Jump to ContentJump to Main Navigation
A History of Roget's Thesaurus$

Werner Hüllen

Print publication date: 2003

Print ISBN-13: 9780199254729

Published to Oxford Scholarship Online: January 2010

DOI: 10.1093/acprof:oso/9780199254729.001.0001

The Topical Tradition in English Lexicography

(p.277) 7 The Topical Tradition in English Lexicography
A History of Roget's Thesaurus

Werner Hüllen

Oxford University Press

Abstract and Keywords

This chapter presents a detailed history of the topical tradition in English and Continental lexicography. It includes a conspectus of John Wilkins's ‘Tables’, that is, the semantic part of his universal language scheme which, in fact, is a thesaurus of words ordered according to philosophical later principles. Around 1700, this topical tradition came to a halt, but started anew within a different framework inspired by the innovative spirit of John Locke. Works discussed in this chapter include Parsigraphie by Jean de Maimieux (1797a, h), which is a new version of Wilkins's universal language scheme, and David Booth's abandoned plan of an analytical dictionary (1835).

Keywords:   topical tradition, John Locke, lexicography, Jean de Maimeux, John Wilkins, David Booth

  1. 7.1. Onomasiology between 800 and 1700

    1. 7.1.1. Lexicographical aspects: macrostructure

    2. 7.1.2. Lexicographical aspects: microstructure

    3. 7.1.3. Historical aspects

    4. 7.1.4. Wilkins’s Tables’: macrostructure and word selection

  2. 7.2. The classification of knowledge

  3. 7.3. Onomasiological lexicography between 1700 and 1852

    1. 7.3.1. Andree(1725)

      1. The preface

      2. Macrostructure

      3. Microstructure

    2. 7.3.2. Booth (1835)

      1. History of the book

      2. The plan of an analytical dictionary

      3. The dictionary as a histroy of ideas

    3. 7.3.3. Maimieux (1797)

      1. Pasigraphy

      2. Pasigraphy Maimieux’s semantic system

      3. Examples

7.1 Onomasiology between 800 and 1700

Roget’s Thesaurus rests on two pillars, the lexicographical traditions devoted to synonymy and to onomasiology.1 The former has been treated extensively in this book; the latter will be surveyed in its final two chapters. As I have already dealt extensively with (p.278) onomasiology in a comprehensive monograph,2 a concise conspectus will suffice for our purposes here.

7.1.1. Lexicographical aspects: macrostructure

Formally speaking, dictionaries are nothing more than the sum of their entries. These entries range from a few hundred (in pocket editions) to many hundreds of thousands (for example in the OED). Dictionaries are not normally meant to be read in toto. They are consulted for information on specific lexemes. Dictionary entries should therefore be arranged in a convenient order. This order is called the macrostructure. The best known is the alphabet: it is easy to use and utterly dependable. Difficulties which stem from homonyms (homographs) are minor. In itself, however, the alphabet is meaningless.

Less well known, but just as well established, is a different, non-alphabetic type of macrostructure with entries sequenced not according to spelling but semantic affinity. Words which have ‘similar’ meanings are placed close to each other. Unfortunately, these dictionaries have been given a plethora of names. They are called ‘ideological’, ‘ideographical’, ‘analogical’, ‘semantic’, ‘thematic’, ‘conceptual’, ‘topical’, or ‘onomasiological’. In spite of some pertinent arguments in favour of one or the other, these names mean the same insofar as they characterize a dictionary type which deviates in its macrostructure from alphabetical order. For reasons given elsewhere (Hullen 1999a: 3–27), I prefer the terms ‘topical’ and ‘onomasiological’ (used here as synonyms) in the present context. Their antonyms are ‘alphabetical’ and ‘semasiological’.

However, the two different macrostructures are not just two ways of doing the same thing, i.e. ordering the entries of a dictionary. They are in fact indicative of two different but natural ways of processing language. Ultimately, they are grounded in the nature of the linguistic sign with its two inseparable aspects, form and content (expression and meaning, signifiant and signifié).

(p.279) The two ways of processing are apparent in language learning. Learners of any age see something, point at it, gesticulate or somehow show that they want to know its name. They go from the meaning of a word—as the most natural unit of the linguistic sign3—to its expression. The same learners may hear a word which they do not understand. In learning its meaning, they go from the form of a sign (word) to its content. When we sit down to write a letter or a paper or to explain something to our interlocutor, we also start from a meaning which is, in some pre-linguistic way (see below), in our minds, with the intention of finding the signs which express it. However, when we listen to a speaker or read a text, we again have to travel from the forms of the words to their meanings. Scientific analyses of linguistic structures can proceed in either of the two directions. We can investigate, for example, how a language expresses invitations or concessions. We can also investigate the meaning, in the same language, of ‘Why don’t you sit down?’ or ‘although’ or negative questions in a conversation.4 In all these cases, one aspect of the linguistic sign is the unknown (the explanandum) whereas the other is the known, which serves as an explanans. If we wish to move from the form to the meaning, the dictionary with an alphabetical macrostructure serves our needs. We readily find the formally familiar, yet semantically obscure, word and are then informed about its meaning. If, however, we wish to move from the meaning to the form, it is the topical dictionary which facilitates this, because its entries are to be found in a pre-existing and, ideally, well-known semantic system.

When engaged in conversation, we take it for granted that we are constantly able spontaneously to change the direction of our language processing, because we alternate between speaking and listening. It is only when difficulties in communication arise that we become aware of this technique. As happens often, linguistic principles become visible in problematic cases.5

(p.280) However, the dichotomy between language processing and its parallel macrostructure in dictionaries is not as symmetrical as it may seem in this chain of lexicographical arguments. Alphabetical order is perfectly straightforward presupposing no more than the knowledge of the alphabet. Topical order, however, is far from straightforward, and never unanimously accepted. It depends on encyclopaedic knowledge of the world and a classification of that knowledge which ultimately rests on philosophical principles. Contrary to the alphabet, it is meaningful in itself, but this mean-ingfulness is more often than not controversial. In consequence, topical dictionaries serve their purposes best in speech communities and in thematic domains where such controversies rarely or never occur, unless the ordering principle is based on generalities that nobody can reasonably challenge, as in certain modern topical dictionaries (e.g. the Longman Lexicon of Contemporary English, 1981 with later editions). The former can be assumed to have been the case in earlier centuries, when our outlook on the world was much more unified by theology and philosophy than it is today.6

Ultimately, the onomasiological principle rests on an epistemo-logical axiom which separates objective and subjective reality on the one hand—i.e. the world of natural things and material objects and that of concepts in the mind—from language—i.e. a system of signs that gives names to these things and concepts but is essentially not a part of them. But almost all the relevant discussion since the end of the eighteenth century has pointed out that this is too simple a view, that our way of looking at the world is dependent not only on the world and the human mind but also on language itself. One of the most influential originators of this idea was John Locke, one of its best-known representatives in the philosophy of language was Wilhelm von Humboldt (1767–1835), and one of its generally acknowledged linguistic proponents was Benjamin Lee Whorf (1897–1941). The direct influence of language on thought is generally (p.281) assumed to be culture-specific. More recent experts in the cognitive sciences (or cognitive linguistics), however, are inclined to find the general condition of human cognition to be a genetically conditioned faculty. But according to all these thinkers, language is not a neutral instrument of communicating about reality but one of its most essential means. Therefore, the assumption that the macro-structure of onomasiological dictionaries is language-independent is, philosophically speaking, too simple.

However, dictionaries are not treatises on logic or epistemology. They serve our natural ways of processing language, as explained above, and a dictionary is useful for this purpose if it is ordered macrostructurally according to topical classifications which are, in the framework of a given culture, generally accepted or at least intelligible. Such dictionaries are undoubtedly possible, but not always feasible. They are more useful in certain domains of reality and corresponding vocabulary than in others, such as the world of nature (e.g. plants, animals, humans) and man-made objects (e.g. houses, tools, vehicles) in older dictionaries and, additionally, the world of technical achievements in more recent ones. Emotions (e.g. the four humours), logical concepts (e.g. the various degrees of abstraction), and ideological assumptions (e.g. religious dogmas or basic political assertions) have often been seen as bestowed by nature, but appear to a more sceptical view as cultural artefacts and thus also dependent on linguistic possibilities. But it is the layperson’s (though not perhaps the philosopher’s) opinion that meanings can always be constructed by using pictures and images, gestures and indications, metaphors and metonymies, etc., i.e. communicative means independent of words used literally. And laypersons, after all, consult topical dictionaries.

Even lexicographically speaking, the division between alphabetical and topical dictionaries is not that clear-cut. In their techniques of explanation, alphabetical dictionaries in many cases cannot do without onomasiological elements. Definitions of species by reference to their genus (e.g. the cow as a mammal), illustrations of emotions in certain situations (e.g. expression of sorrow at a funeral), or explanation of an activity in a specific situation (buying and selling in shops)—all presuppose phenomena and structures of reality which are not linguistically given in the explanation. The inevitability of using encyclopaedic knowledge for the determination of word (p.282) meanings, a topic so frequently debated in cognitive linguistics, is nothing more than the inevitability of using onomasiological assumptions for semasiological determinations. These assumptions appear in pragmatic tags such as ‘terminology’, ‘medical language’, ‘nautical term’, or ‘poetical’.

Nevertheless, in spite of these complexities, the differences between onomasiological and semasiological principles are perfectly clear in dictionaries. It is the distinction between topical and alphabetical order which distinguishes the needs of speakers and allocates dictionaries clearly to one class or the other.

7.1.2. Lexicographical aspects: microstructure

Whether a dictionary has the one macrostructure or the other also has repercussions on the microstructure, i.e. the order within its entries. Semasiological dictionaries first list the lemma (keyword), i.e. the lexeme as a linguistic form, and then the explaining elements. Depending on the type of dictionary, the latter include indications of pronunciation, of grammatical behaviour, of pragmatic features concerning style, register, etc. and, above all, indications of meaning in the shape of definitions, paraphrases, quotations, and so on. The precise form of a microstructure—to the level of punctuation and abbreviations—is specific to each dictionary. It is important, however, that the lexeme to be explained (the explanandum) comes first, and the explaining elements (the explanantia) follow. Thus, the reader’s eye proceeds from the formally known, but otherwise unfamiliar, element to the explaining ones.

In topical dictionaries, however, the sequence is reversed. Microstructurally, the entry starts with the known lexeme (the explanantium), such as ‘mammal’, which serves as an explanation of the following lexemes (the explananda), such as ‘lion’, ‘mammoth’, and ‘bear’. It is obvious that this arrangement imposes natural restrictions on microstructures. Whereas, for example, in the semasiological microstructure it is natural to have a definition follow the headword of the entry, it would be rather curious (though systematically justifiable) to begin an entry in an onomasiological microstructure with a definition and then list the subsequent lexemes. This further corroborates the statement that onomasiological dictionaries prove more useful in certain domains of communication (p.283) than in others, for example, in those that have obvious hierarchical networks of meanings. The reader’s eye here proceeds from the semantically known to the formally unfamiliar elements.

It also follows from the nature of onomasiological dictionaries that entries are grouped together because of their semantic affinities. This creates the so-called pragmatic structures. They are highly interesting, but frequently neglected, because they often reflect the lexicographer’s idiosyncrasies. In a semasiological dictionary, each entry starts afresh, so to speak, and without any relation to the previous or following entry.7 In onomasiological dictionaries, entries like ‘mammal’, ‘fish’, and ‘bird’ will appear close to each other; in semasiological dictionaries they will be far apart.

Where is the dictionary of synonyms to be located in this system?

