6 Dec 2010

The semantics of giving

Avid polyglots may eventually notice that so many innocent things can get lost in translation. Translation is a comical affair at times. I believe Etruscanists are also getting lost in translation when they try to wrap their heads around the syntactics and semantics of a quite innocent-looking verb like tur 'to give'.

First an example of bilingual confusion

Before I explain the Etruscan problem, let me illustrate a living example of confusion between, say, English and Chinese speakers. The verb "to rain" seems like such a simple little verb. How can anyone misunderstand it? When waking up to a rainy day, an English speaker may announce, "It's raining today!" while a Mandarin speaker may say, "Jintian xia yu le!" (今天下雨了!) Roughly this means the same thing, but more specifically the Chinese sentence might be translated back into English literally as "Today (jintian) has started (le) descending (xia) rain (yu)!"

What can be confusing in this language clash is not just the slightly different idiom used but also the way in which "rain" may be perceived by the speaker by way of the words and syntax chosen. In English, it's a continuous action marked in the present tense and the focus is on the state of raining. In Chinese, the action is inchoative and punctual; the verb is also unspecified for tense. What is being focused on in Chinese is the very split-second it's started raining, something which can naturally only occur abruptly. So this is why the Chinese sentence ends in the punctual marker le, leaving many Anglophones perplexed (especially when they further mistake the aspect marker as a past-tense marker).

My intended warning here is to avoid forcing one's own native semantics on foreign vocabulary. Sometimes the most straight-forward equations between tongues turn out to be a little less than exact. When Etruscanists who write books on language and who are otherwise well-versed in linguistic science linger on about the "mystery" of why the genitive is used to express the recipient of tur 'to give', I get the impression that even they are having a hard time stepping out of the boxes of their own native speech patterns.

The moment of gifting

So why is the recipient of tur marked in the genitive instead of, say, the dative? Instead of trying to force our assumed patterns on Etruscan, let's begin by just accepting what we see: What could the Etruscans be trying to express by this? Since the genitive elsewhere is so regularly used to mark the possessor, taken as is, it means simply that the recipient must be understood as equal to the possessor in these instances. Suggesting otherwise presumes added, unnecessary roles for the case ending. We should heed Occam's Razor and hold back the temptation to assume.

Personally I reason that perhaps the aspectual focus of tur, unlike in English 'give', focuses on the final result of the transfer rather than on the moment of transfer. When is the recipient equal to the possessor? When the act of giving is complete. Once the action is complete, the donor is naturally no longer the possessor. Thus, with tur treated as a punctual and resultative action, there's no need to speak of a special "dative usage" for the genitive. The genitive instead can now be seen to consistently mark the possessor, even in cases where "give" is used.

Recipient as possessor

I've just noticed that John Newman has written the perfect book to understand cases where the recipient is equal to the possessor in "give"-constructions called Give: A cognitive linguistic study (1996). Page 98 begins with a relevant example from Australian Dyirbal and then compares with a Chinantec example from Southern Mexico that employs a relative clause signifying "which X has" to express the recipient X. So with that in mind, it suggests that examples like mi tn Arnθal turuce "I gave that to Arnth" are short for mi tn [ipn cei] Arnθal [ama] turuce "I gave that [which is now] of Arnth." My perception of inherent punctual aspect in this verb can coexist with Newman's interpretation and gives a secure answer to a "riddle" that Etruscanists have been sitting on far too long.


  1. This analysis of the structure makes sense to me.

    It all coems down to the argument structure of the verb. sometimes that's logically determined - a verb of transfer is just falt out going to have to involve thre arguments.

    But a language (community) is free decide which argument is going to be core or not. In South Sierra Miwok 'give' takes the recipient as the DO and the thing given is coded as an instrument.

    In fact we have this in English in out verb 'present'. So you get "I presented him with a sword." (Although English is actually ambiguous here, as "him" can reflect either old dative or modern "accusative".

    And then we also get "I presented a sword to him." Loaned verbs have often come into English with their case requirement stripped off or jumbled.

  2. I don't believe your Miwok example is unusual and isn't incompatible with other languages, I find. On pages 71 and 106 of his book that I cited here, Newman offers West Greenlandic Niisi aningaasanik tunivaa 'He gave Niisi (with) money' with an instrumental-declined gift.

    In Etruscan too, TLE 685 contains the phrase eit viscri ture with the Latin loan *viscra 'viscera' (< Lat vīscera) declined in the locative to express that the haruspex gives with viscera. (The thought of Etruscan-Latin bilingual interference comes to my mind as well.)

  3. It comes down to which arguemnt is getting the focus. If you are talkiing about moving an object, then the recipeint gets dative marking; if you are talking about interacting with someone, then the thing gets instrumental marking. If anything the second pattern would seem more likely across languages.

    English probably has loads of verbs of transfer that can go either way, and not just loans either. "Pay" is an obvious one.

    The whole business of how theta roles map to case marking systems is always going to be a tangle. There doesn't have to be any kind of logical, smooth fit, so you are bound to get lots of idiomatic conventional strategies.

  4. In Russian there is no verb "to have". The same concept is implemented using a prepositional clause in the genitive.

  5. I can only suspect this same lack of "to have" in Etruscan.