Other Formats:
PDF - 263k

Volume 12 Issue 2 (2014) DOI:10.1349/PS1.1537-0852.A.441

Note: Linguistic Discovery uses Unicode characters to represent phonetic symbols. Please see Optimizing Display for requirements to accurately reproduce this page.

Arguments and Adjuncts Cross-Linguistically: A Brief Introduction

Søren Wichmann

Max Planck Institute for Evolutionary Anthropology

1. Background

This volume is intended to clarify the utility of the notions of arguments and adjuncts for linguistic theory. It brings together papers that reflect current thinking on the distinction and brings cross-linguistic evidence to bear on its relevance.

The papers represent a selection of contributions to the 2011 SLE workshop ‘The argument/adjunct distinction cross-linguistically’ organized by myself, Iren Hartmann, Andrej Malchukov, Martin Haspelmath, and Bernard Comrie (all from the Max Planck Institute for Evolutionary Anthropology) under the auspices of the Leipzig Valency Classes Project, which has been funded by the Deutsche Forschungsgemeinschaft.

2. The Notions of Arguments and Adjunct

‘Argument’ and ‘adjunct’ are notions that are deeply entrenched in contemporary grammatical theories. Several questions attach to these notions, however: Are they typologically useful? Are they universally applicable? Do we need them for describing individual languages? This volume addresses these questions by bringing together papers discussing a number of typologically diverse languages.

The question of how to distinguish arguments (complements) and adjuncts has been debated since the 1970’s (e.g., Vater 1977), yet it is hardly resolved. One challenge is to find tests that are applicable to all languages. For instance, a popular cross-linguistic test such as verbal anaphora is not necessarily universally applicable. If universally applicable tests are not forthcoming, it would be hard to make a strong claim for the relevance of arguments and adjuncts cross-linguistically. A weaker claim, however, can be made according to which arguments and adjuncts are cross-linguistically relevant, even if the criteria for distinguishing them are specific to individual languages. Moreover, instead of requiring a sharp distinction we may satisfy ourselves with a gradient one (Langacker 1987, Croft 2001).

The papers in this volume converge on the view that arguments and adjunct are useful for describing individual languages, but the papers also demonstrate a great variety of criteria for making the distinction, so it is far from obvious how to apply it typologically.^[1] Different ways of operationalizing the gradient view of arguments-adjuncts is another trend uniting the papers.

3. Individual Contributions

Haspelmath thematizes the difference between the conceptual and empirical domains of arguments and adjuncts through his distinction between ‘comparative’ and ‘descriptive’ categories. He argues that the difficulty of applying the argument/adjunct distinction in a consistent way cross-linguistically is shared with many other linguistic categories. According to Haspelmath, this difficulty should not cause us to abandon the argument/adjunct distinction, which is highly useful for describing languages, but he argues against treating the distinction as a universal category.

Schaefer and Egbokhare discuss the West Benue-Congo language Emai. They apply Croft’s (2001) definition according to which an argument is a semantic argument of a head and an adjunct is a predicate whose argument is the event described. For instance, in Randy chased the dog in the park, the adjunct in the park can be semantically viewed as a predicate whose argument is the chasing event. Emai adjuncts interestingly provide syntactic evidence for this type of analysis, since some adjunct types combine with verbs such as ‘be at’ or ‘take’. A greater propensity for such ‘latent’ verbs to occur increases adjuncthood in the authors’ analysis, allowing them to posit the following scale: Manner < Result < Temporal < Locative < Reason, where Reason is most adjunct-like.

Forker’s strategy towards defining a language-specific cline, in this case for the Nakh-Daghestanian language Hinuq, is to define canonical instances of argumenthood and adjuncthood and then to place non-canonical instances (e.g., agents in non-canonical agent constructions) at points in between.

Creissels applies Haspelmath’s approach in a typological study of beneficiaries, showing that these elements can have a more argument-like or more adjunct-like status, depending on the language.

In his study of Balinese locatives, similarly to the other authors, I Wayan Arka also observes a cline between arguments and adjuncts. A further step is taken by proposing a numerically expressed index to measure how argument- or adjunct-like a given verb behaves when tested for a number of different diagnostics.

References

Baker, Mark. 1996. The polysynthesis parameter. Oxford: Oxford University Press.

Croft, William. 2001. Radical construction grammar. Oxford: Oxford University Press.

Jelinek, Eloise. 1984. Empty categories, case, and configurationality. Natural Language and Linguistic Theory 2.39–76. doi:10.1007/bf00233713

Langacker, Ronald. 1987. Foundations of cognitive grammar, Vol. 1: Theoretical prerequisites. Stanford: Stanford University Press.

Vater, Heinz (ed.). 1977. Valence, semantic case and grammatical relations. Amsterdam, John Benjamins.

Author’s Contact Information:

Søren Wichmann
Max Planck Institute for Evolutionary Anthropology
Deutscher Platz 6
D-04103 Leipzig
Germany
wichmann@eva.mpg.de

^[1]It has been suggested that the distinction might correlate with other typologically significant parameters. There is, for instance, the claim that NPs show an adjunct-like behavior in radically head-marking languages (Jelinik 1984, Baker 1996). Such a claim is difficult to sustain if arguments and adjunct are not universally distinguishable.