Synonyms are lexemes with partially overlapping and partially differing meanings. Their special linguistic feature is that they can be used in both of these semantic properties, depending on the context and the information which is intended to be conveyed. Human languages cannot work without them. Usually, a row of synonyms starts with a well-known lexeme and complements it with several others standing in a synonymous relation to it and to each other. As the boundaries of synonymy vary between small groups and large semantic fields, the number of lexemes in these rows can vary enormously too. They can, but need not, be accompanied by explanations. If they are, we call this row of synonyms explaining; if not, they are cumulative. This entry structure corresponds to the microstructure of an onomasiological dictionary. However, most synonym dictionaries follow the alphabet in the listing of ‘first’ lexemes which begin the rows of synonyms. Thus synonym dictionaries have a semasiological macrostructure, but an onomasiological microstructure. This is the special property of almost all of them. They consist of a sequence of content-oriented word lists arranged in an alphabetical succession of keywords. The user of such dictionaries must, therefore, engage in both ways of processing language when using them.

It is the outstanding property of Roget’s Thesaurus to combine the onomasiological macrostructure with an onomasiological microstructure, i.e. to be onomasiological in both respects. The term (p.284) ’thesaurus’ can rightly be reserved for this sort of arrangement. Before Roget, no other dictionary, properly speaking, had been organized in this way, although at least two works were also thesauri in this sense, namely John Wilkins’s ‘Tables’ and Jean de Maimieux’s list and nomenclators (see below). However, neither are dictionaries in the usual sense, but are rather the semantic part of universal language schemes. The other synonym dictionaries are hybrids located between semasiology and onomasiology.

7.1.3. Historical aspects

Historically speaking, the onomasiological tradition is older than the semasiological one. At the beginning of European written culture north of the Alps, the translation of Latin words into Anglo-Saxon ones in so-called glosses and their collection in glossaries laid the foundations of English lexicography. At first, they were exclusively devoted to semantically related vocabulary. Listing the Latin terms (for plants, fish, precious stones, etc.) and finding vernacular equivalents for them was not merely a linguistic exercise but also the collection of encyclopaedic knowledge. In spite of focusing on words, these glossaries approached the status of encyclopaedias, an important genre in the early and high Middle Ages. Admittedly, topical glossaries were soon reorganized into alphabetical ones and thus made more usable, but the tradition of topical glossaries and, later, dictionaries was never disrupted. It became important for surveying the knowledge of the world (e.g. in anatomy and botany), for the training of craftsmen (e.g. in husbacndry and seafaring), and for the learning of languages, most of all Latin. Classical studies in the wake of the Renaissance and Humanism gave them a strong boost. Although finally outstripped by alphabetical dictionaries, the topical tradition pervaded cultural life for centuries and came to an end only around 1700, with John Wilkins’s ‘Tables’ the last work of this kind.

Topical glossaries and dictionaries played their part in the encyclopaedic and also in the pedagogical tradition of language learning down to the seventeenth century. Their philosophical background is the assumption of a pre-existing reality which is to be mirrored in the ordered arrangement of words. This philosophical background (p.285) changed towards the end of the seventeenth century, mainly because of the epistemological deliberations of John Locke.8 This can be illustrated in this (and only in this) respect by an analysis of John Wilkins’s Tables’. This is the last great onomasiological work of the seventeenth century and of the old tradition, and it was mentioned by Roget, albeit critically, as a forerunner of his own work. Only a few other works in the onomasiological tradition were published between Wilkins and Roget (see below), and none of them would be so closely related to Roget’s Thesaurus as Wilkins’s Tables’.

7.1.4. Wilkins’s Tables’: macrostructure and word selection

In paragraphs II and III of chapter V, part I, of the Essay (1668: 19–21), John Wilkins explains his onomasiological programme. Although the author’s use of terms is unfortunately not very precise, this programme is sufficiently clear. Its pivotal point is the conviction that people generally agree on the principle(s) of reason and the apprehension of things. Concerning the latter, Wilkins distinguishes between reality, the mind, and the expression of the mind for the sake of communication. Reality is the realm of ‘things’. They are ‘natural’: they have their own ‘nature’, ‘shape’, and ‘use’. In the mind, notions correspond to things. Notions are called ‘mental’ and ‘internal’, which allows things to be called ‘real (= natural)’ and ‘external’. Notions are also called ‘conceits’, ‘apprehensions’, and ‘conceptions’. The expression of notions (etc.) is by sounds which we find in ‘articulate voice’ and ‘words’. Writing, which is also in words, gives a ‘figure’ or ‘picture’ of the sounds. Words which express the notions of things are also called ‘names’, ‘marks’, and ‘characters’.9

(p.286) In spite of this bewildering abundance of terms, it seems clear that Wilkins has in mind a world of things parallel to which there is a world of notions, governed by the ‘predicaments’ as ordering principles. These predicaments are given in the first three tables (called ‘General’, ‘Relation mixed’, and ‘Relation of Action’). Language is added as table IV. The title of this table reads ‘Of Discourse, Or the several notions belonging to Grammar or Logick’, but the introductory text speaks of ‘The several things and notions belonging to discourse’ (44). According to its nature as ‘external expressions, whereby men do make known their thoughts to one another’ (44), language is located between notions and things. What are formally termed ‘words’ are functionally ‘names’ which are made up of marks, notes, or characters.

The main philosophical assumption of the onomasiological lexicographer is that there is order in this world, that this order defines the nature of things and is discernible by the human mind. It is Wilkins’s programme to arrange the names of things in such a way that the sequence itself reflects the arrangement of notions which reflects the arrangement of things:

The first thing to be considered and enquired into is, Concerning a just Enumeration and description of such things or notions as are to have Marks or Names assigned to them. The chief Difficulty and Labour will be to contrive the Enumeration of things and notions, as that they may be full and adequate, without any Redundancy or Deficiency as to the Number of them, and regular as to their Place and Order. (20)

It is the order of entries in the ‘Tables’ which ‘contribute[s] to the defining of them, and determining their primary significations’ (22).

If we look at the macrostructure of the ‘Tables’, we find that this order follows the general principles of traditional logic and ontology. The ultimate source of both sets of principles is Aristotelian, just as the differentiation between the thing, the mental image, and the word is Aristotelian. This does not mean that Wilkins actually used Aristotle or some commentary like Porphyry’s as his model.

(p.287) He probably depended on one or more books by the so-called schoolmen which were available at his time and which represented the mainstream of traditional philosophy, often via a discussion of the works of Peter Ramus.10

After the epistemological introduction in chapter I (i.e. genera and tables I to IV), there follows chapter II (genera and tables V and VI) on God and the world as a whole, which functions as a bridge. After this, the macrostructure of the Tables’ is determined by the five predicaments (categories) substance, quantity, quality, action, and relation as the headings of the relevant chapters of word lists, and then by the differentiation between genus, difference, and species within each of them. In chapters III to VI (genera and tables VII to XX), governed by the predicament substance, we find entries pertaining to natural history, i.e. the elements and the various kingdoms of nature. In chapter VII (genus and tables XXI to XXIII), governed by the predicament quantity, we find entries pertaining to geometry and algebra. In chapter VIII (genera and tables XXIV to XXVIII), governed by the predicament quality, we find a mixture of entries which have in common that they pertain to certain states and faculties of human beings. They deal with the dispositions of characters and the body, with the senses, manners, habits, etc. Chapter IX (genera and tables XXIX to XXXII), governed by the predicament action, deals with human behaviour and activities in the world, including games and sports, the arts, crafts, and trades. Finally, chapters X and XI (genera and tables XXXIII to XL), governed by the predicament relation, deal with houses and everything belonging to them, and with the great societal systems, i.e. the law, military forces, the navy, and the Church, in terms of people, offices, objects, instruments, activities, etc. The macrostructure can be fully set out as follows:

  • Chapter I: predicabilia, predicaments, language

    • genus I: General, genus II: Relation Mixed, genus III: Relation of Action,

    • genus IV: Discourse (= language)

  • (p.288)
  • Chapter II: God and world as creation

    • genus V: Creator, genus VI: World

  • Chapters III-VI: substance, nature

    • genus VII: Element, genus VIII: Stone, genus IX: Metal, genus X: Herb, Leaf, genus XI: Herb, Flower, genus XII: Herb, Seed Vessel, genus XIII: Shrub, genus XIV: Tree, genus XV: animal Exsanguinous, genus XVI: Fish, genus XVII: Bird, genus XVIII: Beast, genus XIX: Peculiar Parts of animals, genus XX: General Parts of animals

  • Chapter VII: quantity, geometry, algebra

    • genus XXI: Magnitude, genus XXII: Space, genus XXIII: Measure

  • Chapter VIII: quality, states and faculties of human beings, souls and body

    • genus XXIV: Natural Power, genus XXV: Habit, genus XXVI: Manners, genus XXVII: Sensible Quality, genus XXVIII: Sickness

  • Chapter IX: action, human behaviour and activities, sports, crafts

    • genus XXIX: Spiritual, genus XXX: Corporeal, genus XXXI: Motion, genus XXXII: Operation

  • Chapters X-XI: relation, societal systems

    • genus XXXIII: Oeconomical, genus XXXIV: Possessions, genus XXXV: Provisions, genus XXXVI: Civil, genus XXXVII: Judicial, genus XXXVIII: Military, genus XXXIX: Naval, genus XL: Ecclesiastical.11

All these areas of reality and their corresponding areas of lexemes are well known from the tradition of glossaries, nomenclators, and topical dictionaries.

Wilkins calls the entries in the tables radicals. By far the greatest number come in pairs, which are either antonyms or otherwise related to each other. This prefigures Roget’s arrangement in opposing columns. The meanings of the radicals are provided by their position in the table, i.e. by the categories above them. Most of these entries are complemented by one or (sometimes) more lexemes which are obviously thought to be synonymous.

(p.289) The predicaments on the level of the macrostructure are necessarily abstract. But what is the nature of the vocabulary which is carefully and neatly distributed into these abstractly defined semantic compartments?

In most of the chapters, though not all, it is notable for its semantic concreteness, for the visual, palpable, audible, easily imaginable character of word meanings. For Wilkins, reality truly seems to consist of things with their own characters and shapes, and also with their natural uses. No proof of this is needed concerning the entries under substance (genera VII to XX). The word lists here swarm with names (in Wilkins’s sense) of the objects of nature. His lament that he is unable to enumerate them exhaustively itself shows that this was actually his aim. Even the headings of the genera and species, i.e. abstract terms, are introduced with quite concrete features. Elements, for example, are ‘the great Masses of natural Bodies, which are of a more simple Fabric than the rest’ (56); fire is ‘the hottest and lightest’ element (57); air is known ‘for its Levity and Warmth’ (58) and water for ‘its Gravity and Moisture’ (58); finally ‘the Coldest, Thickest, Heaviest, of any of those Bodies counted Elements, is called Earth’ (59). The essentially descriptive taxonomy of biology and related disciplines at Wilkins’s time supported his method. The whole taxonomy of minerals, stones, plants, and animals, including their exterior and interior parts, is represented. The entries under quantity (genera XXI to XXIII), divided into ‘magnitude’ and ‘measure’, also contain lexemes with concrete meanings, because they denote the simple and complex figures of geometry, the numbers of algebra, the units of value (money), of quantity and duration. Only the chapter ‘Of Space’ (genus XXII) is different in that it does not enumerate imaginable chunks of reality but ideas (see below).

With the tables under quality (genera XXIV to XXVIII) the picture is slightly different. Lexemes denoting the powers of the soul (e.g. understanding, judgement) and the body (e.g. the senses), habits (e.g. emotions, virtues), manners (e.g. candor, patience), and sensible qualities (e.g. sweetness, fattiness) indicate ideas rather than things or actions, mostly expressed by nouns derived from adjectives (like equity, vigilance, peaceableness, condescension, submission). But then there follow entries with the names of diseases (p.290) (genus XXVIII), which denote fairly concrete phenomena and states (e.g. ulcer, wart).

