3rd Conference Abstracts |
ENS Fontenay/St-Cloud , Institut Universitaire de France
UMR CNRS-ENS 8503 " Analyses de corpus linguistiques"
marchell@ccr.jussieu.fr
abstract
The term "grammaticalization" is generally used to designate changes which occur at the level of concrete units of language: morphemes or syntactic constructions. The goal is to see how a new unit enters into a given language's grammar within a pre-existent paradigm, how a noun becomes a preposition, or a verb a modal auxiliary, or again, how a certain word order has become fixed.
This term does have another definition, introduced at the start by A. Meillet (1912), and well illustrated in particular by J. Bybee (1984) as well as by J. Bybee and W. Pagliuca (1994). It deals with the appearance of new paradigms to mark distinctions that grammar had not expressed until then : as such, this research concerns itself for example with the progressive "grammaticalization" of modal, or temporal, distinctions.
In one case as in the other, the question why a new phenomenon appears isn't explicitly posed. We can, indeed, ask ourselves if the facts of linguistic change are aleatory and dispersed,or if, on the contrary, they , at least in some instances, motivate and relate to each other. In short, is it possible to trace a causality behind these processes - whatever the level might be?
We will defend the following hypothesis : that certain changes at least, although they appear unmotivated, are in fact governed by the emergence or development of distinctions which are situated at the level of grammatical semantics.
This kind of path of development was hypothesized on the basis of the analysis of a very important change which occurred in French language : that is to say, the radical transformation which affected the demonstrative system between the 13th and the 16th century, including in particular the appearance of a new paradigm (ce / ces), and the grammaticalization of the adverbs "ci" and "là" into suffixes ("ceux-ci" and "ceux-là").
We proposed arguments for explaining that change by the extension of a syntactic hierarchised distinction between pronouns and determiners : the old forms were pronouns as well as determiners, the new forms are only determiners (ce / ces) or only pronouns (celui-ci...). A new distinction can indeed generate new units in the old paradigms which it contributes to destabilizing.
In order to test this hypothesis, we examined the changes which occurred at the same period in the grammar. All of them in the nominal system, lead to generalize the same morpho-syntactic distinction between two levels : such as the apparition of the determiner chaque alongside chacun (which till then was pronoun and determiner), of quelqu’un alongside quelque, etc..
Furthermore, phenomena of change which seem to have no link with these may in fact be linked to the same general movement.
This is the case with the replacement in French of the quantifying / intensifying adverb moult (which could modify all kind of words), by two terms, très which is used only with adjectives and adverbs, and beaucoup which modifies the noun and the verb. This phenomenon is to be related to the generalization of the distinction we mentioned above, between two morpho-syntaxtical levels.
A last point to conclude our analysis. As far as we can see, all these changes seem to be "unidirectional" - what is one of the most important charasteristics of grammaticalization.
Thus we propose, lastly, to specify the concept of "grammaticalization" by distinguishing three levels of analysis, which will enable us to create a place for the type of process at issue here. The term "grammaticalization" can designate either the dynamic process by which a new unit enters into grammar, or the constitution of a new paradigm "grammaticalizing" a distinction expressed until then by the lexicon. It may also designate a change situated at the level of the grammatical system, and integrating into it a distinction of abstract, purely grammatical order - as in the case studied here the distinction between the level of nouns, verbs and pronouns (heads), and that of determiners, adjectives and adverbs ; and the referent of this distinction is in the grammar itself.
The advantage of this conception is double.
On the one hand it enables us to explain why certain forms, otherwise perfectly "viable"- as in the case of moult -, suddenly disappear at a specific point.
But also and most of all, it permits us to shed light on fundamental movements of grammatical systems, and to see certain regularities that govern their evolution on a very long span of time : in this case, the emergence and unidirectional development of grammatical - and cognitive – distinction.
Conference site: http://www.infres.enst.fr/confs/evolang/