A catalogue of categories?

Linguists have their catalogue of languages, biologists are working on their catalogue of species, and chemists have their catalogue (“periodic system”) of elements. Full inventorization is clearly a desirable goal of science, and some kind of order in the phenomena is a prerequisite for deeper understanding. But do we also want a catalogue of linguistic categories?

The appendixes of linguistics textbooks have glossaries with a bewildering array of technical terms (aorist, associative, conative, equative, etc.), and often even experienced linguists are confused by diverging usages. So sometimes especially comparative linguists feel that some terminological unification would be helpful. Wouldn’t it be nice if we all used the term “converb” for the things that have gone by various other names (e.g. gerund, absolutive, adverbial/conjunctive participle)? (See Haspelmath 1995.) Eventually, we would perhaps end up with a well-ordered catalogue of all the categories that are found in the world’s languages, and fieldworkers would have the manageable task of matching the phenomena they observe to that catalogue.

The recently introduced rubric “category check” in the journal Linguistic Typology can be understood in this way: The editor invites submissions to “acquaint the public with the more unfamiliar categories…”, and “to explain categories…that are polysemous or homonymous or just unclear”. This is followed by a list of about 50 category labels, from “aorist” to “trills”. If we could agree on how to use these terms, would we have a solid foundation for future typological and descriptive work, and kind of GOLD standard for the field of diversity linguistics?

Well, when I proposed the “converb as a cross-linguistically valid category” in 1995, I was rather naive: I assumed that the identification of aprioristic cross-linguistic categories was not an error of naive Cartesian universalism, but rather a prerequisite and goal of typological research. But the experience of countless category-assignment controversies (“Is phenomenon X a converb or not?” – see Bickel 1998), as well as reading Dryer (1997) and Croft (2001), taught me otherwise: Categories are language-specific, like words and constructions, so the complete catalogue of categories will contain millions of categories (like the complete catalogue of species).

But still, wouldn’t it be useful to have journal articles on terms like “collective”, “inverse”, “oblique” and “supine”, which are used differently by different people? Absolutely – and this is what the Glottopedia project originally addressed. Many of our technical terms are confusingly polysemous or seem to have the same meaning as others. For example, “oblique” can mean “non-nominative” (in ancient Indo-European case inflection), or it can mean “nominal participant that is coded differently from A, S and P”. This kind of polysemy is confusing, so a “term check” would be very helpful, especially for junior linguists.

I have sometimes wondered why terminological confusion arises again and again, despite our best efforts to be rigorous and comprehensible to our colleagues. It seems that it’s a combination of neophobia (avoidance of neologisms, for fear of appearing extravagant) and naive categorial universalism. When Fillmore (1968) developed a theory of the syntactic realization of semantic roles, he used the old term “case”, which was previously used in a much more restricted way. Why didn’t he introduce a novel term for a novel concept? Fillmore apparently thought that people would be more willing to accept his new ideas if he didn’t scare them by using a new term (and maybe he was right). But he also thought that “deep case” in his sense was a universal innate category, and that the surface variations in its realization were not important. So from this universalist position, it makes perfect sense to ask: “How is case expressed in English, or in Mandarin?” The term “mirative”, introduced by Scott DeLancey in the first issue of Linguistic Typology, has suffered a similar fate: In a recent talk at MPI-EVA, Tyler Peterson equated “mirarivity” with any kind of expression of surprise, made the assumption that surprise is a universal concept, and that all languages express it in some way. As semantic categories appear to be more universal than formal categories, there is a natural tendency for terms originally denoting some formal category to come to be used for a semantic category (see Lehmann 2007 for more on this).

Since different languages have different categories (Haspelmath 2007), a “term check” proposing cross-linguistically applicable definitions of terms would only be useful for comparative concepts (Haspelmath 2010), not for language-specific categories. But when a category is quite similar to a category in another language, there is of course no reason not to use the same term, and using totally new terms for every new language would be quite cumbersome. So in actual fact, the contributions to the “category check” rubric will be “term check” articles, I think, and it is to be hoped that they will be clear about this. Questions such as “What is an X?” (where X is a category label) are not empirical questions, but questions about convenient terminological usage.

While categories are parts of the structures of languages, comparative concepts are just convenient tools for linguists, and it makes little sense to draw up a catalogue of tools.

References

Lehmann, Christian. 2007. On the upgrading of grammatical concepts. In Fons Moerdijk, van Santen, Ariane, & Rob Tempelaars (eds.), Leven met woorden: Opstellen aangeboden aan Piet van Sterkenburg., 409–422. Leiden: Brill.

Bickel, Balthasar. 1998. Review article of Haspelmath & König, eds., Converbs, Berlin: Mouton de Gruyther 1995. Linguistic Typology 2. 381–97.

Croft, William. 2001. Radical construction grammar: syntactic theory in typological perspective.. Oxford: Oxford University Press.

DeLancey, Scott. 1997. Mirativity: The grammatical marking of unexpected information. Linguistic Typology 1(1). 33–52.

Dryer, Matthew S. 1997. Are grammatical relations universal? In Joan L. Bybee, John Haiman, & Sandra A. Thompson (eds.), Essays on Language Function and Language Type: Dedicated to T. Givón, 115–143. Amsterdam: Benjamins.

Haspelmath, Martin. 1995. The converb as a cross-linguistically valid category.. In Martin Haspelmath & Ekkehard König (eds.), Converbs in cross-linguistic perspective: Structure and meaning of adverbial verb forms—adverbial participles, gerunds, 1–56. Berlin: Mouton de Gruyter.

Haspelmath, Martin. 2007. Pre-established categories don’t exist: consequences for language description and typology. Linguistic Typology 11. 119 – 132.

Haspelmath, Martin. 2010. Comparative concepts and descriptive categories in cross-linguistic studies. Language 86(3). 663–687.


3 thoughts on “A catalogue of categories?

  1. Because the use of a category is dictated by the use of those categories with which it is in opposition, ‘Categories are language-specific’ ought to be an obvious result of Saussurean structuralism, but for reasons I at least do not understand, this is a result that many are reluctant to embrace. We are lucky to have Martin Haspelmath to state the case as eloquently as is possible.

    On the journey of ‘mirativity’ from a description of the morpheme ‘dug in Lhasa Tibetan to an apparent synonym for ‘surprise’, I have penned a recent study, which was honored with rebuttals by DeLancey and Aikhenvald among others.

    Hill, Nathan W. (2012) ‘“Mirativity” does not exist: ḥdug in “Lhasa” Tibetan and other suspects.’ Linguistic Typology, 16 (3). pp. 389-433.
    http://eprints.soas.ac.uk/14858/

    • A sort of reply to DeLancey’s reply. Not specifically about mirativity though, but rather about other points of disagreement between DeLancey and Hill vis a vis Tibetan evidentials.

      Hill, Nathan W. (2013) ‘Contextual semantics of ‘Lhasa’ Tibetan evidentials.’ SKASE Journal of Theoretical Linguistics, 10 (3). pp. 47-54.
      http://eprints.soas.ac.uk/17855/

  2. This is a little web page I made for an etymology class many years ago; it merely gives Latin etymologies for some of the obvious terminology, without recommending anybody use it (except possibly to avoid coining yet another synonym). Not a solution, but a footnote to a solution.

Leave a Reply to Nathan W. Hill Cancel reply

Your email address will not be published. Required fields are marked *