A History of the Chambers Dictionary 9783110312737, 9783110312508

Citation preview

Mariusz Kamiński A History of the Chambers Dictionary

LEXICOGRAPHICA

Series Maior

Supplementary Volumes to the International Annual for Lexicography Suppléments à la Revue Internationale de Lexicographie Supplementbände zum Internationalen Jahrbuch für Lexikographie

Edited by Rufus Hjalmar Gouws, Ulrich Heid, Thomas Herbst, Oskar Reichmann, Stefan Schierholz, Wolfgang Schweickard and Herbert Ernst Wiegand

Volume 

Mariusz Kamiński

A History of the Chambers Dictionary

DE GRUYTER

ISBN 978-3-11-031250-8 e-ISBN 978-3-11-031273-7 ISSN 0175-9264 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.dnb.de abrufbar. © 2013 Walter de Gruyter GmbH, Berlin/Boston Druck: Hubert & Co. GmbH & Co. KG, Göttingen ♾ Gedruckt auf säurefreiem Papier Printed in Germany www.degruyter.com

Acknowledgements I would like to express my profound gratitude to Professor Tadeusz Piotrowski for suggesting the topic for my PhD thesis, on which the current book is based, as well as supervising the work, and sharing with me his deep knowledge of lexicographic theory and practice. Special thanks are due to the reviewers of the original manuscript, Professor Arleta Adamska-Sałaciak and Professor Rafał Molencki, for their insightful critical remarks and valuable comments which laid the foundations for the revision of the work.

Contents Introduction | 1 The object of the study | 1 The structure of the book | 2 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 1.11 1.12

Methodological and theoretical foundations of the analysis | 3 Numbering of editions | 3 Sampling method | 5 The arrangement of entries | 6 The selection of morphological forms | 7 The selection and growth of vocabulary | 8 Defining meaning | 9 The order and discrimination of senses | 12 Etymology | 13 Usage labels | 14 Syntagmatic and paradigmatic information | 15 Pronunciation | 15 Outside matter | 17

2

The founders of W. & R. Chambers publishing house | 18

3 3.1 3.2

The origin of the dictionary | 21 The sociohistorical background | 21 The intellectual background: an overview of possible sources of the Etymological Dictionary of the English Language (Ch-A) | 26

4 4.1 4.2 4.3 4.4 4.5 4.6 4.7

Biographical notes on the editors | 31 James Donald | 31 Thomas Davidson | 32 William Geddie | 32 Agnes Macdonald | 33 Betty Kirkpatrick | 33 Catherine Schwarz | 34 Ian Brookes | 35

5

Educational beginnings: Chambers’s Etymological Dictionary of the English Language (Ch-A, 1867) | 36 Meaning and etymology in Ch-A | 36 Primary meanings | 36 Secondary meanings | 40

5.1 5.1.1 5.1.2

viii | Contents

5.1.3 5.1.4 5.2 5.2.1 5.2.2 5.2.3 5.2.4 5.2.5 5.2.6 5.2.6.1 5.2.6.2 5.2.6.3 5.2.7

Etymological groups | 42 Etymology | 44 Other descriptive and pedagogical considerations in Ch-A | 50 The selection of vocabulary | 50 The selection of morphological forms | 50 Defining meaning | 52 Usage labels | 57 Syntagmatic and paradigmatic information | 57 Pronunciation | 58 Notation system | 58 Suprasegmental features | 60 Choice of variety | 61 Outside matter | 62

6 6.1 6.2 6.3 6.4

Towards the optimal arrangement of entries | 64 Ch-B | 64 Ch-C | 65 Ch-1—Ch-3 | 67 Ch-4—Ch-12 | 68

7 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8

The selection of morphological forms | 71 Ch-B | 71 Ch-C | 73 Ch-1 | 73 Ch-2 | 74 Ch-3 | 74 Ch-4 | 74 Ch-5—Ch-7 | 74 Ch-8—Ch-12 | 75

8 8.1 8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9 8.10 8.11

The growth of vocabulary | 77 Ch-B | 78 Ch-C | 79 Ch-1 | 82 Ch-2 | 82 Ch-3 | 84 Ch-4 | 84 Ch-5 | 85 Ch-6 | 86 Ch-7 | 87 Ch-8 | 88 Ch-9 | 88

Contents | ix

8.12 8.13 8.14

Ch-9new | 89 Ch-10 | 90 Ch-11—Ch-12 | 90

9 9.1 9.2 9.3

Defining meaning | 97 Definitions: general characteristics | 97 Encyclopaedic definitions | 100 Idiosyncratic definitions | 102

10 10.1 10.2 10.3 10.4 10.5 10.6

The order and discrimination of senses | 111 Ch-B | 111 Ch-C—Ch-1 | 112 Ch-2—Ch-3 | 112 Ch-4—Ch-7 | 113 Ch-8—Ch-9 | 113 Ch-9new—Ch-12 | 114

11 11.1 11.2 11.3 11.4 11.5 11.6 11.7 11.8

Etymology | 116 Ch-B | 116 Ch-C | 117 Ch-1 | 119 Ch-2— Ch-3 | 119 Ch-4— Ch-5 | 121 Ch-6—Ch-7 | 121 Ch-8 | 121 Ch-9—Ch-12 | 122

12 12.1 12.2 12.3 12.4 12.5 12.6 12.7 12.8

Usage labels | 124 Ch-B | 124 Ch-C—Ch-1 | 125 Ch-2—Ch-3 | 126 Ch-4—Ch-5 | 128 Ch-6—Ch-7 | 129 Ch-8 | 130 Ch-9 | 131 Ch-9new—Ch-12 | 131

13 13.1 13.2 13.3

Syntagmatic and paradigmatic information | 143 Ch-B—Ch-1 | 143 Ch-2 | 144 Ch-3—Ch-12 | 145

x | Contents

14 14.1 14.2 14.3 14.4 14.5 14.6

Pronunciation | 147 Ch-B | 147 Ch-C | 147 Ch-1 | 148 Ch-2—Ch-3 | 149 Ch-4—Ch-5 | 150 Ch-6—Ch-12 | 151

15 15.1 15.2 15.3 15.4 15.5 15.6 15.7 15.8 15.9

Outside matter | 153 Ch-B | 153 Ch-C | 155 Ch-1 | 156 Ch-2—Ch-3 | 157 Ch-4—Ch-5 | 157 Ch-6 | 158 Ch-7 | 158 Ch-8—Ch-9 | 159 Ch-9new—Ch-12 | 159

Conclusions | 162 Appendices | 166 Selected pages from the early editions of Chambers | 166 Sample 1 | 170 Sample 2 | 233 References | 286 1. Dictionaries | 286 1.1. Editions of the Chambers Dictionary, in chronological order | 286 1.2. Other dictionaries and encyclopedias mentioned in the text | 286 2. Internet sources | 288 3. Other literature | 289 Index | 294

Editions of Chambers | xi

Editions of Chambers, in chronological order Ch-A Donald, James (ed.) (1867): Chambers’s Etymological Dictionary of the English Language. – Edinburgh: W. & R. Chambers. Ch-B Donald, James (ed.) (1872a): Chambers’s English Dictionary. – Edinburgh: W. & R. Chambers. Ch-C Davidson, Thomas (ed.) (1898): Chambers’s English Dictionary. – Edinburgh: W. & R. Chambers. Ch-1 Davidson, Thomas (ed.) (1901): Chambers’s Twentieth Century Dictionary of the English Language. – Edinburgh: W & R. Chambers. Ch-2 Geddie, William (ed.) (1952): Chambers’s Twentieth Century Dictionary. – Edinburgh: W. & R. Chambers. Ch-3 Geddie, William (ed.) (1959): Chambers’s Twentieth Century Dictionary. Revised Edition with Supplement. – Edinburgh: W. & R. Chambers. Ch-4 Macdonald, Agnes. M. (ed.) (1972): Chambers Twentieth Century Dictionary. – Edinburgh: W & R Chambers. Ch-5 Macdonald, Agnes. M. (ed.) (1977): Chambers Twentieth Century Dictionary. (with Supplement). – Edinburgh: W & R Chambers. Ch-6 Kirkpatrick, Betty (ed.) (1983): Chambers 20th Century Dictionary. – Edinburgh: W & R Chambers. Ch-7 Schwarz, Catherine et al. (eds.) (1988): Chambers English Dictionary. (7th edition). – Edinburgh: W & R Chambers. Ch-8 Schwarz, Catherine et al. (eds.) (1993): The Chambers Dictionary. – Edinburgh: Chambers Harrap Publishers. Ch-9 Schwarz, Catherine et al. (eds.) (1998): The Chambers Dictionary. – Edinburgh: Chambers Harrap Publishers. Ch-9new Brookes, Ian et al. (eds.) (2003): The Chambers Dictionary. (new 9th edition). – Edinburgh: Chambers Harrap Publishers. Ch-10 Brookes, Ian et al. (eds.) (2006): The Chambers Dictionary. (10th edition). – Edinburgh: Chambers Harrap Publishers. Ch-11 Marr, Vivian. et al. (eds.) (2008): The Chambers Dictionary. (11th edition). – Edinburgh: Chambers Harrap Publishers. Ch-12 Marr, Vivian. et al. (eds.) (2011): The Chambers Dictionary. (12th edition). – Edinburgh: Chambers Harrap Publishers.

xii | List of tables and figures

List of tables Table 1. Numbering of Chambers editions. | 4 Table 2. A comparison of definitions in Ch-A, Webster-Mahn, and Worcester, based on the selection of entries from Sample 2. | 53 Table 3-1. Dictionary size. Number of characters in the main body. | 92 Table 3-2. Continuation from the previous page. | 93 Table 4. The number of main entries in successive editions. | 95 Table 5-1. Analysis of definitions (based on sample 2) | 107 Table 5-2. Continuation from the previous page. | 109 Table 6. The number and types of labels. | 136 Table 7. A comparison of the macrostructures. | 170 Table 8. A comparison of the microstructures. | 233

List of figures Fig. 1. A treatment of orthographic variants. | 65 Fig. 2. The changing dictionary size: the number of characters in the main body. | 94 Fig. 3. The number of main entries in successive editions. | 96 Fig. 4. The number of labels in sample 3. | 134 Fig. 5. Proportions of labels in sample 3 | 135 Fig. 6. The illustration of the heraldic sense of supporter. | 154 Fig. 7. The illustration of machicolation. | 154 Fig. 8. The illustration of pillory. | 154 Fig. 9. An illustration and the legend under the entry for ship. | 156 Fig. 10. An illustration of altar. | 156 Fig. 11. A page from the 1867 edition (Ch-A). | 166 Fig. 12. A page from the 1872 edition (Ch-B). | 167 Fig. 13. A page from the 1898 edition (Ch-C). | 168 Fig. 14. A page from the 1901 edition (Ch-1). | 169

Introduction The object of the study In the mid-nineteenth century two Scottish brothers William and Robert Chambers started a publishing firm in Edinburgh. Having found a niche in the market, they began to bring out dictionaries and other educational works, making a name for themselves as “publishers for the people” (McArthur 1986: 134). In 1867 they published Chambers’s Etymological Dictionary (henceforth Ch-A), which was edited by James Donald. Five years later, in 1872, a larger version of the dictionary appeared under the title Chambers’s English Dictionary (henceforth Ch-B). For nearly 140 years, the Chambers Dictionary was revised and expanded, with subsequent editions building on the success of the preceding ones. The 1901 edition, titled Chambers’s Twentieth Century Dictionary (henceforth Ch-1), became widely acclaimed by its users and “over the decades grew into the flagship of the Chambers fleet” (McArthur 1988: vii). The latest edition came out in 2011. It is worthwhile to trace the history of the Chambers dictionary for several reasons. The history spans the long period of an undoubtedly successful tradition of native lexicography of English, which numerous editions and reprints testify to. This popular dictionary has gone through twelve editions, though this number would be higher, had the count included three nineteenth-century dictionaries that belong to the tradition of the Chambers dictionary1. A systematic comparison of Chambers editions may give us insights into the work of the lexicographers, and reveal the strengths and weaknesses of each edition. This study, it is believed, will also throw light on the development of British lexicography in general, especially in the period of the latter half of the nineteenth century, that is before the emergence of Murray’s New English Dictionary, the period which is considered as “largely unexplored by lexicographical historians” (Simpson 1990: 1958). The aim of the present book is to trace how the Chambers dictionary has developed over the years, with regard to the following aspects of the structure: the arrangement of entries, the selection of morphological forms, the selection of vocabulary, definitions, the order and discrimination of senses, etymology, usage labels, syntagmatic and paradigmatic information, pronunciation, and outside matter. These aspects will be analysed on the basis of the samples taken from the microand macrostructure of each edition.

|| 1 The nineteenth-century dictionaries were published in 1867, 1872, and 1898. In the current study the author includes them in the count of the Chambers editions, using the following abbreviations: Ch-A, Ch-B, Ch-C, respectively. If these dictionaries had been treated as Chambers editions by the publisher, the most recent edition would have been the fifteenth.

2 | Introduction

The structure of the book This book consists of fifteen chapters. Chapter 1 introduces the theoretical and methodological basis of the research. It explains the method of taking the samples, and fundamental tenets to be followed in the lexicographic analysis. Chapter 2 briefly presents the founders of the Chambers publishers, William and Robert Chambers. The origin of the dictionary is dealt with in Chapter 3 by outlining the sociohistorical factors that stimulated the growth of educational publishing in Victorian Britain. This chapter also addresses the issue of sources of the dictionary, and presents intellectual ideas that might have influenced the structure of the original edition of Chambers. Chapter 4 provides a biographical account of the editors. The design of the first edition (Ch-A) is discussed in detail in Chapter 5. This chapter is entirely devoted to the analysis of Ch-A, within the framework presented in Chapter 1. This original edition is given special attention as it serves as a point of departure for the discussion of subsequent editions. The evolution of the dictionary is the subject of the remaining chapters, 6-15, each of which analysing how a particular aspect of the dictionary structure has changed over the years. The thematic approach adopted in these chapters will ensure a proper perspective on the changing dictionary. Chapters 5-15 are largely based on a systematic analysis of the samples, though references are also made to parts of the dictionary that fall outside the sample range. Each chapter is followed by a summary of the main points. Tables are usually given at the end of the relevant chapters. Complete lists of tables and figures are provided at the beginning of the book. Chapter 15 is followed by Conclusions. At the back of the book, in the Appendices, there are pages selected from the early editions published from 1867 to 1901. They are followed by Sample 1 and 2, which contain excerpts from all editions of Chambers. The former sample represents the macrostructure and comprises six sections of the word-list, while the latter represents the microstructure and covers a selection of full entries. The samples were analysed in detail for statistical purposes. The method of sampling is explained in detail in Chapter 1. Throughout the analysis cross-references pointing to particular entries in sample 2 are used. The cross-references are given in an abbreviated form (e.g. A-3; B-22; 1-4, etc.), in which the first letter (or number) indicates the edition from which the entry is taken, and the following number indicates the number of the entry in the sample. This formula is further explained in Chapter 1. The list of all editions of Chambers, with their abbreviations as used in the present book, is given both at the beginning and at the end of the work.

1 Methodological and theoretical foundations of the analysis This chapter provides a framework for the analysis of the Chambers dictionary. It explains the method of taking the samples and the major theoretical points according to which successive editions were analysed. Each point will be discussed in the context of the original edition, in Chapter 5, and of the subsequent editions, in Chapters 6—15. The framework of the analysis has in part been inspired by Hausmann and Wiegand (1989: 328-359) and their followers, for example Frączek (1999). One of the recurring themes in this scheme of analysis is the major tenet of the theory of lexicographical functions, according to which dictionaries are considered as objects of use produced for a specific group of users in mind (Tarp 2008; Wiegand 1987).

1.1 Numbering of editions Not all editions of the dictionary were numbered. Of all editions only five appeared with a number, either on the imprint or on the title page: the 7th (published in 1988), “the new 9th”(2003), the 10th (2006), the 11th (2008), and the 12th (2011). The 1901 edition did not appear with a number, but it was referred to as the first one on the imprint page of some other editions, for example in the 1959 edition, where it was mentioned in the following note: “first published as Chambers’s Twentieth Century Dictionary in 1901”. What is more, the nineteenth-century editions were excluded from the count by the publisher, while the two editions with supplements (1959 and 1977) were counted as separate2. All in all, it is only recently that the publisher has begun numbering editions in any systematic way. Since most editions were not numbered, the present author has marked them with abbreviations, trying not to interfere with the numbers assigned by the publisher. In this book, Ch-A, Ch-B and Ch-C represent the earliest, nineteenth-century editions, while Ch-1 to Ch-12 refer to subsequent editions. Thus, the numbers overlap with those assigned by the publisher, including the 2003 edition, which is referred to here as “Ch-9new”. The table below shows all Chambers editions, with their corresponding numbers as given by the publisher (the middle column), and as used in the book (right-hand column).

|| 2 This finding has been confirmed by Katie Brooks, the assistant editor of the 2006 edition (private correspondence).

4 | Methodological and theoretical foundations

Table 1. Numbering of Chambers editions.

Chambers editions, in chronological order

1867. James Donald. Chambers’s Etymological Dictionary of the English Language 1872. James Donald. Chambers’s English Dictionary 1898. Thomas Davidson. Chambers’s English Dictionary 1901. Thomas Davidson. Chambers’s Twentieth Century Dictionary of the English Language 1952. William Geddie. Chambers’s Twentieth Century Dictionary 1959. William Geddie. Chambers’s Twentieth Century Dictionary. (Edition with the Supplement). 1972. Agnes M. Macdonald. Chambers Twentieth Century Dictionary 1977. Agnes M. Macdonald. Chambers Twentieth Century Dictionary. (Edition with the Supplement) 1983. Betty Kirkpatrick. Chambers 20th Century Dictionary 1988. Catherine Schwarz. Chambers English Dictionary 1993. Catherine Schwarz. The Chambers Dictionary 1998. Catherine Schwarz. The Chambers Dictionary 2003. Ian Brookes. The Chambers Dictionary 2006. Ian Brookes. The Chambers Dictionary 2008. Vivian Marr. The Chambers Dictionary. 2011. Vivian Marr. The Chambers Dictionary.

||

∗ Mentioned for the first time on the imprint page of Ch-2.

