2006 ThePhonologyAndMorphOfFuncWordContrInGer

From GM-RKB
Jump to navigation Jump to search

Subject Headings: Function Word Contraction, Function Word, Inflected Preposition, German Language, Cliticization.

Notes

Quotes

  • Key words: analogical extension, cliticization, function words, German dialects, inflected prepositions, phonological fusion, Prosodic Phonology, reanalysis

Abstract

  • The observation that adjacent function words are typically contracted has been made for a number of languages including German. The problems this phenomenon poses for phonology still need to be treated, however, and the extent to which function word contractions set the stage for the evolution of inflected function words from simple cliticization remains to be explored. This paper first addresses issues with regard to the prosodic representation of function word sequences preceding lexical words (i.e., [Fnc Fnc Lex]). Investigating such contractions in a number of German dialects, it will be shown that a sequence of Fnc Fnc constitutes a special phonological unit of its own, which is arguably a trochaic foot that adjoins to the neighboring prosodic word. It will be argued that this analysis accounts for several phonological processes that are peculiar to function word contractions, and also predicts phonological fusion in which the second function word is reduced. Second, the paper suggests that the phonological fusion of function words sets the stage for a number of diachronic processes such as reanalysis and analogical extension, which arguably lead to the evolution of paradigms of inflected function words.

1. Introduction

  • This paper investigates the phonological and morphological properties of certain grammatical constructions that contract function word sequences in various dialects of German. It examines the different developmental paths such constructions have taken in the course of reduction. The function word contractions under investigation involve the fusion of determiners with preceding prepositions, or pronouns with complementizers (e.g., Standard German: auf’m Auto < auf dem Auto ‘on the car’, weil’s < weil es ‘because it’, etc.). The phenomenon of function words fusing with other function words poses a number of theoretical questions for both phonology and morphology.
  • From the viewpoint of morphology, the case of function words such as prepositions fusing with other function words constitutes an instructive case for understanding the emergence of inflectional morphology. When determiners, for instance, fuse with prepositions, the complex grammatical construction carries various grammatical categories such as definiteness, number, gender, etc. The research question for morphology thus remains: how do such function words acquire grammatical categories that are normally attributed to lexical words? The fusion phenomenon we observe in various dialects of German provides us with a unique way to track the development of inflectional morphology from simple cliticization in the course of grammaticalization.

2 The phonology of [Fnc Fnc Lex]

2.1. Statement of the problem

  • Previous work on individual languages has shown systematic differences between lexical words (Lex) and function words (Fnc) with respect to their phonological behavior (e.g., Selkirk 1984 for English; Nespor and Vogel 1986 for various languages; Zec 1993 for Serbo-Croatian; Booij 1996 for Dutch; Hall 1999 for German; Viga´ rio 1999 for Portuguese).1 As prosodically weak and subminimal, function words are vulnerable to phonological reductions and therefore undergo cliticization to a prosodically stronger host, namely lexical words (e.g., I saw [m] ‘I saw him’). For the most part, previous research has focused on the prosody of morphosyntactic strings consisting of elements of the class Fnc or Lex, bringing forth various proposals about the different ways in which [Fnc Lex] sequences are prosodized. For instance, Fnc can be grouped with Lex in the same Prosodic Word (PW), or it can be adjoined with Lex at the next possible domain in the prosodic hierarchy (e.g., Selkirk 1995; Booij 1996; Hall 1999). The Syntactic structure (S-structure) can also produce sequences where multiple function words co-occur, yielding various types of [Fnc Fnc Lex] complexes (e.g., auf dem Auto ‘on the car’, mit dem Auto ‘with the car’, weil das Auto ‘because the car’). Such adjacent function words in German may undergo phonological fusion (e.g., auf’m Auto, mit’m Auto, weil’s Auto), which is arguably initiated by the cliticization of determiners and pronouns to preceding functional heads such as prepositions and complementizers. Contractions of this nature provide an instructive example for the extensively illustrated nonisomorphism between phonology and syntax: while the determiner and the noun form a constituent at the S-structure, the determiner fuses with the preposition before adjoining to the noun at the Phonological structure (P-structure). The prosodic representation of such [Fnc Fnc Lex] combinations, however, has largely been ignored in the literature. The question as to which prosodic domain covers the phonological reduction rules motivating such contractions remains open.
    • 1 We use ‘function word’ as a cover term to refer to free morphemes that mainly (though not exclusively) convey grammatical information and relations such as determiners, adpositions, pronouns, complementizers, etc.
  • The fusion of function words with other function words poses a problem for one of the fundamental assumptions in the theory of Prosodic Phonology with regard to the phonological status of function words. The mapping rules that derive the prosodic constituency from S-structure crucially rely on the difference between function words and lexical words. For instance, Selkirk’s (1984, p. 343) Principle of the Categorial Invisibility of Function Words holds that the principles of syntax-phonology mapping are blind to the presence of functional constituents. Subsequent work in Optimality Theory (OT) restates the same restriction as a series of alignment constraints (Selkirk 1995; Peperkamp 1997). When prepositions and complementizers host other function words such as determiners and pronouns, the fundamental mechanism for building prosodic structure in the syntax-phonology mapping is violated. In particular, as function words, prepositions and complementizers do not form prosodic words themselves, but must be prosodized with a prosodic word as their host. The question, then, remains as to why and how the mapping allows two function words to fuse at the cost of violating a principled restriction.

