The anomaly of the verb ‘give’ explained by its high (formal and semantic) transitivity

Abstract

Please cite this paper as: Kittilä, Seppo. (2006) The anomaly of the verb 'give' explained by its high (formal and semantic) transitivity. Linguistics 44: 569-612.Abstract 'Give' is a very atypical trivalent verb in many ways. In the present paper, an explanation for this anomaly will be proposed. The goal of the paper at hand is to show that the anomaly of 'give' follows from its high formal transitivity, which also has a semantic basis. This means that 'give' shares a number of features with highly transitive two-actant clauses and thus outranks other trivalent verbs in formal transitivity in languages in which there are at least two syntactic classes of trivalent verbs. The discussed features comprise the number and marking of arguments, traits of verb morphology and verbal agreement and the application of operations like passive and dative shift. The analysis is not applicable to all languages, but the uniformity languages display in this respect cannot be a mere coincidence. After the formal analysis, some reasons that underlie the high formal transitivity of 'give' will be examined.

Alignment preferences in basic and derived ditransitives

“…, show that while caused motion verbs (like 'throw') and verbs with beneficiaries (e.g., 'build for smb') show a preference for indirective alignment and verbs of impact (e.g., 'hit') have a preference for secundative alignment, canonical ditransitives (as well as verbs of dispossession like 'steal') show a preference for neutral alignment, unless they align themselves with the neighboring semantic types. On different grounds, a similar conclusion has been reached by Kittilä (2006), who noted that in languages with closed class of verbs appearing in a double object construction, 'give'-verbs almost invariably belong to this class.…”

“…Haspelmath 2005a,b). Yet, it has long been noted that 'give' may be an atypical ditransitive verb, with exceptional properties, and not representative for its class (Borg & Comrie 1984, Kittilä 2006. This also suggests that when one looks beyond prototypical ditransitives, such splits may be pervasive cross-linguistically.…”

“…Recent typological work on ditransitive constructions (Kittilä 2006, Conti 2008) has revealed that ditransitives show a preference for a double object construction. Although this preference cannot be stated in absolute terms (in fact, Haspelmath 2005a,b finds for a sample of languages represented in WALS that 'give' verbs show a preference for indirective alignment in flagging2), this preference can be detected by comparing ditransitive verbs to other, related, verb types (see Margetts & Austin 2007 for discussing typology of three-argument verbs from a cross-linguistic perspective).…”

Defining prototypical transitivity

“…We should bear in mind that Naess states explicitly (p. 167) that her prototype does not need to frequent or least marked, quite the contrary actually, since prototypical transitivity seems to be rare (see also DuBois, 1987). Naess's prototype is defined semantically, and it does not aim at presenting statistical predictions (see also Kittilä, 2006).…”

Ditransitive constructions in Kashibo-Kakataibo and the non-distinguishable objects analysis

“…Prototypically, ditransitive verbs relate to the semantic domain of (physical, but also mental) transfer and the best exemplar of a ditransitive verb is the verb 'to give' (see also Kittilä 2006 andNewman 1996). According to Malchukov et al (2010), this semantically-oriented definition is the one to be used in cross-linguistic comparison (and is what Haspelmath 2010 calls a "comparative concept"), since ditransitive constructions in individual languages exhibit very heterogeneous formal properties, which make them non-comparable in this respect.…”