The number of the edition as given on the title or imprint page

The abbreviation used in this book

-

Ch-A

-

Ch-B

-

Ch-C

1∗

Ch-1

-

Ch-2

-

Ch-3

-

Ch-4

-

Ch-5

-

Ch-6

7

Ch-7

-

Ch-8

-

Ch-9

new 9

Ch-9new

10

Ch-10

11

Ch-11

12

Ch-12

Sampling method | 5

1.2 Sampling method In order to ensure a consistent comparison between editions, the same sections of the alphabet were examined. Basically, there were two samples: 1 and 2. The former represented the macrostructure, while the latter the microstructure. The former, which is shown in Table 7, covered six sections of the main word-list: bourn – brandish, gestation – gin, Mab – magic, person – pestilent, sand – sarcenet, waist – warfare. As seen, the headwords were selected from six letters dispersed through the alphabet, so as to represent different parts of the dictionary word-list: the beginning, the central, and the ending part. Given the fact that lexicographers rarely work in a systematic way, this method is preferable to a single-section sampling. The method reduces consequences of “alphabet fatigue”, that is a situation when lexicographers are less thorough and systematic as they work on subsequent letters of the alphabet (Coleman and Ogilvie 2009: 4; Osselton 2007; Piotrowski 2001: 107). The above section boundaries were established in all editions of Chambers. They served as reference points against which the editions were compared. In Ch-A, the whole sample covered 300 headwords, with 50 headwords taken from each letter above. Sample 1 allowed for the analysis of changes in the macrostructure, showing in particular headwords that had been added to the word-list. These headwords were highlighted in Table 7. Sample 2, which is shown in Table 8, represented the microstructure. It covered 42 full entries in Ch-A. The entries were selected on the basis of their position in sample 1 taken from Ch-A, where they appeared as successive entries: boy, boyhood, boyish, brace, bracing, bracelet, bracket, gibbon, gibbose, gibbous, gibe, giblets, giddy, gier-eagle, mad, madcap, madhouse, madman, madwort, madden, madam, persuade, persuasible, persuasion, persuasive, pert, pertain, pertinent, sandwich, sandy, sane, sanable, sanative, sanatory, sanity, wand, wander, wane, want, wanting, wanton adj., wanton n. Thus, each letter in the sample was represented by 7 entries. In order to obtain more reliable results, certain aspects of the microstructure were analysed by looking into the microstructure of entries belonging to sample 1 rather than 2. In this case, the entries are referred to as sample 3. For lack of space, this sample is not included in the book. Sample 3 was used in the chapters on usage labels and on the growth of vocabulary. The three samples served as a major source of information on the evolution of the dictionary. In addition, entries from other parts of the dictionary were sometimes taken into consideration, especially when a given feature, for example an idiosyncratic definition, did not appear in the sample. In order to illustrate changes in the order of senses more vividly, certain entries that fell outside the sample range were selected for further analysis. Moreover, the examination of changes in the outside matter required that parts of the dictionary outside the samples be studied. These included the front, the middle, and the back matter.

6 | Methodological and theoretical foundations

In the course of the book, the reader is often referred to entries from sample 2 by means of symbols. These are given in the form of numbers and sometimes letters, the latter being used to mark the nineteenth-century editions. The first number (or letter) indicates the number of the edition, and the second number refers to the number of the entry in the sample. For example, the notation A-1, which marks the entry boy, means that the entry is from Ch-A and is given as the first entry in sample 2. The entry for boy in Ch-B is coded with B-1, the one in Ch-C with C-1, the one in Ch1 with 1-1, etc. The remaining sections of this chapter will outline major theoretical points relevant to the analysis of the structure of the Chambers dictionary.

1.3 The arrangement of entries The discussion of the arrangement of entries basically involves the description of the macrostructure, i.e. the word-list, which provides initial access to information. However, since many dictionaries place lexical items in the microstructure, i.e. the internal structure of the article, it is also necessary to examine this structure in order to identify the principle whereby the items are ordered. In Chapter 6 and in the section “Etymological groups” in Chapter 5, we will identify the type of entry arrangement, and consider the consequences it could have for the dictionary use. We will also see how the entry arrangement related to the purpose of the dictionary. The organisation of entries depends on the orientation of semantic information. Two types of the presentation of information can be distinguished: semasiological and onomasiological. In the former, the linguistic form, as expressed by spelling, is the starting point for the lexicographic description of the meaning of the form. For the user engaged in decoding, it is also the initial point of departure for the search of an unfamiliar meaning of the word encountered in the text being read. By contrast, in onomasiologically-organised dictionaries, the orientation of information is reversed, that is, it goes from concept to form, for example in a thesaurus, where a concept provides an access point to various word forms expressing that concept. One may expect to find such organisation in dictionaries that are designed to serve productive needs of their users. While the division of dictionaries according to the orientation of semantic information seems to be absolute, in practical lexicography one type of approach does not exclude the other. For example, in semasiologicallyoriented dictionaries, the relations of synonymy, antonymy, and hyponymy, which are typical of onomasiologically-oriented books, can be made explicit on the microstructural level (Louw 2000: 121; Wiegand 1999: 103). The decision as to which approach to take depends on the function of the dictionary, and involves considerations of, among other things, the situation in which the dictionary is to be consulted (Wiegand 1999; Tarp 2008). Thus, reception and production may be regarded as the two main types of language situations that determine the organisation of entries.

The selection of morphological forms | 7

Traditionally, semasiological dictionaries arrange entries in strict alphabetical order. The obvious advantage of this order is that it enables the user to locate an entry easily, according to the position of letters in the alphabet, but it does not do justice to the morphological and semantic relations between words. This problem, however, can be remedied by the so-called nesting of entries, in which items that are morphologically, and sometimes semantically, related are clustered under a shared root or a constituent element, for example handled, handler, handling which may be given under handle. Nested items, which are referred to as subheads, are usually followed by independent explanatory information and other relevant data. Nesting can be revealing as to which items are derived from the same etymological root. It has another advantage, that of saving space, which in any print dictionary is at a premium. On the other hand, the nested type of structure is obtained at the expense of the broken alphabetisation. This is not the case in niching, in which subheads are clustered or listed following the alphabetical order. Nevertheless, clustering of entries, whether by nesting or niching, under a morphologically simpler form may be a hindrance to an inexperienced user, who has to peruse long entries in order to find a given lexical item. Thus, the lexicographer’s decision as to which type of arrangement of entries to use has significant implications for the process of look-up. The alphabetisation of headwords and subheads may be letter-by-letter or wordby-word. In the former type, boundaries between words or constituent free morphemes are ignored, for example: bow, ..., bowsprit, bow-window; sand, ..., sandal, ..., sand-eel; wall, wallflower, wall-fruit (Ch-A). In the word-by-word alphabetisation, in which word boundaries are respected, the lexical items just mentioned would be arranged in the following order: bow, bow-window, bowsprit,; sand, sand-eel, sandal; wall, wall-fruit, wallflower. All these points will be taken into consideration in the analysis.

1.4 The selection of morphological forms There is a wide range of potential candidates for headwords. On the level of morphology, they can be classified as simple forms, derivatives, compounds, bound morphemes, combining forms, and abbreviations. In addition, the main entry status may be assigned to items whose classification goes beyond the morphological analysis and enters into the domain of the phraseology, for example free combinations of words. Again, the decision concerning the selection of morphological forms depends on the function of the dictionary. It requires the consideration of needs of potential users, and the identification of their linguistic problems that the dictionary is supposed to solve (Tarp 2008: 97). The decision as to which forms should be included depends not only on the function of the dictionary but also the availability of space. In order to use space

8 | Methodological and theoretical foundations

efficiently, general dictionaries tend to be limited to the presentation of those language facts that are of irregular character. This is of course justified by the wellknown division of functions: while the grammar describes regularities, the dictionary deals with irregularities. It follows that editors compiling general dictionaries for native speakers tend to exclude regularly inflected forms from the word-list. By the same token, derivatives showing transparent semantic structure are discarded from the word-list on the grounds that their meaning can be deduced from the meaning of constituent morphemes. Whether the Chambers editors limited themselves to the description of irregularities will be investigated in the relevant section in Chapter 5 as well as Chapter 7.

1.5 The selection and growth of vocabulary Chapter 8 and the section on “The selection of vocabulary” in Chapter 5 are concerned with the range of vocabulary selected for lexicographical treatment. They also attempt to determine some aspects of the function of the dictionary, such as the purpose of the dictionary and the intended user’s profile. In order to estimate the growth of vocabulary in successive editions, the editions were compared with regard to size. Depending on the users’ needs, dictionaries may be more or less selective in coverage of vocabulary. A specialised dictionary addressed to adult readers of English literature collects words and meanings that are typical of literature. It may also contain extremely rare words, including the so-called hapax legomena, that is items used only once in a corpus. To what extent such vocabulary should be included depends on various decisions regarding the profile of the target user, the scope of the dictionary, financial resources etc. Yet even if the decisions are made, the realisation of the lexicographical project may not be straightforward, and the lexicographers need to make individual judgements about which items to include. In general dictionaries, the problem of vocabulary selection seems to be even more difficult to solve, as the dictionary is to serve not one but many different groups of users. In such a case, it is very difficult to aim at all-inclusiveness. The print form constraints prevent dictionaries from being exhaustive. What seems to be more realistic is the coverage of various areas of lexis in a more or less extensive way. Selection decisions must have certainly been more difficult in the past, especially when lexicographers worked alone, with no access to a reliable corpus. Dictionaries can be analysed in terms of the direction of use, that is whether they are designed to assist the user in decoding or encoding texts. In native speakers’ general dictionaries, the latter function is rarely the focus of the lexicographers’ attention, as it requires a great deal of space. Typically native speakers’ dictionaries offer relatively little productively-useful information, such as collocations, notes on grammar, the specification of selectional restrictions of words, and related senses.

Defining meaning | 9

However, dictionaries vary in this regard, and certain dictionaries for native users are more useful in encoding than others. The problem of vocabulary coverage is related to dictionary size. We may assume, with a good deal of simplification, that the larger dictionary is, the more vocabulary it contains. Thus, it may be worthwhile to examine the size of the Chambers dictionary in order to gain an idea of how the vocabulary coverage has changed over years. One of possible methods of assessing the dictionary size is to count the total number of entries. Dictionary publishers usually quote numbers of entries to prove that their dictionaries are larger than competitors’ but this method is unreliable, as dictionaries differ from each other in a variety of ways. For example, as Piotrowski remarks (2001: 125), dictionaries do not agree as to what counts as a headword. While some dictionaries assign the main entry status only to one-word lexical items, others may also include multiword lexical items. The decision as to which items should be lemmatised may also be based on whether the potential candidate for the main entry is semantically opaque, but consistency in this regard is hardly achieved as opaqueness, or non-compositionality, is a matter of degree. Taking the above considerations into account, the size of the Chambers dictionary in this book was calculated by counting the total number of characters in the main body, a method which is by far more reliable than the one based on the entry count (ibid.). Nevertheless, for the purpose of illustrating how Chambers editions differed from each other with regard to the macrostructure, the present author also compared the editions in respect of the number of entries. The method of counting characters consisted in scanning ten pages from the main body and using the Statistics tab in the OpenOffice Writer document to obtain the number of characters. The number was divided by ten in order to obtain the number of characters on a page, which was then multiplied by the number of pages in the main body. Chambers editions were also compared in terms of physical dimensions, as measured on the edges of the cover. Assuming that the dimensions of the cover roughly correspond to the dimensions of a page, the surface area of the page was calculated (see Tables 3-1 and 3-2). By dividing the number of characters that appeared on a page by its surface area, the density of the text was obtained. This value reflected the degree of textual condensation on a page. It can safely be assumed that lower values of the text density correspond to better clarity of presentation of information.

1.6 Defining meaning The aim of the discussion below is to raise some points pertaining to the analysis of Chambers definitions, specifically to outline those definition types that typically appear in dictionaries for native users. Definitions can be divided into major classes.

10 | Methodological and theoretical foundations

Although this classification is not exhaustive by any means, it serves as a starting point for the quantitative analysis presented in Chapter 9 and the section “Defining meaning” in Chapter 5. In addition, the chapter/section will deal with such aspects as the definition style, the use of key-words, and the defining vocabulary. These will be briefly discussed below. The definition is a traditional and widely used means of explaining meaning. Following Geeraerts (2003: 84-93), we distinguish several types of definitions, depending on the type of meaning, lexicographic approach, and the target user’s profile. In what follows, major definition types will be presented, in particular those which typically appear in native speakers’ dictionaries and which pertain to the current analysis. In monolingual dictionaries, the most prevailing type of definition has a form of the analytical phrase, which explains the meaning of a word in terms of its genus, or hyperonym, and differentia specifica, that is a selection of differentiating features (e.g. mouse “a little rodent animal found in houses and in the fields” Ch-1). Although analytical definition is most common major definition type, it is criticised for being imprecise (Hanks 1987: 191), unnatural and incomplete (Piotrowski 1989: 95), and subjective or even unintelligible for certain groups of users. Some of these shortcomings are due to the traditional requirement that the definition be substitutable for the word defined. In order to meet this demand, lexicographers often omit complements and objects of verbs from definitions with the effect that the definitions are difficult to understand for less experienced users. The imperfections of analytical definition and the awareness of specific users’ needs have given rise to the development of innovative defining styles involving the use of full sentences. In the sentential definition (Geeraerts 2003: 93), the item being defined is used in the definition itself. One obvious merit of such a definition is that it breaks away from the tradition of substitutability. However, the application of full-sentence definition in lexicography for native speakers is limited, its merits being rather appreciated more in lexicography for EFL learners. Some analytical definitions are special in that they reveal a morphological structure of the item defined. The so-called morpho-semantic definition (Geeraerts 2003: 90) is often used in defining derivatives and compounds. The language of the definition typically includes morphemes or constituent elements of the item (e.g. breakfast “a break or breaking of a fast” Ch-A), but it may also include inflected forms or derivatives of the root of the item (e.g. giddy “that causes giddiness” Ch-A). Another type of definition frequently used in monolingual dictionaries has a form of a single near-synonym or a string of items of similar meaning. For the lexicographer, such a definition is convenient because it is easy to construct, with no need to take the meaning apart. What is more, of all definition types, this one occupies the least space, which is an enormous advantage in a general dictionary aiming at exhaustiveness of word-stock. On the other hand, true synonymy hardly exists in language and a definition by near-synonym offers but a vague idea of the word

Defining meaning | 11

meaning. Yet in a typical situation of dictionary use, when the native user encounters an unfamiliar word in a text and then turns to the dictionary for help, a short instruction or a general hint by means of a near-synonym may suffice for the purpose of appropriate interpreting of the text (Piotrowski 2001: 44). A characteristic feature of many analytical definitions is the coordination of words and phrases, using either the conjunction “or” or a comma. These devices are used to broaden the range of meaning, and are often supported by the abbreviation “etc.”, which is an indicator of an open class of objects (Bańko 2001: 107). Irrespective of the practical utility of these devices, they are considered by Wierzbicka (1996: 266) as nothing more than a sign of the lexicographer’s “defeat” and “resignation” in the attempt at the delineation of meaning. As Wierzbicka remarks (1996: 266), their use testifies to the lack of precision of definition. What is more, frequent use of the above devices results in a highly complex structure of definitions, which may hinder their reception by the user, especially when the coordinated items or phrases are semantically unrelated and their syntactic patterns differ (Bańko 2001: 104). As opposed to analytical definition, which embodies the intensional approach to meaning, lexicographers may choose to define lexical items by extension, that is by enumerating particular objects or concepts denoted by the word rather than describing their semantic features (Geeraerts 2003: 90). In practice, however, as Geeraerts (ibid.) observes, a pure extensional definition is hardly ever used because it is often impossible to list all the members of the category being defined. In effect, lexicographers mention only those members which are typical of the category, using the words such as “especially”, “such as”, “e.g.”, “usually”, “typically”. These keywords, which narrow the range of meaning, are not limited to extensional definition but are used in the analytical definition as well. Following Geeraerts (2003: 91), definitions in which such key-words appear will be referred to as prototypical, as they provide tangible evidence for the protypical organisation of lexical items, emphasising the central part of the category and implying the existence of boundary cases (ibid.). Definitions in dictionaries for native speakers often include a fair amount of encyclopaedic information. The encyclopaedic component may appear in a variety of forms: the inclusion of Latin taxonomic terms in definitions of plants and animals, the inclusion of biographical data, the description of a historical event or a geographical place, the detailed description of objects, processes etc. All such information refers to the extra-linguistic world rather than the language, though in practice it is difficult to establish a clear line between the two (Hartmann and James 1998: 49). The use of encyclopaedic definitions is open to criticism by certain scholars who emphasise that a definition should reflect a conceptual rather than a scientific view of the world (Wierzbicka 1985; Apresjan 1980: 81). This opinion is strengthened by the fact that the scientific view is best described in the encyclopaedia, which is where one should expect to find the information about the world (Piotrowski 1988: 58).

12 | Methodological and theoretical foundations

The process of constructing definitions involves making decisions as to which words to use as part of definitions. Neglecting this problem may result in the ignotum per ignotius type of definition. It is obvious that the employment of rare and technical words may hinder comprehension of definitions and discourage novices from using the dictionary. The problem has been successfully addressed in EFL lexicography by the use of a restricted defining vocabulary. However, in native speakers lexicography, the same model cannot be adopted uncritically for a number of reasons. Firstly, definitions constructed within restricted vocabulary take much more space than traditional ones, which affects the volume and the cost of the work. Secondly, the use of specialist vocabulary is unavoidable in certain definitions (Landau 2001: 166), especially of scientific and technical terms, which constitute a relatively large proportion of entries in native speakers’ dictionaries. Thirdly, lexicographers may be tempted to use a specific word rather than its more general, or psychologically salient, hyperonym because only the former adds to the precision of the definition. However, the above arguments are not sufficient to discourage lexicographers from experimenting with innovative methods of defining, especially when understanding definitions is at stake.

