaluative morphology and if sounds create feelings, the first poem containing diminutives should be matched with the first picture and the second poem with the second picture. The reason why this task was included in the experiment is the presence of context. The words were not presented in isolation. In addition, each of the poems includes more than one diminutive/augmentative. Task 4 The last task was aimed at examining the acquisition of derivation. Two drawings were used. In both cases only a default-size object was given its Slovak name. Neither the smaller-than-default object in the first drawing, nor the bigger-than-default object in the second drawing of this task were labelled with a Slovak word. The subjects’ task was to create a diminutive word in the former case and an augmentative word in the latter case by relying on the sound form of the Slovak word for the default-size object. The experimenter’s task was to play the recorded words three times and ask the subjects to form the required words. It was presumed that the age group of 4-5 might experience difficulties in resolving this task. This is why the experimenter was instructed to leave the answer sheet clear if there was no answer by a child. It should be noted that the default-size words were not only recorded but also presented in their written forms, thus providing the subjects an additional stimulus. The experimenter was asked to write down the words created by the 4-5 age group. There are two houses in the picture – a common one and a small one. Give a name to the small house on the basis of the word ‘dom’.

Chapter Three

160 Figure 3-4 Drrawing for Taskk 4

do om And therre are two treees – a normal one and a bigg one. Give a name to the bigg tree on the basis b of the wo ord ‘strom’. Figure 3-5 Drrawing for Taskk 4

strrom Recording: dom, dom m–strom, stroom The expeected answerss are either iconic forms orr morphologiccal forms. Iconic form ms corresponnd with the idea of soound symbollism. By


161

implication, the diminutiive form of dom d could bee dim or dem m and the augmentativve forms of sttrom are strum and stram.. Morphologiccal forms should consiist of dom/stroom plus a dim minutive/augm mentative mark ker, either a Slovak-soounding markker, or a marrker influenceed by the infformant’s mother langguage. In 3.5.5 each languagge is analysed and commeented on acco ording to individual aage groups. Inn addition, thee results for eeach task in individual age groups are comparedd. The analysis employs thhe labels ‘icon nic form’ and ‘morphological form m’ where an iconic form iss one that corrresponds with the iddea of phonnetic iconicity y. A morphoological form m is any morphologiccally compleex form th hat expressees evaluation n by a morphologiccal marker. The possib bility of phhonetic icon nicity in morphologiccal forms is taaken into conssideration too..

3.5.5 Disccussion SPANISH Age group 4-5 There weere 19 childreen in the youn ngest age grouup of Spanish speakers, consisting mostly of five f year old ds. None off them matcched the augmentativve medvediskoo with the bigg gest bear. In the seecond task, 133 children chose the iconic form ‘min’ fo or a small mon and 6 oof them opted for the morph hologically forrmed ‘moníčeek’: Chart 3-1 Spaanish, Age grouup 4-5, Task 2, Diminutives D

A TINY MON monníček 322%

minn 68% %

162

Chapter Three

The choice of the augmentative form was similar. 16 children chose the iconic form, the majority of which (13) picked up ‘mun’, while ‘man’ was chosen by 3 children. The morphological form ‘monisko’ was found only in three cases: Chart 3-2 Spanish, Age group 4-5, Task 2, Augmentatives

The results of Task 3 are unfit for analysis. Unfortunately, the conditions for the conducting of the experiment were unsatisfactory, and children were tested as one large group instead of being interviewed in groups of three or four. This situation led to one of the children suggesting a solution and the rest of the group followed his decision. This is why the results obtained for this task were not taken into consideration. A similar problem was encountered in Task 4. However, as it is the youngest age group, no conclusive data were anticipated. The most interesting form given by these children is the iconic ‘din’ for a small house. Age group 6-11 The group consisted of 22 Spanish children aged 11. The majority of them (82%) matched the augmentative form of medveď with the smallest bear. Only 18% correctly matched medvedisko with the big bear.


163

Chart 3-3 Spaanish, Age grouup 6-11, Task 1

MEDV VEĎ medvedisko 18%

meedvedík 82%

o informants chose the ico onic min, In the seecond task, thhe majority of while a morpphological forrm was proposed in only 7 cases (32%). Chart 3-4 Spaanish, Age grouup 6-11, Task 2, Diminutives

A TINY Y MON m moníček 32%

m min 688%

t augmentative forms iss somewhat different. The situuation with the Slightly morre than a halff of children chose c the iconnic form (eitheer man or mun) and the second half of children ch hose the morpphological form m.

Chapter Three

164

Chart 3-5 Spaanish, Age grouup 6-11, Task 2, Augmentativess

A HUGE MON man 9% m monisko 46% munn 45% %

T 3. 50% oof children maatched the Similar rresults were iddentified in Task poems withh the correct pictures, p whille the remainning fifty perccent were wrong in theeir assessmentt. In the piie chart for thhis task the bu unny+bunny label suggests that the informants m matched the picture p of the bunny with th the poem abou ut bunny. Bunny+wolff, on the otherr hand labels the t match of tthe bunny piccture with the poem abbout wolf. Chart 3-6 Spaanish, Age grouup 6-11, Task 3

POEM

bunny+ wolf 50%

bunny+ bunny 50%


165

In the first part of Task 4 children were asked to name a small house. I anticipated some influence from the previous two tasks here. This influence could be either iconic or morphological. Children provided various words for a small house and a big tree. Altogether 16 various forms were collected. Table 3-1 Spanish, Age group 6-11, Task 4, proposed words dom

strom

dim, dam, domidved, dorms, domín 2x, domisko, domi, doming, domiogen, domiginte, domest, domedi, doming, tiny dom stram, stromisko, stromid, stromer, strondved, stromks, huge strom, stroman, stromi, stromsko, stromón, stromogen, stromgest, stromgs, stromked, stromdom, strombón,strongisko, stroming

Two iconic forms were proposed, however, one of which was augmentative (dam). This is why only dim can be considered a diminutive form of dom. None of the morphological forms resembled the Slovak words. domín is analysable as dom-ín, a Spanish suffix. One of the expressions (tiny dom) is analytic, combining English and Slovak. Therefore, it was excluded from further analysis. In one case, the Slovak augmentative domisko was found. An interesting case is that of domveď, a blend of dom and ‘medveď’, i.e., a form that occurred in Task 1 of the experiment. The rest of the proposed words are quite difficult to analyse. However, the informants used suffixes in all cases, and the majority of them are iconic (-edi, -ing, -inte). Similar results were obtained for the ‘big tree’. No diminutive iconic forms were proposed. There was one augmentative iconic, stram. The rest of the coined words were morphological constructions. Stromisko and stromsko are forms with the Slovak augmentative suffix -isko. The word strombón resembles Spanish in that -ón is an augmentative suffix.9 In addition, this word may have been motivated by the personal name Ramón (meaning originally ‘large branch’). The rest of the words were difficult to analyse. One expression was analytic and was excluded from further analysis. In analogy to domved above, the expression stromved was proposed for a big tree. This excludes the possibility that the informant wrongly identified the form -ved’ as a diminutive or an augmentative 9

However, as indicated by one of the reviewers, #sC clusters are illegal in Spanish and are turned into #esC clusters.

Chapter Three

166

suffix. Evenn though we caan (crosslingu uistically) findd suffixes form ming both diminutives and augmeentatives depending on tthe context (like the Lithuanian --ùt),10 one cann hardly expeect this kind oof knowledgee from an informant frrom the age grroup of 6-11. Nevertheless,, some of the proposed suffixes cann be considereed as iconic au ugmentative ssuffixes (-dom m; -gisko), others as icoonic diminutivve suffixes (-in ng; -id). Age group 12-15 In total, there were 24 informantts in the age group of 12 2-15. The majority of them were aged a 13 (21 informants). 2 informants were 14 years old annd one was 122 years old. The T majority oof informants failed to choose the ccorrect answeer in Task 1. 62% 6 of them picked the diiminutive form insteadd of the augmeentative. For 38% 3 the soundd suggested th he size. Chart 3-7 Spaanish, Age grouup 12-15, Task 1

MED DVEĎ medvediskoo 38%

m medvedík 62%

% of informan nts preferred tthe morpholog gical In the seecond task, 71% diminutive fform moníček.

10

For exampple, balt-ùt-is ‘vvery white’ can n be used in aaugmentative meaning m as well as in dim minutive/amelioorative meaning g (Bonifacas Stuundžia, pers.com m.).


167

Chart 3-8 Spaanish, Age grouup 12-15, Task 2, 2 Diminutives

The choiice of the auggmentative forrm was not soo unequivocal. Half of the informannts chose a morphological m l form and thhe other half an iconic form. 37% pproposed mun, while only 13% proposedd man. Chart 3-9 Spaanish, Age grouup 12-15, Task 2, 2 Augmentativves

A HUGE MON m man 113%

monisko 50% munn 37% %

mants associatted the poemss with the In Task 3, only one-thhird of inform mants failed to choose the rigght poem. correct pictuure. 16 inform

168

Chapter Three

Chart 3-10 Spanish, Age group 12-15, Task 3

In Task 4, 9 informants failed to come up with any word for a small house. 15 informants created various words from which dim and dem are iconic. Dum is iconic too, however, in an augmentative meaning, and so it refers to a big house rather than to a small one. The rest of the proposed words are morphological forms. The form domníček was probably influenced by moníček as one of the options in Task 2. The suffix -ike in domike appears to be its distorted variant. Domín contains the Spanish diminutive suffix -ín. Domisko is an augmentative form. Even though the remaining words are difficult to analyze, they were coined by means of suffixation and the suffixes are mostly iconic. In all, the informants were able to match diminutive means, either iconic or morphological, with diminutive meaning in 8 cases. Thus, the morphological markers prevail. 10 informants proposed no augmentative form of strom. The iconic form stram was identified by two informants. While morphological forms prevailed, only stromisko (used twice), stroman and stromako can be identified as the augmentative forms of strom. Stromček and stromíček are clearly diminutive and estrog is not analysable. Morphological forms prevailed, but only 4 out of 8 forms are augmentatives (stromako, stromisko 2x and stroman).


169

Table 3-2 S Spanish, Age group g 12-15, Task 4, prop posed words dom strom

dim, dem, dum, domníčeek, domisko 22x, dominiz, dimi, d domin, dom mnisko, domee, domike strem 2x, stram, s stromak ko, stromček, stromíček 2x,, strome, stromisko 2x, 2 estrog, stroman

Age group 16-19 This group includes 19 informants, with 13 at thhe age of sixteeen, and 6 at the age oof eighteen. Inn Task 1, 68% % of the infoormants picked out the diminutive fform instead of o the augmen ntative. Iconiciity, consequen ntly, does not seem to play any role in this age group. Chart 3-11 Sppanish, Age grooup 16-19, Task k1

MEDV VEĎ medvediskko 32%

meedvedík 668%

n the iconic aand the morp phological In Task 2, the difference between forms is verry small, onlyy 6%. Whereaas 53% of thee informants preferred the iconic foorm 47% of thhem opted for a morphologiical form.

Chapter Three

170

k 2, Diminutivess Chart 3-12 Sppanish, Age grooup 16-19, Task

A TINY Y MON

moníček 47%

minn 53% %

me results weere obtained for ‘huge moon’, with 53% % of the The sam informants ppreferring prefferred a morphological form m, and 47% of them an iconic one. O Only one resppondent chose the form mann. Chart 3-13 Sppanish, Age grooup 16-19, Task k 2, Augmentatiives

A HUGE MON man 5%

munn 42% %

monisko 53%

nformants (58 %) did not maatch the In Task 33, more than a half of the in picture withh the poem corrrectly.


171

Chart 3-14 Spanish, Age group 16-19, Task 3

Four informants did not propose any word in Task 4. The rest of the informants proposed both iconic and morphological forms, as follows from Table 3-3. Table 3-3 Spanish, Age group 16-19, Task 4, proposed words dom strom

dim 4x, dum, dam, domíček 4x, domišík, domín domisko 3x strum 7x, stromisko 4x, stromón, stromig, stromk, stromonisko

Out of three iconic forms, only dim corresponds with Universal #1926. Dum and dam are iconic augmentatives. Among the morphological forms only domisko (used by 3 informants) is an augmentative. Apparently, analogy with the preceding tasks may have played a role here. The morphological form domín appears to be a combination of the Slovak dom and the Spanish diminutive suffix -iin. Domišík can be analyzed as dom+išík, a suffix resembling the Slovak diminutive suffix -ik. Domíček, proposed by 4 informants, is close to the actual Slovak diminutive domček. It may have been influenced by moníček. In total, 10 out of 15 informants proposed one of the possible diminutive forms; in 4 cases it was an iconic form and in 6 cases a morphological form. 15 out of 19 informants completed this task for a big tree. The iconic form strum was proposed by 6 informants. Only 2 morphological forms

Chapter Three

172

can be analyysed as forms with diminutiive meaning; both of these resemble the Slovak ssuffix -ik (stroomig, stromik)). The rest werre augmentativ ve forms: the form stroomisko appeared 4 times, which w is a com mbination of th he Slovak elements sttrom and -issko. The forrm stromonissko can be analysed analogicallyy. Stromón seeems to be a combinationn of the Slov vak word strom and thhe Spanish suuffix -ón. In to otal, the augm mentative meaaning was expressed bby 13 informaants, in 6 casses by iconiccity and in 7 cases by means of moorphology. Adults aabove 19 There w were 19 subjects in this ag ge group. Surrprisingly, alll of them matched thee biggest bear with the diminutive form m medvedík in Taask 1. The resuults of Task 2 are unambigu uous – 79% off the informan nts picked up the morpphological dim minutive insteaad of an iconicc form. Chart 3-15 Sppanish, Age grooup above19, Taask 2, Diminutiives

A TINY MON m min 221%

moníčeek 79% % The pie chart for the second part of Task 2 alsso provides co onclusive results, how wever, in this case c the inform mants preferreed iconic form ms.


173

Chart 3-16 Sppanish, Age grooup above 19, Task T 2, Augmenntatives

A HUGE MON maan 5% %

m monisko 26%

mun 69%

In Taskk 3, 79% of the informants matched the picture with the nteresting thaat the same nu umber of correspondinng poem corrrectly. It is in informants oopted for the morphological m l form mon. Chart 3-17 Sppanish, Age grooup above 19, Task T 3

POE EM bunnny+ wolff 21%

bunny+ buunny 79%

T 4 are sum mmarized in T Table 3-4: The worrds coined in Task

Chapter Three

174

Table 3-4 Spanish, Age group above 19, Task 4, proposed words dom strom

domíček 10x, domček 2x, dim, domisko 2x, dim 2x, domicki, domi strum 9x, stromun 2x, stromum, strumum, stromisko 5x, stromu

All informants coined a diminutive form of dom. Only one form (proposed by 5 informants) has a Slovak augmentative suffix -isko, and no form was analysed as iconic augmentative. 17 informants coined a form that could be analysed as the diminutive form of dom. In 3 cases, the principle of iconicity can be identified, in particular, in the proposed word dim. The most frequent morphological form was domíček, analysable as dom-íček (including a Slovak diminutive suffix). The same holds for domček. Domi combines both morphology and iconicity. While the suffix -i is iconic, it does not exist in Slovak or Spanish. In all, this group of informants was very successful in this task: 17 out of 19 informants preferred a diminutive form. Morphological forms prevail, and, generally, the Slovak suffixes or ‘Slovak sounding’ suffixes were used. The augmenative of strom was coined by all informants. Three strong tendencies can be observed. First, all forms are evidently augmentative. Second, there is only one iconic form strum, but it was used by as many as 9 informants. Third, the morphological forms can be divided into two groups, one group employing the Slovak augmentative suffix -isko (stromisko proposed 5 times) and another group apparently motivated by the Spanish augmentative suffix -on (stromun, stromum, strumum, stromu). Summary by tasks Medveď In the youngest age group, none of the children matched the form for a big bear with the relevant picture. Interestingly, the same situation occurred in the age group above 19. The highest percentage of those who matched the sound form referring to a big bear with the picture of a big bear was in the age group of 12-15. However, it was only 38%. The hypothesis suggested by Universal #1926 was not confirmed by Task 1.


