How different are head-marking constructions?

In the recently published festschrift for Johanna Nichols (Bickel et al. 2013), Robert Van Valin updates his earlier treatment of head-marking constructions in Role and Reference Grammar (RRG, cf. Van Valin 1985).

Bickel&al2013

Van Valin starts by noting that in head-marking constructions, such as (1) from Lakhota (apparently based on his own data), syntactic rules target the bound person forms (3rd person plural index wíčha-, 1st person singular index wa-), not the optional conominals (here mathó ki ‘the bear(s)’).

(1) Mathó ki wo-wíčha-wa-t’į-kte
bear the shoot-3PL.UNDERGOER-1SG.ACTOR-die-FUT
‘I will shoot the bears to death.’ (p. 108)

With reference to Nichols (1986), he says that only full nominals in dependent-marking constructions, but not the conominals in head-marking constructions are governed by the verb. The latter are in some kind of “apposition” relationship. The new development in the new paper (Van Valin 2013) is that he makes it clear what the position of the conominal is: It is not in the periphery, but in an “extra-core slot”, within the clause, but outside the core, as in his Figure 4.

VanValin2013

By contrast, the nominals in dependent-marking constructions occur directly in the argument positions in the core. Van Valin’s analysis is similar to my proposal about what I call cross-indexing constructions (verbs with bound person forms and optional conominals, Haspelmath 2013) in one respect: He says that both the cross-index wíčha- (3rd plural undergor) and the conominal mathó ki ‘the bear’ “instantiate the undergoer argument”. Thus, this argument is doubly expressed.

But are the two types of constructions really so radically different? Siewierska (1999, 2001, 2004) has studied person forms in great detail across many languages, finding that there are many intermediate cases (Van Valin does not mention this work). For one thing, different arguments tend to behave differently. Lakhota has both actor and undergoer person forms, but in three-participant constructions, there are usually only two person forms – even though the third argument is not dependent-marked.

Moreover, what about European languages with subject cross-indexing (Haspelmath 2013) like Spanish or Serbo-Croatian, where the subject conominal is just as optional as the undergoer conominal in Lakhota? This is regarded by Nichols (1986) as head-marking as well. But Nichols does not see the cross-indexing head-marking constructions as radically different from head-marking constructions with classical person agreement, as in German (du komm-st ‘you come’). At least in Nichols (1992: 59), she adopts the virtual-agreement (or “pro-drop”) view, which is also implied by the very term “head-marking”: The relationship between the dependent (which may be zero) and the head verb is “expressed on the head”. If the person forms were themselves the dependents, then they would not express the relationship.

Van Valin does not mention this disagreement with Nichols (maybe understandable in a festschrift contribution for her), but he does treat Serbo-Croatian like Nichols, and differently from Lakhota. Since Serbo-Croatian also has nominative and accusative marking, apparently he does not want to say that the subject is appositional to the verb (“in an extra-core slot”). But this leads back to the problem of what to do with the subject person-marking: He ends up adopting the “ambiguous agreement” view of Bresnan & Mchombo (1987) and Siewierska (1999): If an overt conominal is present, the person forms are agreement-markers, otherwise they “count as the instantiation of the argument” (p. 119). But this gives radically different structures for ‘Maria bought the book’ and ‘She bought the book’, which is very strange.

One type of language that does not figure at all in this construction is the Japanese type, where nominals are just as optional as in Lakhota, even though there are no person forms. What is going on here? Are they also in an “appositional” relationship with the verb, or in an extra-core slot? How could we tell?

It seems to me that we don’t need a radical syntactic difference between the two types of constructions. In all languages we have a verb and arguments, but languages differ in two salient ways: (i) whether or not nominals are obligatory, and (ii) whether or not there are bound person forms (= person indexes). Languages with bound person forms quite typically allow conominals with these person forms, i.e. cross-indexing is the most common case (and pro-indexing is less common). When a language has bound person forms, the conominals are very rarely obligatory (as shown by Siewierska 1999), but when there are no bound person forms, this seems to be the case more frequently (as with English objects, and Swedish subjects and objects). There is probably also a tendency for constructions with person-marking to lack overt case on the conominals, but I do not see a good reason to say that dependent-marking and person-indexing constructions are radically different (this has also recently been stressed by Kibrik 2012 (discussed here), another paper that Van Valin does not mention).

