HISTLING posting

Mon Jul 14 20:55:55 UTC 1997

----------------------------Original message----------------------------
Dear HISTLINGers,

Here are the results (below) from my inquiry from 6/23 on the notion of
grammar competition as an approach to the problem of gradiency in
historical change.  Sincere thanks to everyone who took the time to
respond to my questions, some at considerable length: Henning Andersen,
Alice Harris, Paul Hopper, Anthony Kroch, David Lightfoot, Niki Ritt,
Eric Schiller, and Barbara Vance. 

The tenor of the responses I received was by and large sympathetic to
the ideas advanced by Kroch/Fontana/Pintzuk/Santorini/Taylor.  A number
of people also responded by pointing out congruent points from their own
work: see the references below for Harris and Campbell (1995), Lightfoot
(1991), Andersen (1973), and Need and Schiller (1994)

My summary has grown rather long, so I’ve arranged it under subheadings,
which you’ll find below along with complete references; I haven’t tried
to provide an exhaustive bibliography on the problem of periphrastic
_do_ discussed below, merely the titles I’ve managed to get at thus far
-- apologies for anything I’ve overlooked.  

Many thanks again for all the engaging responses.

Michael Getty

------------------------------------------------------------------------

A. Precursors of the Phrase-Structures-in-Competition (PSC) approach.
B. Critical responses to PSC
C. Indeterminacy: Support for PSC?

A. Precursors of PSC

I was not terribly surprised to find out that the idea of accounting for
variation and change through co-existing, discrete grammars was around
before Kroch (1989). What did surprise me was the explicitness with
which the idea of conflict or competition between such grammars is
spelled out in earlier work -- at least in the study of phonemic systems
-- with scattered references to Bloomfield and various Prague School
theoreticians, including Jakobson. Within the post-WWII era, the first
important study in this area seems to have been Fries and Pike’s study
of Spanish loan phonology in Mazatec ([1949]; I was alerted to this
paper by Weinreich et al. [1968], who are cited by Fontana [1994] along
with Hankamer [1977]). Isolating some apparent phonemic contrasts in
Mazatec, which they found only in Spanish loan words, Fries and Pike
argued that the best way of describing this situation was to posit that
Mazatec in fact had two phonemic systems: one native, one borrowed. 
They went on to extend this idea to cases of change in phonemic systems
brought about by dialect contact, at one point coming to a very
prescient formulation “In the process of change from one phonemic system
to a different phonemic system of the same language, there may be a time
during which parts of the two systems exist simultaneously and in
conflict within the speech of two individuals.” As Weinreich et al. (p.
161) pointed out, however, the impact of this idea seems to have been
deflected onto the study of bilingualism and dialect contact, whereby
“it does not seem to have occurred to anyone that the theory could serve
as a socially realistic basis for the investigation of language
change.”  

The more recent formulations of the idea of grammar competition differ
from these precursors in making no crucial reference to dialect contact
(though Kroch and Taylor [in press] is an important exception). Rather,
the source for the introduction of innovative grammatical options is
thought to come from UG (‘endogenous optimization’ as suggested in
Kiparsky [1996]), as a consequence of independent changes (Kroch 1989,
Fontana 1994) or is reserved for separate investigation.

A different sort of prelude to the PSC model can be found in Andersen
(1973). His theory of deductive and abductive changes in phonology -- if
I’ve understood it correctly -- seems to hold that speakers within a
community in which a change has been actuated can hold two ‘grammars’
inside their heads, at least to the extent that they are thought to have
phonemic inventories which diverge from the evidence provided by the
speech community, a divergence which is masked over by the effects of
separate, ‘adaptive’ rules. These adaptive rules are diacritic to
individual lexical items; underlying changes manifest themselves more
freely and more often, in turn, as the adaptive rules are lost in a
process of lexical attrition. If extended onto syntax, this would put
Andersen’s theory squarely on the side of lexical diffusion as a
mechanism of syntactic change, which is the next point I’ll address.

B. Critical responses to PSC

In the citations I found as a result of replies to my posting, there
were only two sources, viz. Ogura (1993) and Hudson (1997), which
directly addressed the PSC approach, both of them taking issue with
Kroch (1989), and specifically with his Constant Rate Hypothesis.  Both
papers, I believe, argue for approaches to the rise of auxiliary _do_ in
English which would fall under the rubric of ‘lexical diffusion,’ namely
to the extent that the gradient nature of the change involved is thought
to result from the change being shunted through the lexicon, proceeding
item by item (Harris and Campbell [1995: 106-115] have a good overview).
Ogura’s article seems most in line with other recent work making the
case for lexical diffusion as accounting for both the gradual nature of
the rise of periphrastic _do_ and even its persistent s-shaped curve
over time, e.g. Denison (1985), Tottie (1991), Stein (1990). At present,
the idea of lexical diffusion seems to be the most prevalent approach to
syntactic changes of this type. In the two papers addressing Kroch
(1989), in particular, the advance of _do_ periphrasis is thought to be
constrained by word frequency -- as an emergent property of individual
lexical entries (Ogura) -- or by independent developments towards a
well-defined class of syntactic auxiliaries (Hudson; cf. Kroch’s reply
in the same volume).  