175

Mon Table 3-5 Task 2 A tiny mon, Diminutives, overview by age groups 4-5 6-11 12-15 16-19 above 19

moníček (%) 32 32 71 47 79

min (%) 68 68 29 53 21

In the first two age groups, the iconic form is clearly preferred to a morphological construction. A completely different picture is provided by the age group 12-15 where the morphological from is preferred by 72% of informants. In contrast to this, the iconic form prevails in the 16-19 age group, although the difference between the iconic and the morphological forms is just a single informant: 9 informants chose the morphological form and 10 informants proposed the iconic form. The results of the adult informants (> 19) roughly correspond with those of the 12-15 age group. In principle, the assumption based on Berko’s experiment was confirmed. If small children can choose between a morphological and an iconic form, they prefer the latter. Table 3-6 Task 2 A huge mon, Augmentatives, overview by age groups Age group

monisko %

4-5

16

6-11

46

12-15

50

16-19

53

above 19

26

Iconic forms mun % man% 84 68 16 54 45 9 50 37 13 47 42 5 74 69 5

Only 16% of the youngest age group proposed the morphological form monisko. 84% preferred the iconic form instead. The same tendency, although not so strong, can be observed in the age group of 6-11. The following two age groups provided us with almost identical results: about

Chapter Three

176

50% of the informants prefer a morphological form. The data for the oldest age group is very similar to that of the youngest informants: 75% of them prefer an iconic form. There were two iconic forms, mun and man. The u vowel is pronounced farther back compared to a, and as such it should better correlate with the idea of bigness than a. This was confirmed in each of the age groups. Poem Table 3-7 Task 3 Poem, overview by age groups

4-5 6-11 12-15 16-19 above 19

bunny+bunny % 50 33 42 79

bunny+wolf % 50 67 58 21

The proportion of correct and incorrect answers is balanced in the age group of 6-11. The majority of informants in the next two age groups failed to provide a correct match. The only age group with predominantly correct matching were the adult informants. By implication, this task – despite the involvement of context – did not give much support to the hypothesis of the role of feeling in evaluative morphology. dom/strom The results indicate that the higher age groups of informants were more systematic in completing this task and that the preference for morphological forms grows with the age of the informants. Last but not least, this group of informants, in contrast to the younger ones, clearly prefers words analysable as dom/strom + Slovak suffix. HUNGARIAN Age group 4-5 This group includes 18 children: 4 of them were 4-years old, the rest were 5-year-old children. In the first task, 72% of informants correctly matched the diminutive form of medveď with a small bear.


177

Chart 3-18 Huungarian, Age group g 4-5, Task k1

MEDV VEĎ meddvedík 228%

medvvedisko 772%

wever, this In Task 2, 56% of infformants prefeerred the iconiic form. How preference iis not concluusive; the diffference of 8% % is made by only 2 subjects. Chart 3-19 Huungarian, Age group g 4-5, Task k 2, Diminutivess

A TINY Y MON

minn 56% %

m moníček 44%

Chapter Three

178

In the augmentative part of task 2, the prooportion betw ween the informants who chose an iconic fo orm and thosse who prefeerred the morphologiccal form is fiffty-fifty. From m the two iconnic forms man n is more common. Chart 3-20 Huungarian, Age group g 4-5, Task k 2, Augmentatiives

A HUGE E MON

man 44%

m monisko 50%

m mun 66% matched the picture In the thiird task, only 40% of subjects correctly m with the poeem. Chart 3-21 Huungarian, Age group g 4-5, Task k3

POEM bunnny+bunny 39% bunnny + wolf 61%


179

Words coined within Task 4 are summarized in the following table: Table 3-8 Hungarian, Age group 4-5, Task 4, proposed words dom strom

kis dom 5x, kicsi dom 3x; domikó 7x ; domodík 2x; domisko 2x; domocska 2x nagy strom 14 x; lombos strom 2x; stromisko2x; stromiček 2x

All informants created both diminutive and augmentative forms of dom and strom. None of the diminutive forms of dom were iconic. Two of them were analytic (kis dom, kicsi dom). The rest were morphological forms, combining both Slovak and Hungarian elements. In two cases the augmentative domisko was used. Domikó and domocska could be analysed as dom-kó/-cska (both are Hungarian diminutive suffixes). Domodík is probably a combination of the Slovak dom and an element extracted from medvedík in the first task. The Hungarian word for ‘bear’, medve, is very similar to Slovak medveď. This means that the analysis of medvedík for a Hungarian speaker could be medve-dík. Analogically, a small house is dom-dík. In all, no iconic forms were used, analytic forms were preferred, and 9 informants combined a Slovak word with a diminutive marker (sounding either Slovak or Hungarian). There are no morphological augmentatives in the Hungarian language. This may be the reason for the prevalance of analytic forms in the informants’ proposals (nagy strom ‘a big strom’, lombos strom ‘deciduous strom’). No iconic forms were used. Out of two morphological forms, one is a combination of Slovak strom and a Slovak diminutive suffix -iček, and the other one is an augmentative Slovak form of strom. The word was coined by the same informant as domdved in the first part of the task. In summary, there is a considerable dominance of analytic forms, with total absence of iconic forms. Two informants proposed the Slovak augmentative word stromisko. Age group 6-11 This age group was made up of 29 children, including one eight-yearold child, 20 nine-year-old children, and the remaining informants were of the age of 10. Only 17% of the informants matched the picture of a small bear with the sound medvedík. The rest, 83%, matched a sound referring to a small bear with a picture of a big bear.

Chapter Three

180

g 6-11, Tassk 1 Chart 3-22 Huungarian, Age group

MEDV VEĎ medvedíík 17%

medveddisko 83% %

f min for a tiny mon waas preferred by y 69% of In Task 2, the iconic form subjects. Chart 3-23 Huungarian, Age group g 6-11, Tassk 2, Diminutivves

A TINY MON monníček 311%

minn 69% %

m monisko wass preferred to denote a In contraast, the morphhological form big mon.

Phonetic Iconicity in Ev valuative Morphhology Chart 3-24 Huungarian, Age group g 6-11, Tassk 2, Augmentat atives

A HUGE MON man 17%

mun 21%

moonisko 662%

m a correct match in Taskk 3. 41% of tthe subjects made Chart 3-25 Huungarian, Age group g 6-11, Tassk 3

l of words proposed p withiin Task 4: Table 3-9 provides a list

181

182

Chapter Three

Table 3-9 Hungarian, Age group 6-11, Task 4, proposed words dom strom

domíček 2x, domisko 3x, dim 8x, dem 3x, tom, domo, domd, dom, bom, dame, dobe, sdrom, mod, drom, , dum, stromisko 3x, stram 5x, strem 2x, strim 3x, strum, stromíček, trom, drom, somt, stro, drom, otrom, dromb, strom, trom, lohd, romro, stremo

Two informants did not coin any word for a big tree and a small house. In comparison to the previous age group, no analytic forms were introduced. On the other hand, many opaque forms, for example, dobe, sdrom, mod, trom, drom, somt, otrom, dromb, were proposed, which may indicate better creativity of this age group. Both iconic and morphological forms were used for a small house: dim occurred 8 times and dem three times. The morphological form domíček, based on the Slovak language, occurred twice. Some augmentatives (instead of diminutives) were proposed, too: the iconic dum occurred once and the morphological form domisko 3 times. Opaque expressions were also proposed for the augmentative of strom, for example, trom, drom, somt, stro, drom, otrom, dromb, strom, trom, lohd, romro, stremo. The iconic stram was the most frequent with 5 occurrences. The morphological form stromisko was proposed 3 times. Iconic diminutives were also found among the proposed words: strem occurred twice and strim three times. Age group 12-15 There were 23 pupils in this age group, three of which were at the age of 13, and the rest were 14-year-old children. In the first task, only 35% of the informants made a ‘correct’ choice and matched the augmentative form of the word medveď with the biggest bear in the picture. The difference of a mere 4 % between the iconic min and the morphological moníček is the difference of one informant, and therefore, it is insignificant. From this point of view, the proportion of iconic and morphological forms is roughly equal.


183

Chart 3-26 Huungarian, Age group g 12-15, Taask 1

MEDV VEĎ medvediskoo 35%

meedvedík 665%

Chart 3-27 Huungarian, Age group g 12-15, Taask 2, Diminutiives

A TINY MON

min 52%

m moníček 48%

The situuation in the augmentative part of Taask 2 is diffeerent. As indicated abbove, Hungariian does not have morpho logical augmentatives. This may bbe the reason for only 35% % of the info formants prov viding the morphologiccal form monisko. The restt chose an icoonic form, preevailingly the man optiion.

Chapter Three

184

g 12-15, Taask 2, Augmenttatives Chart 3-28 Huungarian, Age group

A HUGE MON moonisko 335%

man 39%

mun 26% c matcched the pictu ures in the More thaan a half of thhe informants correctly poem task. g 12-15, Taask 3 Chart 3-29 Huungarian, Age group

POE EM

buunny + wolf 48%

bbunny+bunny 52%

The worrds coined in Task T 4 are sum mmarized in thhe following table: t


185

Table 3-10 Hungarian, Age group 12-15, Task 4, proposed words dom strom

domdisko 4x domisko 6x; domeska 2x; dim 2x; domosko 2x; domodík 5x; domík 2x stromdík 4x; stromed 2x; stromík 3x; stram 4x; stromisko 6x, stromíček 2x; stromdisko 2x

The iconic form dim was used twice. Domodík and domík might be analyzed as dom plus a Slovak suffix, either -ík or o+dík, where -o- can be a linking element inserted due to the vowel harmony rule in Hungarian; and -dík could be a result of secretion from medvedík, similar to the proposals in the age group of 4-5. The rest of the words are not easy to analyse for two reasons. First, in Hungarian orthography the letter s corresponds with sound /ʃ/. This being a palatal sound it is more iconic than the alveoral /s/. Consequently, domisko, for example, pronounced with /ʃ/ could be analyzed as a diminutive form. However, if the Slovak pronunciation of s is applied, the form is clearly augmentative. -isko is a Slovak augmentative suffix. Second, in the Hungarian language vowel harmony plays an important role. In fact, it overrides the phonetic iconicity principle. Given these circumstances, domosko is a diminutive form. For all these reasons, the conclusion is that all the informants proposed a diminutive form of dom: two of the words were iconic, and the rest morphological forms. There are four occurrences of augmentative iconic forms of strom. One of the morphological forms is beyond analysis–stromed. Stromdík, stromík, and stromíček occurred nine times in total. They may be analyzed as strom and a Slovak diminutive suffix. Stromisko and stromdisko contain the Slovak augmentative suffix -isko. 12 out of 23 informants coined an augmentative form of strom, either iconic or morphological. Age group 16-19 The group consisted of 19 subjects, 9 of them aged 16, and 10 of them aged 17. In the first task, 74% of informants matched the big bear with the augmentative form of medveď.

Chapter Three

186

g 16-19, Taask 1 Chart 3-30 Huungarian, Age group

MEDV VEĎ medvvedík 266%

medvediskko 74% nformant. In the seecond task, thee differnce of 6% is causedd by a single in The share off the iconcic and a morpholog gical forms is,, thus, roughly y equal. Chart 3-31 Huungarian, Age group g 16-19, Taask 2, Diminutiives

orm of mon w was proposed by b only a The morrphological auugmentative fo quarter of the informannts. The rest chose eitheer man or mun, m with preference ffor mun.


187

Chart 3-32 Huungarian, Age group g 16-19, Taask 2, Augmenttatives

A HUGE MON moniisko 26% %

mann 32% %

mun 42% % m the ppoem with dim minutives 63% of the informantts correctly matched with the bunnny picture. g 16-19, Taask 3 Chart 3-33 Huungarian, Age group

POE EM buunny+wolf 37% buunny+bunny 63%

T 4 are sum mmarized in thhe following table: t The worrds coined in Task

Chapter Three

188

Table 3-111 Hungarian n, Age group p 16-19, Task 4, proposed words w dom strom

dim 3x, dom mič, domíček 3x, domček 33x, domík, domičke 2x, domičkko, domico, do omici, domiskka, domisko2x x strum 4x, stram s 4x, strom moko 2x, strom mičko, stromiisko 4x, stromod, sttromdík, strom míček, stromkoo 2x

The iconnic form dim occurred o threee times. Howevver, all the oth her forms are morphollogical. Whilee all informan nts proposed a word, the an nalysis of the words pproposed is, however, diff fficult for thoose reasons previously p mentioned iin the 12-15 age group. In n spite of thi s, the words domíček, domček, dom mík, domičke,, domičko cou uld be analyzzed as dom + a Slovak diminutive ssuffix (or one resembling Slovak). 8 iconic augmentativee words of strrom were prooposed in this group of S respondents. The remainiing 11 words are morpholoogical forms. Stromdík, stromíček, sstromko and stromičko s cou uld also be aanalyzed as strom + a Slovak dim minutive suffixx. The only unequivocal unambiguouss form is stromisko. bove 19 Age group Adults ab The grouup of adults inncludes 22 sub bjects, all of w whom were un niversity students. In the firrst task, 18 infformants (out of 22) matcheed correctly th he Slovak diminutive m medvedík withh the picture of o a small bearr. Chart 3-34 Huungarian, Age group g above 19 9, Task 1


meddvedík 882%


189

In the ssecond task, the iconic fo orm min wass preferred, while w the morphologiccal form monííček was chosen by only 23 % of the inforrmants. Chart 3-35 Huungarian, Age group g above 19 9, Task 2, Diminnutives

A TINY MON moníčček 23% %

m min 777% Even thhough there are no morrphological aaugmentativess in the Hungarian llanguage, the morphologiccal form mon isko was proposed by the majorityy of the inform mants (55%). Chart 3-36 Huungarian, Age group g above 19 9, Task 2, Augm mentattives

A HUGE MON man 9%

mun 36%

m monisko 55%

Chapter Three

190

In Task 33, 77% of the informants su ucceeded in thhe matching taask. Chart 3-37 Huungarian, Age group g above 19 9, Task 3

bunny+woolf 23%

POE EM

bunnny+bunny 77%

mmarized in thhe table below w: The dataa collected forr Task 4 is sum a 19, Tassk 4, proposed d words Table 3-122 Hungarian, Age group above dom strom

domiček 8xx, domček 2x, domi 2x, dom meka, domeček, domik, dom misko 7x stromisko 9x, 9 strum 3x, stram s 3x, strom mego, stroma 2x, stromiček

An attem mpt to coin a diminutive form of dom m was made by each informant. None of thhe words pro oposed resem mbles the Hungarian H language. A All of them weere morpholog gical forms, aand with the exception e of the augm mentative dom misko (occurrring 7 times)), all forms could be analysed as morphologiccal diminutivees. Two subj ects failed to o coin an augmentativve form for strom. Aparrt from strom miček, which h can be analysed as morphological diminutive, all forms arre rather augm mentative. Both morphhological and iconic means were used. T The most freq quent was stromisko (99 times), follow wed by the ico onic strum annd stram (3 occcurrences each).


191

Summary by tasks Medveď The results obtained from the Hungarian informants indicate that there is a tendency for the lower age groups (4-5, 6-11) to match the sound form representing a small bear with a picture of the big bear, which contradicts the hypothesis. In contrast to this, as many as 65% of the informants of the 12-15 age group and 82% of the group of adults managed a “correct” match The data for the age group 16-19 do not contribute to the “Hungarian tendency” of stronger iconicity with increased age. As a result, the age-based data obtained from Task 1 is too heterogeneous to manifest any transparent tendency with regard to the role of phonetic iconicity in evaluative morphology. Mon Table 3-13 Task 2 A tiny mon, Diminutives, overview by age groups moníček % 44 31 48 53 23

4-5 6-11 12-15 16-19 above 19

min % 56 69 52 47 77

In two age groups (above 19 and 6-11) the preference for an iconic form is significant. This fact and the other data in Table 3-13 thus do not affirm the hypothesis that children, unlike adults, prefer iconic forms. Table 3-14 Task 2 A huge mon, Augmentatives, overview by age groups Age group

monisko %

4-5

50

6-11

62

12-15

35

16-19

26

above 19

55

Iconic forms mun % man% 6

40

17

21

26

39

42

32

36

9

Chapter Three

192

The fact that there are no morphological augmentatives in the Hungarian language must be taken into consideration when commenting on Task 4. The numbers in the table show that the age groups of 4-5, 6-11 and >19 display no significant differences – their choice between the iconic and morphological forms is fifty-fifty. On the other hand, the remaining two age groups (12-15 and 16-19) show a very strong preference for iconic forms. The data on iconic forms show interesting results, too. While the younger age groups prefer man to mun, the preference for mun increases with the age of the informants. This result may be blurred by the fact that the vowel a in the Hungarian language is an open (low), back vowel. Poem Table 3-15 Task 3, overview by age groups 4-5 6-11 12-15 16-19 above 19

bunny+bunny% 39 41 52 63 77

bunny+wolf% 61 59 48 37 23

The table shows an unambiguous tendency: the number of correct matches positively correlates to the age of the informants. While in the youngest age group only 39% of matches were correct, the percentage for the adult group was twice as much. dom/strom Remarkably, while in the youngest age group analytic forms occurred, they were absent in the remaining age groups. In the case of a small house, morphological forms are preferred. The augmentative morphological form stromisko competes with the iconic forms strum and stram. Proposals like strim, strem for a big tree and domo, bom for a small house contradict the hypothesis of the association between the sound form representing an object and its meaning. If it were a sort of association, the choice for a small house would be dim or dem, and that for a big tree strum or stram.


193

GERMAN Age group 4-5 In the youngest age category, there were 14 children, 6 of them at the age of four, and 8 children at the age of five. In Task 1, only 2 out of 14 children matched the sound representing a small bear with the picture of a small bear. 12 informants matched the picture of a big bear with the augmentative form of the word medveď. Chart 3-38 German, Age group 4-5, Task 1

It follows from Chart 3-38 that nearly half of the informants chose the derived form for a small mon.

