Specializations > Typology and Descriptive Linguistics

Is animacy distinction more common in interrogative than in personal pronouns?


It seems to me (though I could be wrong) that the distinction between "who" and "what" is, for some reason, much more common that the distinction between "it" and "he/she". The only language I know about that does not differentiate between "who" and "what" is Lithuanian, while there seem to be plenty of languages that do not differentiate between "it" and "he/she" (Turkish, many dialects of Chinese, many dialects of Finnish...). Is my observation correct? And, if so, what is the reason for it?

That's an interesting question.

With "why" questions in Linguistics, the answer is generally historical. What are the grammaticalization paths that these words follow? There may also be a usage based explanation, that humans speaking languages are more likely to distinguish question words for animacy than pronouns, but I'm not sure why that would be. My instinct here is that pronouns typically develop in parallel from forms marked with whatever (or no) genders or noun classes, while "who" may be more likely to develop from a subject (i.e. nominative) form (as humans are more likely to be actors) and "what" from an object (i.e. accusative) form (as inanimate objects are more likely to be acted upon)*, resulting in a potential different form. Other possibilities for 'who' etymologies would of course include 'which/what person', etc. Third-person pronouns tend to develop from demonstratives, so it's less likely that extra forms would be created during that grammaticalization process. There's also the fact that Wh-words are to some degree inherently emphatic (according to some derivational syntactic theories they're focused or topicalized when fronted), while pronouns are often background information.

(*An interesting pattern in the Indo-European languages is nominative-accusative syncretism [systematic identity] for neuter nouns, because of this same frequency effect, that neuter [typically inanimate] nouns are more likely to be objects than subjects.)

It could be revealing to look at the etymologies of these words in a number of languages (but you'd want to look beyond Indo-European, where etymological dictionaries would be harder to find) to see if there are clear patterns.

My gut reaction is that surely this is a sampling error, but in my quick mental survey of a pile of languages, I can't think of a single counterexample. So if true, congratulations, this is a very striking observation. Although I generally aim to explain apparent universals via non-grammatical factors influencing the path of grammaticalization, that kind of explanation doesn't work well with absolute and robustly attested patterns. E.g. why of [p t k] is [p] the most likely missing consonant – but most languages have all three. If having a word "who/what" were marginal, I could see that being an evolutionary accident, but every language I know of has both who and what. I would suggest turning attention to the truthiness of the observation (seems true to me, let's make sure it is true).

The pattern about who/what being ubiquitous is reasonably robust. I recall a recent conversation about that, with how very few exceptions there are (including Lithuanian), so that's on the right track. (It wouldn't hurt to confirm this for less familiar languages, but it's very hard to come up with any exceptions.)

As for pronouns, impressionistically, this is widely attested. Of course it becomes a definitional problem to decide what counts as a "pronoun": languages may still have ways of distinguishing animacy even if there isn't a typical pronoun (contrast) for that function. So it may not be so much about whether such a distinction is useful, but how that usefulness translates into grammaticalization patterns, for example demonstratives vs. pronouns. In many languages omission of a subject ("pro-drop") is possible and would tend to occur in inanimate contexts, so that could be another factor.

But I agree that the next step would be to gather more data about this, and I think, as I said, it would be helpful to check on etymologies (to the extent possible) to look for patterns there.

This night, I had a weird dream. In that dream, an angel appeared to me and asked me if I want him to teach me the language of the angels. And I agreed. The angel taught me a lot of the words from the language. "Hvar" (incidentally the same as the name of the island in Croatia) was a word meaning both "who" and "what". And if you wanted to say precisely "who", you would say "hvar hartsa", because the word "hartsa" meant "person". It was a common phrase, and a lot of angels started saying "hvar hvartsa", similar to how Slavs started saying "devet deset" (nine ten) instead of "nevet deset". A lot of angels also said "hvartsa" for "person". I still remember a word that meant "cat", it was "kitida". However, I forgot other words from that language in my dream.


[0] Message Index

Go to full version