The “typology vs. theory“ mistake: Why “comparative linguistics“ is the best label

There is a misleading but widespread stereotype in linguistics: that “language typology” is somehow opposed to “linguistic theory”. Dryer (2006) has explained why this is wrong, but the stereotype keeps being repeated, so I feel that I need to write again about it. And in the end, I think that the “typology” label has outlived its usefulness and should be replaced by “comparative linguistics” – also because in other disciplines, the label has completely different associations.

Here are some examples of book titles and journal paper titles that use “typology” and “theory” as if they were somehow on a par:

  • Hengeveld, Kees. 1992. Non-verbal predication: Theory, typology, diachrony. Berlin: Mouton de Gruyter.
  • Hewson, John & Bubenik, Vít. 1997. Tense and aspect in Indo-European languages: Theory, typology, diachrony. Amsterdam: Benjamins.
  • Polinsky, Maria & Robert Kluender. 2007. Linguistic typology and theory construction: Common challenges ahead. Linguistic Typology 11(1). 273–283.
  • Van Langendonck, Willy. 2008. Theory and typology of proper names. Berlin: De Gruyter Mouton.
  • Zimmermann, Malte & Féry, Caroline (eds.) 2010. Information structure: Theoretical, typological, and experimental perspectives. Oxford: Oxford Univrsity Press.
  • Rolle, Nicholas. 2018. Grammatical tone: Typology and theory. PhD dissertation, UC Berkeley.

It would be easy to enlarge this random collection, as all experienced linguists will be able to confirm. So what is wrong with the opposition between “theory” and “typology”?

The term “theory” generally contrasts either with “practice/application” (cf. Leibniz’s motto “theoria cum praxi”), or with “experiment/data”. So we could talk about experimental methods or corpus data vs. theoretical interpretation of the data, and linguists often do, of course. But in what sense could language typology be subsumed under either “practice” or “data”? The term “typology” implies classification, and any kind of data classification is theoretical. So language typology is by definition a theoretical enterprise. (Not all of theoretical linguistics is typological, though: Language-particular studies are also theoretical to the extent that they go beyond data collection and do not have applied goals.)

If this is so, what could the authors who use “typology and theory” in their titles have in mind?

I suspect that this confusing terminological usage derives from the fact that a very prestigious type of theory in linguistics tries to explain cross-linguistic similarities with reference to a hypothetical genetically specified grammar blueprint (“universal grammar”). Such nativist theories have been advanced since the 1960s, when it was proposed that most of the features and architectures of grammars are given in advance and are thus largely the same in all languages (e.g. Chomsky 1965; Chomsky & Halle 1968). And since the 1980s, such proposals have increasingly been applied to worldwide linguistic diversity (e.g. Baker 2001; Holmberg 2017), often under the label of “Principles and Parameters” (Huang & Roberts 2016). From this perspective, one may say that the classification of languages is a superficial first step, which must then be followed by “deeper analysis” in terms of a grammar blueprint theory. One could then say that a comparative study consists of “typology” (= classification) and “nativist theory” (= explanation of the classification by the grammar blueprint).

However, grammar-blueprint theories have had very limited success (cf. Newmeyer 2005; Haspelmath 2008; Boeckx 2014). They were very popular in the 1990s, but over the last two decades, most linguists seem to have realized that the kinds of broad patterns that have become famous since Greenberg (1963) cannot be explained by a theory that is based on innate elements of grammar (analogous to the chemical elements; Baker 2001). In any event, there have not been any major proposals that are widely considered successful, and none of the proposals of the 1980s and 1990s has been universally accepted. The proposals are getting less and less restrictive (thus less and less explanatory), and many generative linguists of the younger generation even tell me that they see generative syntax merely as a descriptive tool. Thus, there is very little clarity around the theoretical status of the grammar blueprint idea, and we certainly do not have specific claims that could be tested if we had enough funding.

Of course, there are many other ways in which cross-linguistic distributions could be explained (e.g. Moravcsik 2011; Schmidtke-Bode et al. 2019), and in particular the efficiency theory of asymmetric coding has been quite successful (to my mind). But none of these theories can be seen as distinct from “typology” – we do not somehow first classify and then explain, but these activities are part of a single theoretical enterprise. (This CAN be seen as distinct from language-particular description, in dictionaries and grammars, which need not make reference to the insights of comparative studies, although they often do.)

So what should be the name of this enterprise? I have long been unhappy with the label “typology”, not only because there is no agreement whether the subdiscipline of linguistics should be called language typology or linguistic typology. I have suggested earlier that comparative grammar, or comparative linguistics, might be better labels. And I have started using these terms (e.g. here and here, and in my 2014 paper). The nice thing about this term is that it is completely transparent, so it is immediately obvious to outsiders what we are doing. Other disciplines have similar subdisciplines, e.g. comparative law, comparative biology, comparative religion.

Now it must be admitted that there is another sense in which the term “typology” is sometimes used: to refer to a particular community of linguists, e.g. those that attend the meetings of the Association for Linguistic Typology (cf. Croft 2003: 2). But it is unclear what exactly unites this community, other than shared personal connections (e.g. common descent from prominent authors like Greenberg and Givón). One may need informal labels for scholar communities, but there seems to be no reason to have a formal term for them that is used in book titles and journal titles.

