Naomi Nagy

Linguistics at U of T

Modeling contact-induced language change

I demonstrate how to quantitatively analyze social factors which are relevant in the study of contact-induced language change. Focusing on intensity of contact as the primary correlate of contact-induced change, I propose that, in order to make progress, a number of comparable case studies is necessary. I set out a paradigm for conducting such studies, building on the factors which have been shown to be pertinent in earlier studies. I establish a set of factors which should be addressed in all studies and indicate how they are to be aligned along three axes (amount of contact, cultural identity, linguistic factors). I give examples of how metrics can be established to evaluate the factors, and propose a method for combining the effects of these factors, using a logistic equation. Once a set of such equations is available from a series of similarly conducted studies, the set of equations is, in principle, solvable, and we will have a model of how social factors affect language change in contact communities. In combination with work on structural effects and typological effects, a complete model of language change will be closer within reach.

This paper is a response to numerous calls for structured analysis of social factors involved in contact-induced language change, including Kiparsky, Coteanu, Paul, Schuchardt, Saussure, Weinreich, Labov, Gumperz, and Gal. The importance of examining social factors in contact situations has been highlighted in recent work by Mufwene, Mougeon et al., and Siegel. Motivation for this type of study includes: claims which depend on a (currently non-existing) testable measure of intensity of contact (cf. Thomason & Kaufman 1988), frequent references to social factors and "clines of borrowability" as if there were a standard way of computing them, and claims that changes begin in particular sectors of the community (Kroch 1978, Labov 1980, 1981, Guy et al. 1986); all of which can be examined via the proposed model.

While many extensive and detailed studies of the interaction of social and linguistic factors have been conducted, each has been undertaken in an independent framework, making comparison across cases difficult or impossible. However, I show that there is a finite (and relatively small) number of factors which should be included in reports of language change. There are two types of factors: those which describe the community and those which describe an individual who is a member of the community. The community level factors form a large, disparate and unwieldy set. My approach, therefore, is through the more constrained group of individual level factors.

The model of contact-induced language change is similar to a Labovian variable rule: I = ax + by + cz... [1] I is the level of intensity of contact for an individual. {x, y, z ...} are the weights of the factors contributing to intensity of contact, such as attending school in the source language or marrying a member of the source language community. These values are empirically obtained. {a, b, c ...} are constants indicating the importance of the factor group in a given community. Their values are obtained by simultaneously solving equations for as many individuals as there are social factors.


[1] The additive model is used as a first approximation and for ease of exposition, although logistic equation is more appropriate.

Return to list of Naomi's research interests.
email: naomi dot nagy at utoronto dot ca | Return to my home page