Derived lexemes denoting ideas are also to be found among the entries of the tables under action (genera XXIX to XXXII), in this case mostly nouns derived from verbs (like blessing, assurance). But they are to be understood as ‘actions of (blessing, assurance, etc.)’, which again gives them a concrete appearance. Whereas for the ‘spiritual actions’ it is sometimes difficult to decide whether an abstract idea or its concrete realization is meant, there is nothing equivocal about entries pertaining to ‘corporeal actions’ (like drinking, reading, or laughing, including gestures), ‘motions’, and ‘operations’. To the latter belongs everything people can do with the help of instruments (like forging, casting, kneading, or turning), i.e. the whole world of crafts and labour. Finally, the tables under relation (genera XXXIII to XL), probably the most abstract of all the predicaments, nevertheless almost exclusively consist of the names of concrete objects. ‘Oeconomical [i.e. domestic] relations’ are defined by degrees of consanguinity (like parent and child), by possessions (like farms or bridges and everything pertaining to buildings), and finally by provisions (i.e. edibles). ‘Civil relations’, ‘judicial relations’, ‘military relations’, ‘naval relations’, and ‘ecclesiastical relations’ are merely the umbrella terms for the names of people, offices, objects, actions, processes, etc. in families, law courts, the army, the navy, and the Church.

Admittedly, the distinction between concrete and abstract may in a number of cases be open to debate. When put to the test, the difference between an abstract term and the name of a concrete thing (in Wilkins’s sense) is not always that clear. But it is still safe to say that, beginning with genus VII, the vocabulary of Wilkins’s ‘Tables’ is dominated by lexemes expressing well-delimited, clearly imaginable, concrete things, presenting to readers a universe of experience, though classified in a traditional system of thought. The exceptions in genus XXII and genera XXIV to XXVII (‘space’, and then ‘natural powers’, ‘habits’, ‘manners’, and ‘sensible qualities’) may find an easy explanation. For most of the entries in these genera there are adjectives in common use. Elsewhere in the tables, however, the vast majority of entries are nouns for which no common adjectives are available. Wilkins may have changed most of the adjectival entries of genera XXII and XXIV to XXVII to (p.291) nouns on purely formal grounds. He may have written ‘gentleness’ instead of gentle, ‘prosperity’ instead of prosperous, and ‘redness’ instead of red, without wishing to indicate that he was changing the level of abstraction and hence speaking of the idea of gentleness, prosperity, redness, etc. and not of the states themselves. This is also why verbs are generally, though not exclusively, rendered in the nominalizing -ing form.

Wilkins’s ‘Tables’ are probably the last specimen of onomasio-logical lexicography, which combines three essential properties that all depend on a traditional philosophical background: (i) intentional universalism, i.e. the wish to bring all of the lexical means of language use into one system; (ii) strict (in Wilkins’s case rather philosophical) classification; and (iii) an attitude towards reality which is largely governed by experience and observation. The words in this system are like names tagged to ‘things’ or ‘notions’, i.e. to objects or mental representations.12

7.2 The classification of knowledge

The purpose of dictionaries is first and foremost to explain the meaning of lexemes. The semasiological and the onomasiological types do this in their own ways. Onomasiological dictionaries come close to being encyclopaedias whose main purpose is not the explanation of words (with the help of things) but the explanation of things (with the help of words). On the surface, this is evident in their textual structure. The more lexemes that are embedded in explanatory and narrative prose, as is the case, for example, in Wilkins’s ‘Tables’ or in Booth’s so-called analytical dictionary,13 the more a work of this lexicographical structure distances itself from a dictionary proper. The parallelism between world and dictionary (and even encyclopaedia) was characterized by the metaphor speculative, i.e. ‘in the manner of a mirror’ (see n. 12).

(p.292) In the eighteenth century, this division changed. There appeared books whose aim was to explain a certain type of ‘things’ but which did this in the manner of (semasiological) dictionaries, i.e. by an alphabetical arrangement of key words and fairly short articles.14 These works certainly continued the older tradition of assembling legal or medical terms, names for plants or minerals, etc. in dictionaries. However, their focus now was not the explanation of terms (i.e. the meaning of words) but the presentation of matter. Though these two intentions are close to and presuppose each other, there is a marked difference between them, lexicographically speaking.

It was undoubtedly the tremendous development of the arts and sciences in the eighteenth century and the concomitant pedagogical interest in them which made it appear feasible that human knowledge in these areas could be presented in a conspectus to anyone who cared to read it. For this aim, alphabetical order proved most practicable and easy to use.

Conducive to this decision was the fact that, in the meantime, epistemological analyses, such as, for example, those published by the most influential philosopher of the time, John Locke, regarded human knowledge of the world no longer as a mirror image of reality (i.e. as ‘speculative’) but as a cosmos of simple and complex ideas, originating from experience (sensation) and thinking (reflection) and developed into an ideational world which was best expressed by signs. The most important property of this world is its having been constructed by the human mind according to general communicative needs. It is certainly not inappropriate to find here the beginnings of what is today called ‘mental’ lexicography.

With this new approach to the status of human knowledge, the onomasiological order of dictionaries and encyclopaedias became outdated. It no longer made any significant contribution to understanding and could be replaced by the far handier alphabet. The only devices needed were the explanation in words of the various connections between ideas—for example, the ways in which simple ideas make a complex one—and references to relevant information.

(p.293) In accordance with the pedagogical endeavours of the Enlightenment, the wish to encapsulate many areas of knowledge in dictionary form broadened into the wish to do this with human knowledge in toto and for everybody. The term encyclopaedia adopted this new meaning—‘the presentation of all knowledge available at a certain era in a form which was intelligible to everybody’ (OED). The huge projects undertaken for this purpose were borne along by an enthusiasm for teaching and the firm belief that the arts and sciences are helpful for mankind. In this respect, the encyclopaedists rightly saw their ancestor in Francis Bacon.

In 1728, Chambers’ Cyclopaedia appeared for the first time. The name ‘cyclopaedia’ (which remained unique to this book) was chosen to stress the image of a circle of knowledge. The plan was to inform ordinary people by means of fairly short articles about anything worth knowing. Note the full title (of the second edition), which best describes the intentions of the whole undertaking:

Cyclopaedia: Or An Universal Dictionary Of Arts and Sciences; Containing An Explication Of The Terms, And An Account Of The Things Signified Thereby, In The Several Arts, Both Liberal And Mechanical; And The Several Sciences, Human And Divine: The Eigures, Kinds, Properties, Productions, Preparations, and uses of Things Natural And Artificial: The Rise, Progress, And State of Things Ecclesiastical, Civil, Military, And Commercial: With the several Systems, Sects, Opinions &c among Philosophers, Divines, Mathematicians, Physicians, Antiquaries, Critics, &c. The whole intended as a Course of ancient and modern Learning. Extracted from the best Authors, Dictionaries, Journals, Memoirs, Transactions, Ephemerides, &c in several Languages. By E. Chambers, F.R.S. […]. The Second Edition, Corrected and Amended; With Some Additions. In Two Volumes. […] London: Printed for D. Midwinter […]. M.DCC.XXXVIII.

BL L 40/14.

Chambers’ Cyclopaedia became the direct model for the French project which was to establish the generic meaning of encyclopaedia all over Europe. Edited by Dennis Diderot (1713–1784) and Jean d’Alembert (1717–1783), the seventeen volumes of the Encyclopedic appeared between 1751 and 1765 after many political difficulties had to be overcome in the course of its preparation. Eleven volumes with illustrations were to follow. The book has a reputation for being the most elaborate and extensive undertaking of (p.294) its kind and of its time.15 It also has the reputation of being the joint effort of a great number of contributors of unique scientific eminence.

Encyclopédic, ou Dictionnaire raisonné des Sciences, des Arts et des Métiers, par une Société de Gens des Lettres. Mis en ordre et publie par M. Diderot […] et quant á la Partie mathématique, par M. D’ Alembert […]. A Paris, chez Briasson, David, Le Breton, Durand. 1751–1765.

Although acknowledging Chambers’ Cyclopaedia as the seminal forerunner of their Éncyclopedic, Diderot and d’Alembert deviated from this book in one most important point: they were not content with giving information for the general reader in short articles, but were ambitious enough to present a cross-section of knowledge in all of the sciences and arts in extensive articles written and organized according to generally accepted style and methods. The editors and authors of the Encyclopaedia Britannica adopted this project whole-heartedly and produced the counterpart of the Encyclopedic in Britain. Its first edition appeared in 1771; its seventh edition, with substantial changes and amendments, in 1842, with Peter Mark Roget among its authors. The additions to the seventh edition consisted of historical and biographical articles, which had been absent in the Encyclopédie. According to the ‘Preface’ by its editor, Macrey Napier, each science was treated in a systematic form, and an extensive, in fact book-long, ‘Dissertation’ showed the progress that had been made between the first and the seventh edition.

The Encyclopaedia Britannica. Or Dictionary of Arts, Sciences, and General Literature. Seventh Edition, With Preliminary Dissertations On The History Of The Sciences, And Other Extensive Improvements And Additions […]. Adam and Charles Black, Edinburgh M.DCCCXL11.

In the prefaces to all three books, the authors go out of their way to explain that the articles are interconnected. The information, though not arranged according to the alphabet, is linked by scientific discipline and by the cross-references to other articles. The old (p.295) order of onomasiological dictionaries, which had some fundamentum in re, is now replaced by a classification of human knowledge according to methodological principles. The preface to Chambers (1738: ii-xxv), for example, sets out a system of 47 scientific disciplines broken down into six groups according to the following argument. All knowledge is either natural or artificial (technical). Natural knowledge is either (i) sensible or (ii) rational. Artificial knowledge is either (iii) internal or external, which is either (iv) real or (v) symbolic. Consequently (i) comprises the disciplines belonging to physiology or natural history, (ii) the disciplines belonging to mathematics and quantifying procedures, (iii) logic and ethics, (iv) all the other sciences, and (v) language-related disciplines.

The system of human knowledge which is appended to the preface of the Encyclopedie has three main categories under which all the sciences and arts, both liberal and mechanical (crafts), are to be subsumed: (i) memory, (ii) reason, and (iii) imagination, (i) is for everything historical, be it man-made or natural, (ii) is for everything philosophical dealing with being in general, God, humanity, and nature, and (iii) is for the liberal arts. The whole system contains more than 150 disciplines, of which many carry the note ‘etc.’, indicating that the system is not yet perfect.

The system of the Encyclopedie has an obvious affinity to that of Francis Bacon, who distinguished between historia (naturalis, civilis), philosophia (de deo, de natura, de homine) and poesis (narrativa, dramatica, parabolica).16 Chambers’ system has an obvious affinity to the last chapter of John Locke’s Essay, called ‘Of the division of the sciences’:

All that can fall within the compass of human understanding being either, first, the nature of things, as they are in themselves, their relations, and their manner of operation; or, secondly, that which man himself ought to do, as a rational and voluntary agent, for the attainment of any end, especially happiness; or, thirdly, the ways and means whereby the knowledge of both the one or the other of these is attained and communicated. (Locke 1961: ii, 308–9)

Locke does not give his division as a stemma but as an argument. To the first group belong what was then called theology, metaphysics, (p.296) mathematics, natural philosophy, and natural history. The second group is ethics, and the third what was then called logic and would today be called linguistics or semiotics.