Chapter Three

194

Chart 3-39 Geerman, Age grooup 4-5, Task 2,, Diminutives

A TINY MON

m moníček 43% min 57%

The second part of Task 2 was aimed at thhe augmentatiive form. Monisko is ccoined on the basis of rules of the Slovakk language: mu un makes use of phoneetic iconicity, man is a kind d of distractor.. The derivved form monnisko was also o preferred in tthe augmentattiveoriented parrt of Task 2. Chart 3-40 Geerman, Age grooup 4-5, Task 2,, Augmentativees

A HUGE MON man 14%

mun 22% moonisko 664%


195

In Task 3, 57% of the subjects wrongly matched the picture of a wolf with the poem about a bunny. Chart 3-41 German, Age group 4-5, Task 3

As with the Spanish and Hungarian children, Task 4 proved to be the most difficult for this age group. 9 out of 14 responses were “don’t know” for the picture of a house and 7 for the tree picture. Table 3-16 German, Age group 4-5, Task 4, proposed words dom strom

bek, dom, pju, tom, gela bem, báume, spek, strom, minudika, dom, grossbaum

Bek, dom, pju, tom, gela were the words created for the small house. In one case, the default-size word dom was used for a small house. Interestingly, none of the words proposed by the informants are derived. With the exception of tom, they do not copy or imitate the sound form of the original dom. Bek, pju and gela seem to be nonce-formations based on the creativity of children. It may be assumed that these words were motivated by the pictures and feelings the pictures engendered in children. Similar conclusions can be drawn on the second part of Task 4. Strom is a repetition of the original word standing for a default-size tree. It was proposed by the same informant as dom above. Bem, báume are apparently motivated by the German Baum ‘tree’. Groβbaum is a compound based on German constituents. Spek and minudika are words made up by children.

Chapter Three

196

What makess minudika intteresting is thaat it starts witth min, which indicates something ssmall, and ennds with -ka. This last eelement resem mbles the Slovak dimiinutive suffix -ka (e.g. in vranka v ‘crow--DIM’). By im mplication, there are tw wo diminutive techniques ussed in this woord to evoke something small. For obvious reasonns, the informaant could not bbe aware of th he Slovak diminutive system. Whatt makes this word even m more interestin ng is that despite the strong diminuutive-oriented d iconicity strructure, it stan nds for a big tree. Age group 6-11 The secoond youngest age group co onsisted of 2 nnine-year-old subjects, 10 informannts were 10 years old, and 8 informants were 11-years old. In the first taskk, 85% of the subjects correectly matchedd the sound referring to a small bearr with a picture of the small bear. Chart 3-42 Geerman, Age grooup 6-11, Task 1


meddvedík 885%

In Task 2, a half of thhe informantss selected the iconic form min for a m preferred thhe morphological form small mon, and the otherr half of them moníček.


197

Chart 3-43 German, Age group 6-11, Task 2, Diminutives

The morphological form monisko was preferred by 70% of the informants. The remaining 30% chose either of the iconic forms: 20% man and 10% mun. Chart 3-44 German, Age group 6-11, Task 2, Augmentatives

In Task 3, 75% of the subjects correctly matched the poem with diminutives to the picture of a bunny.

Chapter Three

198

Chart 3-45 Geerman, Age grooup 6-11, Task 3

POE EM w wolf+bunny 25%

buunny+bunny 75%

The folloowing table suummarizes thee words propoosed by the infformants in Task 4: posed words Table 3-17 German, Agee group 6-11,, Task 4, prop dom strom

na, duum 3x, dim 4xx, domiček 3x,, domnjeski, ddomimi, domin doomisko 3x, dam msta, domiaom m, strrim 2x, stram 2x, stromísko o 3x, stromus, stromnjuwsko o, strromovič, strom mistcek, strom mko, stromoniss, stromiček, strosem, s strromamasima, big strom 2x

19 varioous words foor the diminu utive of dom were proposed. One informant diid not proposse any word. Two T iconic foorms include dum d with three occurrrences and dim d which occcurred four ttimes. Dim meets m the conditions ffor phonetic iconicity. i Thee rest of the w words were, however, morphologiccal formations. Domiček (o occurring threee times) and d moníček may have beeen inspired by b Task 2. A similar situatiion can be ob bserved in the group of augmentaative formations for stroom. Stram meets m the conditions oof phonetic iconicity; i striim (occurringg twice) is diiminutive rather than aaugmentative in its phonetiic shape. Stroomisko (occurrring three times) is thhe correct Sloovak word fo or the augmeentative of strrom. The analytic form m big strom was w proposed twice. t


199

Age group 12-15 The group consisted of 26 informants. 7 of them were 14 years old; 19 were 15 years old. In Task 1, 20 subjects correctly matched the picture of a small bear with the corresponding sound form. Chart 3-46 German, Age group 12-15, Task 1

In the first part of Task 2, 62% of the subjects chose the iconic form for the denotation of a small mon. Chart 3-47 German, Age group 12-15, Task 2, Diminutives

200

Chapter Three

Although with a smaller difference in preference between the iconic and morphological forms, the situation in the second part of Task 2 mirrors that for diminutive forms. 46% of subjects chose the morphological form monisko and 54% preferred one of the iconic forms, with mun occurring more frequently. Chart 3-48 German, Age group 12-15, Task 2, Augmentatives

69% of the subjects correctly matched the poem about a bunny with the ‘bunny’ picture. It can be assumed that the sound of the diminutive forms also suggested the meaning. Chart 3-49 German, Age group 12-15, Task 3


201

The words coined in Task 4 are given in the following table: Table 3-18 German, Age group 12-15, Task 4, proposed words dom strom

domiček 8x, domisko, 11, domi, dim 5x,duffe, duffi, stromisko 12x, stromo, strum 6x, stram3x, stromdig, strim, strabimn, strôm

The base dom provided with the same diminutive suffix as that introduced in Task 2 (-íček) occurred 8 times. The corresponding augmentative suffix from Task 2 (-isko) used incorrectly for the diminutive form of dom occurred as many as 11 times. The iconic diminutive dim was proposed 5 times. Two interesting forms were proposed in this age group, duffi and duffe. They are interesting in that they combine a stem vowel alternation, including a high back vowel, and suffixation, including a high front and central front vowel. By implication, these respondents attach greater weight to the role of suffix as a diminutive marker. At the same time, they do not seem to perceive upper-backness of vowels as contradictory to the diminutive meaning. Strom with the augmentative suffix -isko had 12 occurrences. The iconic words strum and stram occurred 9 times. However, the diminutive iconic strim in the augmentative meaning was proposed, too. The word strôm indicates that the informant tried to imitate the Slovak language by using in the stem one of its characteristic (even though rarely occurring) phonemes. Age group 16-19 The group consisted of 17 students at the age of 18, and 5 informants aged 19. The first task brought interesting results. As the pie chart below illustrates, the proportion between the correct and the incorrect matches was fifty-fifty.

Chapter Three

202

k1 Chart 3-50 Geerman, Age grooup 16-19, Task

MEDV VEĎ

medvedisko 50%

m medvedík 50%

This agee group of Gerrman informaants preferred to derive a diiminutive of mon by suuffixation. Onnly 27% of theem chose the iiconic min. Chart 3-51 Geerman, Age grooup 16-19, Task k 2, Diminutive s

me situation occurred o in th he case of huuge mon: 73% % of the The sam subjects choose the morphological form m. 27% preferrred the iconic form, but only 1 subjeect chose man..


203

Chart 3-52 Geerman, Age grooup 16-19, Task k 2, Augmentatiives

A HUGE MON man 4%

munn 23% %

moniisko % 73% ociated the firsst poem about a bunny In Task 3, 55% of thee subjects asso with the corrrect picture off a bunny. Chart 3-53 Geerman, Age grooup 16-19, Task k3

POE EM

wolf+bunny 45%

bunny+bunny y 55%

w proposed d within Taskk 4: The folloowing words were

Chapter Three

204

Table 3-19 German, Agee group 16-19 9, Task 4, prooposed wordss dom strom

doomíček 10x, dim d 3x, domisko 2x, domicii 3x, domí misko 9x, strom mig, stromestiiko, stromicek k, sttrum 3x, strom sttromosko 2x, stromo, s strom miko

As the table shows, the diminutiive form of dom with th he Slovak diminutive ssuffix -íček apppeared 10 tim mes; the iconiic dim was prroposed 3 times. The S Slovak augmeentative suffix x -isko was inccorrectly used d twice in the diminutiive function. The T same aug gmentative sufffix -isko was correctly used 10 timees as an augm mentivizing sufffix of strom. The iconic fo orm strum occurred 3 times. bove 19 Age group Adults ab The grouup was comprrised of 36 su ubjects. The rresults of Task k 1 show that 56% off the informantts correctly matched m the souund form refeerring to a small bear w with the picturre of a small bear. Chart 3-54 Geerman, Age grooup above 19, Task T 1

MEDV VEĎ

medvedisko 44%

m medvedík 56%

a balanced, where w 53% thhe of subjects preferred The dataa for Task 2 are the morphollogical diminuutive form of mon. m


205

Chart 3-55 Geerman, Age grooup above 19, Task T 2, Diminuttives

A TINY MON

min 47%

m moníček 53%

T 2, the difference betw ween the morp phological In the seecond part of Task and iconic fforms lies in favour fa of the morphological m l form, chosen n by 64% of the inform mants. Either of o the iconic forms f was choosen by 36%. Only 8% of the inform mants picked up u the iconic form f man. Chart 3-56 Geerman, Age grooup above 19, Task T 2, Augmenntatives

mann 8% %

A HUG GE MON

mun 28% moonisko 664%

o the inform mants incorrecctly matched the t poem In Task 3, only 8% of about a bunnny with a pictuure of a wolf.

206

Chapter Three

Chart 3-57 German, Age group above 19, Task 3

The words coined within the last task are summarized in the following table: Table 3-20 German, Age group above 19, Task 4, proposed words dom strom

domíček 21x, dim 8x, domisko 5, domin, domí stromisko 21x, strum 5x, stram 4x, stromicek, stromik, stromasko, stromonsko, stromesk, stromo

For the diminutive of dom, a morphological form is preferred. Two Slovak suffixes were used for this purpose: the diminutive suffix -íček occurred 21 times and the augmentative suffix -isko occurred 5 times. Similar results are provided by the analysis of augmentatives proposed by this age group. In 21 cases the Slovak augmentative suffix -isko was used. In 9 cases an iconic form was proposed, either strum or stram. The variety of words proposed for the augmentative of strom is much greater than that for the diminutive of dom. Summary by tasks Medveď The results for this task are rather inconsistent. While the sound form referring to a small bear was matched with the picture of a small bear by only 14% of the youngest informants, the percentage for the same task in the second youngest age group was 85%. In the following age groups this


207

tendency declined to just 56 % of the informants who matched medvedisko with a picture of a small bear. A tiny mon Only the age group of 12-15 shows a strong tendency towards iconic forms, and only the age group of 16-19 towards morphological form. In the other age groups the distribution of iconic and morphological forms is more or less balanced. Table 3-21 Task 2 A tiny mon, Diminutives, overview by age groups 4-5 6-11 12-15 16-19 above 19

moníček % 43 50 38 73 53

min % 57 50 62 27 47

A huge mon The age group of 12-15 differs from the other groups also in the second part of Task 2. It is the only age group with a slight preference for iconic forms (54%). The other age groups prefer the morphological monisko to the iconic mun or man. In the case of iconic forms, mun is preferred to man in all age groups with the exception of the 6-11 age group. Table 3-22 Task 2 A huge mon, Augmentatives, overview by age groups Age group

monisko %

4-5

64

6-11

70

12-15

46

16-19

73

above 19

64

Iconic forms mun % man% 22

14

10

20

39

15

23

4

28

8

Chapter Three

208

Poem Even though the results are not homogeneous, they suggest a relationship between the age of the subjects and the tendency to associate the sound with meaning. The most successful group in matching the ‘diminutive’ poem with a nice and neat picture of a bunny was the group of adults, while the least successful group was the youngest age group. This observation is, however, violated by the data for the second oldest age group. Table 3-23 Task 3, overview by age groups 4-5 6-11 12-15 16-19 above 19

bunny+bunny% 43 75 69 55 92

bunny+wolf% 57 25 31 45 8

dom/strom The German informants almost uniformly completed the task, with the exception of the youngest age group. Both iconic and morphological forms were equally distributed. The most frequent words for the diminutive dom were either the morphological domiček or the iconic dim. Very often a form with the Slovak diminutive suffix -isko occurred. Interestingly, forms that most probably should resemble the Slovak language occurred quite often (e.g. domnjeski). The most frequent augmentative form for strom was the morphological stromisko and the iconic strum. With the higher age of the informants, the frequency of iconic forms increased.

3.5.6 Summary by languages, tasks and age groups Task 1 vs. age groups Age group 4-5 Table 3-24 below summarizes the results for the age group 4-5 across the languages. As mentioned above, no data was gained from the Spanish language. The results for Hungarian and German lead towards a similar conclusion: the sound of the diminutive medvedík does not establish any association with the picture of a small bear.


209

Table 3-24 Summary by languages, Task 1, Age group 4-5 Language Spanish Hungarian German

medvedík 28% 14%

medvedisko 72% 82%

Age group 6-11 Table 3-25 gives the results for the age group 6-11. Two interesting facts can be observed. First, the results for the Hungarian language are very similar to the previous age group. On the other hand, the results for the German language are opposite to the age group 4-5. Second, they are nearly the same as the results for the Spanish language. The difference between Hungarian and the other two languages, is striking. While the sound form of medvedík was associated with a small bear by German and Spanish speakers, Hungarian speakers did not make such an associaiton. Table 3-25 Summary by languages, Task 1, Age group 6-11 Language Spanish Hungarian German

medvedík 82% 17% 85%

medvedisko 18% 83% 15%

Age group 12-15 The results for all three languages are very similar. The sound designating a small bear was matched with a picture of a small bear in all three languages by more than 60% of informants. Table 3-26 Summary by languages, Task 1, Age group 12-15 Language Spanish Hungarian German



Age group 16-19 As illustrated in Table 3-27, the Hungarian language again deviates from the results of the German and Spanish languages. This repeatedly occurring deviation from the percentages obtained for the other two languages raises principled questions. First, is it the nature of language

Chapter Three

210

(e.g. its genetic class or morphological classification) that influences the sensibility for phonetic iconicity? Or are the inter-lingual differences caused by the degree of use (saturation) of evaluative morphology in a particular language? The Hungarian language has no morphological augmentatives; in contrast to this, the saturation value of augmentatives in German is 1.67, and for Spanish it is as much as 2.67. Table 3-27 Summary by languages, Task 1, Age group 16-19 Language Spanish Hungarian German



Age group above 19 The results for the oldest age group are summarized in Table 3-28. In the age group of adults, the correspondence between sound and meaning was significant. The Slovak diminutive suffix -ík was matched with a small bear in all three languages by the prevailing number of subjects; in Spanish it reached the absolute value of 100%. Table 3-28 Summary by languages, Task 1, Age group above19 Language Spanish Hungarian German

medvedík 100% 82% 56%


Conclusion 1 The analysis of the cross-linguistic data for various age groups in Task 1 suggests that the sensitivity to phonetic iconicity in evaluative morphology is very low in the youngest age groups. The sensitivity to phonetic iconicity seems to increase in Hungarian speakers with advancing age, although the 16-19 age group violates this age-conditioned growth. Disregarding the youngest group of respondents, the sensitivity to phonetic iconicity is very high in the other two languages, as early as the 6-11 age group of speakers. While high percentages across the individual Spanish age groups are consistent, culminating with the adults at 100%, the tendency for German speakers seems to be opposite, with the percentage data being in indirect proportion to the increasing age of


211

German speakers. Given these substantial differences across the individual languages, no clear tendencies for phonetic iconicity can be proposed on the basis of Task 1. Task 2 (part 1) vs age groups The results of the first part of Task 2 are summarized in Table 3-29. In the youngest age groups (4-5; 6-11) the iconic min is preferred in each of the three languages. The only exception is the German 6-11 group, with equal distribution of the iconic and the morphological forms min and moníček. If the two age groups are taken as a whole, the most significant dominance of the iconic form min can be observed for Spanish. The share of the iconic form in the older Spanish groups significantly drops, even if this type of diminutive slightly prevails in the 16-19 age group. In any case, Spanish seems to be the only of the three languages that appears to give support to Berko’s hypothesis of the preference for iconic formations in the youngest age groups and the growing preference for morphological constructions with the increasing age of speakers. The evidence provided by the German informants does not seem to substantiate this assumption, with only minor differences between the individual age groups and with the exception of deep drop of iconic formations in the 16-19 age group. In Hungarian, the iconic forms prevail in the age-distanced groups of 6-11 and >19, with no clear tendency in either direction. Table 3-29 Summary by languages, Task 2, Diminutives, all age groups

Language

min (%)

moníček (%)

Spanish

68

32

Spanish

29

71

Hungarian

56

44

Hungarian

69

31

Hungarian

52

48

German

57

43

German

50

50

German

62

38

(%)

32

moníček

min (%)

68

(%)

Spanish

moníček

Language

Age group 12-15

min (%)

Age group 6-11

Language

Age group 4-5

Chapter Three

Spanish Hungarian German

21 77 47

79 23 73

(%)

47 53 73

moníček

53 47 27

min (%)

moníček


(%)

min (%)

