The term ‘terminology’ displays the same kind of ambiguity as many other linguistic terms: It designates both a subfield of the object area of the discipline – in this case, the set of terms – and a subdiscipline devoted to this subfield – here, the scientific study and formation of terms.

A term (also called terminus technicus ‘technical term’) is an expression designating a scientific concept. Commonly, two constraints are imposed on the concept ‘term’:

• A term is normally a noun, like modification and modifier. The terminological status of verbs like modify or adjectives like obligatory is considered derivative.
• A noun qualifying for terminological status must be a common noun. This is because proper nouns designate individuals; but the meaning of a term is a concept, not an individual.

Terminological conservatism

Probably the majority of scientists neglect terminology in both of the senses of this term. Something like the following reasoning prevails: Scientific theories are made of concepts; what matters is to get these right. Once you control the correct concepts, it does not matter how you call them. In certain circles, a rather contemptuous attitude is widespread:

1. Every author is free to use the terms he prefers.
2. Terms cannot be right or wrong; they must just be used consistently.
3. Therefore, controversies about terminology are proverbial disputes over words.

If questioned, advocates of #1 would readily add “as long as he defines them, of course.” In this lies the first error committed by these people: It is impossible to define all the terms of a scientific treatise. The far majority of the concepts being used, with their associated terms, must be presupposed, in science just as in any rational communication. Moreover, precisely those authors who advocate a position of terminological liberalism seldom define their terms properly. To the extent that they use terms which are not well-established in the discipline, the result is confusion.

The precise opposite of terminological liberalism, viz. terminological conservatism, is necessary in a scientific discipline: If an author uses a concept already existent in the discipline, then standard scientific behavior is to use the term hitherto associated with it. As long as he so does, he does not need to define the term. Depending on his hypotheses about the familiarity of the term to the readership, he may just use it without further ado. For example, 99% of the uses of the term ‘preposition’ in the discipline work in this way. Otherwise, a bibliographical reference may suffice. For instance, if the author is not sure whether readers will know the term ‘phone’, he may say “a phone (Clark et al. 2007: 34)”.

Needless to say, deviations from the default case may be justified. It is possible that an existent term does not fit the concept very well. Then the author may explicitly replace it by a better term.

If one distinguishes properly between term and concept, it is clear enough than terminological conservatism does not entail conservatism concerning scientific concepts. We understand the world surrounding us in terms of concepts and relations among them. It is one of the tasks of science to enrich our view and control of the world by providing useful concepts. It may then be necessary to find new terms for them. Terminological conservatism is concerned about continuity in the use of established and useful terms. Terminological liberalism ignores these – partly willfully –, entertaining the mistaken idea that, since research has to be innovative, earlier terms can be ignored. As a consequence, if the liberal author coins a new term for an existent concept, the reader will feel unsure whether something different is being meant; and if the liberal author uses an existent term in a new sense, the reader will likewise be disoriented. The result is confusion and stagnation of research instead of scientific progress.

Classical terminology

A considerable portion of American structuralists and their heirs have never developed a composed attitude towards the grammatical terminology that has come down to us from ancient grammar. Two attitudes prevail:

1. Either the terms designating categories of the classical languages are banned from newly described languages, on the supposition that every language has its own grammar and that the use of terms apt for the classical languages in the description of another language would imply that this has the same categories as those. This attitude has led to a horror termini classici. For instance, terms like ‘dative’ or ‘perfective’ are not found in such descriptions, and instead the authors say things like ‘recipient marker’ and ‘completive’, resp.
2. Or else the classical terms have been upgraded (Lehmann 2007): Since modern European languages like English lack most of the morphological categories that the classical terms apply to, these terms can be reused to designate constructions fulfilling a similar function. For instance, ‘dative’ is used instead of ‘indirect object’, and ‘passive’ designates a diathetic construction which demotes the underlying subject.

Both of the above practices lead to confusion. A rational attitude towards received grammatical terminology is as follows:

1. Presuppose that both the semiotic functions fulfilled by grammatical categories and their structural properties have to be captured by prototypical concepts. This is true even in the description of one language and a fortiori so whenever such a concept is used in the description of more than one language (Lehmann 2018).
2. Analyze the structure and function of the categories of the classical languages.
3. By typological methods, find the language-independent function and the language-independent structure under which the language-specific structure and function of the category in question can be subsumed.
4. Based on the result of #3, redefine the concept of the category in question in a more generic and prototypical way.
5. Keep the traditional term for the concept thus defined.

If this procedure were straightforward, it would probably have been applied more often. The most problematic step is #3. One of the important points here is the distinction between (the value of) a morphological category and a syntactic construction. Voice is a morphological category, diathesis is a (kind of) syntactic construction. ‘Dative’ is a morphological category; ‘indirect object’ is a syntactic function. These two levels have to be distinguished in the description of every language (more on this). One cannot, therefore, widen the scope of a received term for a morphological category to designate the syntactic construction that it helps to form. For instance, however usefully expanded, the term ‘dative’ has to designate a case marker; it cannot designate a component of a syntactic construction fuflilling a function otherwise comprising such a case marker.

Subjunctive and irrealis

The use of the term ‘irrealis’ observable since, roughly, since the 1980s may illustrate the misguided practice #1 above. In order to understand what has been happening, a general point may have to be premised: Mood is a morphological category; modality is a cognitive/communicative functional domain of languages. Observe that both of these concepts have already been widened with respect to their classical use, since additional moods (values of the category) have been subsumed under the former concept, and the latter concept no longer is limited to the functions of modal verbs, but comprises any function of the same kind.

Presupposing this, then the subjunctive is a mood whose basic function it is to indicate that the speaker is not asserting the proposition thus marked. Irrealis is the modality of a proposition marked as counterfactual. As usual with morphological categories, some languages like Latin, Yucatec and Cabecar have a subjunctive; others like Russian, Hebrew and Mandarin lack it. And as is the case by definition, every language has a means to convey irrealis modality (even if by a paraphrasis). There is no need to call the subjunctive (and much less the future, for that matter) ‘irrealis’. This is just sloppy terminology.

References

Clark, John et al. 2007, An introduction to phonetics and phonology. Malden, MA etc.: B. Blackwell (Blackwell Textbooks in Linguistics, 9).

Lehmann, Christian 2007, “On the upgrading of grammatical concepts.” Moerdijk, Fons & Santen, Ariane van & Tempelaars, Rob (eds.), Leven met woorden. Opstellen aangeboden aan Piet van Sterkenburg ... Leiden: Instituut voor Nederlandse Lexicologie; 409-422. [download]

Lehmann, Christian 2018, “Linguistic concepts and categories in language description and comparison”. Chini, Marina & Cuzzolin, Pierluigi (eds.), Typology, acquisition, grammaticalization studies. Milano: Franco Angeli (Materiali Linguistici); 27-50. [download ]