Clitics in LFG: Prosodic Structure and Phrasal Affixation

The c-structure treatment of clitics in LFG has tended to follow the approach of Grimshaw (1982), recent examples being Bresnan (2001) and Schwarze (2001). In this work clitics are daughters of a node labelled 'CL', suggesting that all clitics can be grouped into a unified c-structure category.

Cross-linguistically elements as disparate as pronominals, auxiliaries, discourse and grammatical particles are expressed as clitics. Such differing category membership constitutes a strong argument against treating clitics as a single c-structure category. Moreover, clitics do not behave in a unified fashion: some are distributed differently to words or phrases of similar function --- others are not; some correspond to phonologically similar full forms with similar meaning and function --- others do not.

However, the requirement for a prosodic host is common to all clitics, but is not a property on the basis of which clitics can be claimed to belong to a single syntactic category. Instead, the parallel architecture of LFG suggests a different approach. I propose that p(rosodic)-structure is the appropriate dimension in which to group clitics together as a unified category, that category being the syllable (s). P-structure contains the categories of the prosodic hierarchy in (1), from Selkirk (1978), along with conditions and constraints on the well-formedness of prosodic constituents.

The p- to c-structure mapping which I develop in this paper is based on mapping algorithms such as those in Nespor & Vogel (1986), Selkirk (1986) and Inkelas & Zec (1995). Developments along these lines, including Selkirk (1995) and Truckenbrodt (1999), reinterpret the mapping as a series of alignment constraints, like those in (2).

Imperfect correspondence between p- and c-structure can be represented through violation of such constraints. Thus, the branching p-structure constituent in the Serbian example (3b), [[Marija]PRWD [mu]s]PHONP, corresponds to a non-branching c-structure constituent, [Marija=mu]NP.

Here, I represent the clitic, mu, in c-structure as an affix, reflecting proposals by Anderson (1992, 2000) and Legendre (2000), among others, that cliticization is a form of affixation, but one that operates at the phrasal, rather than word, level. Interactions between the affixal and prosodic properties of clitics have been shown by O'Connor (2002) to influence clitic placement in Serbian.

In LFG terms, then, clitics (or phrasal affixes), like word-level affixes, contribute featural information to f-structure without having to be represented under distinct c-structure nodes. Further, within constructive morphology (Nordlinger & Bresnan 1996, Nordlinger 1998) pronominal clitics contribute grammatical function information, and auxiliary clitics contribute tense/modality information, in a manner analagous to case and tense affixes, that is via inside-out function assignment.

One major advantage of this approach is the avoidance of a 'CL' c-structure node for all clitics, irrespective of their 'real' category. In this paper I provide an analysis within this framework of the Serbian data like that in (3). This data set provides an interesting challenge to any theory of clitics since it illustrates a number of difficulties associated with clitics, for instance different positions for clitics compared with full phrasal equivalents (3a, b), different categories of clitic appearing in the same sentential position (3c), as well as alternating clitic placement (3d, e).

References

• Anderson, S. (1992). A-morphous morphology. Cambridge: CUP.
• Anderson, S. (2000). Toward an optimal account of second-position phenomena. In Dekkers, J. et. al. (eds.), 302-333.
• Bresnan, J. (2001). Lexical-functional syntax. Oxford: Blackwell.
• Dekkers, J., van der Leeuw, F. & van der Weijer, J. (eds.) (2000). Optimality theory: syntax, phonology and acquisition. Oxford: OUP.
• Grimshaw, J. (1982). On the lexical representation of Romance reflexive clitics. In Bresnan, J. (ed.) The mental representation of grammatical relations. Cambridge, MA: MIT Press. 87-148.
• Inkelas, S. & Zec, D. (1995). Syntax-phonology interface. In Goldsmith, J. (ed.) The handbook of phonological theory. Oxford: Blackwell. 535-549.
• Legendre, G. (2000). Morphological and prosodic alignment of Bulgarian clitics. In Dekkers, J. et. al. (eds.), 423-462.
• Nespor, M. & Vogel, I. (1986). Prosodic phonology. Dordrecht, Netherlands: Foris Publications.
• Nordlinger, R. (1998). Constructive case: evidence from Australian languages. Stanford, CA: CSLI Publications.
• Nordlinger, R. & Bresnan, J. (1996). Nonconfigurational tense in Wambaya. In Butt, M. & King, T. The proceedings of the LFG '96 conference. http://cslipublications.stanford.edu/LFG/1/lfg1.html.
• O'Connor, R. (2002). The placement of enclitics in Bosnian, Croatian and Serbian. Ms., University of Manchester. http://lings.ln.man.ac.uk/.
• Schwarze, C. (2001). On the representation of French and Italian clitics. In Butt, M. & King, T. The proceedings of the LFG '01 conference. http://cslipublications.stanford.edu/LFG/6/lfg01.html.
• Selkirk, E. (1978). On prosodic structure and its relation to syntactic structure. In Fretheim, T. (ed.) Nordic prosody II. Trondheim: Tapir. 111-140.
• Selkirk, E. (1986). On derived domains in sentence phonology. Phonology Yearbook 3. 371-405.
• Selkirk, E. (1995). The prosodic structure of function words. In Beckman, J., Walsh Dickey, L. & Urbanczyk, S. (eds.) (1995). Papers in optimality theory. Amherst: University of Massachusetts. 439-469.
• Truckenbrodt, H. (1999). On the relation between syntactic phrases and phonological phrases. Linguistic Inquiry 30. 219-255.