Age group adults

Language

Age group 16-19

Language

212

Conclusion 2 Based on Berko’s wug experiment, it was expected that the preference for the morphological form would rise with age regardless of the language examined. Berko’s hypothesis was only partly supported by Spanish informants and refuted by the German and Hungarian informants. Even though it can be generally stated that the youngest age group prefers an iconic form of diminutives, the growing tendency towards morphological constructions has not been confirmed in the other two languages. Moreover, the clear tendency for the growth of morphological constructions in Spanish is interrupted in the age group 16-19. The iconicmorphological forms in the German language are more or less balanced; the Hungarian language, however, shows a counter-tendency. This result, as in Task 1, correlates with the saturation value of evaluative morphology in these languages. The saturation value for the Spanish language is 6.00, for the German language it is 4.34, and for Hungarian it is 3.33. The highest saturation value appears to positively correlate with the frequency of occurrence of morphological diminutive forms with the growing age. Chart 3-57 illustrates the preference for morphological forms in relation to individual age groups. In the youngest age group (4-5), the starting point for Hungarian and German languages is roughly the same: 43% of German and 44% of Hungarian informants chose morphological forms. Spanish, the language with the highest EM saturation, starts with a different value: only 32% of Spanish informants aged 4-5 prefer morphological forms. In other words, the preference for iconic forms is stronger. Even though in Hungarian and German this tendency is weaker, all of these three languages show preference for iconic forms in the youngest age group (cf. Table 3-29). Regardless this similar starting point and various falling vs. rising tendencies in the older age groups, the results for the two oldest age groups in my research provide solid evidence of the correlation between the EM saturation and the preference for morphological forms. As the full line in the chart suggests, in the


213

Hungarian language (the language with the lowest EM saturation), the preference for the morphological forms decreases with the growing age. The Spanish language (language with the highest EM saturation) shows an opposite tendency. The German language assumes a position between Hungarian and Spanish in both respects – from the point of view of EM saturation and the preference for morphological forms. Chart 3-58 Morphological forms vs age groups

Morphological forms vs. age groups 100 preference in %

80 60

Spanish

40

German

20

Hungarian

0 age age age age adults group group group group 4-5 6-11 12-15 16-19

Task 2 (part 2) vs. age groups The following table gives an outline of the results for the second part of Task 2: Table 3-30 Summary by languages, Task 2, Augmentatives, all age groups

monisko (%)

mun (%)

man (%)

Language

monisko (%)

mun (%)

man (%)

Age group 6-11

Language

Age group 4-5


16 50 64

68 6 22

16 44 14


46 62 70

45 21 10

9 17 20

Chapter Three

214

man (%)

69 36 28

5 9 8


53 26 73

42 42 23

man (%)

13 39 15

monisko (%) mun (%)

man (%)

37 26 39

Language

mun (%)

50 35 46

Age group 16-19

mun (%)


monisko (%)

Language

Age group 12-15

5 32 4


monisko (%)

Language

Age group Adults

26 55 64

The diversity of the results for the augmentative part of Task 2 precludes the drawing of any significant generalizations. The hypothesis was that (besides the preference of the iconic mun and man by the younger age groups) the speakers of the Hungarian language (a language without augmentatives) would prefer the iconic forms. This hypothesis has not been confirmed by the experimental data. The results show that in the Spanish language the preference for the morphological form monisko rises with the age, with the exception of the adult age group where its frequency of occurrence significantly drops. In any case, the youngest Spanish informants are the only group to give support to Berko’s hypothesis. The German percentages for the morphological form monisko are almost equally distributed across the individual age groups (with the exception of the age group 12-15), ranging from 64% to 73%. In this respect, they roughly map the situation in the first part of Task 2. Even the youngest age group prefers the morphological form. The data for the Hungarian language oscillate considerably: the younger age groups have no significant preferences, the age groups 12-15 and 16-19 prefer iconic forms, and the adults use both iconic and morphological forms without any significant preferences.


215

Conclusion 3 As soon as the above data is correlated with the saturation values, a tendency can be observed. The augmentative saturation value for the Spanish language is 2.67, for the German language it is 1.67, and for the Hungarian 0. The highest saturation value suggests the rising preference for morphological formations. The non-existence of augmentatives in the Hungarian language means a sort of gap in the linguistic system of Hungarian language speakers in terms of the non-existence of any morphological rule that would prompt the informants to form augmentatives by way of suffixation rules. Nevertheless, this system-level situation has not been reflected in the ‘behaviour’ of each of the age groups. Based on the results, another conclusion can be drawn. According to the hypothesis, back vowels are associated with the idea of bigness. The [u] vowel could be specified as more back than [a], and this idea is also reflected in the recording of the words mun and man. From this it follows that mun should be preferred to man. Very strong support for this hypothesis was given by Spanish speakers. In Hungarian, this expectation was unambiguously met only in the two oldest age groups, while the data for the youngest group of Hungarian informants totally contradict the expectations. The hypothesis of the preference for mun to man has also been unambiguously supported by the three oldest age groups of German informants. The data thus enable us to conclude that, while not being a universal feature, the preference for the u-iconocity to a-iconicity is a strong tendency. Surprisingly, however, it is a tendency manifested especially in the older age groups of language users. Task 3 vs age groups Table 3-31 Summary by languages, Task 3, all age groups

bunny+ bunny (%)

wolf+ bunny (%)

Language

bunny+ bunny (%)

wolf+ bunny (%)

Age group 6-11

Language

Age group 4-5


39 43

61 57


50 41 75

50 59 25

wolf+ bunny (%)


bunny+ bunny (%)

Language

79 77 92

21 23 8

wolf+ bunny (%)


bunny+ bunny (%)

Spanish 33 67 Hungarian 52 48 German 69 31 Age group above 19

Language

Age group 16-19 wolf+ bunny (%)

Age group 12-15 bunny+ bunny (%)

Chapter Three

Language

216

42 63 55

58 37 45

The Hungarian language shows an evident tendency: the older the subjects are the higher the number of those who matched a poem dominated by morphological diminutives with the picture of a smiling bunny. This age-related growth of correct answers is impressive. The same tendency can be identified for the Spanish language from the age group of 12-15. Even though the results for the German language are more heterogeneous, the youngest age group has the lowest success rate, and the age group above 19 the highest success rate. With the exception of the 1619 age group, the German success rates are highest from among all three languages. Conclusion 4 The main goal of this task was to examine the significance of context in perceiving phonetic iconicity in evaluative morphology. As soon as the words were not presented in isolation and the task consisted of more than one morphological diminutive, the subjects’ correct identification increased with age. While the diminutive and augmentative forms of medveď ‘bear’ in Task 1 were tested as isolated units, the diminutive and augmentative forms in this task were set in a context. The result gained is in accordance with the results obtained in Task 1 – the sensitivity to


217

phonetic iconicity in evaluative morphology increases with the progressing age of language users. Context seems to even strengthen this tendency. Task 4 vs age groups This task aimed to evaluate the following questions: first, the presence of iconic forms was checked. Two types of size-related iconic forms are distinguished: the forms that are associated with the concept of smallness (dim, dem) and the forms that are associated with the concept of bigness (dam, dum). The morphological forms can be threefold. The evaluative markers used can resemble or be identical to Slovak markers under the influence of the preceding tasks of the experiment (especially Task 2); they can be taken from the mother tongue of the informants; or they can be creatively “manufactured”. If the marker was adopted from the Slovak language, it could be either diminutive or augmentative. If the marker was Slovak and diminutive for the word dom or Slovak and augmentative for the word strom, the result could be compared to the results of Berko’s observations (1958) suggesting that children acquire morphological rules at the age of 7. Spanish language The youngest Spanish respondents failed to complete the task. The second youngest age group was more successful and various forms were proposed for both dom and strom – iconic, morphological and also (undesired) analytic formations. Iconic forms are rare and inconsistent: next to dim (front vowel) one can also find dam (back vowel). The morphological markers were not adopted from Slovak but for the augmentative marker -isko. In the age group 12-15 the Slovak diminutive marker -íček occurs in the correct diminutive meaning, and the suffix -isko is used properly for the augmentative meaning. Various iconic forms were proposed, but the use of front and back vowels did not always reflect their expected iconicity. Front vowels were also used to indicate augmentative meaning (strem), and back vowels served for the diminutive meaning (dum). The results that best comply with the expectations were obtained from the age group of 16-19: the prevailing iconic diminutive forms contain the front vowel [i] (dim) and the prevailing augmentative forms the back vowel [u] (strum). Moreover, the adopted Slovak diminutive/augmentative markers were used correctly. These tendencies are stronger in the age group above 19, which at the same time prefers the morphological over the iconic forms.

218

Chapter Three

Hungarian language Forms for the diminutive dom in the youngest age group are either morphological or analytic. The morphological forms are coined with both diminutive and augmentative Slovak evaluative markers. Similar results can be observed in the youngest age group for strom. The second age group displays a higher number of the iconic forms: dim occurred 8 times; strum and stram 6 times each. Front vowels in the proposed iconic forms were also used in the augmentative meaning. If the Hungarian data is compared to Spanish, the number of different forms (very often unanalyzable) is much higher. This does not apply to the third age group, where the morphological forms start to prevail. This is interesting, especially in the case of the augmentative forms of strom, as there are no morphological augmentatives in the Hungarian language. In the 16-19 age group, the iconic form for dom occurs only twice (in both cases it is dim). Morphological forms are more frequent. The morphological markers essentially resemble the Slovak language suffixes, however, the augmentative suffixes are prevailingly employed to express the diminutive meaning. Interestingly, the opposite tendency is observed for strom: the prevailing morphological forms make use of diminutive markers. In the category of adults no iconic forms for dom were proposed. The morphological forms usually have a properly used Slovak diminutive marker in compliance with the expected meaning. In the case of augmentatives both strum and stram were proposed 3 times each. Stromisko occurred 9 times. A Slovak evaluative marker with diminutive meaning occurred only once. Keeping in mind the fact that Hungarian morphology has no augmentative markers, it may be concluded that the higher age prefers morphological markers to iconic as these are more transparent. German language As in the Spanish language, children aged 4-5 failed to complete the task. The forms for dom in the 6-11 age group are diverse, but two tendencies can be observed–the preference for the iconic dim and the morphological domiček. This tendency is strengthened in the age group of 16-19 and is unambiguously present in the age group of adults, with 21 occurrences of morphological forms and 8 occurrences of iconic forms. These observations are also confirmed for the augmentative meaning. While in the age group of 6-11 the words proposed are fairly heterogeneous, the age group of 12-15 proposed the augmentative stromisko 12 times and the iconic forms (strum/stram) 9 times. In the adult


219

group, the proportion is 21 morphological formations and 9 iconic formations. Conclusion 5 This task gives further support to the hypothesis of the influence of the saturation value on phonetic iconicity. The age at which we can observe the growing preference for morphological forms is indirectly proportional to the EM saturation value characterizing the language. While in the Spanish language (EM saturation 6) the preference for morphological forms is obvious as early as the 6-11 age group, the same tendency in the Hungarian language (EM saturation 3.33) can be first observed in the age group of 12-15. In addition, the preference for the morphological forms strengthens with the growing age of language users. This observation is also evident in the Hungarian language. Although there are no morphological augmentatives, the subjects preferred morphological augmentatives to iconic forms. Despite the already mentioned gap in the system-level rules for augmentative formation, one might hypothesize that this gap can be compensated for by analogical diminutive formation rules. The preference for morphological formations is further strengthened by the transparency of these kinds of diminutives and augmentatives. The iconic forms, on the other hand, as pointed out by Sapir, rely on “feelings”, intuition. The experimental results suggest that informants of a higher age prefer to apply linguistic knowledge instead of feelings. Of course, various factors such as education, occupation, gender etc. must be taken into consideration, too.

CONCLUSIONS

Several objectives were pursued in this monograph. First, this monograph aims to be another contribution to research into the intriguing topic of Standard Average European, in particular, by examining the role of another parameter, the parameter of evaluative morphology. It goes without saying that the more linguistic parameters are examined and evaluated, the better the picture of the geographical delimitation and the internal structure of SAE. Certainly, this kind of research can hardly be expected to deliver significant data without the contrastive analysis of two extensive samples, one of them covering the SAE languages and the other the languages of the world. It is only by way of this kind of comparison that one can justify the relevance of the parameter examined and its importance to the study of SAE. A comparison of data of the two language samples has proved the significance of the parameter of evaluative morphology to this area of research. In addition, it has been demonstrated that different parameters may yield different internal structuring of the SAE territory in terms of its core and periphery. It has been demonstrated that the EM saturation parameter shows, first of all, clear and unambiguous differences between SAE and global samples, and, second, that the internal structure of the SAE territory differs from that obtained on the basis of some other parameters. Specifically, unlike the former assumptions, postulating the core of the SAE area to be represented by French, German and one or two other languages (Italian, Dutch, Polish, Russian) depending on the author and/or the linguistic category examined, my research has moved the centre eastward (primarily Slavic languages). Similar differences can be found in the identification of the SAE territory border: thus the position of Celtic, eastern Uralic languages, Maltese, Basque, and the Eastern European languages seems to be uncertain, heavily dependent on the methodology of research. This entails serious methodological and theoretical questions concerning the homogeneity of the SAE territory, the interrelation of the conclusions obtained from the evaluation of various linguistic parameters, and the identification of the number of relevant as well as key parameters that should, no doubt, range over all linguistic levels. As a by-product, several important assumptions have been made. Most importantly, this research has re-acknowledged the validity of the almost


221

absolutely universal observation that there is an implicational relation between augmentatives and diminutives. In other words, if a language has morphological augmentatives it may be assumed to have morphological diminutives. This gives rise to the question of why morphologically formed diminutives are much more common across languages of the world than morphological augmentatives. This fact has a strong diachronic support in the existence of diminutive suffixes in some of the examined proto-languages of SAE. Thus, for example, the PIE *-ko- is preserved in Celtic (Matasović 2005); the ProtoNakh *-ik’ / *-k’ is found in all three Nakh languages (Chechen, Ingush, Batsbi) and is very productive in Ingush (Nichols 2011); Turkic velar dimunitive -Ak (Erdal 1991, 40-58) indicating PTK *-Ak diminutive (Martine Robbeets, pers. com.). Contrary to this, the data obtained do not provide us with the corresponding augmentative suffixes. The reason for this asymmetry may only remain a matter of speculation, but one reason may stem from physiological grounding and cognitive processes in the human mind as discussed in Prieto (forthcoming). An additional reason may be due to cultural background of human society, protecting their offspring and thus the future of a community, and expressing intensive feelings in communicating with babies and small children. From the diachronic point of view, this can be accounted for by the fact that “diminutives emerge prior to augmentatives; augmentatives tend to be recent inovations” (Mutz, forthcoming). As manifested by Mutz, Latin, for example, had a number diminutive suffixes but no augmentative ones. It has been demonstrated that evaluative morphology is far from being a self-evident and integral part of the languages of the world. A relatively high proportion of the languages of my global sample have no diminutives, and many more of them no augmentatives. It might be an interesting topic for parole-oriented research to assess the extent of stylistic correspondence of morphological and analytic diminutives and augmentatives. The EM saturation parameter, employed to assess the role of evaluative morphology in SAE, seems to be applicable to other linguistic features and might become a unifying methodological device in research into SAE. The phonetic iconicity part of the monograph has also yielded, inter alia, two brand new observations. First of all, my experimental cross-linguistic research indicates that there is correlation between EM saturation value and phonetic iconicity. The age at which we can observe the growing preference for morphological

222

Conclusions

forms is indirectly proportional to the EM saturation value characterizing the language. Second, the data obtained seems to justify the assumption that the preference for the morphological forms is proportional to the age of language users. Despite numerous interesting observations arrived at in this area of research, there are many tasks remaining, including more extensive research into phonetic iconicity in terms of the number of languages, the number of informants, their internal structuring by age and also by linguistic background in the case of bilingual speakers, mental types of the informants in terms of their way of thinking (e.g., the opposition between students preferring humanities, on one hand, and students preferring sciences, on the other), biological gender differences, etc. I believe that this monograph has provided an impetus for research in this direction.

APPENDIX A

Evaluative morphology data sheet Informant/source of data: Geographical area: Language: Morphological type Language family: Word order: Genus: A. DIMINUTIVES Does the language form DIMINUTIVE morphologically? WF process

Example

Literal translation of constituent morphemes

English equivalent

Prefixation Suffixation Infixation Root-and-pattern Compounding: Incorporation1 Reduplication Conversion2 Other

1

A verb-forming process, whereby a nominal stem is fused with a verbal stem to yield a larger, derived verbal stem. 2 Including cases of change of class without adding any explicit morpheme, with inflectional paradigm functioning analogically to zero derivational morpheme.

224

Prefixation Does your language have more than one DIMINUTIVE prefix? Is more than one DIMINUTIVE prefix possible in a word? Can one and the same DIMINUTIVE prefix be used more than once in the same word? Are there DIMINUTIVE prefixes that can have other than diminutive meaning? Suffixation Does your language have more than one DIMINUTIVE suffix? Is more than one DIMINUTIVE suffix possible in a word? Can one and the same DIMINUTIVE suffix be used more than once in the same word? Are there any DIMINUTIVE suffixes that can have other than DIMINUTIVE meaning? Reduplication Complete reduplication Partial reduplication Preposing reduplication Postposing reduplication Infixing reduplication

Appendix A


225

B. AUGMENATIVES Does the language form AUGMENTATIVE morphologically? WF process

Example

Literal translation of constituent morphemes

English equivalent

Prefixation Suffixation Infixation Root-and-pattern derivation (transfixation) Compounding Incorporation3 Reduplication Conversion4 Other: Prefixation Does your language have more than one AUGMENTATIVE

prefix? Is more than one AUGMENTATIVE

prefix possible in a word? Can one and the same AUGMENTATIVE

prefix be used more than once in the same word?