1.7 The order and discrimination of senses The analysis of changes in the arrangement of senses will be carried out in Chapter 10 and in the section on “Secondary meanings” in Chapter 5. Special attention will be paid to graphic ways of discriminating senses. At the outset, it should be clarified that the word sense is used here in the strictly lexicographic sense as a term for a unit of division of word meaning within the microstructure (Piotrowski 1994: 21). This unit roughly corresponds to a definition. The clear-cut lexicographic division of meaning into numbered senses does not necessarily do justice to the polysemous structure of lexical items as postulated by linguists (Adamska-Sałaciak 2007: 186; Lew 2009: 237; Lewandowska-Tomaszczyk 2007: 155-156). The reason behind specific treatment of meaning in dictionaries is strictly to do with their purpose, target users, and specific language situations they are designed for (Lew 2009: 237). In principle, these conditions also determine the order of senses. Kipfer (1984: 101) distinguishes three major ways of sense ordering: by arranging them according to frequency or usage, by distinguishing groups of senses clustered around the core one, and by arranging senses in chronological order. Each of them will be briefly discussed below. In the first method, a sense which is most frequent in the language is given first, and the other, less common senses follow. A systematic execution of the usage principle presupposes that the lexicographers have access to a reliable source of such information, otherwise they are at the mercy of their intuition, which is a rather

Etymology | 13

poor guide to sense ordering. It appears that such an arrangement of senses is of great benefit to learners of foreign languages, for whom mastering common word meanings should be a priority. The second method is based on the lexicographer’s logical inference rather than citation evidence. In this method, senses that are related in meaning are grouped together around a core sense without actual investigation of the order in which they occurred first in the language. As Kipfer (1984: 103) remarks, the method is justified in the case of gaps in the citation evidence, especially when one attempts to show how metaphorical meanings fit into the chronological tree of sense development. Obviously, this method cannot be used as a sole means of establishing the historical development of senses but should rather be combined with the historical principle as supplementary one to the latter (ibid.). Recent studies show that it is worthwhile to exercise logical inference in dictionaries for foreign learners, even when it is made at the expense of the violation of the usage principle. As Van der Meer (1999) demonstrates, the practice of citing a literal sense first, and then the metaphorical one, despite the former being less frequent than the latter in the language, not only shows explicitly the motivation for the extension of the figurative sense but also contributes to raising awareness of such relationships among learners. This finding appears to have implications for native speakers’ lexicography as well. In the last method, senses are ordered in the way that shows the chronological and semantic changes of the word (cf. Hartmann and James 1998: 125). The sense that is given first in the entry is presumed to be the one that appeared first in the language. However, establishing the true path of meaning development is not always easy, even when the lexicographer has access to historical corpora. As mentioned earlier, the textual evidence may be patchy and incomplete. In such a case, it is reasonable to support the historical principle by the logical one (Kipfer 1984: 104), though in the periods with insufficient textual record the inferred order may not necessarily overlap with the actual one. Readers of old classical texts seem to benefit most from historically arranged senses. From this discussion it follows that lexicographers use different criteria for establishing the primary sense, that is the one that should come first in the entry.

1.8 Etymology Chapter 11 and the section “Etymology” in Chapter 5 will consider the contents of the etymological part of the entry and the way etymological information was presented from the user’s point of view. In the context of the study of language, the purpose of etymology is to show changes in form and meaning that a word has undergone over time, from its origin, as far as it can be established or reconstructed from the textual evidence, down to the present day (Durkin 2006: 261). The account of a word history may be of great

14 | Methodological and theoretical foundations

interest for historical linguists and other experts in the field. It may be beneficial for learners of foreign or second languages by raising their awareness of relations between words (Szczepaniak and Lew 2011: 329). However, it is not the type of information that native speakers turn to their dictionaries for most frequently (Béjoint 2000: 144). Thus, it should come as no surprise that general dictionaries for a native audience tend to provide a limited account of etymology. In addition, space constraints are at play too, allowing only for sketchy treatment of the subject, often with a brief statement of the immediate source language and form. Sometimes dictionaries may attempt to outline in more detail the path of the word through the history, indicating intermediate stages of the form and meaning, major cognates in related languages, language periods, and the ultimate form and language, whether known from a corpus or reconstructed (Drysdale 1989: 527).

1.9 Usage labels The relevant section in Chapter 5 and then Chapter 12 will trace how the system of usage labels developed with regard to the type and number of labels. Special attention will be paid to the problem of transparency and clarity of presentation of labels. The analysis will be carried out on the basis of sample 3, which is large enough to illustrate the changes. The data are presented in Figure 4 and 5, and Table 6, at the end of the chapter. We will take into consideration major types of usage labels, excluding the grammatical type, such as v(erb), n(noun), pl(ural), etc. The system of usage labels provides information on the limitations on the usage of a word or a sense by means of special markers, known as labels (Hartmann and James 1998: 40). According to Hartmann and James (1998: 151), labels can be classified according to various dimensions of usage such as currency, emotionality, frequency of occurrence, assimilation, mediality, normativity, formality, style, technicality, textuality, and regionality. Atkins and Rundell (2008: 227-230) distinguish the following types of labels: domain, register (which includes subsets of slang and jargon, and of offensive terms), style, time, attitude, and meaning. Yet these classifications are not exhaustive: in practical lexicography one may find instances of labels that go beyond the pre-established categories or fall into more than one. In this work, we will use a combination of the two classifications above, with the following types of markedness being distinguished: time (e.g. archaic, obsolete), frequency (e.g. rare), attitude (e.g. jocularly, ironically, contemptuously), register and style (e.g. formal, informal, slang, colloquial, poetical), assimilation (e.g. German, French), field (e.g. banking, chemistry, medicine, photography), region (e.g. American, dialectal). In addition, for the purpose of our study, we will add two groups of labels: semantic type (e.g. literal, figurative), and source type. The latter indicates the source text in which the labelled word has appeared in a given sense. Such a

Pronunciation | 15

label appears either as the name of the source (e.g. Bible) or the writer’s name (e.g. Shakespeare). The use of a developed system of marking may testify to the professionalism of the editor and his or her lexicographic skills (Walczak 1991: 32). The lexicographer specifies restrictions on the usage, on the basis of their awareness of the language. Although the choice of a label is often a matter of subjective judgement, the more dimensions of usage is specified in the dictionary, the more valuable the information is (ibid.).

1.10 Syntagmatic and paradigmatic information Chapter 13 and the relevant section in Chapter 5 are concerned with the identification of syntagmatic and paradigmatic information in the dictionary. Syntagmatic relations hold between words in the same sequence or construction (Matthews 1997: 368). Being analysed on the basis of sequential order, they may appear in a variety of forms such as collocations, syntactic phrases, set expressions, and sentences (Hartmann and James 1998: 109). They are distinguished from paradigmatic relations, which hold between words that can replace each other in a given sequence. Examples of paradigmatic relations are antonymy, hyponymy, synonymy, paronymy. Paronyms are items which are similar in form or meaning; hence the alternative name is confusable items (Hartmann and James 1998: 106). The provision of syntagmatic and paradigmatic information is related to the function of the dictionary. In many native speakers’ dictionaries, which are designed chiefly as an aid in reception, one is not likely to find many examples of this productively-useful information, unlike in dictionaries for EFL learners, which in principle are designed as an aid in production.

1.11 Pronunciation The problem of the presentation of pronunciation information involves taking several decisions as to which notation system to use, which accent to represent, how detailed the presentation should be, and which lexical items should be provided with the information. Before discussing the above points (in Chapter 5 and 14), it will be useful to make a brief overview of pronunciation systems used in dictionaries. Respelling is a traditional way of representing pronunciation. It consists in retyping of the headword by means of letters and digraphs. In addition, the system can be supported by the use of different typefaces of the same letter in order to distinguish different phonemes. Furthermore, it can be combined with diacritics over vowel letters to indicate vowel quality, for example a macron, which usually stands

16 | Methodological and theoretical foundations

for a long vowel (e.g. “ā”), and others such as a breve (e.g. “ǎ”), a stroke (e.g. ó), and a double dot (e.g. ä). Diacritics may appear either on the respelled form of the headword, which is often the case, or on the headword itself. The latter used to be favoured by Oxford dictionaries (Fraser 1997: 184), notably COD, the editions from 1911 to 1982 (Kamińska 2010). Yet in this Oxford system, since headwords had to be printed according to the normal rules of spelling, with no change of letters being possible (unlike in respelling), the system necessitated expanding the inventory of diacritics so as to represent the whole set of phonemes. Rather than using diacritics, some early pronouncing dictionaries, such as Kenrick’s New Dictionary of the English Language (1773) used numbers above vowel letters in order to represent the vowel sound. This numerical system dominated most pronouncing dictionaries of the eighteenth century (Beal 2009: 160). On the whole, respelling combined with the above symbols, whether diacritics, numbers or typographic distinction, has developed as a consequence of the fact that the number of letters of the alphabet was insufficient to represent all English phonemes (Fraser 1997: 184). The gap was filled by the use of aforementioned extra devices extending the inventory of symbols. In respelling, the aim is usually to always represent each phoneme with the same letter or a combination of letters, the system being called phonemic (Fraser 1997: 184). In another version of respelling, which Fraser refers to as non-phonemic, a phoneme may be represented by more than one letter or combination of letters (ibid.). In this type of respelling, systematicity between phonemes and symbols is not the lexicographer’s goal (ibid.). Rather the system is based on “ordinary spelling conventions” that the reader is accustomed to. A relatively recent notation system is based on the symbols of the International Phonetic Alphabet (IPA). The IPA transcription has been widely adopted in EFL and bilingual lexicography (Piotrowski 1987: 40), and recently has even found its way to some British dictionaries for adult native speakers, for example COED (Kamińska 2010). Transcription is not dependent on conventional spelling in the way respelling is, though it uses letters of the alphabet. In addition to letters, it employs a number of symbols outside the alphabet. Yet the use of the non-alphabetic symbols causes that the reader has to learn the symbols by heart. On the other hand, an advantage of this system is that it is systematic, with symbol-to-sound and sound-to-symbol correspondences. Other questions that need to be answered in the analysis of pronunciation concern the indication of suprasegmental features such as stress and syllable division. The problem involves the description of typographic symbols and the identification of the rules followed in syllabification. With regard to the latter, two principles of syllable division may be of help. The first one says that the syllable boundary falls between certain indivisible affixes and stems, for example “mis.treat, un.able, free.dom, work.ing” (Kreidler 1992: 84-85). The other says that the place of the syllable boundary depends on the vowel length in such a way that if a consonant letter

Outside matter | 17

follows a short vowel letter, the consonant ends the syllable, as in “rad.ic.al, sav.age” (ibid.). On the other hand, if a consonant follows a long vowel letter, the vowel ends the syllable and the consonant starts the following syllable, as in “ra.di.al, sa.vor” (ibid.). It should be emphasised that the principles, which have been followed by convention in dictionaries since the eighteenth century, are totally arbitrary and text-bound, and do not overlap with the rules of speech (ibid.).

1.12 Outside matter In Chapter 15 and in the section “Outside matter” in Chapter 5, we will look into changes in three parts of the outside matter: front, middle and back matters. Special attention will be paid to the type of information offered in each part. In what follows, we will briefly present the contents of each part of the outside matter. Front matter is the lexicographic data that precedes the central word-list (Burkhanov 1998: 168). Apart from conventional components such as title page, imprint page, preface and table of contents, it provides information essential for the user, namely a list of abbreviations including labels, user’s guide or explanatory information, and pronunciation symbols. Middle matter may appear in the form of illustrations, diagrams, usage notes, and other types of information completing entries. It is strictly speaking not part of the word-list, but it is placed close to the relevant entries (Hartmann and James 1998: 94). Back matter is the lexicographic data which follows the central entry-list (Burkhanov 1998: 168). It may include personal and place names, weights and measures, quotations and proverbs, list of abbreviations, prefixes and suffixes, and chemical elements (Hartmann and James 1998: 12). Back matter in a general dictionary usually contains a high component of encyclopaedic information.

2 The founders of W. & R. Chambers publishing house 3

In this chapter we shall outline the biographies of William and Robert Chambers, throwing light on their efforts to establish the publishing house. The brothers were born in Peebles (Scottish Borders): Robert in 1800 and William in 1802. The former attended Peebles parish and grammar schools, and the latter received “a legendary Scottish education: dame-school for reading; burgh school for reading, writing, and arithmetic; and grammar school for Latin” (Cooney 2004: 1). Their father, James Chambers, cultivated in his children a high respect for education by encouraging them to read books (Hannay 2006: 1). In school Robert showed more enthusiasm for learning than William, who had “a more practical mind” (Scott 1997: 4). William admitted that his brother surpassed him in education: “Indisposed to indulge in the boisterous exercise of other boys – studious, docile in temperament, and excelling in mental qualifications – he shot ahead of me in all matters of education.” (ibid.). Nevertheless, in later life William began to appreciate learning when he saw real benefits from study. On the other hand, William had qualities that predisposed him to succeed in business; the skills proved useful in his working life when he started the publishing firm. The family did fairly well at the time while the father earned a living as a commission agent for cotton weavers, but soon life turned hard. During the Napoleonic wars weaving became no longer profitable and the father, James Chambers, moved into the drapery trade. At that time many French prisoners-of-war stayed in Peebles. Being rather naive, James Chambers gave them cloth on credit, hoping that the Frenchmen would repay their debts as soon as they went home. However, when the French soldiers failed to do it, the Chambers family got into serious financial trouble and had to face poverty. Soon they decided to leave Peebles for Edinburgh. Only Robert remained in the home town to complete his school. Before long it turned out that living on the textile trade in Edinburgh was even more difficult than in Peebles. William, who was then 13, had to help support the family, working as an apprentice in a bookshop. He liked the job, particularly because he could borrow books for reading in the hope that “any knowledge he gained would help him to better himself and his family” (Hannay 2006: 2). The job gave him considerable experience in the book trade, which proved useful in his later life. Robert, who had completed his education in Peebles, had hopes of going to university but the family could not afford it. For some time, he had no regular job,

|| 3 This chapter is based chiefly on a biographical account of W. and R. Chambers that was kindly provided to me by Rosemary Hannay, the curator of the Tweeddale Museum of the Chambers Institute in Peebles.

The founders of W. & R. Chambers publishing house | 19

which made him depressed. He decided to rent a shop in Leith Walk and started his own business as a second-hand book seller. First, he traded in books that he had at home, mostly from his father’s collection. With the money earned, he bought more stock and soon he could support himself. Nevertheless, the life was still difficult for the family. The father lost his job due to his drinking and the mother had to maintain the family on her own. She opened a small tavern, doing most of the work herself. Shortly afterwards, William moved into Robert’s shop and shared a room with him. The living conditions were rudimentary, as William wrote: so miserably was the place furnished, that at first we had no bed, but lay on the floor with a rug for covering and a bundle of books for pillow. Afterwards a bed stuffed with chaff made things a little easier, and rolled up during the day the bed with its rug made a convenient sofa. (Hannay 2006: 2)

Soon William opened a little shop and started to run his own business as a bookseller. He did not confine himself to selling books: he bought them unbound and bound them himself in order to reduce expenses (Hannay 2006: 3). When he had enough money, he purchased a second-hand printing press and taught himself how to typeset and print. Yet the printing machine was difficult to operate and the impression was of poor quality. The brothers’ first attempt at printing was a pocket edition of the Songs of Robert Burns, which brought them financial success (Brookes 2001: 4). The next endeavour was a cheap satirical journal called Kaleidoscope. The price was low because the journal did not raise political matters; otherwise, a tax would be applicable. However, Kaleidoscope turned out to be rather unsuccessful and was ended after three months. In 1824 the brothers published Traditions of Edinburgh, a historical account of the old city of Edinburgh, its people and customs. It went into the third edition, bringing the brothers a financial success. At that time, a number of cheap periodicals intended for a wide audience were available in Scotland. Yet not all of them had a large readership. The publishing business required skills, experience and, above all, knowledge of the market, which William had: “selling books, printing notices, and generally huckstering among the lower middle classes of Edinburgh had given him a nose for what would succeed” (Hannay 2006: 5). Having discovered what the competing papers lacked and what the public needed, William came up with the idea that a successful paper should offer its readers a means of social improvement: With no purpose but to furnish temporary amusement, they [i.e. the papers] were, as it appeared to me, the perversion of what, if rightly conducted, might become a powerful engine of social improvement. Pondering on this idea, I resolved to take advantage of the evidently growing taste for cheap literature, and lead it, as far as was in my power, in a proper direction (Scott 1997: 25).