In his search for a comprehensive framework in which to place the synonyms of English, Roget could not go back to Wilkins, whose work he knew and mentioned in the preface of his book. Though he would not have expressed himself in precisely these words, he would have agreed with Jeremy Bentham that the semantic principle of the Tables’ was outmoded: ‘In the character of a practical project fit for use, this work, with all its ingenuity, failed in its design: being written before the discoveries made in the field of Pscycology [!] by Locke’ (Bentham 1983: 323).17 It was as simple as that. Instead, Roget would turn to the various classifications of knowledge which had taken the place of onomasiological order. Such classifications were in demand, especially in the rapidly growing natural sciences. William Whewell, the founder of mineralogy and a highly influential natural scientist, had devised a complex system. Roget himself worked in this field,18 and as a contributor to the seventh edition of the Encyclopaedia Britannica he was used to this kind of thinking. In the preface of his book, he acknowledged having read a pamphlet by J. W. Lubbock (1838) which discussed classificatory systems at great length, from the Porphyrian Tree to Jeremy Bentham, in the context of constructing adequate systems for cataloguing the great libraries, which were growing rapidly at this time. In 1816, Bentham, with whom Roget had worked in his youth,19 had drawn up an elaborate ‘chrestomathic table’ (Bentham 1983) intended as a curriculum for a so-called Lancaster School. This was a secondary day-school with a special tutorial system, in which an able boy was paired with a less able one. The teacher taught the former, who taught the latter. Bentham became very enthusiastic about the project and, in addition to the curriculum, wrote many papers about its foundation and implementation. However, it never got off the ground, or at least it is not known that this classificatory system was ever used for teaching. It is elaborated in a complex stemma, leading eventually (p.297) to eighteen fields of knowledge. The umbrella term for the entire curriculum is ontology. I reproduce the entire stemma here, rearranged as a numbered list.20 Bentham’s terminology is strictly organized but very idiosyncratic, so I have not followed his usage in all cases. Nevertheless, the denominations of the sciences are all taken from his stemma.


  1. 1 regarding properties common to all beings: metaphysics

  2. 2 regarding properties peculiar to different classes of beings:

  3. 2.1 related to the body

  4. 2.11 related to quantity [mathematics]:2.111 not regarding form

  5. 2.1111 arithmetic

  6. 2.1112 algebra

  7. 2.112 regarding form: geometry

  8. 2.12 related to quality

  9. 2.121 related to nature [natural history]:

  10. 2.1211 related to the heavens: astronomy

  11. 2.1212 related to the earth

  12. 2.12121 not possessing life: mineralogy

  13. 2.12122 possessing life

  14. 2.121221 botany

  15. 2.121222 zoology

  16. 2.122 related to man [natural philosophy]:

  17. 2.1221 mechanics (experimental and technical)

  18. 2.1222 chemistry (experimental and technical)

  19. 2.2 related to the mind

  20. 2.21 related to the intellectual faculty

  21. 2.211 logic

  22. 2.212 related to communication

  23. 2.2121 grammar

  24. 2.2122 rhetoric

  25. 2.22 related to the sensitive faculty2.221 related to volition [ethics]:

  26. 2.2211 expository ethics

  27. 2.2212 censorial ethics

  28. 2.22121 related to the state

  29. 2.221211 related to internal government

  30. 2.2212111 legislation

  31. 2.2212112 administration

  32. 2.22122 related to the individual: private ethics

  33. 2.222 related to the senses: aesthetics

(p.298) This is a system of 22 scientific disciplines which Roget would remember when making a blueprint for his thesaurus: metaphysics; [mathematics], arithmetic, algebra, geometry; [natural history], astronomy, mineralogy, botany, zoology, mechanics, chemistry; logic; grammar, rhetoric; [ethics], legislation, administration, politics, private ethics; aesthetics.

7.3 Onomasiological lexicography between 1700 and 1852

John Wilkins’s Tables’—in essence a thesaurus of the English language though incorporated into his sketch of a universal and philosophical language—was the last attempt at compiling the complete lexical potential of the English language in an onomasiological manner. The two outstanding intentions of the century-old genre of topical dictionaries—knowledge-based (foreign) language learning and the classification of human knowledge for any other purpose—were now catered for by alphabetical dictionaries and by alphabetical encyclopaedias. The convenience of the former, for example in school classes, outweighed the educational benefits provided by the semantic cohesion of vocabulary. As noted earlier, the growing habit of publishing dictionary pairs—English-foreign language and foreign language-English—gave dictionary users the opportunity of first learning the meaning of a foreign-language word in their mother tongue by translation and then of using this knowledge to proceed further in the foreign language itself.

The availability of alphabetical encyclopaedias on many levels—from the Encyclopaedia Britannica to the topical alphabetical dictionary for everyone—made accessing information on any kind (p.299) of human knowledge much easier, since the idea that encyclopaedias should structurally mirror the world was abandoned. As soon as words were taken to indicate ideas and not things or notions, it no longer mattered how they were arranged.21

Between (roughly) 1700 and 1852, the year of publication of Roget’s Thesaurus, only a few major works in the old genre appeared. Of course, this does not mean that such books, in particular the smaller ones, vanished altogether. Textbooks for school use continued to collect vocabulary topically in appendices, for example, as did glossaries appended to editions of literary texts. The three works discussed in this section represent in various ways the category of general topical dictionaries. Andree (1725) is fairly traditional and can be regarded as a late-comer of the old genre. Booth (1835) is most astonishing because it instrumentalizes an enormous list of keywords by arranging and connecting them textually so as to create a general European history of culture. It is heavily influenced by diachronic studies as they evolved during the nineteenth century out of Indo-European philology. It is a dictionary because it deals with words, although it does not display them as a list. (But John Wilkins can also be said to do this because he combines his entries, which are displayed as a list, with an explanatory text.) Maimieux (1797) presents the lexical part of his universal language schema, a perfect parallel to John Wilkins. Maimieux (and also Wilkins) were mentioned by Roget in the preface to his Thesaurus. We cannot ascertain whether he knew the two other works.

7.3.1 Andree (1725)

[R. John Andree [Andrée]:] A Vocabulary, In Six Languages; Viz. English, Latin, Italian, French, Spanish, and Portugues [!], After A New Method, To Shew the Dependance of the four last upon the Latin, and their mutual Analogy to each other. With Proper Rules for several Pronunciations; and a Dissertation upon their Origin, Change, and Mixture; besides many other Advantages, not to be met with in common Dictionaries and Vocabularies. […].

London: Printed and Sold by P. Vaillant, in the Strand, and W. Meadows, at the Angel in Cornhill. M.DCC.XXV.

(p.300) No mention of author; dedication to John, Lord Garteret signed ‘R. John Andree’. 4°, 178 pp., i.e. 256 pp. because opposite pages have an identical number. BL RB. 23.a. 14332.

There is no entry for the author in the DNB. The preface

In his preface To The Reader’, the author presents a curious mixture of rather traditional ideas and more recent deliberations on language learning. On closer inspection, this mixture reveals itself to be the most distinctive property of the entire book. Rather traditional is the reference to the universal language which was lost in Babel and which forces us to learn national languages; rather new, however, is the observation that climate, diet, education, disposition, and the genius of a people are also responsible for the differences between them. This, and the remark that some notions can be expressed in one but not other languages, reminds us, for example, of James Harris. However, it amounts to only a somewhat timid espousal of what was later called ‘linguistic relativity’, without consequences for the dictionary itself.

It is a Lockean concept that foreign languages have to be learnt in order to communicate ‘[i]deas of things, present, past, and absent’. The choice between the available languages is guided by the criteria of usefulness: Latin is the language of the learned, Italian the language of princes, French is used everywhere in Europe, Spanish and Portuguese are useful for trade in the West and East Indies,. There is no mention of German. English is a mixed language and, because of this property, superior to the others:

In this they have so well succeeded, that their Tongue is capable of all the lively Strains of true Eloquence, nothing inferior to the most fluent Orations pronounced of Old by the Romans; And as that Language ought to be valued most, by which Men are best enabled to explain themselves clearly and concisely, I make bold to add, that the English has as fair a Claim to Preference as any. (p. xi)

This kind of linguistic pride is, again, well known from authors of the previous century. In spite of his attitude to English, Andree compiled his dictionary for learning other idioms. Quite new in this respect is his systematic exploitation of the similarities between Latin and the Romance languages for the purpose of memorization.

(p.301) At least two of them, he maintains, can therefore be learnt at once. This is to be done with the help of lists of nouns being ‘brought in by way of a natural Discourse’ (p. xiv), i-e. in semantically related groups—again a very old idea. The advice given for learning in general and for mastering pronunciation in particular shows that the author is a man of practice rather than of rules (pp. xiv, xxii). He does not like what he calls ‘tedious grammar’. Macrostructure

The chapters of Andree’s book are arranged as follows:

Ch. I: Of Things; Their Mode, Manner, Beginning and End. Of the Sky and Caelestial Bodies.

Ch. II: Of the Elements. Of Fire. Of the Air. Of Rain, Hail, and Snow &c. Of Water. Of Land or Earth.

Ch. III: Of Minerals. Kinds of Stones. Precious Stones.

Ch. IV: Of Plants and Herbs. Eatable Herbs and Roots. Sorts of Corn.

Ch. V: Of Trees and Shrubs. Of Fruits. Of Spices. Of a Plant.

Ch. VI: Of Animals in general. The five outward Senses. Inward Senses.

Ch. VII: Of Insects.

Ch. VIII: Of Birds. Parts of a Bird.

Ch. IX: Of Fishes. Parts of a Fish.

Ch. X: Of Four footed Beasts. Wild Beasts. Tame Beasts. Parts of Beasts.

Ch. XI: Of Man, respecting his Age and Kindred. Male Kindred. Female Kindred. Stations of Man.

Ch. XII: Proper Names of Men. Proper Names of Women.

Ch. ΧΙΠ: Parts of Man’s Body. Humours of the Body. Excrements.

Ch. XIV: Of Diseases. Of Remedies.

Ch. XV: Of the Mind and its Affections. Of Virtues. Vices. Passions of the Mind. Of Actions. Hunger, Thirst, etc.

Ch. XVI: Of Meats and Drinks. Kinds of Meats. Kinds of Meals.

Ch. XVII: Of Apparel in general. Mens [!] Apparel. Womens [!] Apparel.

Ch. XVIII: Of Buildings. Parts of a House. Workmens [!] Tools.

Ch. XIX: Of Household Stuff. Furniture for a Room. For the Kitchin. For giving of Light. Kinds of Vessels.

(p.302) Ch. XX: Of the Country. Of Coaches and Waggons. Bridle, Halter, &c. to govern Beasts.

Ch. XXI: Of Societies, and their several Members. Of Dignities and Honours. The King’s Ministers. Names of Tradesmen. Of music and musical Instruments. Of Diversions and Games.

Ch. XXII: Of a School and Learning. Of Writing. Kinds of Studies.

Ch. XXIII: Of Ecclesiastical Affairs. Of a Church, and what belongs to it.

Ch. XXIV: Of judicial Affairs. Of Crimes. Of Punishments. Of Rewards.

Ch. XXV: Of Military Affairs. Of Peace, Of War. Of Military Persons. Of Weapons, or Arms. Warlike Instruments.

Ch. XXVI: Of Naval Affairs. Kinds of Ships or Vessels. Parts of a Ship.

Ch. XXVII: Of Time. Days of the Week, Chief Holy Days. The twelve Months. The four Seasons.

Ch. XXVIII: Names of Countries and Cities.

Ch. XXIX: Nouns Adjective.

Ch. XXX: Of Cardinal Numbers.

Ch. XXXI: Of Ordinal Numbers.

Ch. XXXII: Of Colours.

Ch. XXXIII: Verbs.

The nouns (I-XXVIII) are arranged according to semantic affinity, the adjectives (XXIX) and verbs (XXXIII), however, according to the alphabet. English is the language mainly used for this purpose. The numbers (XXX-XXXI) are treated seriatim, while the colours (XXXII) have a scarcely intelligible order of their own. Sometimes we find an alphabetical ordering within the topical chapters. There is, for example, an alphabetical list of craftsmen, from armourer to work-man, in chapter XXI.