3

A verb-forming process, whereby a nominal stem is fused with a verbal stem to yield a larger, derived verbal stem. 4 Including cases of change of class without adding any explicit morpheme, with inflectional paradigm functioning analogically to zero derivational morpheme.

226

Are there AUGMENTATIVE

prefixes that can have other than AUGMENTATIVE

meaning? Suffixation Does your language have more than one AUGMENTATIVE

suffix? Is more than one AUGMENTATIVE

suffix possible in a word? Can one and the same AUGMENTATIVE

suffix be used more than once in the same word? Reduplication Complete reduplication Partial reduplication Preposing reduplication Postposing reduplication Infixing reduplication

Appendix A


227

C. SEMANTIC CATEGORIES OF EVALUATIVE MORPHOLOGY Can your language express the following semantic categories? Less than standard– QUANTITY OF Diminutives SUBSTANCE More than standard Augmentatives Less than standard– Diminutives QUANTITY OF QUALITY More than standard– Augmentatives Less than standard– Diminutives QUANTITY OF ACTION More than standard– Augmentatives QUANTITY OF Diminutives CONCOMITANT CIRCUMSTANCE

Yes/No

Example

Augmentatives

D. WORD CLASSES Can your language express Diminutiveness/ Augmentativeness in the following word-classes? Diminutives Nouns Augmentatives Diminutives Verbs Augmentatives Diminutives Adjectives Augmentatives Diminutives Adverbs Augmentatives Diminutive Class-changing Augmentative

Yes/No

Example

APPEND DIX B

Phonetic Iconicity I SHEET ANSWER S

AGE

Task 1

_____

medveď

______

A tiny m mon is

a) moníček

b) min.

A huge m mon is

a) mun

b moniskko

Task 2

m c) man


Task 3

Poem _______________

Poem _______________

229

230

Append dix B Task 4

do om

strrom

BIBLIOGRAPHY

Abondolo, Daniel Mario. “Khanty.” In The Uralic languages, edited by Daniel Mario Abondolo, 358-386. London and New York: Routledge, 1998. Abondolo, Daniel Mario, ed. The Uralic languages. London and New York: Routledge, 1998. Adelaar, Willem. Tarma Quechua: grammar, texts, dictionary. Lisse: Peter de Ridder Press, 1977. Acharya, Jayaraj. Descriptive grammar of Nepali. Washington, DC: Georgetown University Press, 1991. Alhoniemi, Alho. Grammatik des Tscheremissischen (Mari): mit Texten und Glossar. Hamburg: Helmut Buske Verlag, 1993. Amha, Azeb. The Maale language. Leiden: Research School of Asian, African, and Amerindian Studies, Universität Leiden, 2001. Anderson, Earl R. Grammar of iconism. Massachusetts: Fairleigh Dickinson University Press, 1998. Anderson, Stephen. “Where’s morphology?” Linguistic Inquiry, 1982: 571-612. Antunes de Araujo, Gabriel. A grammar of Sabanê, a Nambikwaran language. Amsterdam: LOT, 2004. Aoki, Haruo. “Symbolism in Nez Perce.” In Sound symbolism, edited by Leanne Hinton, Johanna Nichols and John J. Ohala, chapter 2: 15-22. Cambridge: Cambridge University Press, 1994. Apresjan, Yuri D. Lexical semantics: user’s guide to contemporary Russian vocabulary. Ann Arbor: Karoma, 1992. Aronoff, Mark, and Kirsten Fudeman. What is morphology? Oxford: Blackwell, 2005. Aronoff, Mark, and Janie Rees-Miller, ed. The handbook of linguistics. Oxford: Blackwell, 2003. Aronoff, Mark, and Shikaripur N. Sridhar. “Prefixation in Kannada.” In Theoretical morphology: approaches in modern linguistics, edited by Michael Hammond and Michael Noonan. San Diego: Academic Press, 1988, 154-195. Asher, R.E., and T.C. Kumari. Malayalam. London: Routledge, 1997. Atzet, Jon, and Harold B. Gerard. “A study of phonetic symbolism among native Navajo speakers.” Journal of Personality and Social Psychology, 41 (1965): 524-528.

232

Bibliography

Austin, Peter. A grammar of Diyari, South Australia. Cambridge: Cambridge University Press, 1981. Bakema, Peter, and Dirk Geeraerts. “Diminution and augmentation.” Vol. 1, 1045-1052. In Morphologie/Morphology: Ein internationales Handbuch zur Flexion und Wortbildung/An international handbook on inflection and word formation, edited by Geert Booij, Christian Lehmann and Joachim Mugdan. Berlin: Mouton de Gruyter, 2000. Bakker, Dik. “Language sampling.” In The Oxford Handbook of Linguistic Typology, edited by Jae Jung Song, 100-130. Oxford: Oxford University Press, 2013. Ball, Martin J., and James Fife. The Celtic languages. London and New York: Routledge, 1993. Bauer, Laurie. English word-formation. Cambridge: Cambridge University Press, 1983. —. “Evaluative morphology: in search of universals.” Studies in Language 3, no. 21 (1997): 533-575. —. “The function of word-formation and the inflection-derivation distinction.” In Words in their places: A Festschrift for J. Lachlan Mackenzie edited by Hendrik Aertsen, Michael Hannay and Rod Lyall, 283-292. Amsterdam: Vrije Universiteit, 2004. —. Introducing linguistics morphology. Edinburgh: Edinburgh University Press, 1988. —. “No phonetic iconicity in evaluative morphology.” Studia Linguistics. A Journal of General Linguistics, 2, no. 50 (1996): 189-206. Beard, Robert. Lexeme-morpheme base morphology. Alabama: State University of New York, 1995. —. “The plural as lexical derivation.” Glossa, 16 (1982): 133-148. Becker, Judith B., and Sylvia K. Fisher. “Comparison of associations to vowel speech sounds by English and Spanish speaker.” American Journal of Psychology, 1 (1988): 524-527. Bentley, Madison, and Edith J. Varon. “An accessory study of phonetic symbolism.” American Journal of Psychology 1, no. 45 (1993): 76-86. Berko-Gleason Jean. “The child‘s learning of English morphology.” Word, no. 14 (1958): 50-170. Berlin, Brent. “Huambisa sound symbolism.” In Sound Symbolism, edited by Leanne Hinton, Johanna Nichols and John J. Ohala, 76-93. Cambridge: Cambridge University Press, 1995. Bierwisch, Manfred. “The semantics of gradation.” In Dimensional adjectives: grammatical structure and conceptual interpretation, edited by Manfred Bierwisch, 71-261. Berlin: Springer-Verlag, 1989. Böhmerová, Ada. “Suffixal diminutives and augmentatives in Slovak – a


233

systemic view with some cross-linguistic considerations.” In Diminutives and augmentatives in the languages of the world, edited by Pavol Štekauer and Lívia Körtvélyessy, Lexis 6 (2011): 59-84. Bolozky, Shmuel. “On the formation of diminutives in Modern Hebrew morphology.” Hebrew Studies, 35 (1994): 47-63. Booij, Geert. “Inflection and derivation.” In Morphologie/Morphology. Ein internationales Handbuch zur Flexion und Wortbildung/An international handbook on inflection and word formation, edited by Geert Booij, Christian Lehmann and Joachim Mugdan, vol. 1: 360369. Berlin: Mouton de Gruyter, 2000. —. The grammar of words. Oxford: Oxford University Press, 2005. Booij, Geert, and Jaap van Marle. “Inherent versus contextual inflection and the split morphology hypothesis.” Yearbook of morphology edited by Geert Booij, 1-16. Springer, 1996. Brackbill, Yvonne, and B. Little Kenneth. “Factors determining the guessing of meanings of foreign words.” Journal of Abnormal and Social Psychology, no. 54 (1957): 312-318. Brown, Lea and Matthew S. Dryer. “Diminutive as an inflectional category in Walman.” http://wings.buffalo.edu/people/faculty/dryer/dryer/BrownDryerWalma nDimin.pdf (accessed 31 October 2014). Brown, Roger W., Abraham H. Black, and Arnold E. Horowitz. “Phonetic symbolism in natural languages.” Journal of Abnormal and Social Psychology, no. 50 (1955): 388-393. Brown, Roger W., Nuttall, Ronald. “Method in phonetic symbolism experiments.” Journal of Abnormal and Social Psychology, no. 59 (1959): 388-393. Buchholz, Oda. Albanische grammatik. Leipzig: Verlag Enzyklopädie, 1987. Camaj, Martin. Albanian grammar. Wiesbaden: Harrassowitz, 1984. Campbell, Lyle. “Areal linguistics: A closer scrutiny.” In Linguistic Areas: Convergence in historical and typological perspective, edited by April McMahon, Vicent Nigel and Yaron Matras, 1-32. Basingstoke: Palgrave Macmillan, 2006. —. The Pipil language of El Salvador. Berlin/New York: Mouton de Gruyter, 1985. Cardona, George. A Gujarati reference grammar. Philadelphia: University of Pennsylvania Press, 1965. Carlson, Robert. A grammar of Supyire. Berlin/New York: Mouton de Gruyter, 1994. Carstairs-McCarthy, Andrew. Current morphology. New York: Routledge,

234

Bibliography

1992. Chelliah, Shobhana Lakshmi. A grammar of Meithei. Berlin: Mouton de Gruyter, 1997. Child, Tucker S. A grammar of Kisi: a southern Atlantic language. Berlin/New York: Mouton de Gruyter, 1995. Clendon, Mark. A grammar of Worrorra. Kimberley: Kimberley Language Resource Centre, 2000. “Coloring pages tree.” Picornot. n.d. Counts, David R. A grammar of Kaliai-Kove. Honolulu: University of Hawaii Press, 1969. Coupe, Alexander Robertson. A grammar of Mongsen Ao. Berlin/New York: Mouton de Gruyter, 2007. Cowan, Hendrik Karel Jan. Grammar of the Sentani language, with specimen texts and vocabulary. Vols. Verhandelingen van het Koninklijk Instituut voot Taal-, Land- en Volkenkunde, 47. ’sGravenhage: Martinus Nijhoff, 1965. Creissels, Denis. “Niger-Congo.” In The Oxford handbook of derivation, edited by Rochelle Lieber and Pavol Štekauer. Oxford: Oxford University Press, forthcoming. Csúcs, Sándor. “Udmurt.” In The Uralic Languages, edited by Daniel Mario Abondolo, 276-304. London: Routledge, 1998. Davies, John H. Kobon. Amsterdam: North Holland, 1981. Day, Christopher. The Jacaltec language. Bloomington: Indiana University, 1973. Dayley, John Philip. Tzutujil grammar. Berkeley/Los Angeles/London: University of California Press, 1985. Décsy, Gyula. Die linguistische Struktur Europas. Vergangenheit, Gegenwart, Zukunft. Wiesbaden: Kommissionsverlag Otto Harrassowitz, 1973. —. The linguistic identity of Europe. Bloomington: Eurolinga, 2000. Déchaine, Rose-Marie, Raphaël Girard, Calisto Mudzingwa, and Martina Wiltschko. “Shona.” In Edinburgh Handbook of Evaluative Morphology, edited by Nicola Grandi and Lívia Körtvélyessy. Edinburgh: EUP, forthcoming. Derzhanski, Ivan. “On diminutive plurals and plural diminutives.” Morphology and Linguistic Typology. Bologna: University of Bologna, 2005. Diffloth, Gérard. “i: big, a: small.” In Sound Symbolism, edited by Leanne Hinton, John J. Ohala and Johanna Nichols, 107-114. Cambridge: Cambridge University Press, 1994. “Diminutive consonant symbolism in Algonquian.” In Papers of the Sixth


235

Algonquian Conference 1974, edited by William Cowan, 237-252. Ottawa: National Museums of Canada, 1975. Dixon, Robert M.W. A grammar of Boumaa Fijian. Chicago: University of Chicago Press, 1988. —. A grammar of Yidin. Cambridge: Cambridge University Press, 1977. Dokulil, Miloš. Tvoření slov v češtině. Teorie odvozování slov. Prague: Nakladatelství československé akademie věd, 1962. Donaldson, Bruce C. Dutch: a comprehensive grammar. New York and London: Routledge, 1997. Donohue, Mark, and Lila San Roque. I‘saka: a sketch grammar of a language of north-central New Guinea. Canberra: Australian National University, 2004. Dressler, Wolfgang U. “Prototypical differences between inflection and derivation.” Zeitschrift fur Phonetik, Sprachwissenschaft und Kommunikationsforschungen 42, 1989: 3-10. —. “Word-formation in natural morphology.” In Handbook of WordFormation, edited by Rochelle Lieber and Pavol Štekauer, 267-284. Dordrecht: Springer, 2005. Dressler, Wolfgang U., and Lavinia Merlini Barbaresi. “Morphopragmatics of diminutives and augmentatives.” In Perspectives on semantics, pragmatics, and discourse: a Festschrift for Ferenc Keifer, edited by István Kenesei and Robert M. Harnish, 43–58. Amsterdam: John Benjamins, 2001. Dressler, Wolfgang U., and Lavinia Merlini Barbaresi. Morphopragmatics: diminutives and intensifiers in Italian, German, and other languages. Berlin: Mouton de Gruyter, 1994. Dryer, Matthew S. “Large linguistic areas and language sampling.” Studies in Language, 1989: 257-292. Dryer, Matthew S., and Martin Haspelmath. The World Atlas of Language Structures Online. Max Planck Institute for Evolutionary Anthropology, 2013. http://wals.info (accessed 20 November 2013). Dryer, Matthew, and Martin Haspelmath. The World Atlas of Language Structures. Oxford: Oxford University Press, 2011. Dum-Tragut, Jasmine. Armenian. Amsterdam/Philadelphia: John Benjamins, 2009. Dutton, Tom E. Koiari. München: LINCOM Europa, 1996. Erdal, Marcel. Old Turkic word formation: a functional approach to the lexicon. Wiesbaden: Harrassowitz, 1991. Erschler, David. “Ossetic.” In HSK word-formation: an international handbook of the languages of Europe, edited by Susan Olsen, Peter O. Müller, Ingeborg Ohnheiser and Franz Rainer. Berlin/New York:

236

Bibliography

Mouton de Gruyter, forthcoming. Everett, Daniel L., and Barbara Kern. Wari: the Pacaas Novos language of western Brazil. London: Routledge, 1997. Fisher-Jørgsen, Eli. “On the universal character of phonetic symbolism with special reference to vowels.” Studia Linguistica, no. 32 (1978): 80-90. Fischer, Olga. “Evidence for iconicity in language.” Logos and Language: Journal of General Linguistics and Language theory, 2004: 1-19. Fitch, William Tecumseh Sherman. “Vocal tract length and formant frequency dispersion correlate with body size in rhesus macaques.” Journal of the Acoustical Society of America, no. 102 (1997): 12131222. Fleck, David William. “A grammar of Matses.” Rice University Electronic Theses and Dissertations. 2003. http://scholarship.rice.edu/handle/1911/18526 (accessed October 2012). Foley, William A. The Yimas language of New Guinea. Stanford: Stanford University Press, 1991. Fónagy, Ivan, and Judit Fónagy. “Distribution of phonemes in word sets contrasting in meaning.” Mélanges Marcel Cohen, 1970: 69-72. Ford, Lysbeth Julie. Phonology and morphology of Bachamal (Wogait). Canberra: Australian National University, 1990. Foris, David Paul. A grammar of Sochiapan Chinantec. Dallas: Summer Institute of Linguistics, 2000. Fortescue, Michael. West Greenlandic. London: Croom Helm, 1984. Fortin, Antonio. “The Morphology and Semantics of Expressive Affixes.” Unpublished doctoral dissertation. Oxford: University of Oxford, 2011. Frajzyngiern, Zygmunt, and Eric Johnston. A grammar of Mina. Berlin/New York: Mouton de Gruyter, 2005. French, Patrice L. “Towards an explanation of phonetic symbolism.” Word, no. 28 (1972): 305-322. Friedman, Victor A. “Romani in the Balkan linguistic league.” In Valkanike Glossologia: Sygkhronia kai Diakhronia/Balkanlinguistik: Synchronie und Diachronie, edited by Christos Tzitzilis and Charalambos Symeonidis, 2000. 95-105. Gebels, Gustav. “An investigation of phonetic symbolism in different cultures.” Journal of Verbal Learning, no. 8 (1969): 310-312. Genetti, Carol. A grammar of Dolakha Newar. Berlin/New York: Mouton de Gruyter, 2007.Gillis, Steven. “The acquisition of diminutives in Dutch.” Studies in Pre- and protomorphology, 1997: 144-165. González, Hebe Alicia. “A Grammar of Tapiete (Tupi-Guarani).”


