An interview with Dan Slobin on diversity of categories, acquisition, and sign language

Martin Haspelmath: Dan, you have shown an interest in my distinction between comparative concepts and descriptive categories, and you told me that you recently read my new paper “How comparative concepts and linguistic categories are different”. Can you say what you liked about it and how it relates to your own work?

Dan Slobin: I read your paper with great enthusiasm and pleasure. It makes your familiar argument precise, elegant, and, in my opinion, strongly convincing.

You echo an argument that I’ve been having for a very long time with nativist approaches to acquisition, spelled out in most detail in Slobin (1997). I was grappling with the critical distinction you make between “language-particular categories of individual languages, defined within particular systems, and comparative concepts at the cross-linguistic level, defined in substantive terms.” This is especially critical for developmental psycholinguistics who – alas – tend to be both monolingual and linguistically ignorant (at least in my country). And so they tend to think that their familiar grammatical and substantive categories must be universal (thereby simplifying the acquisition problem). And so they naturally fall into the error you so clearly lay out. Perhaps your analogies with biology, geography, economics, and so forth, can help me with my provincial colleagues.

MH: So you think that it makes good sense to say that linguistic categories are more like categories seen in other domains of culture?

Dan Slobin: Yes, linguistic categories are not natural kinds but social categories (and that’s the missing link in understanding the origins of grammar without hard-wiring it in advance). And social categories arise in language use.

MH: I’m not saying that linguistic categories couldn’t be hard-wired (as natural kinds), only that the study of language structures from around the world does not make me confident that we will find truly universal features, and that the variety we find in grammars makes them look much more like cultural features that need to be learned.

Dan Slobin: The universals are not “features” but – as you and others have pointed out – capacities and constraints. These are of several types, including: learning, memory, perception, processing, interpersonal interaction, and more. The human language capacity, then, rests on the ability to create, learn, and use such structured systems (as is true, for example, of music, mathematics, dance, sports–and, indeed, all of our cultural systems). Let me add that I particularly like your observation that “it makes no sense to ask whether Warlpiri has a Latin Accusative because the Latin Accusative is defined with respect to constructions of Latin.” (Ach – memories of old colonial grammars!) I used a similar argument to counter Pinker’s proposal that the child begins with a table of correspondence, such as “patient of transitive action” ~ “Accusative” (Pinker 1984: 40). I pointed out that grammatical categories like “Accusative” do not represent unitary notions and do not lie on a one-dimensional continuum with other case categories, because the subdivisions cut across different types of categories (Slobin 1997, 2001). I countered with the following illustrative examples (which Pinker never responded to):

Some factors influencing choice of grammatical marking of patient:
– definite patient only (Turkish case inflection)
– masculine animate versus other, whole versus partial patient, singular versus plural patient, affirmative versus negative clause (Russian case inflections)
– whole versus partial patient, completed versus non-completed action (Finnish case inflections)
– direct physical action on patient only (Mandarin particle)
– patient marking (direct and indirect conflated) in present tense only (Georgian)
– one marker for patient, goal, recipient, beneficiary (English personal pronouns)This is just a very brief and simplified list, but it makes it clear that the notion of „patient“ or „direct object“ conflates with various other notions from language to language, including such categories as tense, aspect, definiteness, nature of effect, and so forth. (A similar problem arises with regard to Pinker’s alignment of „Aspect“ with „Durativity“, given the range of aspectual categories…)

MH: You have also been very active in sign language research. Do you see similar issues arising here?

Dan Slobin: Yes, I’ve long had a similar running argument with sign language linguists (Slobin 2008), who insist on using generative grammar concepts and categories to describe languages which are radically different in structure. When you point out in your recent paper that “there is no ‘Subject’ concept that would work as a descriptive category in diverse languages,” I’m reminded of my continual objections to sign language grammars full of subjects and objects and pronouns and word-order rules. So you’ve also provided me with a beautifully clear argument from linguistics to bring to their attention.


Pinker, Steven. 1984. Language learnability and language development. Cambridge, MA: Harvard University Press.

Slobin, Dan I. 1997. The origins of grammaticizable notions: Beyond the individual mind. In Dan I. Slobin (Ed.), The crosslinguistic study of language acquisition: Vol. 5. Expanding the contexts (pp. 265-323). Mahwah, NJ: Lawrence Erlbaum Associates.

Slobin, Dan I. 2001. Form function relations: How do children find out what they are? In Melissa Bowerman & Stephen C. Levinson (Eds.), Language acquisition and conceptual development, 406-449. Cambridge: Cambridge University Press.

Slobin, Dan I. 2008. Breaking the molds: Signed languages and the nature of human language. Sign Language Studies 8. 114-130.

Leave a Reply

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