And there is an additional reason why the label “typology” is confusing (not only when wrongly contrasted with “theory”): In biology, it is strongly associated with a famous distinction that was drawn by Ernst Mayr in 1959, between typological thinking and population thinking (see also Sober 1980). Mayr claimed that one of Darwin’s key contributions was to abandon the old thinking in terms of immutable “types”, and to lead biologists to think of living beings in terms of populations and their dynamic interactions. The earlier typological thinking is also called essentialism (discussed by Croft 2000: 13-17).This is how Mayr put it in a popular article later (emphasis is mine):

Darwinism refutes typology. From the time of the Pythagoreans and Plato, the general concept of the diversity of the world emphasized its invariance and stability. This view-point is called typology, or essentialism. The seeming variety, it was said, consisted of a limited number of natural kinds (essences or types), each one forming a class. The members of each class were thought to be identical, constant, and sharply separated from the members of other essences. Variation, in contrast, is nonessential and accidental. (Mayr 2000: 81)

(Typological and essentialist thinking seems to be closely related to the 19th century approach that is also called “transcendental anatomy”, what I called “morphism” (pure-form thinking) in an earlier blogpost. Amundson (1998) calls this approach “typology”.)

While some linguists may think of the language types in the same way as the immutable, pre-Darwinian species concept, this is not how comparative linguists that take a broad perspective think of linguistic diversity these days. Maybe some linguists who work in the Chomskyan tradition regard the parameters of the innate grammar blueprint as immutable, essentialist constraints on grammars, but these scholars typically reject the term “typology” for their work – as I noted above, they are much more likely to call their approach “theoretical” and contrast it somehow with “typology”.

Readers who are confused at this point need not worry: All this is very confusing, so nothing is wrong with them.

The best way to lead us out of the confusion, I think, is to abandon the term “typology” and to replace it with “comparative grammar” (which is always theoretical, except perhaps when it contrasts two languages for purely pedagogical purposes). Within comparative grammar research, there may be different methods (e.g. small-scale vs. large-scale comparison) and different theories (e.g. functional-adaptive theories and biocognitive theories). The methods are probably complementary, though the theories may well be in competition with each other. We may need nativist theories, or we may be able to do without them. If we do not compare the theories systematically, we will not know – and if we keep using confusing terms like “typology”, we may not even know what needs to be compared.

References

Amundson, Ron. 1998. Typology reconsidered: Two doctrines on the history of evolutionary biology. Biology and Philosophy 13(2). 153–177. doi:10.1023/A:1006599002775.

Baker, Mark C. 2001. The atoms of language. New York: Basic Books.

Boeckx, Cedric. 2014. What Principles and Parameters got wrong. In Carme Picallo (ed.), Linguistic variation in the minimalist framework. Oxford: Oxford University Press.

Chomsky, Noam A. 1965. Aspects of the theory of syntax. Cambridge, MA: MIT Press.

Chomsky, Noam & Morris Halle. 1968. The sound pattern of English. New York: Harper & Row.

Croft, William. 2000. Explaining language change: An evolutionary approach. Harlow: Longman.

Croft, William. 2003. Typology and universals. 2nd edition. Cambridge: Cambridge University Press.

Dryer, Matthew S. 2006. Descriptive theories, explanatory theories, and basic linguistic theory. In Felix K. Ameka, Alan Dench & Nicholas Evans (eds.), Catching language: The standing challenge of grammar writing, 207–234. Berlin: Mouton de Gruyter.

Greenberg, Joseph H. 1963. Some universals of grammar with particular reference to the order of meaningful elements. In Joseph H. Greenberg (ed.), Universals of language, 73–113. Cambridge, MA: MIT Press.

Haspelmath, Martin. 2008. Parametric versus functional explanations of syntactic universals. In Theresa Biberauer (ed.), The limits of syntactic variation. Amsterdam: Benjamins.

Haspelmath, Martin. 2014. Comparative syntax. In Andrew Carnie, Yosuke Sato & Dan Siddiqi (eds.), The Routledge handbook of syntax, 490–508. London: Routledge. https://zenodo.org/record/344909.

Holmberg, Anders. 2016. Linguistic typology. In Ian Roberts (ed.), The Oxford handbook of Universal Grammar, 355–376.

Huang, C.-T. James & Ian Roberts. 2016. Principles and parameters of Universal Grammar. The Oxford handbook of Universal Grammar. doi:10.1093/oxfordhb/9780199573776.013.14.

Mayr, Ernst. 1959. Darwin and the evolutionary theory in biology. Evolution and anthropology: A centennial appraisal. Washington, DC: The Anthropological Society of Washington. (Also in: Conceptual issues in evolutionary biology, edited by Elliott Sober, 1994).

Mayr, Ernst. 2000. Darwin’s influence on modern thought. Scientific American 79–83.