237

Unpublished doctoral Dissertation. Pittsburgh: University of Pittsburgh, 2005. Gràcia, Lucia, and Lidia Turon. “On appreciative suffixes.” Acta Linguistica Hungarica 1-4, no. 47 (2000): 231-247. Grandi, Nicola. “Evaluative affixes between inflection and derivation: a typological survey.” Societas Linguistica Europaea - 44th Annual Meeting. Logrono: Universidad de la Rioja, 9 October 2011. —. “Renewal and innovation in the emergence of Indo-European evaluative morphology.” In Diminutives and augmentatives in the languages of the world, edited by Pavol Štekauer and Lívia Körtvélyessy, Lexis 6 (2011): 5-26. —. “Sardinian evaluative morphology in typological perspective.” In Sardinian in typological perspective, edited by Ignazio Putzu, 188-209. Bochum: Dr. Brockmeyer University Press, 2005. Grandi, Nicola, and Lívia Körtvélyessy. Edinburgh handbook of evaluative morphology. Edinburgh: Edinburgh University Press, forthcoming. Grandi, Nicola, and Sergio Scalise. “Semantic Restrictions on Diminutive Formation: Evidence.” In Naturally! Linguistic studies in honour of Wolfgang Ulrich Dressler presented on the occasion of his 60th birthday, edited by Chris Schaner-Wolles, John Rennison and Friedrich Neubarth. Torino: Rosenberg &Sellier, 2000. Greenberg, Joseph H., and James J. Jenkins. “Studies in the psychological correlates of the sound system of American English.” Word, no. 22 (1966): 207-242. Gregová, Renáta. “Evaluative Morphology and (Mor)phonological Changes in Diminutives of Indo-European Languages.” BAS, British and American Studies, A Journal of The Romanian Society of English and American Studies, no. 27 (2011): 333–343. Gregová, Renáta, Lívia Körtvélyessy, and Július Zimmermann. “Phonetic iconicity in the evaluative morphology of a sample of Indo-European, Niger-Congo and Austronesian languages.” Word Structure (EUP) 3, no. 2 (2010): 156-180. Groenke, Ulrich. Die Diminutiva des Isländischen: Ein Beitrag zur isländischen Wortkunde. Göttingen: Göttingen, 1955. Gruzdeva, Ekaterina. Nivkh. London: LINCOM Europa, 1998. Guirardello, Raquel. A reference grammar of Trumai. Rice University Electronic Theses and Dissertations. 1999. http://scholarship.rice.edu/handle/1911/19387?show=full (accessed April 2013). Gustafson, Holly. A grammar of the Nipmuck language. Winnipeg: University of Manitoba, 2000.

238

Bibliography

Haas, Mary. “The expression of the diminutive.” In Studies in Linguistics in Honor of G.L.Trager, edited by Estellie M. Smith, 148-152. The Hague: Mouton, 1972. Hale, Horatio. The Tutelo Language. Bristol: Evolution Publishing, 2001. Hamano, Shoko. The sound-symbolic system of Japanese. Unpublished doctoral dissertation. Florida: University of Florida, 1986. “Happy Rabbit With Carrot Coloring Page.” Savetz Publishing. Free Printable Coloring Pages. 2006-2012. Hardman, Martha James. Jaqaru. München: LINCOM Europa, 2000. Harvey, Mark. A grammar of Gaagudju. Berlin/New York: Mouton de Gruyter, 2002. Haspelmath, Martin. “Comparative concepts and descriptive categories in crosslinguistic studies.” Language, no. 86 (2010): 663-687. —. “How young is Standard Average European?” Language Sciences, no. 3 (1998): 271-287. —. “Pre-established categories don’t exist – consequences for language description and typology.” Linguistic Typology 1, no. 11 (2007): 119132. —. “The European linguistic area: Standard Average European.” In Language Typology and Language Universals / Sprachtypologie und sprachliche Universalien / La typologie des langues et les universaux linguistiques, edited by Martin Haspelmath, Ekkehard König, Wulf Oesterreicher and Wolfgang Raible, vol. 1, 1492-1510. Berlin/New York: Walter de Gruyter, 2001. —. “Word-class changing inflection and morphological theory.” In Yearbook of Morphology, edited by Geert Booij and Jaap van Marle, 43-66. Kluwer Academic Publishers, 1995. Haspelmath, Martin, and Oda Buchholz. “Equative and similative constructions in the languages of Europe.” In Adverbial constructions in the languages of Europe, edited by Johan van der Auwera, 277-334. Berlin: Mouton de Gruyter, 1998. Haude, Katharina. A grammar of Movima. Netherlands: Manta, 2006. Hausenberg, Anu-Reet. “Komi.” In The Uralic Languages, edited by Daniel Mario Abondolo, 305-326. London: Routledge, 1998. Healey, Phyllis M. An Agta grammar. Manila: Bureau of Printing, 1960. Heat, Jeffrey. A grammar of Koyra Chiini. Berlin/New York: Mouton de Gruyter, 1999. —. A grammar of Tamashek (Tuareg of Mali). Berlin/New York: Mouton de Gruyter, 2005. Heine, Bernd, and Tania Kuteva. Language contact and grammatical change. Cambridge: Cambridge University Press, 2005.


239

Heine, Bernd, and Tania Kuteva. The changing languages of Europe. Oxford: Oxford University Press, 2006. Helimski, Eugene. “Ngasan.” In The Uralic languages, edited by Daniel Mario Abondolo, 480-515. London and New York: Routledge, 1998. —. “Selkup.” In The Uralic languages, edited by Daniel Mario Abondolo, 548-579. London and New York: Routledge, 1998. Hercus, Luise Anna. A grammar of the Arabana-Wangkangurru language, Lake Eyre basin, South Australia. Canberra: Australian National University, 1994. —. The Bāgandji language. Canberra: Australian National University, 1982. Hetzron, Robert. “The Agaw languages.” Afroasiatic Linguistics 3, no. 3 (1976): 1-40. —. The verbal system of southern Agaw. Berkeley and Los Angeles: University of California Press, 1969. Hewson, John, and Bernard Francis. The Micmac grammar of farther Pacifique. Winnipeg: Algonquian and Iroquoian Linguistics, 1990. Hicks, Mark A. Little house [clip art].

http://school.discoveryeducation.com/clipart/clip/littleho use.html (accessed 1 November 2014) Hildebrandt, Kristine A. “A grammar and dictionary of the Manange language.” In Tibeto-Burman Languages of Nepal: Manange and Sherpa, edited by Carol Genetti. Canberra: Pacific Linguistics, 2004. Hinton, Leanne, Johanna Nichols, and John J. Ohala. Sound symbolism. Cambridge: Cambridge University Press, 1994. Hualde, Jose Ignacio, and Jon Ortiz De Urbina. A grammar of Basque. Berlin/New York: Mouton de Gruyter, 2003. Hutchinson, John P. The Kanuri language: A reference grammar. Madison: University of Wisconsin, 1981. Hyman, Larry M., ed. Aghem grammatical structure: with special reference to noun classes, tense-aspect and focus marking. Los Angeles: Dept. of Linguistics, University of Southern California, 1979. Ingram, Andrew. Anamuxra. Sydney: University of Sydney, 2001. Iwasaki, Shoichi, and Preeya Ingkaphirom. A reference grammar of Thai. Cambridge: Cambridge University Press, 2005. Jakobi, Angelika. A Fur grammar: phonology, morphophonology, and morphology. Hamburg: Helmut Buske Verlag, 1990. Jespersen, Otto. Language: its nature, development and origin. London: Allen and Unwin, 1922. —. “Symbolic value of the vowel i.” In Linguistica: Selected papers in English, French and German, 283-303. Copenhagen: Levin and

240

Bibliography

Munksgaard, 1933. Johnson, Ronald C. “Magnitude symbolism of English words.” Journal of Verbal Learning and Verbal Behavior, no. 6 (1967): 508-511. Jurafsky, Daniel. “Universal tendencies in the semantics of the diminutive.” Language 3, no. 72 (1996): 533-78. —. “Universals in the semantics of the diminutive.” Parasession on semantic typology and semantic universals. Vol. 19 (1), edited by Joshua Guenter, Barbara Kaiser and Cheryl Zoll, 423 – 436. Berkeley, CA: Berkeley Linguistic Society, 1993. Kachru, Yamuna. Hindi. Amsterdam/Philadelphia: John Benjamins, 2006. Kangasmaa-Minn, Eeva. “Mari.” In The Uralic languages, edited by Daniel Mario Abondolo, 219-248. London: Routledge, 1998. Kay, Carolin J. A grammar of Misantla Totonac. Salt Lake City: University of Utah Press, 1999. Kennedy, Christopher. “Vagueness and grammar: the semantics of relative and absolute gradable adjectives.” Linguistics and Philosophy, no. 30 (2007): 1–45. Keresztes, László. “Mansi.” In The Uralic languages, 387-427. London and New York: Routledge, 1998. Khalilova, Zaira. “Khwarshi.” In HSK word-formation: an international handbook of the languages of Europe, edited by Susan Olsen, Franz Rainer, Müller Peter and Ingeborg Ohnheiser. Berlin/New York: Mouton de Gruyter, forthcoming. Kiefer, Ferenc, and Boglárka Németh. “Hungarian.” In Edinburgh handbook of evaluative morphology, edited by Nicola Grandi and Lívia Körtvélyessy. Edinburgh: Edinburgh University Press, forthcoming. Klamer, Margareta Anna Flora. Grammar of Kambera. Berlin: Mouton de Gruyter, 1998. Klank, Linda J.K, Yau-Huang Huang, and Ronald C. Johnson. “Determinants of success in matching word pairs in tests of phonetic symbolism.” Journal of Verbal Learning and Verbal Behavior, no. 10 (1971): 140-148. Kleparski, Grzegorz Andrzej. “CDs, petticoats, skirts, ankas, Tamaras and sheilas: The metonymical rise of lexical categories related to the conceptual category Female Human Being.” In Papers from the Glasgow Symposium on Linguistic Categories, edited by Christian J. Kay and J.J. Smith, 71-84. Amsterdam: John Benjamins, 2004. —. “Metonymy and the growth of lexical categories related to the conceptual category FEMALE HUMAN BEING.” Studia Anglica Resoviensia (University of Rzeszow), no. 1 (2000): 17-26. Kleparski, Grzegorz Andrzej. Semantic change and semantic components:


241

an inquiry into pejorative developments in English. Regensburg: Friedrich Pustet Verlag, 1986. —. “Semantic change in onomasiological perspective.” In Male and female terms in English, edited by Gunnar Persson and Mats Rydén, 41-91. Acta Universitatis Umensis (Umea), 1996. Kochman-Haladyj, Bozena. Historical derogation of women terms and its social-cultural causes and conditionings. Unpublished doctoral dissertation. Rzeszow: Rzeszow University, 2008. Kong-on, Kim. “Sound symbolism in Korean.” Journal of Linguistics 1, no. 13 (1977): 65-75. Koriat, Asher. “Phonetic symbolism and feeling of knowing.” Memory and Cognition 5, no. 3 (1975): 545-548. Kortmann, Bernd. “Adverbial Subordinators in the Languages of Europe.” In Adverbial Constructions in the Languages of Europe, edited by Johann van der Auwera, 457-561. Berlin/New York: Mouton de Gruyter, 1998. —. “The Evolution of Adverbial Subordinators in Europe.” In Historical linguistics 1997: selected papers from the 13th International Conference on Historical Linguistics, Düsseldorf, 10-17 August 1997, edited by Monika S. Schmid, Jennifer R. Austin and Dieter Stein, 213228. Amsterdam/Philadelphia: John Benjamins, 1998. Laakso, Johanna. “The history of word-formation in Uralic.” In HSK word-formation: an international handbook of the Languages of Europe, edited by Peter Müller, Ingeborg Ohnheiser, Susan Olsen and Franz Rainer. Berlin/New York: Mouton de Gruyter, forthcoming. Lakoff, George. “Hedges: A study in meaning criteria and the logic of fuzzy concepts.” Journal of Philosophical Logic, no. 2 (1973): 458508. Lamb, William. Scottish Gaelic. London: LINCOM Europa, 2001. Langdon, Margaret. A grammar of Diegueno. Berkeley/Los Angeles/ London: University of California Press, 1970. Lanstyák, István. “A magyar nyelv szlovákiai változatainak kutatásáról.” In Tanulmányok a magyar–szlovák kétnyelvűségről, edited by István Lanstyák and Szabolcs Simon, 150-190. Bratislava: Kalligram, 1998. Lapolla, Randy L. “An experimental investigation into phonetic symbolism as it relates to mandarin Chinese.” In Sound symbolism, edited by Leanne Hinton, Johanna Nichols and John J. Ohala, 130-148. Cambridge: Cambridge University Press, 1994. LeSourd, Philip S. “Diminutive verb forms in Passamaquoddy.” International Journal of American Linguistics 1, no. 61 (1995): 103134.

242

Bibliography

Lühr, Rosemarie, ed. Nominale Wortbildung des indogermanischen in Grundzügen. Die Wortbildungsmuster ausgewählter indogermansicher Einzelsprachen. Band 2. Hamburg: Verlag Dr. Kovač, 2008. Maltzman, Irving, Lloyd Morrisett Jr., and Lloyd O. Brooks. “An Investigation of Phonetic Symbolism.” Journal of Abnormal and Social Psychology, no. 53 (1956): 249-251. Maslova, Elena. A grammar of Kolym Yukaghir. Berlin/New York: Mouton de Gruyter, 2003. Matasović, Ranko. Gender in Indo-European. Heidelberg: Winter, 2005. Matisoff, James A. “The mother of all morphemes: augmentatives and diminutives in areal and universal perspective.” In Papers from the first annual meeting of the Southeast Asian Linguistic Society, Tempe, edited by Martha Susan Ratliff and Eric Schiller, 293-349. Arizona State University, 1991. Mattes, Veronica. “One form-opposite meanings? Full reduplication in Bikol.” uni-graz.at. 2006. http://reduplication.uni-graz.at/texte/paper _ICAL.pdf (accessed October 2011). Maurer, Daphne, Thanujeni Pathman, and Catherine J. Mondloch. “The shape of boubas: sound-shape correspondences in toddlers and adults.” Development Science, no. 9 (2006): 316-322. Meinhof, Carl. Die Sprachen der Hamiten. Hamburg: Friederichsen, 1912. Merlan, Francesca C. A grammar of Wardaman. Berlin/New York: Mouton de Gruyter, 1994. Merlini Barbaresi, Lavinia. “Iconicity in language and text.” Paper presented at the ESSE conference, Helsinki 2000.

http://omero.humnet.unipi.it/matdid/372/iconicity.pdf (acessed November 2012) Mohan, Ramesar Peggy. Trinidad Bhojpuri: a morphological study. Michigan: University of Michigan, 1978. “Monsters coloring pages.” Coloring pages for kids. n.d. Morton, Eugene S. “Sound symbolism and its role in non-human vertebrate communication.” In Sound symbolism, edited by Leanne Hinton, Johanna Nichols and John J. Ohala, 348-365. Cambridge: Cambridge University Press, 1994. Mutz, Katrin. “Evaluative Morphology in a Diachronic Perspective.” In Edinburgh handbook of evaluative morphology, edited by Nicola Grandi and Lívia Körtvélyessy. Edinburgh: EUP, forthcoming. Neeleman, Ad, Hans van de Koot, and Jenny Doetjes. “Degree expressions.” The Linguistic Review 1, no. 21 (2004): 1-66. Newman, Paul. The Hausa language: an encyclopedic reference grammar. Yale: Yale University Press, 2000.


243

Newman, Stanley S. “Further experiments in phonetic symbolism.” American Journal of Psychology, no. 45 (1933): 53-75. Nieuwenhuis, Paul. Diminutives. Unpublished doctoral dissertation. Edinburgh: University of Edinburgh, 1985. Nichols, Johanna. Ingush grammar. Berkeley/Los Angeles/London: University of California Press, 2011. —. Linguistic diversity in space and time. Chicago: University of Chicago Press, 1992. —. “Universals and diachrony: Some observations.” In Linguistic universals and language change, edited by Jeff Good, 287-294. New York: Oxford University Press, 2008. —. “What, if anything, is typology?” Linguistic Typology 1, no. 11 (2007): 231-238. Nikolaeva, Irina, and Maria Tolskaya. A grammar of Udihe. Berlin/New York: Mouton de Gruyter, 2001. Nuckolls, Janis B. “The case of sound symbolism.” Annual Review of Anthropology, no. 28 (1999): 225-252. Ohala, John J. “Sound symbolism.” Proceedings 4th Seoul International Conference on Linguistics [SICOL] 11-15 Aug 1997, 1997: 98-103. —. “The frequency code underlies the sound-symbolic use of voice pitch.” In Sound symbolism, edited by Leanne Hinton, Jonanna Nichols and John J. Ohala, 325-348. Cambridge: Cambridge University Press, 1994. —. “The origin of sound patterns in vocal tract constraints.” In The production of speech, edited by Peter F. MacNeilage, 189-216. New York: Springer, 1983. Oliverio, Giulia R.M. A grammar and dictionary of Tutelo. Unpublished doctoral dissertation. Kansas: University of Kansas, 1996. Ormsbee Charney, Jean. A grammar of Comanche. Lincoln and London: University of Nebraska Press, 1993. Osgood, Charles E. “The cross-cultural generality of visual-verbal synesthetic tendencies.” Behavioral Science, no. 5 (1960): 146-169. Oszmianska, Aleksandra. “Sound symbolism as a universal drive to associate sound with meaning: a comparison between English and Japanese.” Poznan Studies in Contemporary Linguistics, no. 37 (2001): 147-155. Palkovič, Konštantín. Slovenské nárečia. Banská Bystrica: Krajské osvetové stredisko, 1981. Payne, Thomas E. Describing morphosyntax: a guide for field linguists. Cambridge: Cambridge University Press, 1997. Peirce, Charles Sanders. “What is a sign?”

