3.0. Typological summary [49]

The history of lexicography is the story of its progress, often faltering, from the explanation of difficult words in the works of Attic poets through its long attachment as an integral part of grammar, to its gradual emergence as a related science with its own rules and principles. The scope of this study does not include a detailed scrutiny of lexicographical history, and we limit our observations in this chapter to: a) the scope and composition of the nomenclature, and b) the evolution of the techniques of presentation.

3.1. Consultability

The formal architecture of the text which is the dictionary is expressed on two planes: the vertical, which is the ensemble of entries (lemmata) making up the nomenclature, also called the macrostructure or macrotext, and the horizontal, which is the ensemble of information concerning the entries and presented in the form of articles, usually called the microstructure or microtext (Collignon/Glatigny 1978: 74; Wooldridge 1977: 97).

By reason of its bidimensional structure, a dictionary is also read in two directions: first, a vertical search of the macrostructure to locate an article, then, a horizontal reading of the article to find information. This bidimensional reading is called 'consultation' and the less time devoted to the search for information, the more 'consultable' is the dictionary (Wooldridge 1977: 97). [50]

The modern user of a general language dictionary expects to locate the entry he is looking for in his vertical search of the macrostructure. In other words, he expects that:

In 3.1.1. we identify the main classes of words represented in the nomenclature of lexica dating after the tenth century, and in 3.1.2. we look at some of the techniques of presentation of the macrostructure.

3.1.1. Classes Represented in the Macrostructure

It is not possible to know the extent of the entire lexicon of any past state of a language, and still less possible to determine the quantity of that lexicon which was recorded in various inventories. Early Greek and Latin glossaries contained only difficult or unusual words, and most of those documents were related to a selected author or a particular field, such as medical terms or flora and fauna. Even the more ambitious collections, although perhaps representative of the general lexicon, could not have been complete, at least insofar as the open classes are concerned (nouns, adjectives, verbs, adverbs). Added to this, of course, is the fact that, in many cases, only portions of the early lexica have come down to us and these, very often, in far from the original form.

Medieval lexicographers appear to have felt a desire to include as many entries as possible in the nomenclature of their dictionaries. In the preface to the Elementarium, Papias explains that he has "undertaken a work which has, to be sure, already been elaborated for a long time by many others and has also been added to and amassed by me more recently as best I could over a period of ten years". [51] Papias does not claim to have included all human knowledge but it is plain that he regards "this work which has been excerpted and compiled from all the writings which I have found" [52] as an extensive and important collection.

John Balbi, on the other hand, announces proudly "Incipit Summa quae vocatur Catholicon", and stresses its comprehensiveness in his preface. [53] John the Grammarian's aim is expressed in the title of his compilation, which he calls Comprehensorium.

Firmin Le Ver calls his work a "dictionary" and claims that it includes "all the words" appearing in the Catholicon, Hugutio, Papias, Brito, and other writers. [54] (This claim is not strictly true; Le Ver omits many entries included by his predecessors, and shortens definitions, as well. His work is an example of the transition from encyclopaedia to dictionary.)

Robins (1989: 212) explains: "In the grammatical analysis of languages words are assigned to word classes on the formal basis of syntactic behaviour, supplemented and reinforced by differences of morphological paradigms, so that every word in a language is a member of a word class". He continues (1989: 214):

Exhibit 26 (adapted from Wooldridge 1977: 83) is a model for subdivision of word classes. It is more appropriate as a tool for analysis of the nomenclature of modern dictionaries, and we have attempted to use it only to test the representative character of our corpus.

Exhibit 26: Word Classes of a Lexicon
Adapted from Wooldridge 1977: 83.
Translation is ours. Grammatical words

Lindemann (1994: 3) contends that the basic difference between dictionaries and glossaries is that dictionaries contain the central vocabulary, and particularly the common words, of a language, or of a specific field of a language, while glossaries contain difficult or unfamiliar words. Her claim is supported by the fact that grammatical terms (which must be considered among the most common of words) are generally absent from the macrostructure of early Latin glossaries.

Grammatical words are not numerous and are difficult to find in the macrostructure of medieval Latin lexica using a derivational approach to their compilation, e.g. the Panormia of Osbern and the Magnae derivationes of Hugutio, due to their basic organizing principle and to the fact that entries are in first-letter alphabetical order.

