The papers in this workshop all deal with the diachrony of grammar and are grouped around the notion of grammaticalization. The term "grammaticalization" has been used in at least two senses. In one sense it refers to the process of becoming part of a grammar, of being entered in a grammar, or of changing in grammatical function; John Hawkins' paper in this Workshop on Grammaticalization is an example of this very general sense of "grammaticalization". Since Meillet's (1912) introduction of the term, "grammaticalization" has been used also in a narrower sense, which has received much attention in the past two decades (see, for example, Heine, Claudi, and Huennemeyer 1991; Heine and Reh 1984; Hopper and Traugott 1993; Lehmann 1995; and Traugott and Heine 1991). Meillet's sense of grammaticalization includes the process by which a word becomes a clitic, a clitic an affix, and an affix a synchronically unanalyzable part of another morpheme; it includes the processes of phonological reduction and semantic bleaching which often accompany the processes named above. The papers by Nigel Vincent, Miriam Butt, and Farrell Ackerman and Irina Nikolaeva discuss grammaticalization in this narrower sense. In the rest of this introduction, I shall focus on the narrower sense of grammaticalization and describe how it fits into the big picture of language change.

A distinct advantage of the grammaticalization approach to syntactic change is that it presents an overall view of what would otherwise be a number of individual changes. Emphasizing the long view, grammaticalization shows how the individual changes fit together and relate to one another to produce an overall change. For example, Latin _cantare habeo_ `I have (something) to sing' was grammaticalized as French _chanterais_ `I will sing'. (See Fleischman 1982 and many other sources for complete descriptions.) This overall change involved syntactic elements (the cliticization of the first person singular form of `have' to the main verb), morphological elements (the creation of new inflectional morphology), phonological reduction (especially _habeo_ becoming -_ais_, pronounced [e]), and semantic "bleaching" (of the `have...to' sense to the future tense). The macrochange can be broken into stages incorporating these elements, but the overall effect of all the diverse elements is still kept in mind.

A distinct disadvantage of many grammaticalization studies has been that they have focused on changes to individual lexical items and the morphemes derived from them, often ignoring the effects on the structures in which they function. For example, in the illustration I used in the previous paragraph, I ignored the question of whether in Latin the verb `have' was in one clause while the lexical verb (there `sing') was in another, and whether a biclausal structure might have been fused into a monoclausal one, as in many languages (see Harris and Campbell 1995: 172-191). Unlike much previous work in the grammaticalization tradition, each of the papers in this workshop does study the relationship of the grammaticalization to changes in sentence structure. These papers thus establish that this is not an intrinsic problem with the grammaticalization approach, but rather simply the way many such studies have been conducted in the past.

The grammaticalization approach has been seen as an epiphenomenon (see for example, Harris and Campbell 1995: 92, Newmeyer, in preparation, Ch. 6). Even most of its supporters agree that an instance of grammaticalization consists of at least one reanalysis, often with extension, phonological change, and semantic change. Borrowing may also be involved (see Butt, this volume). Since reanalysis, extension, and borrowing are the three mechanisms of syntactic change and also occur independently of grammaticalization (Harris and Campbell 1995: 50-52, 61- 150), and since phonological and semantic change also occur independently, it must be admitted that grammaticalization is a derivative notion.

A central claim of the grammaticalization approach is that change in grammar is gradual rather than discrete. This is sometimes related to a "graded scale," a "grammaticalization chain," or a "continuum". The various expressions a language uses for a set of categories are seen as moving gradually along the continuum without changing their relative positions. Lightfoot (1979, 1991) takes the position that change is not gradual and suggests (in Lightfoot 1979) instead that the impression of gradualness is given by a build up of grammatical complexity through many small changes, followed by a single catastrophic change that restructures the grammar. Lightfoot (1991) proposes to deal with this through a "diglossic" approach, setting up two grammars (see Harris and Campbell 1995 :86-88 for commentary). Harris and Campbell (1995) suggest instead that gradualness is in part the result of the fact that speakers maintain multiple analyses of a single construction for some time during an on-going change.

The second central claim of the literature on grammaticalization is that this is a unidirectional process. If this is taken literally, it is surely wrong, since numerous examples of degrammaticalization are documented (see Harris and Campbell 1995: 336-338 for a list and Newmeyer, in preparation, Ch. 6, for additional examples). Looking just in English, we can point to the change of the ancient genitive suffix to clitic status, as seen in the oft-cited example, _the king of England's hat_ (Janda 1981), the degrammaticalization of the suffix -_ism_ to the status of an independent word, and the similar degrammaticalization of the suffixes -_etic_ and -_emic_ to the status of independent words. (While the last two examples may not be familiar to everyone, the words _etic_ and _emic_ are from the linguistic terms _phonetic_ and _phonemic_ and have become technical terms in anthropology.) Many practitioners of grammaticalization, however, do not intend the claim of unidirectionality literally and point instead to the uncontroversial and important observation that the vast majority of relevant changes proceed from less dependent to more dependent status (e.g. word to clitic, clitic to affix). (See Newmeyer, in preparation, Ch. 6 for a proposed explanation of these facts.)

Whatever problems one may recognize in the concept of grammaticalization, it plays an important role in presenting an overall picture of morphological and syntactic change. At the present stage of research on these processes, we need to use all means of investigation at our disposal to increase our understanding of the nature of change in grammar.

Fleischman, Suzanne. 1982. The future in thought and language: Diachronic evidence from Romance. Cambridge: Cambridge University Press.

Harris, Alice C., and Lyle Campbell. 1995. Historical syntax in cross-linguistic perspective. Cambridge: Cambridge University Press.

Heine, Bernd, Ulrike Claudi, and Friederike Huennemeyer. 1991. Grammaticalization: a conceptual framework. Chicago: University of Chicago Press.

Heine, Bernd and Mechtild Reh. 1984. Grammaticalization and reanalysis in African languages. Hamburg: Helmut Buske.

Hopper, Paul J., and Elizabeth Traugott. 1993. Grammaticalization. Cambridge: Cambridge University Press.

Janda, Richard. 1981. A case of liberation from morphology into syntax: the fate of the English genitive-marker -(e)s. Syntactic change, ed. by Brenda B. Johns and David R. Strong, Natural language studies 25.59-114. University of Michigan.

Lehmann, Christian. 1995. Thoughts on grammaticalization. Munich: Lincom Europa. (First published, 1982, as Arbeiten des Koelner Universalien-Projekt 48. Koeln: Institut fuer Sprachwissenschaft der universitaet.)

Lightfoot, David W. 1979. Principles of diachronic syntax. Cambridge: Cambridge University Press.

Lightfoot, David W. 1991. How to set parameters: arguments from language change. Cambridge: MIT Press.

Meillet, Antoine. 1912. L'evolution des formes grammaticales. Scientia 12/26 (Milan). (Reprinted 1951, Linguistique historique et linguistique generale, 130-48. Paris: C. Klincksieck.) [Diacritics have been deleted from this entry.]

Newmeyer, Frederick J. In preparation. Language form and language function.

Traugott, Elizabeth Closs and Bernd Heine (eds). 1991. Approaches to grammaticalization. (Typological studies in language, 19.) Amsterdam: John Benjamins.