In other words, William became aware of the growing need for social changes in Britain and the necessity of political reforms. He also realised that a good publica-

20 | The founders of W. & R. Chambers publishing house

tion should not only be entertaining but also educational. In addition, it was necessary that the publication be sold at a moderate price and thus be available for the “humbler orders” (Cooney 2004: 1). William persuaded Robert to write for a new periodical, as the latter had already made his name as a writer: “His wide ranging interests in literature, philosophy, modern life and manners, arts and science were linked to a homely yet elegant and accessible style” (Hannay 2006: 5). The first issue of the journal brought in a great success which surpassed the brothers’ expectations. Chambers’s Edinburgh Journal first appeared in 1832 at a low price of threehalfpence (Cooney 2004: 1). The success of the journal was tremendous and the circulation rose rapidly, reaching 50 000 issues within a year. Robert was the main contributor and his articles particularly appealed to the working and middle classes of Edinburgh (Hannay 2006: 5). The journal provided texts on various subjects such as literature, science, arts, history, and politics. Although the periodical was supposed to avoid the problem of domestic affairs, some articles touched upon the problem of political reforms, which was a current issue in Britain at that time. The weekly brought the brothers big profits and secured their financial situation to a large extent. When the fifteenth issue of the journal appeared, the brothers founded the company of W. and R. Chambers. In 1834 the Chambers brothers began to publish Chambers’s Instructions for the People, a series of sheets dealing with various subjects, including science and literature. Continuing educational publishing, the following year they began work on a series of schoolbooks and short publications entitled Chambers’s Educational Course. Between 1860 and 1868, they published parts of Chambers’s Encyclopedia: a Dictionary of Universal Knowledge (Cooney 2004: 3). Among publications prepared by Robert himself there were Life of Scott, Life of Burns, and a Cyclopaedia of English Literature (Hannay 2006: 9). He was the author of a bestseller Vestiges of the Natural History of Creation, published anonymously in 1844. Robert’s decision to remain unknown arose from his fear of destroying the reputation of the Chambers company by a storm of criticism that the book would trigger (Sedgwick and Tyler 1939: 427). Indeed, the book aroused a strong reaction from conservative critics and provoked a fierce debate about the origin of language. Inspired by astronomical, geological, and physiological evidence, Robert claimed that human speech developed from the rudimentary and primitive ability of animals to communicate, and that language consequently was not of divine origin (Chambers 1845; Aarsleff 1967: 223-224). However, the book lacked a scientific basis, which is why it was deplored in academic circles (Hannay 2006: 7). Nevertheless, no doubt it brought the evolution of species to the attention of a wider public, predating in this sense Charles Darwin’s Origin of Species by fifteen years (ibid.). Robert died in 1871, a year before the publication of Chambers’s English Dictionary (Ch-B), and William – twelve years later, in 1883. Then the head of the firm was taken over by Robert Chambers (1832-88), William’s nephew.

3 The origin of the dictionary Behind the success of the Chambers brothers in educational publishing there were certain historical and social circumstances which will be outlined below. This chapter consists of two sections. The first sketches the situation in Victorian Britain from the point of the system of education, the state of literacy, social changes, and other relevant factors that contributed to the development of the Chambers publishing house and the early editions of the Chambers Dictionary. The English lexicographic scene in the nineteenth century will also be touched upon. The second section examines major sources that the editor of Ch-A drew on, and outlines the intellectual ideas that might have directly influenced the shape of the first edition of Chambers.

3.1 The sociohistorical background When William and Robert Chambers started up educational publishing, they had a good chance of succeeding in this type of business, especially in Scotland. One reason has to do with historical factors that go back to sixteenth century Scotland, when education was fostered as an inherent aspect of the Scottish Reformation (McArthur 1986: 134; Gibson 2009: 32). Over the centuries, with numerous elementary schools governed by the church, the provision of education in Scotland had become more developed than in England. There is no doubt that the longestablished system of education in Scotland generated a need for educational books in the Victorian age. As McArthur (1986: 134) points out, the brothers became famous as “publishers for the people”, which was the slogan that expressed “a particularly Scottish and Presbyterian desire to spread learning to all men and women”. Yet a number of other factors were related to the situation in nineteenth-century Britain. The Victorian age was characterised by rapid social, economic and political changes, which led to the improvement in the provision of education and the spread of literacy. With regard to elementary education in Scotland, it was provided in 1872, following the Elementary Education Act in England in 1870. In Scotland, schooling was compulsory for all children aged between 5 and 13. What is more, with the increasing power and wealth of Victorian England, there was a growing concern about the current system of secondary education (Trevelyan 1986: 532). Public schools, which flourished in Victorian England, provided the opportunity for the members of the upper and middle classes to take up the governing positions and assume every kind of leadership, which the country demanded (ibid.). Although public schools were fee-paying, a number of parents from the new rising classes could afford to send their sons there to secure their education. There were other factors which contributed to the growth of literacy. The Public Libraries Act of 1850 allowed the establishment of public libraries in England (in Scotland three years later). This new law enabled the man in the street to have ac-

22 | The origin of the dictionary

cess to books other than the Bible. This meant that English literature was no longer restricted to the élite but was now available to lower classes as well. The ground for publishing had already been prepared by a series of legislative decisions. In 1774 the copyright law was finally settled, which made it easier for publishers to interpret the law without fear of being sent to prison (Blake 1996: 273). At the same time publishers had more freedom in expressing their ideas (ibid.: 274). A few decades later, in 1835, the stamp tax on newspapers was considerably reduced, and in 1855 it was abolished altogether. These changes had far-reaching consequences for the circulation of newspapers, which rose rapidly with a lowered price. Papers could now be sold cheaply and more people could afford to buy them (Blake 1996: 274; Reisner 1925: 245). What is more, newspapers became more attractive to the general public, as they had more to offer than before (Newspapers 2010). In the first half of the nineteenth century, the content became more varied and the number of pages in provincial papers increased from one to four (Newspapers 2010). The availability of printed matter and democratisation of knowledge went along with social changes. The middle classes, which had emerged in the eighteenth century, increased in wealth over the years of economic prosperity during the Victorian age. However, money was not everything: in order to enter the influential upper classes, they had to conform to certain well-established socio-linguistic norms. It was common knowledge that people were judged by the way they spoke, and language was a marker of the social position. Those who did not conform in their pronunciation and writing to the language of the élite were thought to be ignorant and uneducated. To speak “correct” language became essential for the members of the middle classes who aspired to “talk like their betters” (McDavid 1979: 24; cited in Béjoint 2000: 103). The growth in literacy among working classes was something that the ruling class could be concerned about. As McArthur points out (1986: 134), the dissemination of information raised the social awareness of the working classes, who were oppressed by factory owners, and of other groups of people who suffered in any way in the capitalist society. The need for change and liberation from this depressing state of affairs could prompt the oppressed to rebel against the long-established order. Hence, education, whether received formally in a school or through selfstudy, certainly helped them to participate effectively in the class struggle (ibid.). One of the educational means available for people at large in Victorian Britain was the dictionary. It was a welcome educational tool, which provided the linguistic norm on pronunciation, spelling and meaning; what is more, it could be consulted as often as the user wished (ibid.). The dictionary was a commodity which best suited contemporary individuals who appreciated self-teaching, self-reliance and self-discipline. These values, which derived from Puritan traditions, were the strengths of Victorian society (Trevelyan 1986: 523).

The sociohistorical background | 23

Still, the dictionaries available in the mid-nineteenth century4 were not tailored to the needs of the layperson. They were not easy for the average user to consult due to their scholarly nature (especially Johnson and Worcester), the difficult language of definitions and a sophisticated system of notation of pronunciation. For Chambers and other newly established5 publishing houses such as Collins, Larousse (in France), and later Funk and Wagnall (in the United States), the lack of dictionaries for people at large meant a gap in the market from which they could make a profit (McArthur 1986: 134-135). Although in the nineteenth century the production of a truly easy-to-use dictionary was wishful thinking, these publishing houses at least made an attempt to target a different group of audience than it had been the case before. The publishing houses began the production of reference works for the mass market, setting out with the aim of spreading universal education. As McArthur (1986: 134) notes, between 1866 and 1876 Larousse published a multi-volumed Grand dictionnaire universel du XIXe siècle6, trying to “encapsulate the world so that everybody would understand it better”. As McArthur (1986: 134) continues, some time later, in 1893, Funk and Wagnall issued Standard Dictionary. The editor of the dictionary simplified some aspects of the presentation of information so as to make the book easy to use for the “average man” (Funk 1893-1895: i, vi). He put current senses first, because these were usually searched for by the “average man”, and placed the etymological information at the end of the entry, explaining that otherwise the presentation would have been confusing and discouraging for the user (Funk 1893-1895: xi, xii; cf. McArthur 1986: 134). The dictionaries just mentioned were multi-volumed but at the end of the century there appeared a number of other popular dictionaries, which came out in a small size. As Simpson (1990: 1964) observes, this newly-emerged lexicographic genre, i.e. the concise dictionary, can be exemplified by Annandale’s Concise Dictionary of the English Language (Annandale 1886), Cassell’s English Dictionary (Hunter 1891), and Collins’s Graphic English Dictionary (Williams 1903). Among the representatives of the genre there was also Chambers’s Twentieth Century Dictionary (Ch-1), published in 1901, which is here considered as the fourth edition of the dictionary under study. Although the first dictionary produced by W. & R. Chambers did not belong to the tradition of the dictionary under study, it was compiled with a clear educational

|| 4 John Ogilvie (1850, i-ii) mentions the following dictionaries which were in use in the midnineteenth century: ‘Johnson’s, first published in 1755; Richardson’s, commenced in 1826; and that of Webster, of America, first published in this country in 1832’ (cited in Simpson, 1990, 1959). To this list, one should add Worcester’s (1846, 1860), and obviously, Ogilvie’s Imperial Dictionary (1850), from which this citation is taken. 5 All of these publishing houses were set up in the nineteenth century. 6 The dictionary was published in fifteen volumes (McArthur, 1986, 134).

24 | The origin of the dictionary purpose7. A Dictionary of the English Language was published in 1861 under the editorship of Arnold J. Cooley. As the editor remarked, the book was designed for students in “the middle and upper forms of schools, and for the general and humbler class of readers” (Cooley 1861: iii). The definitions were framed in such a way as to be “sufficiently clear, full, and accurate to meet the wants of the student and critical inquirer” (ibid.). Cooley devoted over seventy pages of the Introduction to a discussion of English sounds, orthography and word-formation. Following the eighteenth-century orthoepic tradition, he devoted most of the introductory material to pronunciation. Six years later after the publication of Cooley’s dictionary, W. & R. Chambers brought out the Etymological Dictionary of the English Language (Ch-A), beginning a long tradition of the Chambers Dictionary. Edited by James Donald, the book received favourable comments from critics: ... Mr Donald’s work is excellently compiled. Taking into consideration its size and price, we do not know a more compendious or a sounder etymological dictionary (The Westminster Review 1868: 604-605)

Following the publication of Ch-A, a few years later, Donald largely extended the dictionary, with adult users in mind. The new version was published under the title Chambers’s English Dictionary (Ch-B). Although in many respects neither Ch-A nor Ch-B looked like the contemporary editions, they paved the way for further development of the dictionary. By the end of the nineteenth century, W. & R. Chambers had become one of the largest English-speaking publishing houses in the world. In the twentieth century, it assumed a strong position in the market of reference books, with Chambers’s Twentieth Century Dictionary being enormously popular for several decades. However, in recent decades the competition in the dictionary market has become fierce8. In the 1990s, the company was bought by Groupe de la Cité, and obtained the name Chambers Harrap Publishers (Chambers Harrap Press Statement 2009). The company became part of Hachette Livre UK group of publishers. Soon hard times came to many publishers printing reference books. Chambers Harrap was no exception, and in 2009 it announced the reorganisation of the company by closing the Edinburgh office (Churchill 2009: 1). The decision was taken by the parent company, Hachette UK, on the grounds of declining sales of print reference works (ibid.). The blame was also put on the strong competition in the market and || 7 This dictionary differed from Ch-A in a number of ways, including the selection of vocabulary and the treatment of individual words. Although both dictionaries employed respelling over vowel letters, Donald’s used a different set of diacritics from Cooley’s. Nevertheless, both had one feature in common: they were designed for students. 8 New rivals have appeared in the field. One of them was the Collins English Dictionary, first published in 1979, which weakened the position of Chambers (Hanks, personal communication)

The sociohistorical background | 25

the availability of free dictionaries on the Internet. As the spokesman of the company explained: “The digital revolution is changing the way readers consume news and search for information. People are moving away from printed reference books and going online where, generally, they expect to get their information for free.” (Mclaughlin 2009). In this state of affairs, the future of the dictionary in print is uncertain. Nowadays, publishers apparently take the opportunity to rethink their policy with regard to print matter but they also offer dictionaries, though often incomplete versions, on the Internet. The latest edition of the Chambers Dictionary is now available from the publishers’ website in the form of a downloadable application, while other Chambers dictionaries can be searched directly online9.

|| 9 http://www.chambers.co.uk/dictionaries/the-chambers-dictionary.php

26 | The origin of the dictionary

3.2 The intellectual background: an overview of possible sources of the Etymological Dictionary of the English Language (Ch-A) In order to complete the account of the historical background of the dictionary, in what follows, we will provide a survey of potential sources for the compilation of the first edition of Chambers. In the preface, Donald credits several philologists, including Bopp, Pott, Grimm, Curtius, Diez, Donaldson, Müller, Latham, Garnett, Trench, with the investigation of “the Etymology of our language, a study which of late years has made such marked advances” (Preface: vi). Moreover, in the front matter, the editor provides a “Select list of books consulted in the preparation of the etymological portions of this dictionary, and which will be found of value to the student of etymology” (henceforth “List”; Preface: iv). Since these books might have provided the editor with some theoretical inspiration and ready-made material to draw on, it will be worthwhile to take a closer look at some of them. These books, I believe, will give some idea of the intellectual history of the period before the publication of Chambers. Among the books listed, Donald mentioned in the first place Hensleigh Wedgwood’s Dictionary of English Etymology, to which he owed a big debt. Donald valued Wedgwood’s dictionary highly, remarking that “no student of Etymology should be without” it (Preface: vi). Here, it should be noted however, that the selection of Wedgwood as a potential source of etymological information ran counter to Donald’s aim of providing the student with state-of-the-art etymological research. This is because Wedgwood disregarded sound correspondences and displayed misguided ideas about the origin of language in interjections (Liberman 2009: 275-276). As a result, many of Wedgwood’s etymologies were fanciful and intuitive (ibid.). Although Wedgwood did not follow the line of the new philological research, he was a well-known practitioner and theoretician of etymology at the time (ibid.), which earned him some authority among the lay public. However, quite contrary to what Donald implied in the preface, he did not follow Wedgwood faithfully, if at all. Comparing the dictionaries, one can see noticeable differences with regard to etymological information, the choice of cognate words and root forms. Donald was even taken to task by a critic for failing to obey Wedgwood at forest and butterfly: But we think [James Donald] might with advantage have followed Wedgwood even more closely than he has done. For example, Mr. Donald, forgetting his master, tells us that forest is “probably from the Latin foras, fors, out of doors,” which has always appeared to us a fair example of the thoroughly vicious, uncritical, and unphilosophical style of derivation. [...] So, too, Mr. Donald, keeping to old etymological traditions, tells us that butterfly is so called from “the butter colour of one of the species.” (The Westminster Review 1868: 604)

The intellectual background | 27

In fact, this criticism is unjustified, as the etymologies cited do not differ much from those in modern dictionaries10. Indeed, they were not taken over from Wedgwood but from Webster’s American Dictionary of the English Language, which was brought out in 1864, that is three years prior to the publication of Chambers. This source book also figures on the Chambers “List”. Known as Webster-Mahn, the dictionary was prepared with the help of a distinguished German etymologist, C. A. F. Mahn, who cleared the dictionary of fanciful etymologies that had pervaded its previous edition. Webster-Mahn might have been exploited by Donald in the area of etymologies, and, what may become by far more obvious, in the area of definitions. We will return to this point later. Friedrich Max Müller’s Lectures on the Science of Language, which is mentioned on the “List”, was probably the most accessible book on comparative philology available for a general audience (Harris and Taylor 1997: 186). Müller popularised Darwin’s views on language in his Lectures, pointing out that language should be studied in the same way as natural sciences are (ibid.). Discussing the value of written materials for the purposes of comparative philology, he remarked that “we can collect them, we can classify them, we can reduce them to their constituent elements, and deduce from them some of the laws that determine their origin, govern their growth, necessitate their decay; we can treat them, in fact, in exactly the same spirit in which the geologist treats his stones and petrifactions” (Müller 1861; cited in Harris and Taylor 1997: 186). The above passage outlines the main points of the linguistic methodology, emphasising the historical orientation of the study and the opportunity for the discovery of laws from written records. On the “List”, Donald mentions Charles Richardson’s New Dictionary of the English Language (1836-1837). It has to be emphasised that the ideology of Richardson’s dictionary was influenced by John Horne Tooke’s reductionist theory of language. Tooke recognised only two parts of speech, nouns and verbs, which are necessary for communication; all other word classes were merely abbreviations of the former created over time so as to ensure fluent speech (Aarsleff 1967: 46). Tooke saw morphological elements as derived from parts of earlier independent words11. But more importantly, he believed in the indivisibility of etymology and meaning12, and main|| 10 With regard to forest, OED mentions ‘... f. forīs out of doors’. As it comes to butterfly, the dictionary states that the etymology is unknown, though it speculates about the reason of the name by quoting Wedgwood. 11 This conception was also favoured by a German linguist Bopp but he, unlike Tooke, limited its application to a few morphemes (Robins, 1967, 157, 173). 12 Tooke’s assumptions concerning the indivisibility of etymology and meaning were by no means without precedence in the history of philosophical ideas. Even in antiquity a common tenet was that in the etymology of a word, one could discover the underlying meaning, the meaning which was supposed to be ‘true’ because it revealed the ‘true’ nature of the thing. This bias was taken for granted by the seventeenth and eighteenth-century lexicographers, such as Cawdrey, Bullokar, Cockeram, Blount, and Bailey.

