VARIATION CLASSIX

by C.-J. N. Bailey 

© 1996, 2001 by Orchid Land Publications

[updated 2-8-01]

     Most linguists are minilectalists; i.e. they abstract from the variation in grammars--time, age, style, regional, and other--differences and use a virtual grammar that is idiolectal.
Students in one lecture room typically do not agree with one another on the acceptability of a given example.  Haj Ross some years ago advocated an implicational ranking in which an overall grammar included the ones "included in"--i.e. implied by it.  This is the basis of variation theory, which had been in existed a bit earlier than Ross's proposal.  However it depends on a well defined concept of markedness (based on implicates and the direction of natural change--which, since the mid-eighties, makes use of W. Mayerthaler's theory of reversals in marked contexts).  Where there was a complicated set of reversals in marked contexts involving will, shall, and the subjunctive of earlier modern English (the pattern is unknown to the historians and grammarians but analysed in my Essays on time-based linguistic analysis (Oxford University Press, 1996)--which has developed into a reversing pair of be gonna and will.  Cf. reversing be and get in passives, and the exochronous modality of modern English with the present or past.

CLICK HERE FOR VARIATION THEORY  SUMMED UP IN A DIAGRAM

     Except for a few very amateurish approaches to variation, there are two fairly incompatible approaches to variation, apart from Chomsky's parameters, etc. They are conceived in vastly different ways--one looks to what is natural--the one in myVariation and linguistic theory, his Variation in the data, and his Essays on time-based linguistic analysis as well as in other articles and books cited in the last-named and of course listed in the References.. The other rejects naturalness and embraces normality, making use of sophisticated statistical (quantitative, glottometrist, secular, etc.) approach of Labov, illustrated copiously and interestingly in many learned writings. A further differences is that my approach looks at the entire languages, whereas Labov's analyses look at usually fewer than half a dozen variables/variants, and then never synoptically but lect by lect; it is also either behaviorist and avowedly antitheoretical.  But it is not thinkable that statistics are not internalized in the brain to understand variation, whereas implication and other patterns are.  It is what is natural (quality) and learnable vs. the what is normal quantity); but behaviorists ignore how the brain works, and glottometrists like Labov just look at the data in a more or less list-wise and fragmentary fashion.  Nor is he as antitheoretical as he pretends; he accepts the behaviorist paradigm, statistic analysis, and even comes up with reasonable explanations that offer predictions.
     A number of Achilles' heels and other weak points of glottometry, including its use of minilectal (i.e. idiolectal-synchronic) models for variable data--a contradiction in terms that is as incomprehensible and futile as is the application of such models to historical-comparative studies--are laid out in Variation in the data; see chaps. 1 and 5; chap. 2 deals with the ISOGLOSSIST FALLACY, the theoretical DEVELOPMENTAL CRITERION for determining which variation should or should not be annotated in a linguistic analysis, etc. Chap. 3 illustrates dynamic-developmentalist analytical models and techniques; it also discusses many aspects of grammar generally ignored by grammarians and apparently unknown to phonology. (Developmentist analyses of sound systems [beyond the phonetic level] are carried out in phonetology, where the rules [yes; rules still!] are phonetically motivated, and in phonomorphology, when not.  Some Indo-European work is referred to in chap. 5. The Afterword and Appendixes of the book discuss more general (philosopical) issues.