5. Conclusions: a diachronic scenario

  • Investigating various German dialects has provided us a unique perspective to make various observations on the development of inflected prepositions from simple cliticization of function words to other function words. Diachronically as well as synchronically, the fusion of function words starts as a result of the prosodization of [Fnc Fnc] sequences within the domain of the foot, which in turn causes reduction of the second Fnc and thereby motivates its cliticization to the first Fnc. That is, phonology sets the stage for the migration of independent function words into morphology by creating a special phonological context where the affinity between two function words becomes tight as they undergo various phonological rules. As the phonological cohesion between these two elements becomes tighter, the form is more and more likely to be reanalyzed as a complex grammatical word, which prepares the second part of the diachronic scenario. In particular, the complex form that results from phonological reduction is taken as the model of a morphological pattern. This is what we observe in the case of Ruhrdeutsch prepositional phrases, where the pattern has been analogically extended to include all prepositions and to level out a case distinction. With respect to complementizer phrases, we observe essentially the same process, in which the complex forms of complementizers and reduced pronouns have been reanalyzed as one grammatical word, as evidenced in Bavarian dialects. Due to the partial resemblance of these grammatical words with inflected verbs, verbal person/number suffixes have been analogically extended to also be suffixed to complementizers. As clitics gain affix status, their meanings naturally change or become specialized for particular discourse contexts, as evidenced by various differences that exist between the full forms and contracted forms of [preposition determiner] combinations and [complementizer pronoun] combinations with regard to their discourse-pragmatic functions.28
    • 28 Our primary concern here is the phonological and morphological aspects of this development. This does not mean, however, that syntax plays no role in contractions. The cases we discuss systematically show the fusion of functional heads under structural adjacency. In syntactic accounts, it is suggested that only P0 and C0 can host clitic function words in the phonological structure (see Bayer 1984, Weiß 1998 for the case of C0, and van Riemsdijk 1998 for a syntactic analysis of preposition determiner contractions). This is compatible with our observations. Naturally, there are cases where pure adjacency and co-occurrence do not suffice (see van Riemsdijk 1998 for an analysis). Therefore, additional syntactic constraints must be introduced to account for such exceptions. A detailed syntactic account for such cases, however, is beyond the scope of this paper.
  • Several questions remain as to whether our observations on German dialects can also hold for function word sequences in other languages. It is crucial to first understand the phonological nature of function word classes in such languages, because, as noted in the beginning of this paper, function words in several languages do not have the same phonological nature as they have in English or German. Furthermore, our observations and analysis only covered two-member function word combinations. It could be the case that certain languages may allow combinations of three, or perhaps more, such categories. Our analysis makes the prediction that such sequences would be grouped into several feet, if not in a single foot, reflecting the rhythmic composition of the language in question. In the case of trochaic languages such as German, the preference would be to subsume function words under trochaic feet. Indeed, combinations of particles, prepositions, determiners, and object and subject pronouns are common in dialects of German. In such cases, we observe the grouping of these function words in such a way that the material in such sequences forms binary trochaic feet to achieve compliance with the rhythmic properties of German (e.g., Ruhrdeutsch: [(...] schon mal in die Stadt>scho=ma in-e Stadt ‘already in the city’). There are obviously many more questions to ask in order to achieve a full understanding of the nature and representation of function word combinations. We hope that this paper provides a fruitful starting point for future research, which should shed more light on the prosodic properties of function word combinations in languages other than German to capture cross-linguistic generalizations.

,

 AuthorvolumeDate ValuetitletypejournaltitleUrldoinoteyear
2006 ThePhonologyAndMorphOfFuncWordContrInGerBariş Kabak
René Schiering
The Phonology and Morphology of Function Word Contractions in GermanJournal of Comparative Germanic Linguisticshttp://dx.doi.org/10.1007/s10828-005-4533-810.1007/s10828-005-4533-82006