28 | The origin of the dictionary

tained that the etymology of a word can reveal its intrinsic meaning, the meaning that “embraces the totality of the word’s meaning”, including the original meaning as well as metaphorical and consequential meanings (Zgusta 2006: 21). Although the above assumption was also accepted by a few nineteenth-century German philologists (e.g. Jacob Grimm and Franz Passow (Zgusta 2006: 21)), Tooke’s methodology of discovering the roots of words was based solely on superficial resemblance of forms, which resulted in mere speculative etymologies. By contrast, the philologists used a principled method of establishing cognate relationships between words in different languages based on a rigorous study of sound correspondences (ibid.). It is not that Tooke’s views on the nature of language were incorrect in their entirety but they were a result of his a priori theorising, unsupported by empirical evidence. Tooke’s monosemous view of meaning was taken up by Charles Richardson, who in his dictionary noted: The great first principle upon which I have proceeded in the department of the Dictionary which embraces explanation, is that so clearly evolved and so incontrovertibly demonstrated in the “Diversions of Purley”; namely, that a word has one meaning, and one only; that from it all usages must spring and be derived; and that in the etymology of each word must be found this single intrinsic meaning, and the cause of the application of those usages (Richardson 1838: 41; cited in Worcester 1860: iv)

However, despite the above claims, his actual practice deviated from his theory substantially, with the role of etymology being diminished accordingly. In a single entry, Richardson clustered together English words (and their derivatives) that were linked by (often) superficial similarity of form, and that were supposed to be linked by the same intrinsic meaning13. This meaning, which was recoverable from the etymology, was supposed to comprise all the senses of a word entirely or nearly so (Zgusta 2006: 47). In order to explicate the meaning, he collected quotations from literary authors, and arranged them in historical order, typically without support of definitions. Faced with problems of explicit identification of how actual readings related to etymology, he left it up to the user to infer the underlying meaning from the quotations. As Zgusta (2006 24) points out, Richardson failed to show a logical link between the etymological meaning of a word and its subsequent meanings; in fact, while there are entries where such a link is noticeable, in many lexical items with a highly polysemic structure this task is impossible to achieve. It will be noted that Donald intended to achieve a similar aim: to link primary meaning with the secondary ones in a logical way. What made Richardson a valuable source of inspiration for later lexicographers was not the etymological underpinning along the lines of Tooke’s theory but the || 13 As Read (1986, 48-49) points out, Richardson conflated ham and hamlet, explaining that the ‘idea of “uniting” was behind them, a hamlet being a place where people were “united” and the ham being that part of the leg which was “united” to the rest of the body’.

The intellectual background | 29

idea of historical sequencing of quotations. His arrangement of quotations was consistent with the historical trends in linguistics (Zgusta 2006: 23), and thus the chronological principle was worth pursuing by Donald in Chambers too. By the time Chambers was published, Tooke’s ideas had long fallen into disfavour. In 1835, Richard Garnett, whose Philological Essays is also mentioned on Chambers “List”, rightly observed that the adherence to Tooke’s philosophical etymologies is “more likely to impede the improvement of sound philology than to promote it” (Garnett 1835: 18; cited in Aarsleff 1967: 113). One of the most severest critics was John William Donaldson, the author of New Cratylus, who remarked that Tooke’s Diversions of Purley “reflects little credit on English philology that it should have been so regarded then; [...] Its authority and influence have done much harm to us philologists, [...] it has led us upon a false track, and lulled us into a delusive security [...] Let us escape from the slavish fear or silly superstition that has tyrannized over us” (Donaldson 1839; cited in Aarsleff 1967: 113). Like Garnett, Donaldson is acknowledged as a source for Chambers. Richard Chenevix Trench’s On the Study of Words is another source book which is mentioned in the front matter and which might have provided Donald with some theoretical underpinnings for his dictionary. Drawing on natural science, Trench viewed language as “fossil poetry”14, “fossil ethics”, or “fossil history” (Trench 1853: 13). He used the “fossil” metaphor to explain properties of words: “just as in some fossil, curious and beautiful shapes of vegetable or animal life, the graceful fern or the finely vertebrated lizard [...] are permanently bound up with the stone, and rescued from that perishing [...] – so in words are beautiful thoughts and images, the imagination and the feeling of past ages [...] preserved and made safe for ever.” (ibid.: 13). For Trench, the search for etymological meanings was a valuable means of language education. He viewed teaching of etymology as a “didactic art” and a “form of enlightening pedagogy” (Hutton 1998: 193). Quoting the poet Samuel Taylor Coleridge15, Trench points to the educational role of etymological instructions: In a language [...] like ours, where so many words are derived from other languages, there are few modes of instruction more useful or more amusing than that of accustoming young people to seek for the etymology or primary meaning of the words they use. There are cases in which more knowledge of more value may be conveyed by the history of a word than by the history of a campaign. (Trench 1853: 12-13).

|| 14 This is how language had been called by the American poet Emerson (Aarsleff, 1983, 33). 15 Coleridge was a proponent of the Encyclopaedia Metropolitana as a thematically arranged reference work, which began to come out in 1818 (McArthur, 1986, 156).

30 | The origin of the dictionary

The above passage also shows that, for Trench, etymology and primary meaning stood in close relationship. It will be noted that the provision of primary meanings was a distinguishing characteristic of Donald’s dictionary. Trench was interested in investigating the relationships among words, believing that they are essential for the right understanding of the words meaning (Trench 1853: 201). In On the Study of Words, he provided the reader with the analysis of words closely related to each other. His analysis was undertaken on near-synonyms as well as words which were unrelated in contemporary semantic terms but shared “the common root” which could be discovered through etymology (Trench 1953: 203). Trench instructed the reader how to investigate the connections between different senses of a word, using biological and genealogical metaphors: This tracing of that which is common to and connects all its meanings can of course only be done by getting to its heart, to the seminal meaning, from which, as from a fruitful seed, all the others unfold themselves; to the first link in the chain, from which every later one, in a direct line or a lateral, depends. And we may proceed in this investigation, certain that we shall find such, or at least that such there is to be found. For this we may start with, as being lifted above all doubt [...], that a word has originally but one meaning, and that all the others, however widely they may diverge from one another and seem to recede from this one, may yet be affiliated upon it, may be brought back to the one central meaning, which grasps and knits them all together; just as the races of men, black, white, and red, despite of all their present diversity and dispersion, have a central point of unity in their first parents. (Trench 1853: 204-205)

Trench went on to give an example illustrating that the etymology-driven idea grasps all meanings of a word. He explained the methodology of arriving at primary meanings, which might have inspired Donald several years later. Here is the word “post;” how various are the senses in which it is employed; “ post” – office; “post” – haste; a “post” standing in the ground; a military “post;” an official “post;” “to post” a leger. Might one not at first presume it impossible to bring all these uses of “post” to a common centre? Yet indeed when once on the right track, nothing is easier; “post” is the Latin “positus,” that which is placed; the piece of timber is “placed” in the ground, and so a “post;” a military station is a “post,” for a man is “placed” in it, and must not quit it without orders; to travel “post,” is to have certain relays of horses “placed” at intervals, that so no delay on the road may occur; the “post” – office is that which avails itself of this mode of communication; to “post” a leger is to “place” or register its several items. (Trench 1853: 205-206).

The above selected books were among the possible sources available for Donald. They represent the overlap of two traditions: some of them, like Richardson and Wedgwood, are representative of the earlier tradition of philosophical etymologising, while others, like Webster-Mahn and Müller, are indicative of the new epoch in linguistics, i.e. the historical-comparative philology. Obviously not all of these books left traces in Donald’s work. Yet, as we shall see later, Webster-Mahn was exploited extensively by Donald, while Trench’s etymological pedagogy found its practical application in Donald’s work.

4 Biographical notes on the editors The current chapter by no means covers all lexicographers involved in the compilation of the dictionary, and excludes a number of experts and contributors who had their share in the preparation of the work. The chapter is rather sketchy, and is largely based on newspaper archives published on the Internet. The nineteenthcentury issues of Otago Witness, Daily News, The New York Times, The Newcastle Courant turned out to be invaluable sources of information on the earliest editors. One Internet site16 is especially worth noting because it provided comprehensive information on Thomas Davidson, the editor of Ch-C and Ch-1. While the data concerning other editors is fragmentary, thanks to this site we gained a clear picture of the editor’s life. The site, titled “PERSONALIA ET PRINGLEANA”, presented a scrapbook initiated by Davidson’s relative, Alexander Pringle, BSc MA (1842–1934). It included a collection of newspaper cuttings, family photos, and other interesting information submitted over the years by Pringle’s descendants.

4.1 James Donald James Donald was the editor of the first two Chambers editions: Ch-A (1867) and ChB (1872). He was a journalist appointed in Hampshire (Otago Witness 1877) and was also engaged in the preparation of educational works (Daily News 1877). From the title page of Ch-A, we learn that Donald was the editor of Chambers’s Readings in English Literature as well as a fellow of the Royal Geographical Society (“F.R.G.S.”). As the editor of the successful Ch-A, he was praised for being “a ripe scholar and diligent student” (The New York Times 1876). In addition to the editorial work, he seems to have been commissioned to carry out market research for Chambers. In his report to William Chambers, he complained about difficulties in introducing Chambers’s Educational Course in Dundee schools: Speaking generally of Dundee, the books of the Educational Course are not largely used. The books most used are Nelsons’, Mc. Culloch’s, and those of the Scottish Schoolbook Association. [...] There is a pretty general repugnance to change of any kind on the part of Directors of schools, even though the teachers themselves would prefer a change of many of the books. (Donald 1872b)

In 1877 Donald was reported to reside at Rotherwick, Winchfield (The Newcastle Courant 1877). In the same year, at the age of 38, he died tragically in a railway acci|| 16 http://salhun.100webcustomers.com/Main.htm#_Rev_Thomas_Davidson_1. 2007).

(accessed

Dec,

32 | Biographical notes on the editors

dent. Unfortunately enough, the accident happened on the way home from Edinburgh, where he had attended his father’s funeral (ibid.).

4.2 Thomas Davidson Thomas Davidson was the editor of Ch-C (1898) and Ch-1 (1901). He was born on the 27th of April 1856 at Leswalt, as the son of John Davidson and Anna Pringle. Educated at the University of Edinburgh, Davidson was a missionary assistant at Edinburgh Pilrig F.C. He spent several years in Vienna, where he held a vacancy as a Presbyterian chaplain. He engaged in literary work with the Chambers publishers, with whom he collaborated for 17 years. Working on the Chambers dictionary, he was assisted by his brothers, who were also clergymen: Rev. John Davidson (Ch-C), Rev. A. P. Davidson (Ch-C), Rev. Robert P. Davidson (Ch-C and Ch-1), and David G. Davidson (Ch-1). Thomas Davidson was in the same line of work as many other clergymen before him: Reverend James Barclay, the editor of the Complete and Universal English Dictionary (1774), Noah Porter and Chauncey Allen Goodrich, the editors of the American Dictionary of the English Language (1860) (ICON 2008: 158), and Richard Chenevix Trench, Dean of Westminster, who initiated work on the Oxford English Dictionary. Thomas Davidson contributed to the compilation of other works, including Chambers’s Encyclopaedia, Globe Encyclopaedia, and Hastings’s Encyclopaedia of Religion and Ethics. Furthermore, he prepared annotated editions of parts of English literary works of Chaucer, Milton, Byron and Scott. At the outbreak of the First World War, in 1914, he resigned from his post in Vienna and moved to Afton, in Ayrshire, where he served as a minister in a church. That year he left the Chambers publishing house, and the revisions of Ch-1 were taken over by William and Liddell Geddie (Brookes 2001: 17). Thomas Davidson died in 1923 in Afton, at the age of 67.

4.3 William Geddie William Geddie was the editor of Ch-2 and Ch-3, published in the years 1952 and 1959, respectively. He graduated in arts and science from Edinburgh University. For over 20 years, he was the chief editor of Chambers’s Encyclopaedia. Geddie was known for his “scholarly powers in a Scottish Text Society publication on Middle Scots poets” (The Times 1967). While working on the Chambers dictionary, he collaborated with his brother, Liddell. He was also assisted by Agnes Macdonald, the later editor, who was responsible for “a great part of the construction” of Ch-3 (Ch-3: v). William Geddie died in 1967 in Edinburgh. In an obituary, Geddie is remembered as a competent lexicographer:

Betty Kirkpatrick | 33

... he was able to exercise his rare and enviable faculty of being able to work in a peculiarly difficult field at highest speed, yet with the utmost brevity, and with complete accuracy, clarity and objectivity. ... Geddie was of shy and retiring disposition, but he never failed to keep a watchful eye on the world around him. Very often that eye had a twinkle in it, as shown in some of his dictionary definitions, e.g. “baby-sitter, one who mounts guard over a baby to relieve the usual attendant”, or “petting party, a gathering for the purpose of caressing as an organized sport”. (The Times 1967)

4.4 Agnes Macdonald Agnes M. Macdonald, B.A. contributed to Chambers as a chief editor of Ch-4 and Ch5, published in the years 1972 and 1977, respectively. She graduated from St Hilda’s College in Oxford (The Times 1974). Macdonald joined the Chambers publishers in 1944. She was said to be a “cool, formidable woman” (McGlone 2003). When working on the Chambers dictionary, she was assisted by E. Kirkpatrick and P. Kerr. She also contributed to the compilation of several other dictionaries, including Chambers Everyday Dictionary. In 1973 Macdonald was awarded with an OBE (Order of the British Empire) for her work in lexicography (Brookes 2001: 19). She died in Edinburgh in 1974.

4.5 Betty Kirkpatrick Betty Kirkpatrick joined the Chambers publishers in 1966 (Davidson 1983: 14). She worked under the supervision of Macdonald for several years until she took over the editorship of Ch-6, which was published in 1983. Kirkpatrick spoke of herself as an energetic woman who “does the educational books when [she] can find the time” (Davidson 1983: 14). With her husband being a doctor, she could contribute to the dictionary in the field of medicine (ibid.). As Davidson (ibid.) writes, dictionary making at that time was a collective work, with each member of the staff displaying his/her area of interests in which they could offer advice as experts: Anne Seaton is acknowledged as the classics expert and George Davidson is strong on linguistics and phonetics. John Simpson lays claim to sport and jazz, Catherine Schwartz to music and Rachel Sherrard, wife of a minister, knows a lot about religion. (Davidson 1983: 14).

Kirkpatrick has judged the National Scrabble Championship and National Scrabble Club Tournament, for which Chambers is a reference dictionary. In 1980 under her editorship, Chambers Universal Learners’ Dictionary came out, challenging the monopoly of Oxford and Longman publishers in the market of pedagogical dictionaries (McArthur 1998: 143). She has been either an editor or co-editor of a number of reference works for other publishing houses, such as Cassell, Macmillan, Longman,

34 | Biographical notes on the editors

Crombie Jardine. The books published by Cassell include Brewer’s Concise Dictionary of Phrase and Fable, Cassell Paperback English Dictionary, Cassell Concise English Dictionary, Cassell Encyclopedia Dictionary, Cassell Giant Paperback Dictionary, Cassell Student English Dictionary. She has recently compiled The Concise Dictionary of Scottish Words and Phrases, published by Crombie Jardine. Kirkpatrick is also an experienced theoretician of lexicography: she has written several papers, including User’s Guides in Dictionaries, published in An International Encyclopedia of Lexicography (Hausmann et al. 1989). Outside the field of dictionary making, her recent books are Auld Scottish Grannies’ Remedies and Book of Home Remedies.

4.6 Catherine Schwarz Catherine Schwarz was a managing editor of Ch-7 (1988) and Ch-8 (1993) and, together with the lexicographer, Robert Allen, was a consultant editor of Ch-9 (1998). Born in 1933, she studied French and Latin at Edinburgh University. After graduation, Schwarz joined the Chambers publishers. Her parents had also worked for the company: her father, Tom Collocott, was an educational editor and then a managing director while her mother Elma worked as a secretary (White 2006: 219). At the beginning of her career at Chambers, Schwarz prepared encyclopaedic reference books as well as the Chambers Journal, a famous periodical initiated by William and Robert Chambers in 1832 (ibid.). When she set out to compile the Chambers dictionary, she did “much to make it a modern book whilst preserving its character and humour” (ibid.). Although she retired in 1993, she remained a consultant editor until 2003, when Ch-9new was published. She also adjudicated at Scrabble contests, as she once remarked: “once a lexicographer always a lexicographer” (McGlone 2003). Schwarz was a real lexicographer, always in search for new words to be included in the dictionary: I’m forever writing words down that I think should be in the dictionary. At the moment, I’m interested in the word “leafy”, as in: “leafy suburb”, as opposed to “full of leaves”. That is not in the dictionary yet, but I know they’re watching it, which is how words gently slither into the language. As a Scot, I relish the peculiarly Scottish sensibility of TCD [the Chambers dictionary] and its comprehensive coverage of Scottish words, although one of my favourites – “numpty” (“n. an idiot”) – was missing until recently. (McGlone 2003)

Altogether, Schwarz contributed to six editions of the Chambers dictionary (Ch-7: v). She died in 2005.

Ian Brookes | 35

4.7 Ian Brookes Ian Brookes was the editor of Ch-9new (2003) and Ch-10 (2006). He studied Latin at Durham University and received his PhD degree at Newcastle (McGlone 2003). As a lexicographer, he appreciates his educational background in Latin: Latin offered a more distilled version of language than English. Certainly, in terms of understanding etymology, syntax and grammar, a classical education is a useful qualification for a lexicographer – a job that’s a combination of logic, aesthetic and problem-solving. (McGlone 2003)

Brookes is a managing editor of the Chambers Compact Thesaurus, an editor of the Adult Learner’s Dictionary and a co-editor of the Chambers Concise Dictionary. Among his other books on language, there are Chambers Good Writing Guide and Perfect Punctuation. Summing up To conclude this chapter, it seems that recent Chambers lexicographers have had a more professional background in the field of language studies. They studied foreign languages and thus seem to have been more qualified for compiling dictionaries than the early editors, whose academic training was unrelated to lexicography. As a matter of fact, the emergence of linguistics as an academic discipline in Great Britain is a relatively late phenomenon, dating back to the mid-twentieth century (Robins 1967: 213). Suffice it to say that the first professor of linguistics (J. R. Firth) was appointed in the University of London in 1944 (ibid.). More importantly, it is only recently that professional training in lexicography has become available for dictionary compilers (Hartmann 1986). Furthermore, lexicographers’ work has recently become more advanced and efficient: they work together as a team, assisted by consultant editors and professionals in particular branches of science. Finally, it should be noted that since the 1950s, women have joined the Chambers lexicographical staff, with Macdonald, Kirkpatrick and Schwarz taking up leading positions on the editorial ladder. Likewise, in other publishers, like Longman and Collins, more women have recently been involved in the compilation of dictionaries, but it appears that the Chambers publishers have taken the lead in this trend (McArthur 1986: 203). By the time Macdonald took over the editorship in the 1960s, few women had been in charge of lexicographic projects at other publishing houses (ibid.).