Also part of the macrostructure is the arrangement in parallel columns. There are six of them: English, Latin, and Italian on the left-hand pages, French, Spanish, and Portuguese on the right. Lexemes with one (presupposed) meaning are printed in parallel and on one line. However, the discovery of a common etymology is left to the reader. There are no explanations. The author obviously supposes that the graphic and phonetic appearance of words will (p.303) suffice for an explanation. Moreover, the author does not bother about cases where this principle does not hold. Note, for example, the following lines:

  • barley/hordeum/orzo/orge/cevada/cevada

  • foolishness/stultitia/pazzia/folie, sottice/locura/parvoice

  • shop/officina/bottega/boutique/tienda, officina/tenda

This is the deliberate introduction of comparativism to topical lexicography. The principles of topical order, including even its exceptions, had already been used in former dictionaries, and Andree’s topics themselves might have been copied from some earlier work. They show the speculative tendency, i-e. the ambition to mirror the objective world in a list of words, by giving names to things and notions. This function is obviously seen to be the overriding mechanism in language learning. It explains the preponderance of nouns in semantic order and the alphabetical ordering of adjectives and verbs. Nouns are chosen according to subject matter. Adjectives and verbs, however, are chosen in reference to nouns. The parallel arrangement of various languages is modelled on the multilanguage textbooks that had been appearing on the Continent since the early fifteenth century, although in these textbooks no comparison of languages was intended, at least deliberately. Macrostructure

In the leftmost column only lexemes are given, without further information. This column is intended for native speakers. In the Latin column adjoining, the nominal lexeme is followed by gender markers and numbers of declension, the adjectival by the three endings of genders, the verbal by the first-person present ending and number of conjugation. In Italian, Spanish, and Portuguese the adjectival lexemes are followed by gender markers only. These columns are intended for learners.

However, the author does introduce something new to this traditional scheme which, even in the parallelism itself, seems to follow older examples (Hullen 1999a: 305–60). The entries of the English column, though not of the others, are bound into a syntagma in a rudimentary way. Note for example:

  • A PLANT is an Herb

  • (p.304)
  • a Shrub

  • a Tree

  • A[n] Herb is

  • Grass

  • Flax

  • Hemp

  • Some of the most common herbs and Flowers are

  • Anis [and 58 further entries in alphabetical order]

  • Eatable Herbs are

  • Artichoke

  • Sparrow Grass

  • Lettice

  • Colewort or Cabbage

  • Colliflowers

  • Eatable Roots are

  • Beet-Root [8 further entries]

  • Olerocious Fruits are

  • Cucumber [4 further entries]

  • Corn is

  • Barley [6 further entries]

  • Whence cometh

  • Meal or Flower

  • Bran

  • Pulse is

  • a Bean [6 further entries]

  • In Corn is

  • a Beard [7 further entries]

Only the lexemes which are identified by such defining syntagmas have foreign counterparts in the parallel columns. As Andree recommends initially learning the verbs be and have in all the languages, instead of memorizing any grammatical rules, his expectation was obviously that the syntagma of the English column could easily be transferred to the other languages and that this was the core of foreign language learning. Short texts can be created in this way:22

In a School are a Master [ + 1]; a Master teaches Good Manners [+2]. Men declare their Thoughts by a Discourse [+3]. In a Discourse there are (p.305) a Letter [+3]; [words] which compose a Fable [+9]. A Book hath an Author [ + 3]. A writer is a Poet [+0] who writeth one single verse [+1]. For Writing they use a Pen [+7]. A Pen hath a Slit + [0] and is made by a Pen-Knife [+1]. [A line] is made by a Rule [+1]. For Correction the Master useth a Rod [ + 1] and sometimes gives Holy Days. Men apply themselves to a Study [and sciences] such as Arithmetic [+10].

Such stories determine the pragmatic order of entries, i-e. the sequence of words within the various chapters.

This technique is not entirely new either. It looks like a Comenian text put into the format of a multilingual dictionary. Indeed, we find a very similar pattern23 in an adaptation of the Orbis pictus to English. Note the very successful London Vocabulary by James Greenwood (1713; Hüllen 1999a: 422–30). Even the style of the miniature texts has a Comenian ring. The embedding of words into syntagma, if a rudimentary one, was certainly meant as a counterweight to the predominantly paratactic appearance of the languages on account of the long lists of nouns. It is significant that there are no syntactic links in the chapters which list adjectives and verbs.

7.3.2 Booth (1835)

David Booth: An Analytical Dictionary Of The English Language, In Which The Words Are Explained In The Order Of Their Natural Affinity, Independent of Alphabetical Arrangement; And The Signification Of Each Is Traced From Its Etymology, The Present Meaning Being Accounted For, When It Differs From Its Former Acceptation: The Whole Exhibiting, In One Continued Narrative, The Origin, History, And Modern Usage Of The Existing Vocabulary Of The English Tongue; To Which Are Added, An Introduction, Containing A New Grammar Of The Language, And An Alphabetical Index, For The Ease Of Consultation. By […]. London: Printed For James Cochrane And Co., 11, Waterloo Place, Pall Mall. 1835.

[Inner title-page: the former plus] In Two Volumes. Vol-1. London: Printed And Published By J. And C Adlard, Bartholomew Close. 1830.

4°, i-ccxxx + iv + 455 pp. = 689 pp, BL 626 L. 4.

Continuous, narrative text, i-e. not arranged in any dictionary format. In the text, topicalized lexemes (see below) are printed in capitals.

(p.306) History of the book

After a ‘Prospectus of an Analytical Dictionary’ (which seems to be lost) a general announcement of an ‘analytical dictionary’ appeared in 1806:

David Booth: Introduction To An Analytical Dictionary Of The English Language. By […]. Edinburgh: Printed for the Author […]. 1806. 8°, 158 + χ pp. (Index), BL 626.h.9.

The 1835 edition begins with a preface bound before the title-page, dated 1822. It may have been published separately, but is now not traceable. Obviously, the book of 1835 is the final version, whose publication had been planned for 1830 (according to the inner title-page), a plan which however did not materialize. A second volume, as announced in 1830, never appeared. A third title-page indicating ‘Part 1’ is sandwiched between the ‘Introduction’ and the dictionary. It reads: ‘London: Printed For R. Hunter, Successor to Mr. Johnson. No 72, St. Paul’s Churchyard. 1822.’ There is, however, no separate publication traceable to this year. From this complicated and enigmatic history of the book and from the fact that the whole undertaking was eventually abandoned it follows that the author worked under severe difficulties and also, perhaps, that the whole work overtaxed the resources of one man.

David Booth (1766–1846) was a brewer, and later schoolmaster in Newbury, Fifeshire. He went to London (c.1820), where he worked for the Society for the Diffusion of Useful Knowledge. The plan of an analytical dictionary

Booth elaborates on his project in the introduction of 1806 and in the preface and the 230-page introduction to the 1835 edition. The starting point is a direct adaptation of Locke: ‘Ideas are the reflected images of nature. Words are the pictures of ideas. Simplicity of thought will produce simplicity of expression; and hence the individual impulses of the mind will be marked by monosyllabic sounds. Two or more simple impressions form what is termed a complex idea, which is expressed by as many primitive words’ (1806: 11–12). This means that the dictionary pays attention not only to word classes, because each of them is indicative of certain classes of ideas, but also to the regularities of word formation (in (p.307) particular prefixation and suffixation) because they mirror the regularities of ideational processes. The author knows that he is attempting something very new. In the preface to the 1835 edition he derides former dictionaries which suggest that to abandon, give up, resign, quit, desert, forsake all have the same synonymous meaning (1835: ‘Preface’, p. v)—i.e. the normal type of synonym dictionaries—and comments:

A marked feature in the plan of this Dictionary, and that which will distinguish it from every other that has hitherto appeared, is its perfect freedom from the fetters of Alphabetical arrangement. In consequence of this emancipation, the Author is persuaded that he has been enabled materially to improve his definitions, both as to correctness and to perspicuity, while the ease of consultation will be sufficiently provided for by an Index. (1835: ‘Preface’, pp. v-vi)

It was planned as a dictionary to be read in toto, not consulted to resolve local difficulties.

The preface of 1806, like the long introduction of 1835, is dominated by philosophical ideas on human communication and the groundwork of universal grammar. The differences between word classes corresponds to differences between ideas: ‘An investigation, into the composition of language, is resolvable into an enquiry, concerning the classifications and abstractions of the human mind’ (1806: 15). These classifications are nomination, quality, state, or action for nouns, adjectives, and verbs respectively. This is the onto-logical underpinning of grammatical word classes, well known from universal grammar. It is applied to all the classes and to many grammatical structures (genitive, participle, affixes, etc.). It is taken to be axiomatic that linguistic simplicity indicates ideational simplicity and linguistic complexity ideational complexity.

The dictionary proper (if that is the right term—see below) then begins with a reprint of the 1822 preface entitled ‘To The Public’. The main text extends over 455 Arabic-numeralled pages. ‘To discover the thought that each of these words expresses, in the written language of this country [i.e. English], is the object of the present work. […] Whether the thread that shall connect the tale be real or imaginary, we are anxious that the features of the description shall be true to nature’ (1). This dictionary therefore aims to reconstruct the ways in (p.308) which the linguistic mind assimilates the surrounding world and makes it the contents of human consciousness. It is an aim of great philosophical ambition which rests on, among other sources of the history of culture, the methods of comparative philology as it developed from the end of the eighteenth century. The text shows that its author is very knowledgeable in both fields. The dictionary as a history of ideas

What then follows is an account of the history of the basic ideas of mankind, guided by such lexemes as Indo-European languages, and in particular English, have coined for them. The philosophical framework regarding word classes and word formation in general, and also simple and complex ideas, is no longer topicalized but permeates the arguments and explanations. Thus the book is not, properly speaking, a dictionary, but rather a narrative of the history of ideas. The lexemes are its points of reference. The narrative moves along a chain of key terms, departing from them in order to deal with morphologically (word-formational) or seman-tically (associatively) determined extensions (see below), and then turns back to the key terms. It is only the typographical device of capitalizing the lexemes under discussion which allows the reader’s eye to travel over the pages as if they were the pages of a dictionary.

The lexemes on the first 64 pages of the dictionary are, for example, all linked to the word and idea homo. Note, for example, this chain of lexemes (without morphological extensions):

Man-Wife, Woman-Male-female-Homo, Human, Virtue, Virtus-Anthropos, Ape, Baboon, Moon, Luna.

Marriage, Matrimony, Husband, Sponsor, Wed, Gage, Bride, Hymen, Epithalamium, Concubine, Adultery, Bastard, Strumpet (etc.).

Father-Mother-Child, Parents (lat. lexemes), Dame, Matron, Womb.

Boys-Girls, Youth, Adolescence, Infant, Knight, Cavalier, Knave, Esquire, Lad, Nymphs, Brothers and Sisters, Friar, Monk, Hermit, Consanguinity, Affinity, Genus, General, Pedigree, Generation, Ingenuity, Sex, Conception, Brain, Nation.

Nature, Physis, to Get, to Breed, to Hatch, Genitals.

Elder, Minister, Alderman, Pagan, to Pullulate, Chicken.

(p.309) People, Census, Republic, Vulgarity, Lout.

To Bear, to Bring, to Produce.

The following montage of quotations from one particular section, which moves from ‘branch of tree’ to ‘arm’ as a member of the body and then as ‘weapon’, then to ‘army’ and ‘war’, and from there back to ‘measure’ and ‘branch of tree’ (1835: 117–42), illustrates the microstructure and narrative style of this dictionary:

The Latin ramus was oftener written than brachium to denote the Branch of a tree. Ramous, to ramify. The Latin armus, originally denoted the whole of the human arm and shoulder. The Greek harmos denoted a joint or articulation, and particularly the shoulder. Arm, the whole of the limb. The metaphorical usage is unbounded. Armlets, armileated, armillary sphere. Unarmed, disarmed, armipotent, armistice. To alarm is from the Italian alarma, which literally signifies ‘To arms’. To Harm appears to be merely a varied orthography of to Arm. Harmful. Harmless. The Saxon hearm signified Harm, and likewise the Arm. Following the Latin arma, when denoting the materials of war, is always written in the plural. Weapon. This word is common to all Gothic dialects. Weaponed. Weaponsalve. Weapontakes. Firearms. The German Waffen is equivalent to our word Arms. The Swedish Vapen is both singular and plural.