244

Bibliography

http://www.iupui.edu/~peirce/ep/ep2/ep2book/ch02/ch0 2.htm (accessed 1 November 2014) Pentland, David H. “Diminutive consonant symbolism in Algonquian.” In Papers of the Sixth Algonquian Conference 1974, edited by William Cowan, 237-252. Ottawa: National Museums of Canada, 1975. Pillinger, Owen Stephen. Accent, tone and prosodic structure in Rendille. Unpublished docotoral dissertation. London: School of Oriental and African Studies, University of London, 1989. Pitkin, Harvey. Wintu grammar. Berkeley: University of California Press, 1985. Plank, Frans. “Extent and limits of linguistic diversity as the remit of typology – but through constraints on what is diversity limited?” Linguistic Typology 1, no. 11 (2007): 43-68. —. “Inflection and derivation.” In The Encyclopedia of Language and Linguistics, edited by R. E. Asher, vol. 3, 1671-1678. Oxford, Pergamon: Pergamon, 1994. Plank, Frans, Thomas Mayer, and Elena Filimonova. The Universals Archive. 1998. http://typo.uni-konstanz.de/archive/intro/index.php (accessed December 2013). Plato. Plato in Twelve Volumes. Volume 12. Translated by North Harold Fowler. 1952. Poe, Edgar Allan. “The philosophy of composition.” Graham’s Magazine, April 4, 1846: 163-167. Pöll, Bernhard. “Portuguese.” In Word-formation: an international handbook of the languages of Europe, edited by Peter Müller, Ingeborg Ohnheiser, Susan Olsen and Franz Rainer. Berlin/ New York.: De Gruyter Mouton, forthcoming. Price, Glanville, ed. Encyclopedia of the languages of Europe. Oxford: Blackwell, 2005. Prieto, Victor. Spanish Evaluative Morphology: Pragmatic, sociolinguistic, and semantic issues. Unpublished doctoral dissertation. Gainesville, Florida: University of Florida, 2005. Prieto, Victor. “The Semantics of Evaluative Morphology.” In Edinburgh handbook of evaluative morphology, edited by Nicola Grandi and Lívia Körtvélyessy. Edinburgh: Edinburgh University Press, forthcoming. Rácová, Anna, and Ján Horecký. Syntax slovenskej karpatskej rómčiny. Bratislava: Iris, 2007. Ramachandran, Vilayanur S., and Edward M. Hubbard. “Synesthesia – a window into perception, thought, and language.” Journal of Consciousness Studies, no. 8 (2001): 3-34. Ramat, Paolo. “How universal are linguistic categories?” In Universals of


245

language today, edited by Sergio Scalise, Antonietta Bissetto and Elisabetta Magni, 1-12. Dordrecht: Springer, 2009. —. “The notion of Standard Average European. Le Lingue d’Europa.” Lecture. Bologna, 2000. —. “Typological comparison: Towards a historical perspective.” In Approaches to language typology, edited by Masayoshi Shibatani and Theodora Bynon, 27-48. Oxford: Clarendon, 1995. Reesink, Ger P. A grammar of Hatam. Canberra: Australian National University, 1999. Refsing, Kirsten. The Ainu language: the morphology and syntax of the Shizunai dialect. Aarhus: Aarhus University Press, 1986. Reh, Mechthild. Die Krongo-Sprache. Berlin: Dietrich Reimer Verlag, 1985. Rett, Jessica. “Antonymy and evalutivity.” In Proceedings from semantics and linguistic theory XVII, edited by Tova Friedman and Masayuki Gibson, 210-227. Ithaca: Cornell University, 2008. Riese, Timothy. Vogul. London: LINCOM Europa, 2001. —. “Permian.” In The Uralic languages, edited by Daniel Mario Abondolo, 249-275. London: Routledge, 1998. Richard, Ely, Jean Berko Gleason, and Allyssa McCabe. “Why didn’t you talk to your mommy, honey? Parents’ and children’s talk about talk.” Research on Language and Social Interaction 1, no. 29 (n.d.): 7-25. Robins, R.H. The Yurok language: grammar, texts, lexicon. Berkeley: University of California Press, 1958. Romero-Figueroa, Andrés. A reference grammar of Warao. München: LINCOM Europa, 1997. Rood, David S. Wichita grammar. New York: Garland Pub, 1976. Roper, Carolann W., Paul W. Dixon, Elsie H. Ahern, and Verner L. Gibson. “Effect of language and sex on universal phonetic symbolism.” Language and Speech 4, no. 19 (1976): 388-396. Rosch, Eleanor, Carolyn B. Mervis, Wayne D. Gray, David M. Johnson, and Penny Boyes-Braem. “Basic objects in natural categories.” Cognitive Psychology, no. 8 (1976): 382-439. Rowling, Joanne K. Harry Potter and the philosopher’s stone. London: Bloomsbury Publishing, 2001. Rytkonnen, Ahti. Kielentukimuksen tyomaalta III. Eraiden itamerensuomen th-sanojen historiaa. Jyvaskyla, Finland: Privately printed, 1940. Sakel, Jeanette. A grammar of Mosetén. München: Mouton de Gruyter, 2004. Salminen, Tapani. “Nenets.” In The Uralic languages, edited by Daniel

246

Bibliography

Mario Abondolo, 516-547. London and New York: Routledge, 1998. Sammallahti, Pekka. “Saamic.” In The Uralic languages, edited by Daniel Mario Abondolo, 43-95. London and New York: Routledge, 1998. Sapir, David. A grammar of Diola-Fogny : a language spoken in the Basse-Casamance region of Senegal. Cambridge: Cambridge University Press, 1965. Sapir, Edward. “A study in experimental symbolism.” Journal of Experimental Psychology, no. 12 (1929): 225-239. —. Language: an introduction to the study of speech. Whitefish, Montana: Kessinger Publishing, 2005. (First published New York: Harcourt, Brace, 1921.) Saussure, Ferdinand de. Course in general linguistics. Edited by Albert Sechehaye and Charles Bally. Translated by Wade Baskin. New York: The Philosophical Library, 1959. Savickiené, Ineta, and Wolfgang U. Dressler. The acquisition of diminutives. Amsterdam/Philadelphia: John Benjamins, 2007. Scalise, Sergio. Generative morphology. Dordecht: Foris Publications, 1984. Scalise, Sergio, Elisabetta Magni, and Antonietta Bissetto. Universals of language today. Dordecht: Springer, 2009. Seiler, Walter. Imonda, a Papuan language. Canberra: Australian National University, 1985. Seuren, Peter A.M. “The structure and selection of positive and negative gradable adjectives.” In Papers from the Parasession on the Lexicon, CLS 14, edited by Donka Farkas, Wesley M. Jacobsen and Karol W. Todrys, 336-346. Chicago: University of Chicago, 1978. Shakespeare, William. Antony and Cleopatra. http://shakespeare.mit.edu/ (accessed 8 April 2012). —. Romeo and Juliet. Oxford: Oxford University Press, 2008. Sharpe, Margaret C. Alawa phonology and grammar. Canberra: Australian Institute of Aboriginal Studies, 1972. Shetter, William. “The Dutch diminutive.” Journal of English and Germanic 1, no. 58 (1959): 75-90. Shinohara, Kazuko, and Shigeto Kawahara. “A cross-linguistic study of sound symbolism: the images of size.” 36th Annual Meeting of Berkeley Linguistics Society, forthcoming. Shrum, L.J., and Tina M. Lowrey. “Sounds convey meaning: The implications of phonetic symbolism for brand name.” In Psycholinguistic Phenomena in Marketing Communications, edited by Tina M. Lowrey, 39-58. Mahwah, NJ: Erlbaum, 2007. Schneider, Klaus. Diminutives in English. Tübingen: Niemeyer, 2003.


247

Siller-Runggaldier, Heidi. “Ladin.” In HSK word-formation: aAn international handbook of the languages of Europe, edited by Peter Müller, Ingeborg Ohnheiser, Susan Olsen and Franz Rainer. Berlin and New York: Mouton de Gruyter, forthcoming. Silverstein, Michael. “Relative motivation in denotational and indexical sound symbolism of Wasco-Wishram Chinookan.” In Sound symbolism, edited by Leanne Hinton, Johanna Nichols and John J. Ohala, 40-62. Cambridge: Cambridge University Press, 1994. Simoncsics, Peter. “Kamassian.” In The Uralic languages, edited by Daniel Mario Abondolo, 580-601. London and New York: Routledge, 1998. Simone, Raffaele. “The body of the language – the paradigm of arbitrariness and the paradigm of substance.” Edited by Rudolf Engler and René Amacker. Présence de Saussure: actes du Colloque international de Genève, 21-23 mars 1988, 121-141. Paris: Droz, 1990. Slater, Keith W. A grammar of Mangghuer : a Mongolic language of China’s Qinghai-Gansu Sprachbund. London and New York: Routledge Curzon, 2003. “small-house-in-the-village-coloring-page.” Flickr. n.d. Soane, Ely Banister. Grammar of the Kurmanji or Kurdish. London: Luzac, 1913. Sokolová, Miloslava, Martin Ološtiak, and Martina Ivanová. Slovník koreňových morfém slovenčiny. Prešov: Filozofická fakulta Prešovskej univerzity, 2005. Souman, Agnita, and Steven Gillis. “A longitudinal study of the acquisition of diminutives in Dutch.” In Acquisition of Diminutives: A cross-linguistic perspective, edited by Ineta Savickienė and Wolfgang U. Dressler, 183–206. Amsterdam: John Benjamins, 2007. “Sound symbolism in words for proximity and distance.” Language and Speech, no. 17 (1971): 89-94. Stankiewicz, Edward. Declension and gradation of Russian substantives. The Hague: Mouton, 1968. Stenzel, Kristine Sue. A reference grammar of Wanano. Unpublished doctoral dissertation. Ann Arbor: UMI, 2004. Stump, Gregory T. “A paradigm-based theory of morphosemantic mismatches.” Language, no. 67 (1991): 675-725. —. “How peculiar is evaluative morphology?” Journal of Linguistics, no. 29 (1993): 1-36. —. “Word-formation and inflectional morphology.” In Handbook of wordformation, edited by Pavol Štekauer and Rochelle Lieber, 49-72. Dordrecht: Springer, 2005.

248

Bibliography

“Sunshine Coloring Pages.” Lucky Travels. n.d. Szymanek, Bogdan. Categories and categorization in morphology. Lublin: Redakcja Wydawnictwa Katolickiego Uniwersytetu Lubelskiego, 1988. Štekauer, Pavol. “The delimitation of derivation and inflection.” In HSK word-formation: an international handbook of the languages of Europe, edited by Peter Müller, Susan Olsen, Ingeborg Ohnheiser and Franz Rainer. Berlin/New York: Mouton de Gruyter, forthcoming. Štekauer, Pavol, Salvador Valera, and Lívia Körtvélyessy. Word-formation in the world’s languages. Cambridge: Cambridge University Press, 2012. Tanz, Christine. “Sound symbolism in words for proximity and distance.” Language and Speech, no. 17 (1971): 89-94. Tarte, Robert D. “Phonetic symbolism in adult native speakers of Czech.” Language and Speech, no. 174 (1974): 87-94. Terril, Angela. A grammar of Lavukaleve. Berlin/New York: Mouton de Gruyter, 2003. Towsend, William Cameron. “Cakchiquel grammar.” In Mayan studies I, edited by Benjamin Elson, 3-82. Norman: Summer Institute of Linguistics of the University of Oklahoma, 1960. Tozzer, Alfred Marston. A Maya grammar. With bibliography and appraisement of the works noted. Cambridge, Massachusetts: Peabody Museum, 1921. Tsunoda, Tasaku. The Djaru language of Kimberley, Western Australia. Canberra: Australian National University, 1981. Tsuru, Shigeto, and Horace S. Fries. “A problem in meaning.” Journal of General Psychology, no. 8 (1933): 281-284. Ultan, Russel. “Size-sound symbolism.” In Universals of human language, edited by Joseph H. Greenberg, vol. 2, 525-568. Stanford: Stanford University Press, 1978. van der Auwera, and Dónall Ó Baoill. Adverbial constructions in the languages of Europe: empirical approaches to language typology EUROTYP. Berlin/New York: Mouton de Gruyter, 1998. van der Auwera, and Jan Nyuts. “Cognitive linguistics and linguistic typology.” In The Oxford handbook of cognitive linguistics, edited by Dirk Geeraerts and Hubert Cuyckens, 1074-1091. Oxford: Oxford University Press, 2007. van der Auwera, Johan. “Introduction.” In Adverbial constructions in the languages of Europe: empirical approaches to language typology EUROTYP, edited by Johan van der Auwera and Dónall Ó Baoill, 124. Berlin/New York: Mouton de Gruyter, 1998a.


249

—. “Conclusion.” In Adverbial constructions in the languages of Europe: empirical approaches to language typology - EUROTYP, edited by Dónall Ó Baoill and Johan van der Auwera, 813-834. Berlin/New York: Mouton de Gruyter, 1998b. —. “Standard Average European.” In The languages and linguistics of Europe: a comprehensive guide, edited by Bernd Kortmann and Johan van der Auwera, 291-306. Berlin: Mouton de Gruyter, 2011. van der Berg, René. A grammar of the Muna language. Dordecht: Foris, 1989. van der Voort, Hein. A grammar of Kwaza. Berlin/New York: Mouton de Gruyter, 2004. van Driem, George. A grammar of Dumi. Berlin: Mouton de Gruyter, 1993. van Gijn, Rik. A grammar of Yurakaré. Unpublished doctoral dissertation. Nijmegen: Radboud University, 2006. Velten, H. V. “The Nez Perce verb.” Pacific Northwest Quarterly, no. 34 (1943): 271-292. Volker, Craig Alan. The Nalik language of New Ireland, Papua New Guinea. New York: Peter Lang, 1998. Walker, Dick. Central Carrier grammar sketch. Ft. St. James, British Columbia: Summer Institute of Linguistics, 1973. Watsen, Janet. “Arabic morphology: diminutive verbs and diminutive nouns in San’ani Arabic.” Morphology, no. 16 (2006): 189–204. Watters, David E. A grammar of Kham. Cambridge: Cambridge University Press, 2002. Wertheimer, Michael. “The relation between the sound of a word and its meaning.” American Journal of Psychology, no. 71 (1958): 412-415. Westcott, Roger W. “Allolinguistics: Exploring the peripheries of speech.” In The Second LACUS Forum, edited by Peter A. Reich, 497-513. Columbia, SC: Hornbeam Press, 1975. —. “Linguistic iconism.” Language, no. 47 (1971): 416-428. Whorf, Benjamin Lee. “The relation of habitual thought and behavior to language.” In Language, thought and reality: selected writings of Benjamin Lee Whorf, edited by John B. Caroll, 134-159. London/New York: The Technology Press of Massachusetts Institute of Technology and John Wiley & Sons, 1956. Wierzbicka, Anna. “Diminutives and Depreciatives: Semantic Representation for Derivational Categories.” Quaderni di semantica : rivista internazionale di semantica teorica e applicata : an international journal of theoretical and applied semantics, no. 5 (1984): 123-130. “Wind Coloring Page.” Freshcoloring. n.d.

250

Bibliography

Wlodarczyk-Stachurska, Anna. On sociolinguistic variation: the case of English women terms. Unpublished doctoral dissertation. Rzeszow: Rzeszow University, 2010. “Wolf coloring page.” Supercoloring. n.d. Woodworth, Nancy L. “Sound symbolism in proximal and distal forms.” Linguistics 2, no. 29 (1991): 273-300. Woollams, Geoff. A grammar of Karo Batak, Sumatra. Canberra: Australian National University, n.d. Yallop, Colin. Alyawarra. Canberra: Australian Institute of Aboriginal Studies, 1977. Zaicz, Gábor. “Mordva.” In The Uralic languages, edited by Daniel Mario Abondolo, 184-218. London and New York: Routledge, 1998. Zaĭkov, Peter Mefodievich. Karjalan kielioppi 5-9. Petroskoi: Periodika, 2002. Zamponi, Raul. Maipure. München: LINCOM Europa, 2003. Zwicky, Arnold M., and Geoffrey K. Pullum. “Plain morphology and expressive morphology.” In Proceedings of the thirteenth annual meeting, 14-16 February, 1987: general session and parasession on grammar and cognition, edited by John Aske, 330-340. Berkeley: Berkeley Linguistics Society, 1987.