Moravcsik, Edith A. 2011. Explaining language universals. In Jae Jung Song (ed.), The Oxford Handbook of Language Typology, 69–89. Oxford: Oxford University Press.

Newmeyer, Frederick J. 2005. Possible and probable languages: A generative perspective on linguistic typology. Oxford: Oxford University Press.

Schmidtke-Bode, Karsten, Natalia Levshina, Susanne Maria Michaelis & Ilja A. Seržant (eds.). 2019. Explanation in typology: Diachronic sources, functional motivations and the nature of the evidence. Berlin: Language Science Press. http://langsci-press.org/catalog/book/220.

Sober, Elliott. 1980. Evolution, population thinking, and essentialism. Philosophy of Science 47(3). 350–383. doi:10.1086/288942.


OpenEdition suggests that you cite this post as follows:
Martin Haspelmath (December 1, 2019). The “typology vs. theory“ mistake: Why “comparative linguistics“ is the best label. Diversity Linguistics Comment. Retrieved March 16, 2025 from https://doi.org/10.58079/nsva


5 thoughts on “The “typology vs. theory“ mistake: Why “comparative linguistics“ is the best label

  1. <>
    I’m afraid they/we don’t. Guillaume is right in stating that practiotioners of the historical-comparative method (which we use for establishing language relationships and explaining similarities and differences in genetically related languages) most usually call their discipline ‘comparative linguistics’ and the results of their research ‘comparative grammar’ (for instance, of the Indo-Iranian languages). I’ve never heard ‘phylogenetic linguistics’ before. It’s also my experience that the term ‘historical linguistics’ is most often applied to studies within the history of just one language. This (very important) work also includes language comparison (for instance, Old English vs. Middle English) but no application of the historical-comparative method.
    If we want new terminology, we should try to somehow mantain the traditional distinctions between the three sub-disciplines, each of which has different goals and, correspondingly, different methods:
    (a) what is traditionally called ‘comparative linguistics’ (goal: explaining similarities and differencies in genetically related languages, method: historical-comparative method),
    (b) typological linguistics (goal: identifying and explaining similarities between languages which can’t be explained as inheritance from a common ancestor),
    (c) historical linguistics (goal: describing the evolution of a particular language, methods: text philology, corpus linguistics etc.).

    • I see the conceptual division differently:

      (A) historical linguistics, which makes use of older documents (analogous to paleontology in biology)

      (B) comparative linguistics

      (Ba) phylogenetic comparative linguistics (applicable if it has been determined that similarities are due to inheritance; studies the diversification of proto-languages)

      (Bb) comparative contact linguistics (studies borrowing of various kinds)

      (Bc) non-diachronic comparative linguistics

      Of course, diachronically oriented studies have been far more successful than other approaches in explaining similarities between languages – but conceptually, they represent just one kind of comparison. I think that especially when we communicate with other disciplines, it is important to use transparent terminology (also, e.g., to avoid “genetic” when we mean “phylogenetic” or “genealogical”).

  2. “But it is unclear what exactly unites this community, other than shared personal connections.”

    Possibly this is only because I am not following the community actively enough, but it has always appeared quite clear to me what unites the work of linguistic typologists: the study of what’s, well, typical in languages, hence also the name. (By contrast I have not run into much indication that anyone in the field believes in pre-Darwinian immutable “types”.) Usually as can be inferred by relatively shallow observation of a large number of languages, and usually also without proposing causal explanations. The enterprise seems fundamentally statistical, even when run by purely qualitative methods — I could propose amending the new name rather to “statistical-comparative linguistics”. The option of “general comparative linguistics” does not seem to communicate any of these specifics.

    This is all the more so since, besides historical-comparative linguistics, there are also even further types of linguistics that deserve being called comparative! One is contrastive linguistics, the study of differences between often unrelated languages for the purposes of improving language pedagogy; another is semiotic comparison of linguistic signs with non-linguistic signs; a third is the subthread of cognitive and evolutionary linguistics that studies the common biological basis of human linguistic skills, and possible differences in it between individuals and populations…

  3. The proposal to replace “typology” by “comparative linguistics” have been discussed before on this blog (the comments in https://dlc.hypotheses.org/1022). The problem is that the term “comparative linguistics” (as well as “comparative grammar”) already exists and has a precise meaning, and that practitioners of this discipline are not willing to abandon it (see for instance https://dlc.hypotheses.org/1022#comment-103998 ). Many comparative linguistics (including myself) also do typology, but see comparative linguistics and typology as very distinct disciplines, and adopting your suggestion would cause considerable confusion.

    • I don’t think that historical comparative and general-comparative linguistics are “very distinct disciplines” – on the contrary, it’s impossible to do one of them well without knowing a lot about the other. So I think that they are sub-sub-disciplines of the same subdiscipline, just as there are two sub-disciplines of comparative biology (taxonomy and adaptation studies). Historical linguists used to call their subdiscipline “comparative linguistics” when no general-comparative linguistics existed, but nowadays, they generally call ot “historical linguistics” or “phylogenetic linguistics”. In any event, there is less potential for confusion than with the term “typology”.

Leave a Reply

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.