[The narrative then deals with armour and all its parts in the Middle Ages, armée as a ‘Corps of armed men’ with all its organizational divisions and ranks, including musical instruments, and with firearms in their various kinds and handling. It ends with: Ά large force organized and provided with all the materials of Warfare is an Armament.’]

The Saxon eln (Latin ulna) and Old English Ell, signified the Arm, but is now used only to denominate a linear measure. The Elbow (Swedish Armboge) is the bow or bend of the arm, and a chair with two branches or arms. The Cubit (Latin cubitus from cubare, to bend), applied to the Elbow, has been a measure of length among different nations. The English Cubit is eighteen inches.

The stems and branches of trees and shrubs are closely covered and surrounded, as with a skin, by a substance of a softer texture than the wood, called the Bark or Rind.

This montage shows how the author moves from the lexemes to their extensions and the associative chain of the meanings themselves. Also obvious is the historical and comparative attitude with reference to Indo-European languages. Although the arrangement is that of a continuous text, the dictionary structure remains clearly discernible once the reader has got used to this format.

(p.310) 7.3.3 Maimieux (1797) Pasigraphy

It is difficult to ascertain when the artificial word pasigraphy (pasi = for all, graphy = writing) came into use. The relevant OED entry mentions 1796, and the first quotation indeed comes from this year. However, there is a report in existence called Narrative of M. DeChaumereix [Duroys de Chaumereys], Who Escaped From The Massacres Of Aurai And Vannes, London, 1795 (BL 9210.C.37), with an appendix by ‘Mess. De Mémiev and Sicard’ in which the two advertise the ‘Pasigraphy, Or, First Elements Of The Art Of Printing And Writing In Language To Be Understood In All Languages Without Translation’ (Alston 1974: vii, no. 300).24 This points to an earlier origin. Finally, Roget refers in his ‘Introduction’ to an ‘anonymous work’ which appeared in Paris in the year 1797. This must in fact have been the book which is usually regarded as the original of pasigraphy, though it did not appear anonymously. Its author was the Franco-German Jean de Maimieux (see below). In sum, the years 1795 (or earlier), 1796, and 1797 all qualify as possible years of origin for the pasigraphical project and, in consequence, for the name.

Pasigraphy was immediately regarded as a noteworthy project. In a review of the same year (Anonymous 1797: 562–5), it was welcomed as an exciting attempt at universal communication, though not without criticism. The reviewer complained that the new language would ‘supercede so effectually all idiomatic associations, that no work of imagination nor of eloquence could retain any influence in this new dress’. In consequence, it could serve only for the sciences. But ‘science possesses higher principles of classification than are [here] applied’.

In spite of these obvious shortcomings, pasigraphy developed almost at once into a powerful movement. In a later pamphlet (Damm 1876), more than twenty authors and their relevant projects were listed under the heading ‘Praktische Pasigraphy Oder Die Kunst Mit Allen Nationen Der Erde Correspondieren Zu Können, (p.311) Ohne Deren Sprache Mächtig zu Sein’, strangely excluding the best known of them, Maimieux himself. The meaning of the term became quite general in the course of time. The later OED entry (2nd edn) reads:

A name given to a system of writing proposed for universal use, with characters representing ideas instead of words, so as to be (like the ordinary numerals 1, 2, 3 etc.) intelligible to persons of all languages. Applied originally to a system proposed in 1796; subsequently to others having a similar object.

This shows that the term was eventually applied indiscriminately to all universal language projects, including the earlier ones of the seventeenth century, although it would require a more thorough investigation than is possible here to show how the meaning of this term became so broad. Janne Damm (1876) was already using the word for all of the relevant projects between Johann Joachim Becher (1635–1682), who invented a secret writing with numbers (Hüllen 1996b), and 1876; Damm even mentioned Chinese signs and hieroglyphs as alleged forerunners of pasigraphy. Indeed, Maimieux’s undertaking has much in common with the exclusively theoretical reflections on universal languages by Rene Descartes and Gottfried Wilhelm Leibniz and the blueprints by George Dalgarno (Cram and Maat 2001) and John Wilkins. They all intended to create a universal character, i.e. a reading language, by constructing signs which express ideas that are the common property of everyone. Algebra or the signs of the zodiac were the models, and indeed some experiments used the Arabic numbers for this purpose. To this extent, pasigraphy is indeed comparable to the earlier universal language projects. However, the problem was always how this plan could be achieved, i.e. how ideas of things could be indicated by graphical signs and what ‘ideas of things’ actually are. These questions identify the philosophical issues of the early universal language projects, which still attract the attention of scholars today (Subbiondo 1992; Hüllen 1999a; Isermann 2002). However, in pasigraphy this issue is almost tacitly regarded as being resolved. Except for appeals to common sense we do not find any serious discussion of it.

Today’s historiographical research on the seventeenth century no longer uses the term pasigraphy in its extended meaning, but rather, following Wilkins’s terminology, so-called ‘universal’ or (p.312) ’philosophical’ ‘characters’ or ‘languages’. Pasigraphy is used in its original meaning or, if at all, for similar linguistic projects after 1800.25

A quite parallel similarity/difference is to be found in the effects hoped for. Dalgarno and Wilkins had a theological and philosophical background. They wished to reverse the curse of Babel and to restore the early universality of communication to the human race. In order to achieve this, they went back to the philosophical foundations of language use in epistemology. Eventually, this would also provide a chance to propagate the Gospels throughout the world (e.g. in the newly founded American colonies).

The representatives of pasigraphy had a less philosophical and more political intention. Only a few years after the French Revolution, some people began to believe in world citizenship and the common good to be furthered by universal communication (e.g. Vater 1799). De Maimieux himself referred to the atrocities of the Franco-German War, which he had endured and which he wanted to prevent for the benefit of future generations. For Damm, the advantages of pasigraphy were like those of the newly built railways (1876: p. vi). These people had the same humanistic seriousness as the seventeenth-century scholars, but more on the political and economic than on the philosophical level.26

The blueprint of an artificial language—be it in the universal language or in the scheme of pasigraphy—consists of a presentation of the universal signs and of a systematic arrangement of words as ideas of meanings to be expressed by them, and finally a general grammar to combine these signs to form orderly sentences. Our (p.313) interest in pasigraphy is confined to the systematic arrangement of words as ideas of meanings because, taken together, they are a universal thesaurus applicable to all languages. In this sense, pasigraphy is mentioned by Roget in addition to the onomasiological Sanskrit dictionary of Améra Coshá (Hüllen 1999a: 32–3) and John Wilkins’s Tables’. Although Roget criticizes all three of them, his mention shows what was in his mind when he was devising his own ‘Plan of Classification’. Maimieux’s semantic system

Maimieux, J. von [Maimieux, Jean de]: Pasigraphie, XXX oder XXX, Anfangsgrunde Der Neuen Kunst-Wissenschaft In Einer Sprache Alles So Zu Schreiben Und Zu Drucken, Dass Es In Jeder Andern Ohne Überset-zung Gelesen Und Verstanden Werden Kann. Erfunden Und Verfasst Von J. *** von M*** […]. XXX Erste ausgabe, die, wie die franzoesische; originalausgabe ist [!]. XXX. Zu Paris, Im Bureau der Pasigraphie, rue Notre-Dame de Nazareth, numéro 118. 1797.

4, 62 pp. BL 12902.f.l3.27

Identical with:

Pasigraphie, XXX Ou XXX, Premiers Élémens Du Nouvel Art-Science D ‘Écrire Et D ‘Imprimer En Une Langue De Manière A Être Lu Et Entendu Dans Toute Autre Langue Sans Traduction; Inventés Et Rédigés Par J.*** De M***, Ancien Major D’Infanterie Allemande. XXX. Premiere édition, originale comme l’ édition en langue allemande. XXX. A Paris, Au Bureau de la Pasigraphie, rue Notre-Dame de Nazareth, numéro 118, 1797.

4, 63 pp., BL RB 23.b.95. ‘XXX’ indicates pasigraphie text of varying length.

(p.314) In another publication of 1801, the author developed his writing system into a speaking system called pasilaly, just as Wilkins had developed his ‘character’ into a ‘language’.

In its philosophical foundations, the semantic system of pasigraphy is quite similar to that of Wilkins’s ‘Tables’, but it is rather different from them in layout and technicalities. Similar is the recourse to a classification which claims to follow common sense (‘gemeinen Menschensinn’) and the rational mind using analogy (‘ein durch Analogie aufgeklaerter Verstand’, Maimieux 1797 a: 6). For the author, common sense and analogy guarantee the logical path from genus to species and from the known to the unknown. However, this is not knowable a priori as it was for Wilkins. Rather, pasigraphy allows for many semantic systems of lexis, but one must choose a single system, so the rational mind (‘Verstand’, 25–6) can support the memory when learning this language.

The author distinguishes three semantic schemata (‘Hauptrubriken’) into which all possible words are to be placed: (i) the list (‘Verzeichnis’), (ii) the minor nomenclature (‘kleiner Namengeber’), and (iii) the major nomenclature (‘groβer Namengeber’).

(i) consists of two frames (‘Rahmen’) with six columns (‘Kolumnen’) each. Each column is broken down into six paragraphs (‘Abschnitte’) with six lines (‘Linien’) each. Consequently, the list has 432 lines, i.e. entries.28 (ii) consists of twelve frames with six columns each (totalling 72). Again, each column is broken down into six paragraphs with six individual lines, amounting to 2,597 entries in toto. Finally, (iii) consists of twelve classes (‘Klassen’), consisting of six columns each (totalling 432). Their paragraphs and lines amount to as many as 15,552 entries. Although the author does not mention these numbers, they in fact delimit the lexical possibilities of his written language, (i) lists words which combine phrases and complete them emotionally (‘leidenschaftlich’), i.e. conjunctions, adverbials, and interjections, (ii) is a collection of lexemes used for objects, actions, ideas, and inclinations by speakers in families or society and among friends when engaging in business, trade, and transactions in banks (‘Gegenstande, Handlungen, Ideen, Neigungen; Familien, Gesellschaften, Freunde; Geschäfte, Handel, Bank’).

(p.315) (iii), finally, is a collection of terms from the arts and the sciences. According to a footnote (1797a: 7), the word lists of the major nomenclature are merely a fragment of a comprehensive work which ‘scientists, physicists, mathematicians, geometers, astronomers, moralists, and philosophers’ from all the countries of Europe have promised to put together. It is announced for a future volume which, however, never appeared.

The axis of this semantic system is the twelve columns of the list (i). Their semantic determinations reappear in the frames of the minor nomenclature (ii) and the classes of the major one (iii). This means that the columns of (i), the frames of (ii), and the classes of (iii) have the same headings. These are:

material substance, position, change

(i) col. 1, (ii) fr. 1, (iii) cl. 1

the plant kingdom

(i) col. 2, (ii) fr. 2, (iii) cl. 2

the animal kingdom

(i) col. 3, (ii) fr. 3, (iii) cl. 3

the physical human being

(i) col. 4, (ii) fr. 4, (iii) cl. 4

the sensitive and rational human being

(i) col. 5, (ii) fr. 5, (iii) cl. 5

the pious and social human being

(i) col. 6, (ii) fr. 6, (iii) cl. 6

craftsmen, trade

(i) col. 7, (ii) fr. 7, (iii) cl. 7

the arts, agriculture,

(i) col. 8, (ii) fr. 8, (iii) cl. 8

linguistics, arithmetic, the sciences

(i) col. 9, (ii) fr. 9, (iii) cl. 9

time, eras

(i) col. 10, (ii) fr. 10, (iii) cl.10

personal (pro)nouns, possessives

(i) col. 11, (ii) fr. 11, (iii) cl.ll

(pro)nouns of place29

(i) col. 12, (ii) fr. 12, (iii) cl.12

The intentions of this system are obvious, if not always convincing: the list contains structure words, the minor nomenclature the words (p.316) of general communication, and the major nomenclature the terms of the arts and sciences. The first four columns/frames/classes comprise words for objects, plants, and animals, including humans, i.e. the traditional ontological levels of the world. The next three comprise words for the various faculties of humans, the traditional ‘souls’, and their functions in society. Two more columns/frames/ classes are devoted to the classical domains of society, and the last three to linguistic structures, among them deixis. All these semantic areas have their traditions in onomasiological lexicography and encyclopaedic literature, if under different headings.

