Studies in Language. International Journal sponsored by the Foundation “Foundations of Language” - Volume 35, Issue 3, 2011

This paper critically discusses and contrasts some of the different conceptualisations of transitivity that have been presented in the literature, and argues that transitivity as a morphosyntactic phenomenon and effectiveness of an event as a semantic concept should be separated in discussions of transitivity, and also, like many other aspects of grammar, transitivity should be seen as a constructional phenomenon, and so each construction in a language needs to be examined separately, in natural contexts. An Appendix presents some general questions one can consider when analysing language data.

This paper investigates the extent to which the phenomenon of transitivity contributes to syntactic organization in Ao, a Tibeto-Burman language of north-east India, by considering its manifestations from a pragmatic perspective. Agentive case marking is found to be the only formal correlate of transitivity in this language. This serves an obligatory marking function on actor arguments of verbal clauses expressing the habitual activity of a referent, and agentive marking is also consistently used to distinguish the causer argument of causativized predicates. However, in all other types of clause structures its use is pragmatically determined. Transitivity thus appears to be a construction-based phenomenon that has grammaticalized for only some aspects of Ao clausal syntax.

This article describes the morphological and syntactic mechanisms that are involved in the overt and covert realisation of core arguments in Takivatan Bunun. It argues that the interactions between these mechanisms make it difficult to explain Takivatan predicate-argument structure in terms of traditional notions of transitivity and argument alignment. As an alternative, it proposes that argument alignment in Takivatan is realised through the interaction of a number of relatively independent linguistic subsystems.

This paper presents those areas of Iatmul morphosyntax that are relevant to a discussion of transitivity. Evidence for the syntactic status of subject and direct object as core arguments comes from S=O ambitransitive verbs, S/O pivots in complex predicates, switch reference, relative clause formation, agreement marking, and obligatory focus marking. In contrast, there is no evidence for the concept of an “indirect object”. Other relevant phenomena to be explored are case marking, verbs whose morphological make-up correlates with transitivity, zero anaphora, and coalescent nouns in complex predicates. In summary, if languages can be characterized by the extent to which they have grammaticalized the control cline between actor and undergoer, Iatmul can be located in the middle field, with a clear subject category, and a more variable direct object function, whose instantiation is primarily determined by semantic and pragmatic factors.

This paper explores transitivity-related features in Abui, a language with fluid semantic alignment (after Donohue and Wichmann 2008). Many known semantically aligned languages distinguish between two argument roles: actor and undergoer (e.g. Merlan 1985; Durie 1987; Mithun 1991, Donohue and Wichmann 2008 and papers therein). Abui system is unusual; it offers seven coding options for both single-argument and two-argument clauses. A rich set of semantic features (specificity, animacy, individuation, instigation, control, volition, affectedness, change, and change of state) drive the differential realisation of arguments. These features are known to be relevant to transitivity, differential argument marking, and split intransitivity. The paper presents a detailed analysis of these features, and explores their ranking, correlations, and clustering. Based on the Abui system, the paper characterises transitivity as a scalar and constructional phenomenon (refining Hopper and Thompson 1980; Rozwadowska 1988; and Næss 2007) applying only to a subset of two-argument clauses because there is no clear default two-argument construction that contains both actor and undergoer arguments.

This paper presents the analyses of transitivity and questions about transitivity in two languages (Rawang and Qiang) that have been described using very different definitions of transitivity, with a view to showing that each language must be analysed on its own terms, and so the criteria used for identifying transitivity, if it is to be identified at all, might be different between languages. In the case of these two languages it is at least partly due to the two languages differing in terms of the degree of systematicity of the marking, with the Rawang marking being more systematic.

Valence and transitivity in Saliba-Logea can be described with reference to three structural levels, the root, the verb and the clause. Phenomena like clauses with both transitive and intransitive features, which are found across the Oceanic language group can be explained through different possible relationships between these levels.

Transitivity in the Cholim variety of Tangsa is seen to be a grammatical feature with low functional load. It is viewed in terms of three factors: the form of the verb (simplex, middle and causative), the relationship to markers of agentivity and anti-agentive, and the verbal agreement system of Cholim. Using traditional texts, the paper explores in detail the issues raised by an examination of transitivity, and relates the findings to the treatment of transitivity in other Tibeto-Burman languages and beyond.

In this paper I discuss transitivity in Murrinh-Patha, a non-Pama-Nyungan polysynthetic language from northern Australia. I survey the range of bivalent clauses in Murrinh-Patha and their morphosyntactic properties, and consider their analysis in terms of definitions of transitivity in the cross-linguistic literature. I argue that syntactic definitions of transitivity, while compatible with the Murrinh-Patha data, are empirically unrevealing since they provide little account for the varying morphosyntactic properties of different bivalent constructions. Instead, I show that the morphosyntax of bivalent constructions in Murrinh-Patha is sensitive to the semantic features of the participants, supporting a prototype approach to transitivity (such as those proposed by Hopper and Thompson 1980 and Næss 2007).