ment to Benveniste's opening quotation. It is hoped that this chapter helps to provide
the methodological rigour in the semantic aspects of reconstruction that Benveniste
and Meillet sought and demanded.
The agreement in meaning should be as exact and precise as the agreement in phonological form (according to rules of correspondence). This does not mean that meanings
should coincide more than phonological elements; only the differences in the meanings,
if there are any, should be explained not by vague and general possibilities, but by special circumstances ( 1967: 52).
Many of the ideas in this chapter have been clarified and/or developed in collaboration with Nick
Evans. Although initially skeptical, R.M.W. Dixon provided invaluable encouragement and supervision
for the original research which is reported on in section 4 of this chapter. I would also like to thank Cynthia Allen, Balthasar Bickel, Dan Devitt, Matthew Dryer, Patricia Fox, Karin Michelson, David Nash, Fritz
Newmeyer, Robert D. Van Valin Jr. and Barbara Villanova for their advice and suggestions concerning various aspects of this paper. Mark Durie and Malcolm Ross are especially to be thanked for their valuable
editorial insights and their patience. I would also like to thank the Cognitive Anthropology Research Group
of the Max-Planck-Institute for Psycholinguistics in Nijmegen for providing the atmosphere and resources
which enabled me to make the final revisions of the chapter.
Other researchers who have recognised the importance of polysemy in the investigation of semantic change and in the application of the methods of historical reconstruction are Brown and Witkowski
( 1983), Brown ( 1989) and Traugott ( 1986b). Traugott ( 1986b: 539), for instance, shows that 'admitting
polysemy in semantic theory has the added methodological advantage of allowing us to do extensive internal semantic reconstruction'.
Foley (personal communication) observes that in Yimas, another Papuan language, there is also no
word for 'body'. These facts run contrary to Brown ( 1976: 404, 420) and Andersen ( 1978: 352) claim
that the body is labelled in all body-part partonomies.
Note also that it has been commonly observed that many languages treat notions like 'name', 'shadow', and 'tracks' as they do body parts; this tends to make more sense when we realise that a notion like
'name' is not being treated analogically or metaphorically like a body-part, but that it is being treated in
the same way as all other parts which are seen to constitute the unity of a socially placed human being.
Matisoff ( 1990: 111) observes that '[a]lthough this great work [ie. the DED] contains almost 6000
cognate sets, it does not offer any reconstructions either -- but this is simply because, after many decades
of toil, the authors felt they had not yet resolved all the problems of the vowels.' He goes on to note 'the
sets of forms presented are truly cognate, and any irregularities are at least identified and possible explanations are suggested.' With such good data, I was able to reconstruct obvious semantic changes, but it
must be made clear that many of the posited reconstructions for Dravidian are my own, based on Burrow
and Emeneau's work along with other works such as Zvelebil ( 1970).
It is important to remember that, in keeping with the discussion in section 3, a change represented
as X → Y is to be interpreted as X → [X & Y] → Y (or, more precisely, X → 4 [X & YJ and then [X & Y] →
Y), with the intermediate polysemy.
Matisoff in fact deals with the classification of semantic associations rather than semantic changes,
but the principles remain the same, especially given the fact that the cognitive association of semantic notions smooths the path for semantic change, and the results of semantic change (polysemy) are proof of semantic association. Also note that while Matisoff ( 1978: 176) talks of "extra-, inter-, or trans-field association" he opts for the term 'trans-field' while I opt for the term 'interfield'.
In principle metaphoric association and metonymic association are not mutually exclusive. For instance, the Australian English use of the terms 'neck' or 'arm' to refer to the collar or sleeve of a shirt may
be driven by both metonomy and metaphor: when wearing a shirt the 'neck' of the shirt is physically contiguous to the neck of the body and the 'arm' is physically contiguous to the actual arm, thus there is a
metonymic association, but off the body the 'arm' of a shirt and the 'neck' of a shirt are similar to the human body both in terms of spatial arrangement with respect to other parts and gross shape, and thus there
Questia, a part of Gale, Cengage Learning. www.questia.com
Book title: The Comparative Method Reviewed:Regularity and Irregularity in Language Change.
Contributors: Mark Durie - Editor, Malcolm Ross - Editor.
Publisher: Oxford University Press.
Place of publication: New York.
Publication year: 1996.
Page number: 300.
This material is protected by copyright and, with the exception of fair use, may
not be further copied, distributed or transmitted in any form or by any means.