The meanings of individual words are constituted by the system with its three subsystems. The signs of pasigraphy, the so-called Gammen, do not indicate meaning: they indicate the locus in the system which is the meaning. Consequently, words are represented by three, four, or five Gammen: in (i) indicating column, paragraph, and line; in (ii) indicating frame, column, paragraph, and line; and in (iii) indicating class, frame, column, paragraph, and line. In each word, this sequence of Gammen is naturally ordered from the general to the specific and individual.30 Consequently, the system needs twelve signs, which can appear up to five times in one word. For words from the list, however, only six signs are needed, so that the language user can choose among two. What can be called the semantic value of each Gamme, therefore, depends on its shape, which is arbitrarily assigned to the various loci in the system, and on its position in the word (first, second, third, perhaps fourth or fifth). This is an algebraic principle, because the value of a digit in a number depends on the digit itself and on its position relative to other digits, just as 2 may equal two, twenty, two hundred, etc. (see below).

(p.317) The signs designating the position of a word and hence its meaning are prefixed by signs which determine the sense of the meaning and are suffixed by signs which give grammatical markers. The former include signs for action, instrument, genus, etc. They are comparable to Wilkins’s transcendental particles. The latter include signs for gender, number, and, most important, case. Maimieux gives neither a list of prefixes nor one of grammatical markers. His grammatical axiom is that, besides the (structure) words of the list, a language contains only nouns and three basic verbs, namely to become, to be, and to end being. By suffix derivation the nouns can be converted to adjectives and with the help of the three basic verbs all adjectives can be converted to verbs. With tongue in cheek, the author observes that experts in the sciences (‘Liebhaber einer grundlichen Wissenschaft’, 58) can ponder this, but ordinary people should not worry.

Maimieux claims that his system is unalterable. Users of pasi-graphical texts are requested not to run through this flight of logical determinations from the general to the specific, but to look up the words in the various schemata and memorize them mechanically (17, 20). The semantic cross-references between columns, frames, and classes will help them to do so (25–6).

Taken literally, this coherent system on algebraic principles limits the lexis of any language to exactly 18,576 entries, i.e. lexemes. But Maimieux obviously does not believe his system is that perfect. Referring to the list, he admits that lexematic entries may not fit perfectly. Strict exactness, he says, is impossible and would contradict the genius of language.31 Even more telling is this remark:

Wenn die Stufenleiter oder der Maasstab der Begriffe, die den Woertern derselben Sprache entsprechen, nach Entfernungen, nach Groessen, oder nach ganz gleichen Bedeutungen eingerichtet werden koennte, so wuerde es sehr leicht seyn, sie symmetrisch zu ordnen. Wenn sich in ihrem Stammbaume nicht die sich durchkreuzenden Aeste oder die Knoten, die zugleich Zweige und Wurzeln treiben, in einander verschlaengen; so (p.318) koennte man von dem ersten Begriffe zum letzten, von dem Stamme zu den Blaettern und Blüthen übergehen, und das Ganze würde einen sehr regelmässigen Plan darstellen. Aber der Begriff, der auch noch so abgerissen, und noch so vereinzelt zu sein scheint, haengt von alien iibrigen ab, führt zu denselben, entsteht daraus und erzeugt sie wieder. (24)

If the ladder or the measure of terms which corresponds to the words of one language could be constructed according to distance, size, or identical meanings, it would be very easy to give them a symmetrical order. If the branches which cross each other or the nodes which sprout twigs and roots did not intertwine in their pedigree we could move from the first term to the last, from the stem to the leaves and flowers, and the whole would make a most regular plan. But a term, however precisely delimited and individual, depends on all others, leads up to them, is created out of them and creates them in turn, (my translation)

Obviously, the idea of a general semantic interdependence of lexemes is not a purely mathematical notion. It also pertains to the less strictly systematized relations, diachronic and synchronic, between the words of a language, for example between synonyms. Indeed, in the 18,576 loci of his schemata, Maimieux rarely enters just one lexeme; in almost every case he enters several. Thus he admits the existence of synonyms as a natural phenomenon of lexis. Examples

Using a modern analogy, we can say that the list and the two nomenclators are organized like matrices of varying complexity. In the vertical dimension, we have the paragraphs and lines, marked typographically; in the horizontal dimension we have the columns, frames, and classes, all of which have horizontal headings. As an illustration of this abstract schema, note the following examples:


Ninth column: linguistics, arithmetic, sciences [Sprachlehre, Rechenkunst, Wissenschaften]32

First paragraph:

line 1: concerning [In Betreff (dieses)]

line 2: with regard to the question, in doubt [Bey der Frage, im Zweifel]

line 3: in consequence, following [nach Maasgebung, auf d[em] Fuss]

(p.319) line 4: the much more [um so viel mehr]

line 5: in relation to [nach Verhaeltnis]

line 6: most of all, in spite of all this [vor allem, bey allem dem]

Second paragraph:

line 1: after, due to [nach, vermoege]

line 2: on condition, apparently [vorausgesetzt, es scheint]

line 3: on condition, although [mit dem Beding, gleichwohl]

line 4: with (by) [vermittelst]

line 5: and yet, regardless [und dock, ohne Rücksicht]

line 6: consequently, therefore [folglich, daher]

[four more paragraphs]

Minor nomenclature

Ninth frame: linguistics, arithmetic, sciences [Sprachlehre, Rechenkunst, Wissenschaften], second column: writing and reading [Schreiben und Lesen]:

1st paragraph:

line 1: writing, handwriting [Schrift, hand[s] schrift]

line 2: handwriting, flowing handwriting [Current[schrift],fliessende [Schrift]]

line 3: round, middle, clumsy hand [runde., mittel., grobe [Schrift]]

line 4: draft, make out [Ausfertigung, abfassen]

line 5: daubing, scrawling [klek[s]ige Schrift, schmiererey]

line 6: scribbling, daubing [Gekri[t]zel, klecks]

Second paragraph:

line 1: spell, decipher [Buchstabiren, entziffern]

line 2: read, read fluently [lesen, flies[s]end [Lesen]]

line 3: dictate, write after dictation [dictiren, nachschreiben]

line 4: genuine [object], originality, schema [Original, Originalitaet,vorschrift]

line 5: copy, clean, write a fair copy [abschreiben, rein, ins Reine[schreiben]]

line 6: collection, make, wrap [?] [Sammlung, machen, einschlagen] [four more paragraphs]

Major nomenclature

Ninth class: linguistics, arithmetic, sciences [Sprachlehre, Rechenkunst, Wissenschaften], first frame, second column: the arts with reference to light, fire, water, and air [Kunste, die sich auf Licht, Feuer, Wasser und Luft beziehen]:

(p.320) First pararaph:

line 1: optics, dioptrics, catoptrics [Optik, dioptr[en], catoptrfen]]

line 2: vision, brow, see [Sehstrahl, augenbog[e]n, seh[en]]

line 3: incidence, reflect back, diffraction [Einfall, zuru[e]kpra [llen], brechung]

line 4: field, focus, to focus [Feld, Brennpunkt, zusammenlauf[e]n]

line 5: glimmer, to blind [schimmern, blenden] line 6: dial, hand of dial [Sonnenuhr, zeiger der [Sonnenuhr]]

Second paragraph:

line 1: spectacles, sun glasses [Augenglas, daempfbrille]

line 2: magnifying glass, microscope, telescope [Vergroes[serungs] glas, micro[scop], tel[e]scop]

line 3: lens, prism, spectacles [Linse, Prisma, Brille]

line 4: darkroom, fanlight [Dunkle Rammer, Oberfenster]

line 5: optics, magic lantern [Optik, Zauberlaterne]

line 6: spun glass, acromatic [Feinglas, acromatisch] [four more paragraphs]

These examples, if extended, give a bird’s-eye view of the thesaurus as the semantic corpus of pasigraphy. Identifying the meaning of one particular word is, basically, the same logical procedure as observing the system from the opposite perspective. The name England, for example, is to be found in the minor nomenclature (i.e. non-scientific language), twelfth frame (i.e. local pronouns, in this case proper names), fourth column (i.e. description of the earth, world, Europe [Erdbeschreibung, Welt, Europa]), fifth paragraph, line 1: England, GroBbritannien.

As always in schemata which demand a lengthy logical deduction from the general to the specific, it is difficult to deduce the concrete entry from the abstract system. In particular, the order of paragraphs and lines in the three schemata is, in many cases, not easy for the present-day reader to understand. The (linguistically) pragmatic rules which ultimately govern the selection of entries vary from frame to frame, following the inherent character of the various topics. A generic nomination followed by a special one is, of course, the most frequently used arrangement. But even this takes on quite different forms. On average, there are between two and four lexemes on one line, i.e. in one slot. Although Maimieux does not use the term, they must be considered synonyms, because they fill the same locus in the system. Sometimes, however, they are antonyms. In many cases, the (p.321) first lexeme of a line is printed either in capitals or in italics, which indicates that it is a keyword. But there is no strict system in this.

Note the following conspectus of entries concerning weapons, armoury (major nomenclature: sixth class, second frame, fifth column). The sense of its arrangement is in some loose way obvious, but not generalizable. (Only the first lexeme of each line is given.)

First paragraph: rifle, arrow, pike, hammer, catapult, launch [G[e]wehr, Pfeil, Picke, Hammer, Baliste, spannen]

Second paragraph: sword, sabre, dagger, mounting, sword-belt, tassle [Degen, Saebel, Dolch, Griff, Wehrgehaeng, Degenquaste]

Third paragraph: armour, shield, breastplate, spiked helmet, uniform, quiver [Rüstung, Schild, Brustplatte, Bickelhaube, Uniform, Koecher]

Fourth paragraph: musket, sling, pistol, steel, cock, stock [Musketenfeuer, Schleuder, Pistole, Schlos[s]blech, Besatzung, Schaft]

Fifth paragraph: artillery, cannon, small shot, cartridge, to load, powder [Artillerie, Steinstück, Schrot, Ladung, lad[e]n, Pulver]

Sixth paragraph: carriage, fire, to fail, ricochet, grapeshot (?), bayonet [Wagen, Feu[e]r, versagen, Prel[l]schus[s], Alteis[e]n, Bajonett].

It is not known whether a practical test of this onomasiological arrangement of words was ever made. There were certainly books which worked with numeric systems.33 Janne Damm (1876: p. vi) was courageous enough to write, ‘Dictionaries are being made and businessmen in Germany, England, France, Russia and Sweden will be able to commence pasigraphic correspondence this summer.’ That was in 1876. For Roget, pasigraphy was certainly a powerful indication that what he wanted to do was possible, if not in the same way. As he read French and German all his life, the pasigraphic word lists will have given him plenty of inspiration. (p.322)


(1) See Ch. 1.1.

