SEMANTICS AND THE DICTIONARY (LEXICOGRAPHY)

The relationship between linguistic semantics and lexicography, focusing on

what linguists have to offer to lexicographers, and vice versa, in the description of

lexical meaning. It is hard to do justice to all the many strains and positions in one

brief paper, so the reader will have to excuse a hasty generalization here and there

as well as a certain degree of selectiveness.

The relationship between semantics and lexicography lies in the fact that the

term lexicography may variously refer to at least two different activities: firstly,

lexicographic practice (dictionary-making), and, secondly, theoretical

lexicographic reflection and lexicographic research, which currently is often

termed metalexicography.

A. lexicography part of linguistics

The question that probably needs to be asked at this point is whether

lexicography should be seen as part of linguistics. Many linguists seem to

assume without argument that lexicography fits comfortably within

linguistics, and although some have actually argued the point – witness

Hans Meier’s 1969 paper with the telling title “Lexicography as Applied

Linguistics” (Meier 1969) – it may be that the inclusion of lexicography

under linguistics is more of a reflex habit and a matter of academic politics

than well-considered judgement (Burkhanov 1997: 70).

As has been noted (Piotrowski 2001), in this country the profession

1
of lexicographer is not (yet?) well-established as an autonomous one.

Therefore, it is perhaps more common than in, for instance, the US or the

UK for linguists to double as lexicographers. If the above is true, a

tendency

to associate linguistics with lexicography would be understandable.

Piotrowski (1994: 10), as well as Burkhanov (1997), see lexicography

as an autonomous domain, one which employs the findings of linguistics,

but contributes its own methods as well. Bańko (2001) basically upholds

Piotrowski’s view. He points out (2001: 13) that good lexicography must

have a solid grounding in language data, preferably large language corpora

(here he owes much to John Sinclair and the Cobuild project). I think this

is not necessarily a point of difference though, since increasing numbers of

mainstream linguists are also turning to corpora, appreciating the role of

systematic and comprehensive language data in the study of language.

Having said that, however, I also believe that the corpus approach to

linguistic study should rather be seen as a contribution of lexicography to

linguistics than the other way round.

Bańko (2001: 17-18) also points to an interesting inconsistency: some

of those linguists who see lexicography as part of linguistics, expect

dictionaries to be more or less uniform in their description of language

data, which is quite the opposite to what they would expect of non-

lexicographic linguistic works.

2
 Frawley (1993: 1) uses an interesting metaphor when describing the

relationship of (Montague tradition) semantics and lexicography:

“Semantics is to lexicography what economics is to accountancy”.

Further, numerous other positions on the disciplinary status of

lexicography are possible. To mention just two of the more interestin ones,

Spevak (1987) treats both lexicography and semantics as part of

lexicology, while Charles Fillmore describes “the effort to analyze the

frame-setting background of a word’s meaning” in modern English

monolingual learners’ dictionaries as “a kind of ethnography” (Fillmore

2003: 284).

Another complication in interpreting the various positions in the

relationship between semantics and lexicography lies in the fact that the

term lexicography may variously refer to at least two different activities:

firstly, lexicographic practice (dictionary-making), and, secondly,

theoretical lexicographic reflection and lexicographic research, which

currently is often termed metalexicography. The new term is used

precisely to avoid such confusion. The contribution of semantics to

lexicography

B. The contribution of semantics to lexicography

Semanticists have sometimes volunteered suggestions concerning the

description of meaning in dictionaries, but often these suggestions have

been received less than kindly by lexicographers, and have been branded

as unrealistic – unfortunately, not without good reason.

3
 On the other hand, semanticists have been known to accuse

lexicographers of having no theoretical foundation for what they are

doing, such as Wierzbicka’s (1985: 5) much-quoted charge:

"Lexicography has no theoretical foundations, and even the best

lexicographers, when pressed, can never explain what they are doing, or

why".

It was the inadequacy of traditional linguistics as a theoretical

foundation for lexicographic practice that has prompted lexicography into

developing its own theoretical discipline: metalexicography, or theoretical

lexicography. Apart from linguistic considerations, metalexicography

takes account of the extralinguistic context of dictionary-making and

dictionary use, which theoretical linguistics resolutely ignores. Thus, the

rise of metalexicography is likely an indication that traditional linguistic

accounts were generally seen as insufficient and limiting. Wierzbicka

herself makes this very point forcefully: “If modern linguistics were to be

judged by the contribution it made to lexicography, it would be hard to

understand why linguistics is said to have made dramatic advances in

recent decades” (Wierzbicka 1993: 45). 7 On the other hand, Wierzbicka’s

own efforts to offer such a contribution are taken to pieces by Hanks

(1993), who systematically demonstrates how very unrealistic and

impractical the linguists’ view of lexicographic practice may be, if they

have not themselves extensively engaged in real dictionarywriting. In fact,

as Zgusta nicely puts it, “the best practitioners among us are those

4
 lexicographers who have the ability to cope with what is, or theoretically

should be, intractable.” (Zgusta 1993: 136)

To point to but one example of the rift between semanticists’

proposals and real-world lexicography, some semanticists give themselves

the privilege of abstaining from defining the meaning of a subset of lexical

items (usually referred to as semantic primitives or undefinables). Now,

lexicographers are constrained in their work by the expectations of

dictionary users as well as the restrictions imposed by the publishers.

Because of this, lexicographers normally cannot afford the convenience of

leaving out the “troublesome” words from the dictionary.

C. Semantics and lexicography work hand in hand

The relationship between semantics and lexicography: though rarely,

it is probably possible for theoretical semantic insight to go hand in hand

with lexicographic practice: a shining positive example is the original

Meaning <=> Text project of the Moscow linguists (now developed into a

full-blown Meaning-Text Theory), whose relative success suggests that

co-operation, or even integration, between semantics and lexicography is

possible, as evidenced in Igor Mel'čuk’s work (1974; 1981; 1988). One

would be remiss not to mention the corpus revolution, which has had a

major impact on lexicography, resulting in new approaches to

lexicography, meaning, and the advent of Corpus Linguistics, which

however is not a unified theoretical model so much as a bundle of

methodological approaches. Charles Fillmore’s Frame Semantics model,

5
 which combines the insights of the more traditional semantic functions and

the more modern cognitive schemata, also shows good promise for

successful lexicographic application (Atkins 2002). In fact, a number of

cognitive-linguistics approaches do not seem to shy away from the

description of actual lexical meaning, as has been the case in most other

linguistic paradigms (as argued in Section 1 above).

There is, then, some benefit for lexicography to be drawn from

linguistic semantics, but clearly not all strains of semantic research render

themselves to lexicographic application with equal facility.

D. Conclusion

Wierzbicka’s charge that lexicography has no theoretical foundation

has to be refuted: there is now a solid, and growing at a healthy pace,

theoretical foundation to lexicographic practice. This foundation is

metalexicography, which arose to meet those many needs of

lexicographers and dictionary users that were not being properly addressed

by linguists. Although undoubtedly new developments in linguistics bear

some relevance to dictionary-making, metalexicography is less and less

likely for the latter to be seen as just a part of linguistics, as it grows,

branches out, and specializes. It is perhaps unavoidable that a troubled

relationship should end in divorce, but there is no reason this cannot be,

after all, amicable divorce. Let us hope that lexicographers will keep an

open mind to developments in linguistics, and that linguists will continue

to exhibit a healthy fascination with dictionaries.

6
7