5 Educational beginnings: Chambers’s Etymological Dictionary of the English Language (Ch-A, 1867) This early Chambers had a clear educational purpose: “It is confidently hoped that this Dictionary will be found not only eminently suitable for general educational and practical purposes, but also peculiarly adapted for assisting in the higher philological study of the English language in advanced classes” (Preface: vi). The editor, James Donald, believed that his book would “supply the want, so long felt, of a Dictionary based on the etymological relations of words, and exhibiting the results of the latest philological research” (Preface: v). Yet the title was somewhat misleading about the content of the dictionary (cf. Liberman 2009: 269). It was not a dictionary that offered an in-depth account of word etymologies. Designed chiefly as a school dictionary, it covered about 25 000 entries. Chambers devoted relatively little space on etymological information, providing other types of information such as definitions and pronunciation, which were part and parcel of contemporary dictionaries for a general audience. But then the role of etymology in Chambers can hardly be underestimated, considering how it related to meaning, definitions and entry structure. In what follows, we will discuss the role of etymology in the interpretation of meaning, the presentation of senses, and the arrangement of entries in Ch-A.

5.1 Meaning and etymology in Ch-A 5.1.1 Primary meanings A characteristic feature of Donald’s work is the provision of primary, etymological meanings: “The Meanings are based on the root-ideas of the words, a plan not only logically correct, but calculated to give increased vividness to the conception.” (Preface: v). The name “root-idea” may be indicative of a nineteenth-century philosophy of language that accepts the view that meaning is considered as a thought or idea (Geeraerts 1999: 125). Although the editor did not explain to the user what the “root-idea” was, his lexicographic practice shows that it was the meaning of the ultimate etymon, which the editor provided in the etymological part of the entry, and which he used when constructing primary definitions. Root-ideas were typically expressed by direct translations of the etymons (e.g. “to cover” in the entry bark below), which were sometimes supported by near-synonyms or semantically related words (e.g. “rind”).

Meaning and etymology in Ch-A | 37

A characteristic feature of primary definitions was typography. The editor used italic type to highlight root-ideas, leaving the rest of the definition in normal type17: Bark ... the outer rind or covering of a tree ... [Dan. and Sw. bark; Ice. börkr; A.S. beorgan, to cover.] Chocolate ... a beverage made by mixing the cocoa-nut, roasted and ground, with hot water ... [... from Mexican chocolatl – choco, cocoa, latl, water.]

The etymologies, which were placed at the end of the entry, were given to support the primary definitions. Thus, bark was traced back to Anglo-Saxon beorgan, meaning “to cover”, and chocolate, to Mexican choco, meaning “cocoa” and latl, meaning “water”. Donald’s decision to italicise root-ideas may have been inspired by Liddell and Scott’s Greek-English Lexicon (first published in 1843). Following a German scholar Franz Passow, Liddell and Scott used double typography in definitions: italics indicated lexical equivalents to Greek words, and roman type served as “a semantic bridge” to explain how one equivalent related to another (Zgusta 2006: 36). In Chambers, a similar typographic convention was applied: italics indicated English translations of etymons. As a rule, Donald provided primary meanings systematically, regardless of whether they were attested in the language or not. This is the point of difference from Webster-Mahn, whose defining part began “if possible, with the primitive signification, as indicated by the etymology” (Webster: vi; my italics). This means that hypothetical senses, unrecorded in the language, were as a rule omitted in Webster. For example, the etymological meaning of breed which was given in Chambers as “lit. to nourish, or keep warm” was not indicated in Webster as a sense, though it was mentioned in the etymological part (“[A-S. brédan, to nourish, cherish, keep warm ...]”). This also testifies to the fact that Webster was a source of etymological and definitional information for Donald. Comparing Donald with early English lexicographers who had claimed that etymological meaning is “true”, one can note that Donald showed more enlightened views on the relation between etymology and meaning. While many of the earlier lexicographers, such as Cawdrey, Bullokar, Cockeram, Blount, and Bailey, had thought of etymology as synonymous with “true” meaning, Donald did not claim that etymological meaning is “true”. This is a characteristic change in linguistic thought that took place in the nineteenth century: the search for a “true” meaning was no longer treated as a scientific preoccupation, especially when it became obvi|| 17 Actually, in sample 2, 86% of primary definitions were italicised entirely or in part. Apparently, in the remaining 14% of primary definitions, the editor failed to identify ‘root-ideas’. In addition, Donald was quite inconsistent in italicising definitions, for example, compare boyhood ‘state of being a boy’ with sanity ‘state of being sane’.

38 | Educational beginnings

ous that the original meaning, with which the notion of truth was associated, in fact, does not exist (Grygiel and Kleparski 2007: 14; Haßler 2006: 594). Donald drew a dividing line between the senses that had never been used in the language from those which had. The former were indicated by label “lit(erally)” and the latter by “orig(inally)”, for example: bar “lit. a branch”, behold “orig. to hold”, brag “orig. to crack, to make a noise”, wand “lit. a shoot of a tree”, want “lit. the state of having waned or being deficient”. Thus, senses which were indicated by “lit.” were merely hypothetical constructs based on etymology. As the editor explains, this label indicates that the meaning is “according to the letter, though it is not now, and may never have been, in use in the language”, while label “orig.” shows that “the meaning following once existed, though now obsolete” (Ch-A: vii). Inconsistency with regard to the use of the above labels was unavoidable. However, the consequences of inconsistent labelling may have had some implications on the understanding of a word by a student, who may have interpreted such etymology-driven meaning as synonymous with the current or correct usage of the word. This so-called etymological fallacy may have arisen when Donald failed to indicate the status of a primary meaning by label “lit.” or “orig.”, implying that the meaning was current in the language (e.g. breakfast “a break or breaking of a fast”). Donald’s practice of defining morphologically simple words differed from that of complex and compound formations, especially when the latter were semantically transparent. With regard to morphologically simple words, which were morphologically opaque, the editor defined them by recourse to “root-ideas”. Since the resulting etymological definitions were often at odds with the actual usage, Donald marked them with label “lit”. Also words composed of foreign elements that were unanalysable in English were treated in this way. Thus, persuade was defined as “to advise thoroughly” because the word was traced back to the Latin persuadeo, with the morpheme per- meaning “thoroughly” and suadere meaning “advise”. By contrast, derivatives and compounds of transparent structure were defined in terms of their morphological elements, rather than ultimate etymons, for example: madhouse “a house for mad persons”, madman “one who is mad”, madden “to make mad”, persuasible “capable of being persuaded”, persuasion “act of persuading”, persuasive “having the power to persuade”. Thus, by means of primary definitions the meanings of these words were related to the basic words from which they had been derived. This was in line with Passow’s model of lexicography as presented in his Über Zweck, Anlage und Ergänzung Griechischer Wörterbücher, published in 1812 (Zgusta 2006: 27). In Passow’s plan of the dictionary, each derivative was defined in terms of the basic word (ibid.). In the definitions quoted above, which can be classified as a morpho-semantic one, the meanings correspond to the actual usage in a fairly straightforward way. With regard to partly opaque items, such as cranberry and strawberry, definitions were constructed by reference to ultimate etymons, even though the etymological information was tentative or hypothetical:

Meaning and etymology in Ch-A | 39

cranberry ... a red, sour berry growing on a stalk resembling the legs and neck of a crane, much used for tarts, &c. strawberry ... a plant and its berry or fruit which is highly esteemed – prob. so called from its strewing or spreading along the ground.

In both definitions the etymological information motivates the morphological structure of the headwords, and provides a plausible explanation of their meanings, even though this information is not, using cognitive terminology, a well-entrenched part of the headword meanings. For example, in the first definition above, the seeming resemblance of the stalk of a cranberry to the legs and neck of a crane is mentioned for the sake of the immediate applicability of etymology to meaning, rather than for the sake of the description of the concept shared by English-speakers18. Likewise, in the second definition the idea of strewing, which is the meaning of Old English streaw, the ancestor of Modern English straw, is used here in order to corroborate the indivisibility of etymology and meaning19. Donald’s approach to primary definitions fits well into the historically oriented linguistic thought that was developing in nineteenth-century Britain. It should be remembered that Chambers had a particular function to meet in respect of their users, namely to provide them with “the results of the latest philological research” and assist them in “philological study [...] in advanced classes” (Preface: v-vi). Although not all words display a clear etymological motivation, Donald’s overall approach was in accordance with Bréal’s assumption that only insights into a word history can ensure “adequate understanding” of its contemporary meaning (Geeraerts 2010: 11). Similar views were put forward by Trench. The idea that primary meanings served as a logical point of departure for the derivation of secondary meanings was the basic tenet of Passowian lexicography. For Passow, “linguistic derivation [i.e. etymology] gives the primordial, most general meaning, from which the individual applications flow” (Zgusta 2006: 28). The logical derivation of meanings is the only method of establishing senses when language record is patchy and insufficient.

|| 18 There is every reason to believe that few speakers perceive such a resemblence. What is more, few are aware of what exactly the cranberry and the bird crane look like (cf. Durkin 2009: 57-58). 19 Some of the etymologically conditioned definitions had surprisingly long-lasting life in Chambers. They were recorded in later editions of Chambers, which were no longer ‘etymological’ in nature, in spite of the fact that they failed to express concepts shared by English speaking community. Among the long-lasting primary definitions there are breakfast “a break or breaking of a fast”, and illustration “act of making lustrous or clear”, both of which survived in the dictionary in the first position of their respective entries down to 1988. This is despite the fact that the provision of literal meanings (as derived from etymology but unattested in the language record) was, as a rule, abandoned in the nineteenth-century (1898) edition of Chambers.

40 | Educational beginnings

In a larger version of the dictionary (Ch-B), published in 1872, James Donald continued the provision of etymological meanings with root-ideas distinguished in italics. As a rule, the provision of primary (i.e. literary) definitions was abandoned in the edition of 1898 (Ch-C).

5.1.2 Secondary meanings As mentioned earlier, primary meanings were the basis for the derivation of secondary ones: “The primary meaning is given first (in italics), and the secondary meanings in the order of their logical connection with the primary one” (Preface: v). Donald further explained that the secondary senses that were nearer the literal ones were placed first (Preface: vii). The plan presupposed logical relations between all meanings, and the central position of etymological meaning (whether attested in the language record, reconstructed or hypothetical), to which other meanings related. The plan resembled that of Richardson’s in that it placed etymological meaning in the centre of the lexicographic attention. However, unlike Richardson, Donald did not claim that all meanings of a word spring immediately from the etymological meaning. From the quotation mentioned above, it seems that he was aware that some secondary meanings may be related more to the primary one, while others may be less so. In his actual practice, he attempted to sequence senses in a logical order in respect of the primary one, though the logical link with the primary sense was not always clear. For example, in the entry for anchor “lit. that which has an angle; a heavy iron instrument to hold a ship afloat; fig., what gives safety or security”, the last, figurative sense is more related to the second sense than, if at all, to the primary one. The abstract safety sense of anchor is a metaphorical extension of the nautical sense, and is motivated by the fact that the heavy metal object prevents a ship from drifting out to sea rather than by anything that has an angle. It is then clear that the etymological meaning is not present in all meanings of a word. This can be observed in Richardson’s dictionary as well, contrary to what Richardson claimed in the preface (Zgusta 2006: 24). In effect, like Richardson, Donald ended up with a roughly chronological order. A selection of entries from Chambers is given below. The same order is followed in the historical OED, with the exception of the entry crane20. crane ... the bird that croaks or makes the sound cr; a large wading bird ...; a bent pipe for drawing liquor out of a cask, a machine for raising heavy weights

|| 20 In OED the corresponding senses are given in a reverse order: ‘A bent tube, used to draw liquor out of a vessel’ follows the sense ‘A machine for raising and lowering heavy weights’.

Meaning and etymology in Ch-A | 41

cyclopædia ... lit. a circle of learning ; the circle or compass of human knowledge ; a work containing information on every department, or on a particular department of knowledge February ... the month when the ancient Romans offered sacrifices of expiation, because then the last month of the year; the second month fee ... lit. cattle or money; a grant of land for feudal service : an estate inherited : recompense ; price paid for services, as to a lawyer ham ... the hind part or inner bend of the knee; the thigh of an animal, esp. of a hog salted and dried harbour ... lit. and orig. a lodging, station for an army; any refuge or shelter; a port for ships pencil ... lit. a little tail; a small hair-brush for laying on colours; any pointed instrument for writing or drawing, without ink wand ... lit. a shoot of a tree; a long slender rod; a rod of authority, or of conjurors

The above primary meanings give some idea of unity that holds for the secondary senses. The logical relations among senses are not given explicitly: the user is supposed to infer them from the definitions. This parallels the problem in Richardson (Zgusta 2006: 24), in which quotations were collected but rarely commented on and compared. The choice of secondary senses was selective to suit the pedagogical purpose of demonstrating derivation. Entries for highly polysemous words were treated in a brief and selective way, while some senses that showed a weak link to the etymological meaning tended to be omitted. Nevertheless, this treatment appears to be justifed considering the educational aim of the dictionary. As a result, it seems that Chambers was more persuasive with regard to demonstrating the power of etymology in the derivation of meanings than larger dictionaries were, for example Richardson’s. Several remarks on graphic distinction of senses will be made here. As seen in the entries above, senses were separated usually by semi-colons. Sometimes the editor employed a colon, which he used when separating “different classes of meanings” (Preface: v). These must have been the meanings that were more closely related to each other. The indication of the classes of meanings showed that senses were even more interrelated with each other than it might have seemed. However, from the user’s perspective, the colons hardly performed this function successfully, the reason begin that they could easily be confused for semi-colons. This is the only edition of Chambers in which the editor attempted to group senses into classes.

42 | Educational beginnings

5.1.3 Etymological groups As the editor noted in the preface, “The Arrangement of the words is etymological, while the alphabetical order is strictly preserved by means of references.” (Preface: v). The word etymological above implied that the arrangement was derivational, as the words derived from the same basic word appeared in the same group of entries: Words derived from the same root, but with different affixes, are grouped together, the first word of each group being printed with a capital initial, and the derivatives under it with a small letter, while the groups themselves are separated from each other by a space (Preface: v)

Below is an example of a group headed by the headword mad. What the group actually looked like in its entirety can be seen in Table 8 (boxes A-15 to A-20) in the Appendices. Mad ... madly ... madness madcap madhouse madman madwort madden

It should be noted that all the headwords in the group above constituted a selfcontained entry, with its own pronunciation, definitional and grammatical information. The adoption of the alphabetical principle caused that Donald’s etymological groups gave priority to suffixed derivatives. Thus, the group headed by trust contained trustee, trustful, trustworthy but not distrust and distrustful, the latter (prefixed) derivatives belonging to a separate group. In the passage quoted above Donald uses the word “affixes” in the sense of what we now know as suffixes21. Etymological groups were headed by root-words (such as mad in the group above) or base words (such as distrust) that were the basis for deriving other words. As can be seen in the group headed by mad above, some direct derivatives (madly, madness) were nested just under the headword to emphasise their etymological and semantic affinity. Likewise, boyishly and boyishness were under boyish rather than boy; and persuasibleness, persuasibility were under persuasible. This technique not only showed that words were more related to others but also saved space: no defini|| 21 This accords with his definition of affix: ‘a syllable or letter fixed to the end of a word’.

Meaning and etymology in Ch-A | 43

tion was necessary, as the user could infer the meaning of the nested item by referring to the meaning of the headword: When a word stands after another, with no meaning given, its meanings can be at once formed from those of the latter, by adding the signification of the affix: thus the meanings of darkness are obtained by prefixing the meaning of ness, state of being, to those of dark. (Preface: vii)

Prefixes, which were listed separately in the back matter, were also arranged in etymological groups. For example, dis, bis, de, dia, dis, dys were in the group headed by the capitalised Dis, on the grounds that they were traceable to the Latin or Greek form duo, meaning “two”. Suffixes were listed separately under a title “Affixes”, but unlike prefixes, they were not arranged in groups but in strict alphabetical order. The arrangement of entries employed in Ch-A set it apart from most other dictionaries. Compared to the American dictionaries such as Webster-Mahn and Worcester, Donald’s arrangement of entries was extraordinary, though not original, as we shall see later. Both American dictionaries granted transparent derivatives and compounds main entry status, according to the alphabetical order, the result being that etymologically related items were sometimes far away from each other. Also Imperial, which was modelled on Webster, demonstrated such an arrangement. As mentioned earlier, Ch-A was closer to Richardson’s, where such words as madden, maddish, maddingly, madly, madman, madness were clustered under mad. However, by the time Ch-A was brought out, the idea of arranging entries into etymological groups had already been implemented by Alexander Reid in his Dictionary of the English Language (1844), a work of a similar structure and pedagogical function as that of Ch-A. Intended for school students, Reid’s dictionary provided, under a base word, entries for derivatives and compounds, each of which with information on pronunciation, part-of-speech, and one or more definitions. The typography of headwords differed from that of Ch-A: each headword in a family group was printed with an initial capital, the first headword being printed with a somewhat larger typeface, while all other headwords with a much smaller one. Reid’s dictionary was an extremely popular book, reaching the 17th edition in 1863, the edition which may have served Donald as a model for his dictionary. In the remaining part of this section, we will make some remarks from the user’s point of view on the presentation of data and the consequences it had on the retrieval process. Firstly, etymological groups were not easily distinguishable on a page. Although they were “separated from each other by a space” (Ch-A: v), the intervening space was hardly distinguishable from spaces that separated entries. Furthermore, group boundaries were rather blurred, for the simple reason that all headwords in a group were equally protruding, as they were left-justified, giving the impression of being unrelated. Indeed, each of the headwords in a group provided an access point to a self-contained entry, with pronunciation, definitional and

44 | Educational beginnings

sometimes extra etymological information, but a full understanding of how a word related to other words required that the user be aware of the position of the word in the group. One typographical feature that made it possible for the user to identify the beginning of an entry group was the capital initial letter of the first entry-word (for example Mad in the group above). Nevertheless, because of the fact that etymologically related words figured as part of the central word-list, with many of them being thrown out of their alphabetical place, the whole process of retrieval of vocabulary was far from straightforward. Secondly, although the overall word-list was alphabetically arranged, an obvious consequence of etymological grouping was that alphabetisation was often disturbed, for example, the group headed by Mad (quoted above) was closed by madden, the latter being followed by Madam. Looking inside the group, one can see that madword preceded madden. In order to make up for these shortcomings, in the word-list there were extra entries that referred the user to the right entries, like “Madhouse, Madness. See under Mad”. Hence, madhouse and madness were not only entered in the group headed by mad but also appeared as a separate entry (as shown above) in the alphabetical place, after Mademoiselle and before Madonna. This technique of cross-referencing has been widely used in recent editions of Chambers as well as other dictionaries in general. Yet a major pitfall of crossreferencing is that it results in a double look-up. To a certain extent, the problem was alleviated by a relatively small coverage of vocabulary (about 25, 000 entries) and the fact that many cross-referring entries were almost on the same page as their target entries. In sum, the etymological groups in Chambers fit into Passow’s model of lexicography: words which derived from the same root were grouped in the entry for the basic word, with different senses flowing logically from the primordial meaning (Zgusta 2006: 27).