Prepositions, conjunctions and pronouns are included in the macrostructure of Latin and Latin-vernacular dictionaries whose first principle of organization is alphabetical, although the extent of their presence varies widely, depending on the dictionary.

Not surprisingly, given its 'specialist' content (i.e. difficult words from the Bible), few grammatical words are included in the nomenclature of Brito's Summa.

The status of 'glossary' is confirmed for the Latin-French Abavus, since the nomenclature generally includes few grammatical terms although, as may be expected, they occur more frequently in the later manuscripts (Ab4 and Ab5) than in the early ones (Ab1 and Ab2).

In contrast to the other bilingual lexica in our corpus, the language of the macrostructure of the Glossarium gallico-latinum (GGL) in ms. Paris, BN lat. 7684 is French. A substantial portion of its macrostructure consists of locutions, paraphrases and definitions, and a large number of grammatical words appear in them many times without having entry status. Taking the entry 'De quel gent ou de quel pais' as an example, neither 'de' nor 'quel' nor 'ou' occurs as an entry.

Only six conjunctions are included:

as are six prepositions: One pronoun appears as an entry in the macrostructure of the GGL: Soy mesmes - sese - pronomen.

Grammatical terms also occur countless times in the microtext of all medieval lexica without appearing as entries in the macrostructure. It would be long before the notion of a 'closed' text, in which every word occurring in the microstructure also appears in the macrotext, was realized.

Some sixty-five prepositions and seventy conjunctions are found in the macrostructure of the remaining six important dictionaries in our corpus: the Elementarium of Papias, the Catholicon of John Balbi, the Comprehensorium of John the Grammarian, Aalma, MPST, and DLV.

Appendix 3 shows the distribution in these six texts of a representative sample of each of these categories of grammatical terms. The results are much as might be expected: the less extensive dictionaries, Aalma and MPST, have fewer terms, and the DLV, which is the latest and the most highly organized dictionary of the group, has the greatest number, either as entries or sub-entries. The lack of a number of forms from the macrostructure of Part V of the Catholicon, which is the dictionary, may be accounted for by the fact that grammar is dealt with in the first four parts.

Demonstrative, relative, interrogative and indefinite pronouns are well represented, and the intensive ipse, ipsa, ipsum also occurs. Personal pronouns and reflexive pronouns are rare, which is an indication of Latin's synthetic structure. However, the suffix -met is added to personal pronouns as an intensifier, e.g. egomet, memet, semet, sibimet, etc.

The variable presentation of pronouns - from one dictionary to another and within the same dictionary - makes comparison difficult. In some cases, the paradigm of nominative singular forms is shown in one entry and inflected forms also occur as separate entries. In other cases, the nominative singular forms do not appear as one entry, but the masculine, feminine and neuter forms are found as separate entries. Affixes

The nomenclature of each of the six dictionaries discussed in contains a number of affixes, although the quantity varies widely among the texts. Prefixes make up the greater part of the affixes. They include five of the six Latin prepositions which are found only in composition: am, con, di, dis, and se. The sixth, re, occurs very often in the microtext of all the dictionaries in the group but does not have its own entry in any of them.

Several Greek prefixes, which also usually occur only in composition, have entry status, e.g. idro, iper, ipos, macros, sin, etc. Most of the free-standing Latin prepositions having entry status are used in composition as prefixes, for example: ab, ad, ante, circum, de, e, ex, extra and so on.

Suffixes found in the macrostructure of DLV include the enclitic conjunctions: -ne, -que, -ve, as well as the pronoun met. Compounds

Compounds are signalled in most texts in our corpus by the connectors componitur, compositum est and dictio composita. However, in Papias's Elementarium (where the components of compound words are not usually noted), the more common connectors are dicitur, nam, or 0. In Aalma the connector is almost invariably dicitur, and in GGL it is componitur.

Compounds occur in the corpus in two ways. Firstly, they may be included in the macrostructure as separate entries, as in the following examples:

Secondly, compounds appear in the microstructure of most dictionaries as examples of the use of an element of composition:

Wooldridge (1977: 83) points out that the compounds in the second category constitute a potential nomenclature. Whether they also appear in the macrostructure depends on the dictionary. Those in DLV, for example, are included more consistently than in any of the others. Those in the derivationally-organized dictionaries (i.e. Osbern's Panormia and Hugutio's Magnae derivationes) are usually not included.

[Next] -- [Table of Contents]