Complex predicates in Generalized Paradigm Function Morphology

Andrew Spencer

University of Essex, UK

In this talk I address the question of how two familiar types of complex predicate can be described within the model of Generalized Paradigm Function Morphology, a realizational, paradigm-based model of morphology which is a development of the Paradigm Function Morphology model of Stump (2001). One of the features of that model is a generalized paradigm function which relates lexical entries to other lexical entries in terms of subfunctions mapping FORM, SYN, SEM attributes and a lexemic index (LI) value. One feature of this is that attributes such as FORM and SEM are independent. This means that it is easy to express the idea that lexemes are morphologically related even if they have no semantic relationship, an extremely common situation in the world's morphologies.

Next, I describe an enriched model of stem construction, extending proposals of Aronoff, Stump and others. I illustrate this with the very complex verb morphology of the Yeniseian language isolate, Ket (Werner 1997). I then apply this approach to stem formation to the case of standard noun incorporation, in which a verb's internal argument is expressed by compounding an appropriate noun stem with a verb stem to create a complex verb stem. Here we are dealing with a type of compounding process and this turns out to require recursion in stem formation. Illustrations are provided from Chukotko-Kamchatkan languages.

I next turn to complex predicates in which argument structure alternations are expressed by means of peri-verbal adjunct-like elements. I begin with the separable pre-verbs typified by Hungarian and German. In GPFM the exponence of a property (e.g. the identity of the affix which realizes a property) is separated from its placement. This allows us to describe languages in which one and the same cluster of elements can sometimes exhibit the properties of canonical affixes and sometimes exhibit the properties of canonical clitics, notably European Portuguese (Luís & Spencer, 2005a), and Udi (Luís & Spencer, 2005b) . Clitics are handled in GPFM by allowing the morphological rules limited access to syntactic (or prosodic) representations. A similar latitude allows us to treat preverbs as either prefixes or as non-projecting words (Toivonen 2003) positioned with respect to syntactic categories. At the same time we can capture the fact that the preverb is essentially an exponent of derivational morphology (Ackerman and LeSourd 1997). This type of system is compared to verbal prefixation in Russian, which gives rise to semantic alternations comparable to the 'fake resultatives' of English of the type 'drink the pub dry' (Spencer & Zaretskaya 1998). Essentially the same analysis can be applied to verb-particle constructions in Germanic.

Finally, I briefly offer some thoughts on how the GPFM model relates to the converbs of Australian languages (and many others). First, note that semantic opacity is to be expected in these constructions; indeed, semantic compositionality is something of a luxury for lexeme formation processes of this type. Second, the fact that it is generally difficult to assign the converb to a syntactic category might well be made to follow from the assumption that such constructions are essentially morphological. Morphological categories do not need to coincide with syntactic categories and morphological formatives do not necessarily reflect any recognizable category. Finally, converbs and their associated light verbs are often separable in the syntax. However, this does not necessarily mean that they cannot be treated as 'essentially morphological', given the assumptions of GPFM. The morphology of clitics systems and preverbs illustrates the need to distinguish between exponence and placement. Moreover, such morphological processes require (perhaps limited) access to syntactic representations. Seen in this light the converb constructions can be seen fitting in with the general picture of morphology and its relation to syntax.

Selected references

Ackerman, Farrell and LeSourd, Philip 1997. Toward a lexical representation of phrasal predicates. In: Alex Alsina, Joan Bresnan and Peter Sells (eds.) Complex Predicates. Stanford University: Center for the Study of Language and Information, 67—106.

Luís, Ana R. and Andrew Spencer 2005a. A Paradigm Function account of ‘mesoclisis’ in European Portuguese. In Geert Booij and Jaap van Marle (eds.) Yearbook of Morphology 2004, Dordrecht: Springer, 177—228.

Spencer, Andrew 2003. Putting some order into morphology: reflections on Rice (2000) and Stump (2001). Journal of Linguistics 39: 621—646.

Spencer, Andrew and Zaretskaya, Marina D. 1998. Verb prefixation in Russian as lexical subordination. Linguistics 36:1—40.

Stump, Gregory T. 2001. Inflectional Morphology. A Theory of Paradigm Structure. Cambridge: Cambridge University Press.

Toivonen, Ida 2003. Non-Projecting Words. A Case Study of Swedish Particles. Dordrecht: Kluwer Academic Publishers.

Werner, Heinrich 1997. Die Kettische Sprache. Wiesbaden: Harrassowitz Verlag.