5.1.4 Etymology The etymological part of the dictionary “has been prepared with the greatest care, and will be found to embody the very latest researches into this most interesting subject” (Ch-A, Preface: vi). As mentioned earlier, Donald’s understanding of the concept of etymology was in accordance with the philological interest in tracing the development of a word, from its contemporary form, through intermediate stages, to its origin: “the derivation of every word (so far as has been discovered) is given, each word being traced back, step by step, to its ultimate source, and the meaning of each foreign word distinctly told” (ibid.). The theoretical tenets were praiseworthy but difficult to put into practice, especially in a small dictionary like Ch-A where space constraints did not allow for a detailed treatment of etymology.

Meaning and etymology in Ch-A | 45

The depth of etymological treatment varied from entry to entry, ranging from the indication of the source language, as in pert “W.”, the source language followed by one or two forms in this language, as in wane “A.S. wanian – wana”, several cognate languages and forms, as in boy “Ger. bube ; Swiss, bub, bue ; L. pupus”, to the immediate and ultimate source language and form, followed by English translations “Fr. madame – ma, L. mea, my, and Fr. dame, L. domina, lady”. Typically, however, the editor tended to provide translations of ultimate etymons because they were essential for the wording of primary definitions. Etymologies were given between square brackets, at the end of the entry so as to allow “the speedier discovery of the meanings” (Ch-A: vi). As Donald admits, the positioning of the etymological part at the end, rather than the beginning of the entry, was unusual in lexicography at the time (Ch-A: vi). Indeed, in major dictionaries such as Bailey, Johnson (as revised by Todd), Imperial, Richardson, Webster, and Worcester, etymology was given a prominent place, at the beginning of the entry. Given the fact that entries in Ch-A were short, the decision to place the etymological part at the end of the entry probably did not affect the speed of retrieval of this information. Etymologies were normally provided in the entries for capitalised entries, that is for etymological basis. As the editor explained, “When no etymology follows a word, this implies that its derivation is given above, under the chief word of which it is a derivative” (Ch-A: viii). In addition, however, extra etymological information appeared with small-initial entries, explaining further the derivation of a complex word or a compound, for example at bracelet “Fr., dim. of old Fr. brachel, armour for the arm”, madcap “Mad, and Cap”, madwort “Mad, A.S. wurt, plant”, sanable “L. sanabilis―sano, -atum, to heal” (see boxes A-6, A-16, A-19, A-32, respectively). The sources of etymological information have already been outlined earlier. One of them was Webster-Mahn (1864), to which we return now. A brief comparison of a few entries will show that Donald might have referred to that dictionary when working on etymology, though it was not, if it was at all, an exclusive source of this information (see the entries below). It might be that Donald drew on Webster-Mahn at such entries as boy, bracelet, or wander but at other entries, such as brace and pertain, he must have referred to other sources, as the entries contain information not recorded in Webster-Mahn. boy ... [Ger. bube ; Swiss, bub, bue ; L. pupus.] (Ch-A) boy ... [... N. H. Ger. bube, ... Cf. Lat. pupus, boy, child ...] (Webster-Mahn)

brace ... [Fr. bras, a brace, the arm, power ; Gael. brac, W. braich, L. brachium, Gr. brachiōn, the arm, as the type of strength.] (Ch-A)

46 | Educational beginnings

brace ... [O. Fr. brace, brase, ... Sp. & Pg. braza, fathom, from Lat. brachia, the arms ..., pl. of brachium, arm] (Webster-Mahn)

bracelet ... [Fr., dim. of old Fr. brachel, armour for the arm.] (Ch-A) bracelet ... [Fr. bracelet, dim. of O. Fr. brachel, armor for the arm, ...] (Webster-Mahn)

pertain ... [L. pertineo—per, thoroughly, and teneo, to hold, akin to Gr. teinō, Sans, tan, to stretch.] (Ch-A) pertain ... [Lat. pertinere, from the prefix per, through, and tenere, to hold, keep; It. pertenere, Pr. pertener, Sp. pertenecer, Pg. pertencer.] (Webster-Mahn)

wander ... [A.S. wandrian, Ger. wandern, to wander ; old Ger. wantalon, to move to and fro; probably allied to Wend, Bend.] (Ch-A) wander ... [A-S. wandrian, ... N. H. Ger. wandern, ... to wander; ... O. H. Ger. wandalôn, wantalôn ... to be changed] (Webster-Mahn)

By and large, the approach adopted in Ch-A is similar to that in Webster-Mahn with regard to the type of etymological information provided. Donald attempted to show, in a broad outline and within the limits of space available, how a word developed over time by indicating a selected source and/or cognate forms during intermediary stages. Although given in a brief way, the depth of treatment surpassed that of other dictionaries of comparable size, like, for example, Reid’s A Dictionary of the English Language (1844), in which etymological treatment was limited to the indication of the ultimate language and form; and the treatment in Ch-A was even more elaborate than in much larger Boag’s A Popular and Complete English Dictionary (1848). Considering etymological data from the user’s perspective, it should be emphasised that the interpretation of this information could pose a problem for the student, especially when the front matter gave little help in this regard. Although the editor used the em dash “―” to introduce source forms (e.g. pertain “L. pertineo―per...” (A-27)), he did not explain it in the front matter. At times, he used a more descriptive metalanguage: gibe ‘from root of Gabble’ (A-11); want “... from root of Wan, Wane.” (A-39). Likewise, when introducing cognates, he used a selfexplaining metalinguistic expression “akin to”, for example: pertain “... teneo, to hold, akin to Gr. teinō, Sans, tan, to stretch” (A-27); sane “L. sanus, akin to Gr. saos, sōs, sound” (A-31). However, cognates were also introduced, rather inconsistently, by either commas or semi-colons:

Meaning and etymology in Ch-A | 47

brace ... [... ; Gael. brac, W. braich, L. brachium, ...] (A-4) giddy ... [A.S. gydig, Gael. gadoch, giddy ; Norw. gidda, to shake.] (A-13)

mad ... A.S. gemœd ; It. matto, silly ; L. mattus, drunk (A-15) wander ... A.S. wandrian, Ger. wandern (A-37)

The rather indiscriminate use of punctuation marks may have confused the user. It was not until Ch-C that the editor addressed that problem by introducing cognates more systematically by semi-colons, and instructing the user on the correct interpretation of etymology in the front matter. Apparently drawing on available sources, the editor provided hypothetical etymologies using the wording “prob.” and “perh.”, and alternative etymologies, using the conjunction “or”. Sometimes doubtful etymologies were indicated with a somewhat less transparent question mark “?”. When referring to the Old English period, the editor normally used a traditional label “A.S.” (Anglo-Saxon). However, in the dictionary there was at least one entry, in which the editor indicated that period in a less conventional way: brandy: “[old E. brandwine...]” (the label was retained in this entry in Ch-B). Yet it should be stressed that the above pitfalls were nothing extraordinary in lexicography of the nineteenth century. Dictionaries were demanding for their users. But it seems that Donald’s chief concern was not with the way how information was to be presented but rather what type of information was to be offered for the student of etymology. Summing up To round up our discussion on meaning and etymology in Ch-A, it should be emphasised that the etymological approach in this edition was in tune with the historical orientation of nineteenth-century linguistic thought and lexicographic tradition. The aim of the dictionary was to show how lexical items were related by virtue of etymology. The approach adopted in Ch-A fits into what John Considine calls the “genealogical model of lexical relationships”, the model that became popular in nineteenth-century linguistics and that had a profound influence on the construction of the OED (Considine 1996: 367). According to this model, words belong “to families, in which one word begets another, or several others, whose relationships can be traced, not least by their striking family likenesses, and referred to a common ancestor” (ibid.). Likewise, on the level of a polysemous word, its various meanings can be seen as developed from the etymology which provides a “common thread” that “unites” them all (ibid. 366). In Ch-A lexical items which shared a common ancestry were put in groups, or, to use a genealogical term, families. The “root-

48 | Educational beginnings

idea”, that is the meaning of the ultimate etymon, provided the logical origin for secondary meanings, and a common semantic denominator for all words belonging to the family. Donald did not intend to show a detailed picture of such relationships, for this would have required much more financial support and editorial effort, but rather to show the gist of the idea22, in broad outline, and within reasonable limits. The above genealogical model appears in the Proposal for the Publication of a New English Dictionary by the Philological Society, which was published in 1859, eight years before Ch-A, and subsequently endorsed in the dictionary which is now known as OED (Considine 1996: 367). As could be seen in Chapter 3, the model had been popularised by Trench, a member of the aforementioned Philological Society, in his book On the Study of Words, first published in 1851. It should also be pointed out that it was a German scholar Passow who in 1812, i.e. half a century before Donald, had planned a dictionary that was to display etymological, i.e. derivational, relations between words and meanings (Zgusta 2006: 27). Passow’s dictionary, which was published in 1831, became a model for Liddell and Scott’s Greek-English Lexicon (ibid.), which in turn could serve as a guide for Chambers. The model endorsed in Ch-A may seem to be similar to that in Richardson’s, because in both dictionaries etymology and meaning form an indivisible whole. However, unlike Richardson, Donald took advantage of comparative-historical philology, which was based on a systematic comparison of form, rather than superficial resemblance of form and association of meanings. In line with Trench’s pedagogy, the dictionary drew the student’s attention to the role of etymological meaning in the derivation of words and meanings. To this end, the editor identified the ultimate etymon, whose meaning was assigned the prominent position in the entry. It seems, however, that the linear order of presentation of senses and their separation chiefly by semi-colons were a hindrance rather than an aid in showing how the senses related to each other and to the primary one. But this is understandable, given the limited space available and the lexicographic convention of presentation of meanings. In fact, the first dictionary whose entry structure showed genealogical relationships in a quite satisfactory way was yet to come: the monumental NED, with branch numbering and a range of sense levels. Colons which were used in Ch-A to distinguish classes of senses gave only a mere gist of this idea. With this regard, Ch-A followed the time-honoured tradition: available dictionaries, like Johnson and Richardson, indicated the etymological meaning and arranged senses (or only quotations in the case of Richardson) using one-level, roughly chronological order. The lesson to be learnt by the user was that etymology and meaning are interconnected. However, the danger of such an approach may be that the user may have the impression that the etymological meaning is synonymous with the correct or || 22 This point has been suggested to me by John Considine to whom I owe a debt of gratitude.

Meaning and etymology in Ch-A | 49

current usage. This so-called etymological fallacy can arise when the student holds too much belief in the power of etymology and neglects semantic change as a principle of historical-comparative methodology. Donald’s emphasis on etymological meaning and the virtual neglect of how senses of a polysemous item are interrelated with each other could shift the student’s attention from the evolution of forms and meaning, which is the basic aim of the historical-comparative reconstruction, to the problem of naming, that is the appropriateness of current forms and meanings. However, this certainly was not the editor’s intention. Nevertheless, the value of Ch-A rests in its early attempt to popularise among students and the general public a “genealogical model” of language and to provide them with some hints on the methodology of historical-comparative research.

50 | Educational beginnings

5.2 Other descriptive and pedagogical considerations in Ch-A 5.2.1 The selection of vocabulary The editor boasted that the dictionary contained “every English word [!], with the exception of obsolete and very rare words, and technical terms not found in general literature” (Ch-A: v). As a dictionary intended specifically for schools, it included “words in Natural History, Botany, Geology, Physics, Physiology, and other sciences, which of late have become more commonly the subject of study in schools” (ibid.). It also included “all the obsolete words that occur in the Bible” (ibid.). The interest in the Bible was cultivated on a daily basis in Scottish schools, church and at home. The following words in the Biblical senses were found in sample 2: giereagle “in B., a species of eagle” (A-14) and persuade “lit. to advise thoroughly, so in B” (A-22). In the 1872 reprint of Ch-A, the editor added “obsolete and rare words in the Apocrypha” and other religious texts and “the obsolete and rare words and meanings in Milton’s poetical works” (Ch-A: vi). In addition, the extra lists at the back of the 1872 version of Ch-A included “Words and Phrases from the Latin, the Greek, and Modern Foreign Languages”. All these additions were to assist the student in reading classical texts. The usefulness of these words for the student was evident, as grammar schools laid emphasis on teaching Latin and Greek. The knowledge of classical literature and languages was highly valued in the Victorian age. It was considered as the aim of education, even though its practical utility was doubtful for members of lower classes (Blake 1996: 274).

5.2.2 The selection of morphological forms This section provides an overview of morphological forms selected for inclusion in the dictionary. Most of the forms were given main entry status and thus can be found in sample 1. They include, apart from derivatives and compounds, various morphological constructions such as inflected forms, morphemes, abbreviations, and combining forms. Some other morphological forms, which were treated within entries, can be found in sample 2. All of them will be briefly discussed below. As a result of etymological grouping, derivatives and compounds constituted the majority of all headwords: approximately 60 % (11% of compounds and 49% of derivatives). Among them there were self-explaining items which nevertheless were provided with definitions, for example, subject nominalisation with a common suffix -er and –or: getter “one who gets”, gesticulator “one who gesticulates”, gilder “one whose trade is to gild”, personator “one who personates”, waiter “one who waits”, warrior “one engaged in war”; adjectives with the suffix –less: branchless

Other descriptive and pedagogical considerations in Ch-A | 51

“without branches”, sapless “not juicy”; possessive adjectives: sandalled “wearing sandals”; and verbal nouns with the suffix –ing, for example getting “act of getting...”. That the above transparent forms were provided with definitions can be justified by the educational function of the dictionary. Separate entry status was given to zero-derivatives, e.g. wanton adj. and wanton n.; and walk v.i. and walk n. Derivatives and compounds selected for inclusion were written either solid or with a hyphen. With the exception of few phrasal verbs, which were treated in the microstructure of the relevant entries, the editor rather excluded multiword expressions written with a space (e.g. fall from grace, hard money). This shows that the lexicographic description hardly went beyond the word level, which is not surprising given the nineteenth-century interest in the internal structure of words. With regard to inflected forms, “exceptional plurals” (Ch-A: v), which could be a challenge for the user, were treated within the entries for their base forms, for example, phenomena, phalanges. Occasionally, on the word-list one may find regular plural nouns, for example gimbals in its nautical sense: “two rings for suspending the mariner’s compass...” but the lexicographer’s decision to select that particular form was apparently motivated by his intuition that the form prevailed in English texts23. Apart from verbs showing irregular patterns of inflection (e.g. made, sang), which were typically treated as main entries, regularities were also the object of lexicographic description. Following Webster-Mahn, the editor provided present and past participles of verbs. In Ch-A, they appear systematically as run-ons at the end of respective verb entries, for example: bracing, braced (A-4); bracketing, bracketed (A-7); gibing, gibed (A-11); maddening, maddened (A-20); persuading, persuaded (A-22); pertaining, pertained (A-27); wandering, wandered (A-37); waning, waned (A38); wanting, wanted (A-39); wantoning, wantoned (A-42). While Donald consistently provided the user with present and past participles, he was more selective with regard to comparative and superlative forms of adjectives. In sample 2, we can find a comparative and a superlative of mad “comp. madder; superl. maddest” (A-15) but we do not find respective forms for other regular adjectives such as giddy (A-13), sandy (A-30), or sane (A-31). Apparently, such information was limited to words that could pose a problem for the user, even though they showed a regular paradigm (e.g. a double consonant in madder and maddest). Bound morphemes and combining forms figured on separate lists in the back matter. Like entries in the main body, they were provided with definitions. As mentioned earlier, they were also arranged in etymological groups. Prefixes (e.g. dis-, de-, in-, un-, re-) and initial combining forms (e.g. contra-, counter-, hypo-, super-, inter-, intro-, meta-) were on the list of “Prefixes”, while suffixes (e.g. -able, -ate, -en,

|| 23 Indeed, the choice of the form is corroborated by the evidence in OED where only plural form of the word is recorded in all citations of the word in its nautical sense.

52 | Educational beginnings

ing, ish, -ism, ist, -tude, -ty-, -ure) were on the list of “Affixes” (!). Donald’s use of the term affix as a term for a morphological element that we now refer to as suffix was more or less consistent with his definition of affix: “a syllable or letter fixed to the end of a word”. This definition was modelled on that of Webster-Mahn, apparently reflecting the way the word was commonly understood24. It was not until the publication of Ch-C (1898) that the heading “Affixes” was replaced with “Suffixes” at the back of the book. Abbreviations were shown on the “List of Abbreviations” in the back matter. They included initialisms, for example: B.A. “Bachelor of Arts”, B.C. “Before Christ”, F.C. “Free Church of Scotland”, F.R.G.S. “Fellow of the Royal Geographical Society”; and clipped forms, for example: Br. “Brother”, Chap. “Chapter”, Colloq. “Colloquially”, d. “died”, p. “page”, Mad. “Madam”. Summing up, Donald selected a wide range of morphological forms. Even though many of them were semantically analysable, their inclusion in the dictionary fits into the framework of the dictionary that aims at demonstrating the “etymological relations of words”. Furthermore, their inclusion, though drawn on WebsterMahn perhaps indiscriminately, is justified by the pedagogical function of the dictionary.