(2) Hüllen 1999a, on the topical tradition in English lexicography between its beginnings in glossaries (c.800) and 1700. I shall borrow from this monograph without giving detailed references; that work should be consulted for further arguments and for analyses of sample dictionaries. In the lexicographical works treated here and in the major points of my arguments, I do not go beyond the limits of the older work. In the present context, I merely wish to provide the necessary link between my description of the development of the English synonym dictionary and my analysis of Roget’s Thesaurus. This is the only innovation of the present chapter.

(3) See Ch. 3.1.

(4) Note that the meaning of ‘meaning’ is not just ‘reference’ or ‘concept’, but the whole semanticpotential that a word acquires by itself, by the text in which it appears, and by the situation in whichan utterance is made.

(5) For bilingual dictionaries, the dichotomy between these two types is clear only in older works. Inmore recent ones, it is not so sharp; because these dictionaries usually appear in pairs, for example, as both an English-German and a German-English version. The English native speaker can thereforefirst use the English-German version to find a variety of German expressions for some meaning andthen the German-English one to check the unfamiliar German lexemes. Hence there is no need for an onomasiological dictionary.

(6) Obviously, these remarks presuppose a certain type of modern society, namely the liberal state, where significant disagreements between world-views are possible. There are, of course, also societies where this is not the case. The mutual dependency of dictionary types and societal conditions is closer than we usually realize.

(7) There are natural exceptions to this, of course: lexemes with varying derivations or compounds with the same initial element. In so-called nesting dictionaries, they are subsumed under one entry.

(8) See Ch. 5.2.1.

(9) It is confusing that Wilkins speaks of ‘things and notions’, but just as often of ‘things or notions’,leaving it undecided whether notions always stem from things and go together with them, or whetherthere are also notions as such, for example, abstract notions, as independent products of our thinking. This is, of course, the old problem of nominalism vs realism in logic. One would have to study thewhole of Wilkins’s book to sort this vagueness out, and, as Wilkins was not a subtle philosopher, itcould very well be that he is indeed undecided on this point. But we can safely assume that Wilkinsobviously conceived of two kinds of notions, one corresponding to things like trees or horses (hisexamples, 1668: 20), the other one functioning as predicaments (i.e. the categories of Aristotle) andto ‘such matters, as by reason of their Generalness, or in some other respect, are above all those commonheads of things called Predicaments’ (24) (i.e. notions like being, thing, notion itself, moreover genus, difference, and species, hence the so-called predicabilia). In genera I, II, and III of his ‘Tables’, he enumerates this latter kind of ‘Universal notions’ (23), which he then explicitly intends to use as ordering schemata for tables V to XL. Again, nobody so far has investigated whether this was done coherently, i.e. whether the notions of tables I, II, and III do indeed reappear as ordering categories in later tables.

(10) I am thinking, for example, of the books by Dudley Fenner (?1558–1587), Thomas Blundeville (fl. 1599), or Samuel Smith (1587–1620). But there may be others (Hüllen 1999a: 284–92).

(11) I am thinking, for example, of the books by Dudley Fenner (?1558–1587), Thomas Blundeville (fl. 1599), or Samuel Smith (1587–1620). But there may be others (Hüllen 1999a: 284–92).

(12) With reference to this kind of onomasiology, I have spoken several times of a ‘speculativetradition’. This is meant to express a parallelism with so-called speculative grammar. ‘Speculative’means that grammar as much as dictionaries tends to be a mirror (speculum) of the order of the world. In dictionaries, this is to be achieved by the macro structural arrangement. The term ‘speculative’ ishistoriographically descriptive, i.e. no grammarian or lexicographer has ever used it in this sense. However, it was a fixed term in encyclopaedic books.

(13) See Ch. 7.3.2.

(14) For example, John Harris, Lexicon Technicum; or, an Universal English dictionary of arts and sciences, London, 1704; John Barrow, A New and Universal Dictionary of Arts and Sciences, London, 1751; A Society of Gentlemen, New and Complete Dictionary of Arts and Sciences, London, 1754; and Temple Henry Crocker, Thomas Williams, and Samuel Clark, The Complete Dictionary of Arts and Sciences, London, 1764.

(15) In fact, the German counterpart of the Encyclopédie, with its twenty-five volumes is even more voluminous: Johann Heinrich Zedler’s GroBes vollstcindiges Universal-Lexikon aller Wissenschaften und Künste welche bishero durch menschlichen Verstand und Witz erfunden und verbessert wurden, Halle and Leipzig, 1731–50. In 1796, Friedrich Arnold Brockhaus published in Germany his first Konversations-Lexikon für die gebildeten Stánde, whose six rather thin volumes were written by Renatus Gotthelf Lobel. It has since grown into an enormous publication of twenty-five volumes whose last edition was issued in 1996. In German, the term ‘Brockhaus’ now has the generic meaning ‘encyclopaedic dictionary’.

(16) See, for example, the second book of The Advancement of Learning (Bacon 1857–74: iii, 321–491).

(17) The quotation is taken from Bentham’s ‘Appendix Nr. VII’, titled ‘Hints towards a system andcourse of Technology, from Bishop Wilkins’ Logical work, published by the Royal Society, A° 1668,under the title of "An Essay towards a Real Character, and a Philosophical Language"’.

(18) See Ch. 2.3.

(19) 19 See Ch. 2.2.

(20) Designations in square brackets have only a logical function; designations in italics name the scientific disciplines, i.e. the subjects of the teaching envisaged.

(21) Recent so-called cognitive linguistics (semantics) seems to hark back to the old onomasiological pattern by assuming that lexis is stored in a ‘cognitive’, i.e. meaningful, way in the human mind.

(22) Words of the syntagma in Roman fonts, newly introduced words in italics. [+number] indicates the number of words introduced in addition to the one mentioned.

(23) John Wilkins, too, embeds his lexemes and their synonyms in such defining sentences, if on a much higher intellectual level and certainly not for teaching purposes. He explained the configuration of semantic features in this way.

(24) Roch A.C. Sicard was famous for teaching the deaf and dumb. The entry in Alston has notes as to where copies of the publication of 1795 can be found. However, an anonymous review (Anonymous 1796) represents the book as being published in 1796.

(25) In Roget (2002) pasigraphy is placed between ‘diplomatic language, international 1.,International Scientific Vocabulary, Basic English’ and ‘artificial language, Esperanto, Ido, Interlingua[…]’ [¶557, p. 303).

(26) The following is de Maimieux’s conspectus of the advantages of a universal use of pasigraphy. It is interesting not only for his expectation of commercial advantage, but also for his plan ofestablishing an interlingual norm between languages:

1°. Mehr Verbindung in der Gesellschaft und im Handel, sowohl zwischen einzelnen Menschen, als zwischen Voelkern; 2°. Eine Art von Glossometer, der dazu dient, die woertlichen Uibersetzungen in dem gewoehnlichen Briefwechsel zu ergaenzen, die Unrichtigkeiten der gelehrten Uibersetzungen [!] zu berichtigen, und den Sprachen einen allgemeinen Maasstab zu verschaffen; 3°. Mehr Schnelligkeit, Leichtigkeit und Richtigkeit, mehr Sparsamkeit, Geheimnis und Sicherheit in diplomatischen, militaerischen, bürgerlichen und Handels-Geschaeften; denn derselbe Dienstmann, der nur seine Sprache zu verstehen braucht, kann zehn Briefe pasigraphiren, die in ebenso viel Sprachen gelesen werden koennen; 4°. Mehr Einfoermigkeit in der Bekanntmachung der Dinge, die die Menschheit betreffen; 5°. Endlich mehr Mittel, eine grosse Anzahl von Lehrern, Meistern, Schriftstellern, Schriftstechern, Schrifftgiessern und Buchdruckern zu unterhalten, die sich damit abgeben würden, die Elementarbuecher in verschiedenen Gegenstaenden zu vervielfaeltigen; denn diese, wenn sie einmal in einer Sprache erschienen sind, koennen dann in alien andern gelesen werden. Dies sind Zweige von Industrie, die doch von aufgeklaerten Regierungen unterstuetzt zu werden verdienen. (1797, 1–2)

1. More contact in society and in commerce between individuals as well as between peoples; 2. A kind of linguistic metre serving to complement literal translations in the exchange of common correspondence, to correct the faults of learned translations, and to provide a general yardstick for languages; 3. More speed, ease, and correctness, more thrift, secrecy, and security in diplomatic, military, civil, and commercial business; because the same assistant, who needs to understand only one language, can render ten letters into pasigraphy which can be read in as many languages; 4. More uniformity in the publicizing of things of concern to mankind; 5. Finally, more money to maintain a great number of teachers, masters, writers, typecutters, type-founders, and printers who would be busy copying elementary readers on various topics; because they can be read in various languages, once they have been published. These are branches of an industry which deserve to be supported by enlightened governments, (my translation)

(27) The following analysis is based on this edition.

(28) The number of lines/entries does not equal the number of lexemes, as several lexemes can be included in one entry (see below).

(29) Translations mine. The original German terms are: col. 1: Materie, Stellungen und Veränderungen; col. 2: Pflanzenreich; col. 3: Tierreich; col. 4: Der physische [Mensch]; col. 5: Der mit Sinnen und Verstand begabte [Mensch]; col. 6: Der fromme und gesellige Mensch; col. 7: Handwerker, Handel; col. 8: Künste, Ackerbau, Vergnügungen, Spiele; col. 9: Sprachlehre, Rechenkunst, Wissenschaften; col. 10: Zeit, Epochen; col. 11: Persönliche und ‘Besitzungsfurwörter’; col. 12: Fürwörter, den Ort betreffend.

(30) Note the following passage:

Unter pasigraphischer Ordnung [der Wörter] versteht man hier aber eine Classification, die bloss nach dem gemeinen Menschensinn und von einem durch Analogie aufgeklaerten Verstande entworfen ist. Es gehen naemlich hier die so wenig als moeglich angestrengte Aufmerksamkeit und das durch alle Erinnerungsmittel miteinander von der Gattung zur Art, und von der Art zu den Individuen, oder von dem einfachen [!] zu dem Zusammengesetzten, oder von dem bekannten zu dem weniger bekannten [!] ueber, je nachdem die auffallendsten Beziehungen der Ideen aufeinander diesen oder jenen Gang zu nehmen gestatten. (5–6; see also 26–7. The spelling in this book is often inconsistent, as the quotation shows. I shall not indicate this further by exclamation marks.)

The pasigraphic order [of words] is here understood to be a classification which is constructed by common sense and a rational mind using analogy. Here attention, as little strained as possible, and together with all kinds of retrieval, transcends from the genus to the species, and from the species to the individual, or from the simple to the complex, or from the known to the less known, depending on which step the most outstanding relations of ideas allow [us] to take (my translation).

(31) ‘Eine strenge Genauigkeit ist hier unmöglich; denn sie vertraegt sich nicht mit dem Genie aller Sprachen, und würde auf sehr ungleiche Untherabteilungen führen […]’ (‘Strict exactness is impossible because it does not agree with the spirit of all languages and would lead to quite different subsections’; my translation).

(32) I have not tried to imitate the eighteenth-century flavour of the German original in my translations. There are many blemishes in the book.

(33) Note, for example, Bachmeier (1870). The book, ‘Dedicated to all nations. Humanité et patrie’, has one grammatical and two lexicographical parts, all numbered separately. Between them it lists various European and Asian alphabets, even the ‘alphabet of the Japanese’ [!]. The lexicographical part, called ‘English pasigraphical dictionary for composing pasigrams’ (1–188), consists of lists of words complemented with numbers, and then lists of numbers complemented with words, called ‘Pasigraphical English dictionary for deciphering pasigrams’ (1–169). Note the first five examples each: A, An—1, Aback—2402, Abalienate, to—3049, Abandon, to To leave [!]—14, 3619, Abatement, Discount—46; 1—A, An, 2 Eel, 3 Carcass, 4 To alter. Alteration, 5 Variety. Note how the problems of synonymy and polysemy are dealt with. There is no explanation of the selection or the sequencing of lexemes.