Finally, are the person forms referential? Van Valin uses “RP” (= reference phrase) for full nominals, but he also wants to say that they are a kind of “referring expression”. However, they do not express definite reference, as they can combine also with indefinite conominals:

(2) Mathó eyá na-wíčha-wa-xʔu.
bear some stem-3PL.UNDERGOER-1SG.ACTOR-hear
‘I heard some bears.’

So maybe they merely express specific reference, and the definite reference in the absence of a conominal comes about via a Gricean implicature (p. 102). But if this is so, then do they have to be referential at all? Couldn’t it be that they simply express what the gloss says, namely person, number and role? And that their reference is inferred from the context? Again, think of Japanese-type languages with rampant argument omission. Speakers of these languages also manage to talk successfully about referents. It seems to me that the question whether or not person forms on verbs are referential or not is overrated (cf. Mithun 2003). Hearers manage to track reference in a large number of diverse ways (e.g. Nariyama 2003). And indeed, Van Valin notes that Lakhota person forms even combine with nonreferential conominals like ‘no Indians’, or ‘any bears’. Thus, the idea that bound person forms should be referential in cross-indexing constructions seems to derive from the observation that they do not need conominals and thus can function as arguments on their own, plus the clearly wrong assumption that if there is a form expressing an argument and if there is a referent, the argument form itself must express this referent.

References

Bickel, Balthasar, Lenore A. Grenoble, David A Peterson & Alan Timberlake (eds.), Language typology and historical contingency: in honor of Johanna Nichols. Amsterdam: Benjamins.

Bresnan, Joan & Sam A. Mchombo. 1987. Topic, pronoun, and agreement in Chichewa. Language 63. 741–82.

Haspelmath, Martin. 2013. Argument indexing: a conceptual framework for the syntax of bound person forms. In Dik Bakker & Martin Haspelmath (eds.), Languages across boundaries: Studies in memory of Anna Siewierska, 197–226. Berlin: De Gruyter Mouton.

Kibrik, Andrej A. 2012. What’s in the head of head-marking languages? In Pirkko Suihkonen, Bernard Comrie, & Valery Solovyev (eds.). Argument structure and grammatical relations: a cross-linguistic typology, 211-240. Amsterdam: Benjamins.

Mithun, M. 2003. Pronouns and agreement: the information status of pronominal affixes. Transactions of the Philological Society 101(2). 235–278.

Nariyama, Shigeko. 2003. Ellipsis and reference tracking in Japanese. Amsterdam: Benjamins.

Nichols, Johanna. 1986. Head-marking and dependent marking grammar. Language 62(1). 56–119.

Nichols, Johanna. 1992. Linguistic diversity in space and time. Chicago: University of Chicago Press.

Siewierska, Anna. 1999. From anaphoric pronoun to grammatical agreement marker: why objects don’t make it. Folia Linguistica 33(2). 225–251.

Siewierska, Anna. 2001. On the argument status of cross-referencing forms. Challenges and developments in functional grammar, 215–236. Revisita Canaria de Estudios Ingleses.

Siewierska, Anna. 2004. Person. Cambridge: Cambridge University Press.

Valin, Robert D. Van. 1985. Case marking and the structure of the Lakhota clause. In Johanna Nichols & Anthony C. Woodbury (eds.), Grammar inside and outside the clause, 363–413. Cambridge: Cambridge University Press.

Valin, Robert D. Van. 2013. Head-marking languages and linguistic theory. In Balthasar Bickel, Lenore A. Grenoble, David A Peterson & Alan Timberlake (eds.), Language typology and historical contingency: in honor of Johanna Nichols, 91–123. Amsterdam: Benjamins.