5.2.3 Defining meaning The editor pointed out that “The greatest care has been taken to express the meanings in the simplest language, a feature in respect to which this Dictionary will bear favourable comparison with any similar work” (Ch-A: v). Like many other British dictionaries compiled at the time (Hanks 2006: 189), Ch-A was under the influence of American lexicography which flourished in the nineteenth century. Thus, it may be useful to compare Ch-A with the most prominent representatives of the period: Webster-Mahn (1864) and Worcester (1860), taking them as a starting point for the discussion of the definition style. A comparison of Donald’s definitions with the corresponding ones in the American dictionaries shows obvious similarities. The overlapping parts are highlighted in Table 2 below. The exact wording of definitions testifies to the fact that Donald drew extensively on both or either of the sources. Entries A-21, A-23 and A-34 may indicate that Webster-Mahn was copied more extensively than Worcester.

|| 24 By the time Ch-A was published, the modern understanding of the word affix as an umbrella term for prefix and suffix had probably not been widespread in the language, as the earliest attestation of such a meaning in OED is dated 1865.

Other descriptive and pedagogical considerations in Ch-A | 53

Table 2. A comparison of definitions in Ch-A, Webster-Mahn, and Worcester, based on the selection of entries from Sample 2.

Ch-A (1867)

Webster-Mahn (1864)

Worcester (1860)

(A-1) Boy ... a male child ; a lad

Boy ...A male child ... ; a lad.

BOY ... A male child

(A-2) boyhood ... state of being a boy.

Boyhood ...The state of a boy ...

BOYHOOD ... The state of being a boy

(A-3) boyish ... belonging to or like a boy ; trifling

Boyish ... Belonging to a boy; ...; trifling

BOYISH ...1. Belonging to a boy...

(A-6) bracelet ... a piece of defensive armour for the arm ; an ornament for the wrist

Bracelet ... An ornament for the wrist, ... A piece of defensive armor for the arm

BRACELET ... An ornament for the arm ... A piece of defensive armor anciently used to cover the arm;

(A-9) Gibbose, adj. humped ; having one or more elevations

Gibbose ... Humped, ...—said of a surface which presents one or more large elevations.

GIBBOSE ... Humped, protuberant ; gibbous ; convex ; swelling.

(A-11) Gibe, v.t. lit. to wry the mouth ; to mock ; to taunt. —v.i. to cast reproaches ; to sneer : — (...) —n. a scoff or taunt ; contempt.

Gibe ...To cast reproaches and sneering expressions;

GIBE ... to scoff; to jeer; to mock ; to sneer at; ...; to taunt;

(A-15) Mad ... furious with anger : disordered in intellect ; insane

Mad ... Disordered in intellect;... insane ... furious with anger

MAD ... Disordered in the mind; ... insane; ... furious

(A-20) madden ... to become mad ; to act as one mad

Madden ... To become mad ; to act as if mad

MADDEN ... To become mad; to act as mad.

(A-21) Madam ... a courteous form of address to a lady

Madame ... a complimentary or courteous form of address to a lady...

MADAME ... a title of respect for a married lady.

(A-22) Persuade ... so in B.;

Persuade ... To influence by

PERSUADE ... to influence, or

.... to scoff at ; ... to taunt ... An expression of censure mingled with contempt; a scoff ...

54 | Educational beginnings

to influence successfully by argument, advice, &c. ; to convince ; to prevail on

argument, advice, ... To convince by argument, ...

prevail upon, by argument, advice, ...to induce; ... to convince.

(A-23) persuasible, adj., capable of being persuaded.

Persuasible ... Capable of being persuaded ...

PERSUASIBLE ... That may be persuaded.

(A-24) persuasion, n., act of persuading ; state of being persuaded ; ... ; a creed

Persuasion ... The act of persuading ; ... The state of being persuaded ...; a creed

PERSUASION ...The act or the power of persuading... The state of being persuaded ... creed

(A-28) pertinent, adj., pertaining or related to a subject ; fitting or appropriate.

Pertinent ... Related to the subject or matter in hand ; fitted or appropriate ...

PERTINENT ... Relating; ...; pertaining. ... Related to the matter in hand

(A-34) sanatory, adj., healing ; conducive to health

Sanatory ... Conducive to health; healing

SANATORY ... Tending to promote health; healing;

(A-37) Wander ... to ramble with no definite object ; to go astray, lit. or fig. ; to leave home ; to depart from the subject ; to be delirious

Wander ... To ramble here and there without any certain course or object in view; ... To go away; to depart; to stray off ; ...; to go astray. ... To be delirious ; ...

WANDER ...To go without any certain course or object ; to ramble ; ... to go astray ; ... To be delirious ...

(A-41) Wanton ... moving or playing loosely ; ... wandering from rectitude ; licentious : running to excess ; unrestrained ...

Wanton ... Moving or flying loosely; ... running to excess ; loose ; unrestrained. ... wandering from moral rectitude ; licentious;

WANTON ... Wandering ; flying or moving loosely ... Sportive ; frolicsome ; playful. ...Dissolute ; licentious; lewd; ... Loose; unrestrained

(A-42) wanton ... a wanton or lewd person ; a trifler. ; ... to ramble without restraint ; to frolic ; to play lasciviously :

WANTON 1. ... a trifler ... A lewd person; a lascivious man or woman. ... To rove and ramble without restraint ... to play loosely; to frolic

WANTON ... A lewd or lascivious man; ... a trifler ... To play or sport lasciviously

The entries above show that entire definitions or parts were taken virtually unchanged, with numerous difficult words retained (e.g. “elevations”, “cast reproaches”, “to ramble”, “rectitude”, “licentious”, “lasciviously”) in spite of the editor’s claim that the defining language was simple. On the other hand, in sample 2 there are entries where Donald modified the source definitions, apparently to suit the needs of a different group of users and/or

Other descriptive and pedagogical considerations in Ch-A | 55

to meet the requirements of space constraints. Taking Webster-Mahn as a point of reference for the current comparison, one may see that Donald occasionally preferred a more common word than the one in the source dictionary, e.g. bracing “giving strength” (A-5) instead of “Imparting strength”; giddy “that causes giddiness” (A-13) instead of “Promoting or inducing giddiness”. One of Donald’s defining strategy consisted in generalisation. For example, in the entry for gier-eagle (A-14), he used label “in B.” to generalise over a specific book of the Bible, “in Leviticus xi. 18”. Another example can be found in the printing sense of brace (A-4), where a hyperonym “mark” is used instead of a semantically narrower expression “A vertical curved line”. Also the information about the phases of the moon at gibbous is given in more general terms: “convex, as the moon when nearly full” (A-10) instead of “convex; as, the moon is gibbous between the quarters and the full moon”. An extreme form of generalisation may be the use of the abbreviation “&c.”, which not only replaced the specific information in the source dictionary but also indicated that the class of objects was not exhausted, e.g. sandwich “two slices of bread with ham, &c. between” (A-29) instead of “...with a thin slice of ham or other salt meat between them”. The strategy by generalisation brought an obvious benefit for the lexicographer, namely that of saving space. Also for the purpose of saving space, the editor deleted articles, especially the definite ones, before nouns, as in persuasion “act of persuading” (A-24) (instead of “The act ...”), and “state of being persuaded” (A-24) (instead of “The state ...”), though the change of this kind was not carried out in a regular way (cf. A-12). Some changes made the information more explicit by explaining what otherwise might have been less obvious to the student, for example giblets “the internal eatable parts of a fowl, cut off before cooking it” (A-12) instead of “Those parts of a fowl which are removed before cooking”. Yet in some other entries, one may hardly say that the changes lead to a better understanding of the definitions by the student; for example, the omission of the object of the verb (“any thing”) certainly does not make the definition more accessible to the student: brace “anything that draws together and holds tightly” instead of “That which holds any thing tightly” (A-4). By and large, while Ch-A demonstrates certain attempts at simplification of definition language, the changes are rather minor and irregular. Ch-A retains the characteristic features of the source dictionaries, including the traditional style of definitions. This is also true for all subsequent editions. Below we will discuss typical features of the definition style in Chambers. Definitions in Ch-A were brief, typically in the form of one-phrase analytical phrase and a string of synonyms. As seen in Tables 5-1 and 5-2 (see Chapter 9), synonymous definitions account for a very large proportion of definitions (38%). Morpho-semantic definitions, formulated by reproducing morphological elements, represent a relatively high percentage of definitions (18%), compared to the data from the following editions. This type of definition was typically used to express primary meaning. In Ch-A (and Ch-B), the tactic by reproducing morphological elements was

56 | Educational beginnings

also applied to entirely or partly opaque items as a method of establishing the etymological meaning. A pure morpho-semantic definition25 appeared in the entries for such items as carriage “act, or cost of carrying”, strawberry ‘a plant and its berry or fruit which is highly esteemed –prob. so called from its strewing ...”, cranberry “a red, sour berry growing on a stalk resembling the legs and neck of a crane ...”, and honeymoon “the honey or sweet moon or month, ...”. In Ch-A, as well as all subsequent editions, defining words and phrases were typically used in a series. They were typically separated by commas (e.g. giddy “unsteady, dizzy”) or the conjunction “or” (e.g. gibe “a scoff or taunt”). Both devices were frequent; the conjunction “or”, for example, appeared in 20% of definitions. In analytical definitions a string of coordinated words (and phrases) was sometimes followed by “&c”, i.e. a marker of the open class of objects (e.g. persuade “to influence successfully by argument, advice, &c.”). It is somewhat surprising that in the sample from Ch-A (Tables 5-1 and 5-2), there is a complete lack of words indicating prototypical elements, such as “especially”, “e.g”, “usually”, “typically”. This finding may suggest that Donald had a tendency to avoid narrowing of meaning. Since these key-words require specification of central features of the category, it might be that he avoided them for lack of space. Finally, let us briefly survey the strategies that the editor employed when defining major word-classes: nouns, verbs, and adjectives. As a rule, definitions conformed to the traditional principle of substitutability, which means that definitions of transitive verbs did not include the verb objects (e.g. madden v.t. “to make mad”, persuade v.t. “to advise thoroughly”). Intransitive verbs were defined either by other intransitive verbs, as in wander “to ramble”, or by transitive verbs followed by their objects, as in gibe, v.i. “to cast reproaches”. The principle of substitutability required that definitions of verbs include necessary prepositions that complemented the defining verbs, as in persuade “to prevail on”, pertain “to relate to”, want “to be destitute of”. In the definitions of nouns, the editor typically premodified the genus noun by an adjective (e.g. boy “a male child”); postmodified the genus noun by a prepositional phrase with a gerund (e.g. want “state of being without”) or by a relative clause (e.g. want “absence of what is needful...”). Adjectives were defined by other adjectives (e.g. giddy “unsteady, dizzy”), by “ing” and “-ed” participial constructions (e.g. boyish “belonging to or like a boy”, bracing “giving strength”, gibbose “having one or more elevations”, gibbous “humpbacked; swelling”, giblet “made of giblets”, mad “drunk”), by prepositional phrases (e.g. breathless “without or out of breath”), and by relative clauses (e.g. giddy “that

|| 25 A pure morpho-semantic definition is here understood as the one which reproduces all morphological elements of the defined item.

Other descriptive and pedagogical considerations in Ch-A | 57

causes giddiness”). Although the above list is not exhaustive, it shows typical defining strategies used in Ch-A and its subsequent editions. The strategies were taken as a given; they were part of a heritage of the traditional lexicography.

5.2.4 Usage labels The system of labels in Ch-A was rudimentary and geared chiefly towards the indication of the semantic status of primary meanings. As seen in Table 6, the total number of labels in the sample is 66, which is much lower in comparison with subsequent editions. The system was dominated by lit(erally) labels, which occurred 47 times in the sample. Together with fig(urative) labels, the semantic type constituted the largest proportion of labels (74%). The second most frequent type of label indicated the source in which a sense was recorded: B(ible) occurred 10 times and Pr.Bk. (Book of Common Prayer) once in the sample. Other labels were used occasionally; they represented the temporal type: obs(olete), orig(inally); and the field type: geol(ogy), bot(any), archit(ecture). The presentation of information on usage was rather unsystematic. All labels were italicised except for “lit(erally)” and “orig(inally)”. Some labels, such as B(ible) and those representing the field type, were introduced with the preposition “in”, as in gier-eagle “in B., a species of eagle” (A-14), persuade “in B., to use persuasion or advice”, sap “in bot., the part of the wood next to the bark”. Most labels preceded definitions but some could occasionally be found after them, for example at persuade “to advise thoroughly, so in B” (A-22). Sometimes usage information appeared embedded in definitions with no typographic distinction, for example: brace “a mark in printing connecting ...” (A-4) and bracket “in printing, the marks ...” (A7).

5.2.5 Syntagmatic and paradigmatic information The lexicographic description in Ch-A hardly went beyond the word level. This is indicated by the fact that, as mentioned earlier, the dictionary excluded almost entirely open compounds, granting inclusion to only those compounds which were written solid or hyphened. Apparently the concept of word was understood as a sequence of letters bounded at either end by spaces. Thus, the dictionary provided little syntagmatic information. Typically the syntagmatic information appeared in the form of traditional labels indicating in a rough way a sub-category of the verb: “v.t.” stood for the transitive verb while “v.i.” for the intransitive one (e.g. brace “v.t. to tighten ...” (A-4), bracket “v.t. to put within brackets” (A-7), gibe “v.i. to cast reproaches; to sneer” (A-11)). Information on verb complementation by means of prepositions was not given. The editor occasionally provided phrasal verbs, which

58 | Educational beginnings

were run on to the base verbs, for example: the verbs get at, get off, get on, get through, get up were all run on to get, and make away, make of, make out, make over were run on to make. Incidentally, we may observe how the editor’s method of the presentation of phrasal verbs evolved as the work on Ch-A went on. While Donald cited phrasal verbs with get in full form, e.g. get at, get off, he provided the phrasal verbs with make with a space-saving dash: ―for, ―of, etc. The dictionary did not provide illustrative examples or typical context of use, as they were beyond the scope of the work. However, being focused more on the internal structure of words, it provided examples of words with bound morphemes to illustrate that the morphemes carried particular meanings. The examples were given in the entries for prefixes and suffixes in the back matter, for example: counter “against : as counteract”, dis “two, twice, as dissyllable”, hypo “under, as hypotenuse”, able “fit to be, as portable”, ese “belonging to, as Veronese”, ion “being, state of being, as creation”. On the whole, paradigmatic information was confined to words that were related etymologically, as these belonged to the same etymological groups. Other paradigmatic relations were outside the scope of the dictionary, with no antonyms, paronyms or co-hyponyms found in the sample. However, looking at entries that fall outside the sample range, one can see that the editor occasionally provided antonyms, especially for specialist words, for example: macrocosm ... the great world ; the universe: ― opposed to Microcosm pessimist ... one who complains of every thing being for the worst : ―opposed to optimist

The above findings testify to the etymological scope of the dictionary and reflect the lexicographer’s interest in the presentation of the internal structure of words rather than of their potential for co-occurring with other words.

5.2.6 Pronunciation 5.2.6.1 Notation system Donald intended to represent pronunciation “in the simplest manner possible” (ChA: v). He used respelling, whereby “The correct sound of every word is given by being written anew phonetically, thus obviating the use of a confusing array of marks.” (Ch-A: v). Diacritics were used on vowel letters. Their complete list, together with exemplifying words, was shown in a pronunciation key running across the bottom of each page: fāte, fär ; mē, hėr ; mīne ; mōte ; mūte ; m

n ; then

Other descriptive and pedagogical considerations in Ch-A | 59

Transcription by means of respelling and diacritics was provided for all headwords. This was a feature of all subsequent editions of Chambers. As seen in the key above, Donald employed a limited number of diacritics: the macron on one vowel letter (ā, ē, ī, ō, ū) or on two vowel letters ( ), a single dot (ė), and a double one (ä). In comparison to the systems employed in the leading dictionaries, such as Webster (1864) and Worcester (1860), Chambers employed by far fewer diacritics, which definitely made the system easier for the user to decode. For practical reasons, the complex numerical diacritics as used in pronouncing dictionaries of the eighteenth century by Kenrick, Sheridan or Walker were unacceptable. In these dictionaries numeral diacritics were small and could hardly be distinguished from each other (Landau 2001: 67). What is more, they required the user to refer to the pronunciation key every time he attempted to decode the pronunciation, as various combinations of numbers and vowel letters were difficult to memorise (ibid.). One may say that in terms of the provision of simplified pronunciation, Chambers foreshadowed the trend in popular dictionaries towards simplicity and userfriendliness. The trend was clearly marked at the turn of the twentieth century by the publication of Collins’s dictionaries such as Graphic English Dictionaries (Williams 1903), the Home Dictionary (1907), and the Pocket Dictionary (1912) (Simpson 1990: 1964). However, it took a long time for publishers to respond to the needs of the dictionary users efficiently, especially when the dictionaries were used more and more by a wider public, including less experienced and less competent speakers of English (cf. McArthur 1998: 136). The ideal of a user-friendly dictionary blurred on the horizon. With regard to consonantal sounds, the fricative sounds /&/ and /6were coded by digraphs and, following Walker, distinguished from each other by typography: “th” in italic type coded the voiced sound, and “th” in normal type coded the voiceless one. However, the editor’s explanation of this distinction given in the “Explanations to the student” was confusing, owing to the italic typeface of the word “thick”: No consonant used has any mark attached to it, with the one exception of th, which is printed in common letters when sounded as in thick, but in italics when sounded as in then (Ch-A: vii)

This typographic error persisted in Chambers until Ch-1 (1901), which shows that the information was copied uncritically from one edition to another. Other digraphs included “zh”, which stood for the fricative sound /