INDEX

action, 4, 41, 42, 46, 47, 51 affectionate, 22, 29, 40 ameliorative, 22, 44, 166 Amha, A, 60, 61, 62, 134 Anderson, E.R, 145, 146, 148 Anderson, S, 29 Aoki. H, 151 Apresjan, Y.D, 41 areal typology, 1, 2, 18, 52, 140 Aronoff, M, 59, 62, 92 Asher, R.E, 22, 90 attenuation, 41, 44, 46, 47, 112 Atzet, J, 151, 152 augmentative, 1, 2, 3, 4, 20, 22, 25, 32, 33, 34, 35, 36, 37, 40, 44, 46, 49, 51, 54, 55, 56, 59, 61, 62, 63, 64, 65, 66, 67, 73, 74, 75, 79, 80, 81, 85, 87, 88, 92, 93, 94, 100, 101, 102, 103, 108, 109, 110, 111, 113, 114, 115, 116, 117, 118, 119, 123, 124, 125, 126,128, 129, 131, 133, 134, 135, 136, 137, 139, 157, 158, 159, 161, 162, 163, 165, 166, 167, 168, 169, 171, 172, 174, 178, 179, 182, 183, 185, 186, 188, 189, 190, 192, 193, 194, 198, 201, 204, 206, 208, 210, 214, 215, 216, 217, 218, 219, 221 augmentativeness, 1, 3, 30, 41, 46, 132, 134 augmentativisation, 22, 59 augmentativisation, 22 augmentivisation, 126, 133 Bakema, P, 35 Bakker, D, 48 Baoill, D, 2 Bauer, L, 1, 3, 22, 28, 29, 30, 48, 50, 115, 154 Beard, R, 20, 24, 29, 30, 31

Becker, J.B, 151 Bentley, M, 147, 151 Berko Gleason, J, 153, 157, 175, 211, 212, 214, 217 Berlin, B, 150, 151 Bernard, F, 74 Bierwisch, M, 1, 20, 32 Black, A.E, 152 Böhmerová, A, 1 Bolozky,S, 1 Booij, G, 25 Brackbill, Y, 151, 153 Brown, L, 1, 25 Brown, R.W, 152 Buchholz, O, 8 Camaj, M, 119 Campbell, L, 70, 72 Cardona, G, 92 Carlson, R, 59, 134 Carstairs-McCarthy, A, 1, 30 change of paradigm, 58, 59, 61, 62, 100, 103, 133 Chelliah, S.L, 82, 85 chromaesthesia, 149 CIRCUMSTANCE, 4, 41, 42, 46, 47, 51 cognitive categories, 4, 41, 42, 44, 46, 49, 51, 55, 62, 66, 67, 71, 78, 80, 84, 87, 91, 93, 101, 102, 108, 109, 111, 112, 113, 114, 117, 118, 124, 125, 126, 136, 137 compounding, 22, 28, 49, 50, 58, 69, 73, 77, 78, 79, 86, 87, 90, 111, 114, 115, 117, 118, 124, 125, 130, 132 constructional diagrammacity, 147 conversion, 49 Coupe, A.R, 86, 87 Cowan, H, 66 Creissels, D, 25 Dayley, J.P, 43, 69, 73

252 de Araujo, G, 77 de Saussure, F, 4 de Urbina, J.O, 43, 109 Décsy, Gy, 2, 8, 13, 14, 15, 18, 140 Déchaine, R-M, 25 depreciatives, 39 Derzhanski, I, 1 Diffloth, G, 150, 151 diminutive, 1, 2, 3, 4, 20, 22, 25, 29, 32, 33, 34, 35, 36, 37, 38, 40, 41, 44, 46, 49, 51, 54, 55, 56, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 69, 70, 71, 73, 75, 77, 79, 80, 82, 85, 86, 87, 89, 90, 91, 92, 93, 94, 100, 101, 103, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 123, 124, 125, 126, 128, 129, 131, 132, 133, 134, 135, 136, 137, 142, 157, 158, 161, 165, 166, 168, 169, 171, 172, 174, 176, 179, 182, 185, 187, 188, 190, 196, 197, 198, 200, 201, 202, 204, 206, 208, 210, 211, 212, 217, 218, 219, 221 diminutiveness, 1, 3, 21, 22, 30, 33, 34, 35, 41, 46, 71, 132, 134, 135 diminutivisation, 22, 41, 59, 66, 70, 110, 111, 112, 115, 117, 118, 132, 133, 134 distributiveness, 41, 47 Dokulil, M, 32, 51 Donaldson, B, 124 Dressler, W.U, 1, 5, 20, 22, 24, 30, 32, 35, 41 Dryer, M, 1, 25 Dum-Tragut, 22, 119 duration, 46 Erdal, M, 221 Erschler, D, 117 Europeme, 18 EUROTYP, 2, 6, 7 Euroversal, 4, 18, 19, 128, 130, 131, 136, 137 evaluative affix, 1, 27, 41, 55

Index evaluative construction, 21, 23, 38, 39, 40, 45 evaluative formation, 23, 24, 32, 36, 37, 40, 44, 50, 51, 55, 90, 111, 112, 113, 125, 153 evaluative marker, 21, 22, 39, 217, 218 evaluative morphology, 1, 2, 3, 4, 5, 6, 18, 20, 21, 22, 23, 24, 25, 26, 28, 29, 30, 31, 32, 37, 39, 41, 42, 44, 45, 46, 48, 49, 50, 52, 54, 55, 56, 61, 63, 64, 65, 66, 67, 71, 75, 80, 88, 89, 93, 94, 107, 109, 110, 117, 119, 123, 124, 126, 128, 129, 130, 132, 135, 139, 141, 142, 147, 153, 154, 155, 159, 176, 191, 210, 212, 216, 217, 220, 221, 232, 237, 247 Evaluative Morphology Saturation, 4, 6, 55, 144 evaluative suffix, 24, 26, 111, 156 Everett, D.L, 79 Fisher, S.K, 151 Fisher-Jørgsen, E, 147, 152 Fischer, O, 144 Fitch, W.T.S, 150, 151 Fónagy, I, 151 Fónagy, J, 151 Ford, L.J, 65 Fortin, A, 20, 24, 25, 35 French, P.L, 150, 151 frequentativeness, 46 Fries, H.S, 148 front high vowels, 3 Fudeman, K, 59, 62 Gebels, G, 147, 151 Geeraerts, D, 35 Gerard, H.B, 151, 152 Gillis, S, 124 Girard, R, 25 González, H.A, 42, 77 Gràcia, L, 20 Grandi, N, 1, 5, 21, 24, 38, 39, 40, 44, 112, 113, 133 Greenberg, J.H, 152 Gregová, R, 1, 3

Evaluative Morphology from a Cross-Linguistic Perspective Gruzdeva, E, 91 Gustafson, H, 70, 71 Haas, M, 1, 21 Hale, H, 22, 74 Hamano, S, 149 Hardman, M.J, 77 Haspelmath, M, 2, 7, 8, 13, 14, 15, 16, 17, 18, 31, 109, 140, 141, 142 Haude, K, 75, 76 Healey, P.M, 44, 82, 85 Heat, J, 58, 59, 60 Heine, B, 2, 7, 8, 15, 16, 18 Hercus, L.A, 43, 50, 65, 67 Hewson, J, 74 high back vowels, 3 honorific, 29 Horecký, J, 11 Horowitz, A.E, 152 Hualde, J.I, 43 Hubbard, E.M, 146, 150, 151 Hutchinson, J.P, 60 Hyman, L.M, 58 hypocorism, 47 hypocoristics, 39 iconicity, 2, 3, 4, 18, 103, 144, 145, 146, 147, 148, 149, 150, 151, 153, 155, 156, 157, 159, 161, 169, 172, 174, 185, 191, 194, 196, 198, 210, 211, 216, 217, 219, 221, 222 incorporation, 49 infixation, 49, 83, 100, 113, 114 Ingkaphirom, P, 85 Ingram, A, 44, 50 intensification, 34, 39, 41, 44, 47 intensity, 46, 71 isomorphism, 147 iterativity, 41, 46, 47 Iwasaki, S, 85 Jan, K, 66 Jenkins, J.J, 152 Jespersen, O, 150 Jespersen, O, 3, 146, 147, 151, 152 Johnson, R.C, 151 Jurafsky, D, 1, 5, 33, 34, 35, 37, 38, 40, 46, 59, 91, 132

253

Kawahara, S, 150, 151 Kennedy, Ch, 32 Khalilova, Z, 47 Kiefer, F, 133 kinaesthesia, 149 Klank, L, 151 Kleparski, G.A.K, 36, 51 Kochman-Haladyj, B, 36 Kong-on, K, 151 Koriat, A, 150, 151 Kortmann, B, 8 Kumari, T.C, 22, 90 Kuteva, T, 2, 7, 8, 15, 16, 18 Laakso, J, 111, 142 Lakoff, G, 30, 33 Lamb, W, 117 Langdon, M, 71, 73 language sampling, 2, 48 Lanstyák, I, 142 Lapolla, R.L, 147, 150, 151, 152 LeSourd, P.S, 43, 70 Little, B.K, 151, 153 Lowrey, T, 151 Lühr, R, 133 Maltzman, I, 153 Matasović, R, 221 Matisoff, J, 1, 34, 35, 36, 87, 133 Maurer, D, 146, 150, 151 Meinhof, C, 59 Merlini Barbaresi, L, 5, 20, 24, 32, 35, 145, 146 mimesis, 147 Mohan, R.P, 91 Morton, E.S, 148 Mudzingwa, C, 25 Mutz, K, 5, 36, 38, 41, 44, 133, 221 Németh, B, 133 Newman, P, 62 Newman, S, 147 Nieuwenhuis, P, 3 Nichols, J, 221 Nikolaeva, I, 89 Nuckolls, J.B, 149 Nuttall, R, 152 Ohala, J, 147, 148, 150, 151 Oliverio, G.R.M, 51, 69

254 onomatopoeia, 145, 147, 149 Osgood, C.E, 148, 151 Oszmianska, A, 146, 149, 151 palatalized consonants, 3 Palkovič, K, 7 Payne, T.E, 3 Peirce, Ch.S, 4, 144, 145 pejorative, 22, 26, 29, 39, 40, 44, 59, 62, 71, 135 Pentland, D.H, 70 phonaesthesia, 149 phonetic iconicity, 2, 3, 4, 18, 103, 144, 145, 147, 149, 151, 153, 155, 157, 159, 161, 185, 191, 194, 198, 210, 211, 216, 217, 219, 221, 222 phonetic symbolism, 3, 147, 151 phonosymbolism, 145, 147 phonotactics, 146 Plank, F, 3, 29, 30 Poe, A.E, 145 Pöll, B, 39, 112 prefixation, 22, 49, 50, 58, 73, 74, 79, 83, 85, 90, 92, 100, 101, 112, 113, 114, 115, 116, 117, 118, 124, 125, 126, 130, 131 Price, G, 8, 9 Prieto, V, 5, 37, 38, 41, 46, 221 principle of convenience, 48 Pullum, G.K, 20 qualification, 33, 44, 47 QUALITY, 4, 41, 42, 46, 47, 51 quantification, 44 quantity of action, 40, 42, 47 Quantity of Action, 42, 44, 60, 62, 66, 67, 71, 72, 78, 80, 84, 86, 91, 93, 109, 112, 113, 114, 115, 117, 125, 126, 135 Quantity of Circumstance, 42, 44, 60, 62, 66, 67, 71, 73, 78, 79, 84, 86, 91, 93, 101, 109, 111, 112, 114, 115, 118, 125, 135, 136 Quantity of Quality, 55, 60, 62, 66, 67, 71, 74, 78, 80, 84, 86, 87, 91, 93, 101, 104, 108, 109, 110, 111,

Index 112, 113, 114, 115, 117, 118, 125, 126, 135, 136, 137 Quantity of Substance, 55, 60, 61, 62, 63, 65, 66, 67, 71, 72, 74, 78, 79, 80, 84, 86, 87, 91, 92, 93, 101, 103, 104, 108, 109, 110, 111, 112, 113, 114, 115, 117, 118, 119, 125, 126, 135, 137 Rácová, A, 11 radial model, 4, 37, 46 Ramachandran, V.S, 146, 151 Ramat, P, 2, 13, 14, 15, 17, 140, 141, 142 reduplication, 21, 22, 49, 50, 58, 65, 66, 67, 69, 70, 72, 73, 74, 77, 78, 79, 82, 83, 85, 87, 100, 101, 104, 109, 110, 111, 112, 113, 117, 118, 125, 130, 146 Refsing, K, 90, 92 Reh, M, 58 Rett, J, 1, 32 Richard, E, 1 root-and-pattern, 49, 108, 124 Rosch, E, 30 Rowling, J.K, 145 SAE, 2, 3, 4, 6, 7, 8, 12, 14, 15, 16, 17, 18, 19, 20, 48, 50, 52, 54, 55, 88, 104, 107, 108, 109, 113, 119, 126, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 139, 140, 141, 142, 143, 151, 220, 221 Sakel, J, 78, 79 Sammallahti, P, 42 Sapir, E, 3, 4, 146, 147, 148, 150, 151, 153, 158, 219 Scalise, S, 1, 20, 24, 26, 27, 28, 29 Seiler, W, 42 Seuren, P.A.M, 1 Shakespeare, W, 145 Sharpe, M.C, 50, 65 Shetter, W, 123, 124 Shinohara, K, 147, 150, 151 Shrum, L.J, 146, 151 Schneider, K, 1, 5 Siller-Runggaldier, H, 51 Simone, R, 149

Evaluative Morphology from a Cross-Linguistic Perspective Souman, A, 124 sound symbolism, 3, 4, 21, 145, 146, 147, 148, 149, 150, 160 Sprachbund, 8, 14, 16, 19, 141 Sridhar, S.N, 92 Standard Average European, 2, 3, 5, 6, 7, 14, 48, 52, 126, 154, 220 Stankiewicz, E, 1 stem alternation, 74, 100, 101, 103 Stenzel, K.S, 42 Stump, G.T, 1, 22, 27, 28, 30, 116 SUBSTANCE, 41, 42, 46, 51 suffixation, 22, 49, 58, 61, 63, 65, 69, 70, 72, 73, 74, 77, 78, 79, 82, 103, 108, 109, 110, 111, 112, 113, 114, 116, 117, 118, 119, 124, 125, 126, 130, 131, 132, 137, 168, 201, 202, 215 symbolism, 3, 21, 70, 71, 145, 148, 149, 150, 151, 160 synaesthesia, 150 synaesthesia, 148, 149 Szymanek, B, 20, 30, 39 Štekauer, P, 31, 41, 157 Tanz, Ch, 151

255

Tarte, R.D, 151 Tolskaya, M, 90 Tsunoda, T, 43 Tsuru, S, 148 Turon, L, 20 Ultan, R, 3, 147, 151 universal, 3, 18, 19, 36, 39, 48, 70, 125, 128, 132, 135, 147, 215, 221 van der Auwera, J, 2, 7, 8, 9, 10, 13, 14, 15, 17, 18, 19, 140, 142 van der Berg, R, 83, 85 van Driem, G, 84 van Gijn, R, 50, 79 Varon, J.E, 147, 151 Voort, H, 36, 79 Watters, J, 86 Wertheimer, M, 151 Westcott, R.W, 151 Whorf, B.L, 2, 6, 8, 16 Wierzbicka, A, 1, 39 Wiltschko, M, 25 Wlodarczyk-Stachurska, A, 36 Woodworth, N, 151 Zwicky, A.M, 20

Evaluative Morphology from a Cross-Linguistic Perspective

Sentence Processing: A Crosslinguistic Perspective

Processing Syntax and Morphology: A Neurocognitive Perspective

The Language of Memory in a Crosslinguistic Perspective

Islam from a Contemporary Perspective

Philosophy from a Skeptical Perspective

Philosophy from a Skeptical Perspective

Philosophy from a Skeptical Perspective

Computability and complexity from a programming perspective

Attribution theory from a neuropsychological perspective

Consciousness: Essays from a Higher-Order Perspective

From cause to causation: a Peircean perspective

Violence: Reflections from a Christian perspective

Computability and Complexity: From a Programming Perspective

Matrix Algebra From a Statistician's Perspective

Matrix Algebra From A Statiscian's Perspective

Agent Technology from a Formal Perspective

Harville Matrix algebra from a statiscian's perspective

Consciousness: Essays from a Higher-Order Perspective

Computability And Complexity From A Programming Perspective

Statistical Learning from a regression perspective

Statistical learning from a regression perspective

Computability and complexity from a programming perspective

Computational Economics: A Perspective from Computational Intelligence

Violence: Reflections from a Christian Perspective

Malaria: A Hematological Perspective: A Hematological Perspective

Amphibian Metamorphosis: From Morphology to Molecular Biology

Solids far from equilibrium: growth, morphology, defects

Polyurethane Elastomers: From Morphology to Mechanical Aspects

ELF from the programmer's perspective

Evaluative Morphology from a Cross-Linguistic Perspective

Sentence Processing: A Crosslinguistic Perspective

Processing Syntax and Morphology: A Neurocognitive Perspective

The Language of Memory in a Crosslinguistic Perspective

Islam from a Contemporary Perspective

Philosophy from a Skeptical Perspective

Philosophy from a Skeptical Perspective

Philosophy from a Skeptical Perspective

Computability and complexity from a programming perspective

Attribution theory from a neuropsychological perspective

Consciousness: Essays from a Higher-Order Perspective

From cause to causation: a Peircean perspective

Violence: Reflections from a Christian perspective

Computability and Complexity: From a Programming Perspective

Matrix Algebra From a Statistician's Perspective

Matrix Algebra From A Statiscian's Perspective

Agent Technology from a Formal Perspective

Harville Matrix algebra from a statiscian's perspective

Consciousness: Essays from a Higher-Order Perspective

Computability And Complexity From A Programming Perspective

Statistical Learning from a regression perspective

Statistical learning from a regression perspective

Computability and complexity from a programming perspective

Computational Economics: A Perspective from Computational Intelligence

Violence: Reflections from a Christian Perspective

Malaria: A Hematological Perspective: A Hematological Perspective

Amphibian Metamorphosis: From Morphology to Molecular Biology

Solids far from equilibrium: growth, morphology, defects

Polyurethane Elastomers: From Morphology to Mechanical Aspects

ELF from the programmer's perspective

Recommend Documents