I’ll make no attempt to adjudicate on this point, though it should be
pointed out that the PSC approach has been applied -- successfully, it
would appear, especially with regard to the Constant Rate Hypothesis --
to a number of other historical changes, viz. the rise of INFL-medial
and VO syntax in Old/Middle English, Yiddish, Greek (see my original
posting for full citations), as well as to the evolution of the clitic
systems of Modern Spanish (Fontana [1994]) and Bulgarian (Izvorsky, In
press.)  In addition, it seems to me -- at the moment -- that the actual
point of contention between the PSC approach and lexical diffusion
(henceforth: LD) might be difficult to formulate. As far as I’ve been
able to tell, the PSC approach makes no crucial claims on the way in
which competing grammatical options are deployed. In other words, there
is nothing to rule out the kinds of phenomena talked about under the
rubric of LD, i.e. certain words or classes of words preferentially
exhibiting an innovation or systematically resisting it. Rather, in the
strict sense outlined in Kroch (1994), the PSC approach seems only to
hold that the divergent syntactic options evident in a given point of
variation, like morphological doublets, cannot peacefully coexist.
Eventually, so the theory goes, one must displace the other or they must
channeled into separate functional/stylistic domains.  How this happens
would seem to be open to the kinds of favoring/disfavoring factors which
LD theorists have been positing all along -- right down to individual
lexemes -- and which Kroch (1989: 238) sees as expressing themselves in
the initial frequencies of _do_ periphrasis in various contexts.

C. Indeterminacy: Support for PSC?

As a final point, I’m going to go out on a bit of a limb. It strikes me
that one of the resolute properties of a language in which a change is
in progress (e.g. during the rise of modal auxiliaries in English, the
development of epistemic modals, or -- perhaps -- the ongoing change to
prefixal morphology in Colloquial French), is that an appreciable
portion of the utterances one comes across are of indeterminate nature.
In other words, before we find the first iron-clad examples of a new
structure, we see many examples which suggest this new structure but
can’t be unambiguously analyzed as evincing it.  Thus, there are lots of
instances of modal verbs in OE and early ME which might as well be
analyzed as auxiliaries but don’t have to be, or which invite epistemic
readings but don’t demand them. I’m thinking of examples like OE _thonne
maeg hine scamigan thaere braedinge his hlisan_ ‘Then he may be ashamed
of the extent of his fame.’ (Bo 46.5, quoted from Denison in Traugott
[1992: 195]), where it seems to me that nothing forces either a strictly
root or epistemic reading of _maeg_. To the point, then, I think this
kind of indeterminacy (which is put forward programmatically in Hankamer
[1977]) would be somewhat bothersome in a LD account: if change proceeds
item-by-item through the lexicon, then it seems that each item would
have to be either-or with respect to the change in progress. In other
words, _mugan_ in the example just given would have to have a feature
within its lexical entry either licensing epistemic readings or not,
regardless of the ambiguity. In the PSC approach, on the other hand,
(putting aside the possibility of LD-type effects as mentioned above)
indeterminacy would seem to be a natural property: if no contextual or
structural factors in a given utterance force a particular analysis,
indeterminacy would follow from the availability of two grammatical
options applying to the same string (cf. Harris and Campbell [1995:
82-88], who accept the idea of multiple analyses without the notion of
distinct grammars; their point regarding ‘blending’ structures is
particularly relevant). 

 I would particularly welcome responses on this point: the notion of
discrete, co-existing syntaxes is intriguing, but strikes me as being
open only to the kind of indirect confirmation the problem of
indeterminacy suggests.  In addition, there might be a question as to
how well the PSC approach can be extended onto other phenomena: in the
case of the development of epistemic modals in English, the innovating
system does not seem to be in competition with *anything* (cf. Frisch
[1997] for a similar issue in the development of ME sentence negation).

REFERENCES:

Andersen, Henning. 1973. Abductive and deductive change. Language.
49(4): 765-793

Denison, David. 1985. Why Old English had no prepositional passive.
English Studies. 66: 189-204

Fontana, Josep. 19XX. A Variationist account of the development of the
Spanish clitic system. . K. Beals et al. (eds.), Papers from the 30th
Regional Meeting of the Chicago Linguistics Society. Vol. 2: Parasession
on the Variation and Linguistic Theory. 87-100

Fries, C.C. and Kenneth Pike. 1949. Co-existent phonemic systems.
Language. 25: 29-50.

Frisch, Stefan. 1997. The change in negation in Middle English: A NEGP
licensing account. Lingua. 101: 21-64.

Hankamer, Jorge. 1997. Multiple Analyses. Charles Li (ed), Mechanisms of
Syntactic Change. Austin: UT Press. 583-607.

Harris, Alice C., and Lyle Campbell. 1995. Historical syntax in
cross-linguistic perspective. New York: Cambridge. Especially pp. 41,
116-117, 83-88.

Hudson, Richard. 1997. The rise of auxiliary _do_: Verb-non-raising or
category strengthening? Transactions of the Philological Society. 95(1):
41-72. Reply by Kroch in same volume, pp. 140-144.

Izvorski, Roumyana. In press. The Syntax of Clitics in the History of
Bulgarian. To appear in the Proceedings of Diachronic Generative Syntax
4, Universite de Quebec a Montreal, 31 October 1995. Available at
http://babel.lling.upenn.edu/~izvorski/papers.html

Kiparsky, Paul. 1996. The shift to head-initial VP in Germanic. H. s,
S.D. Epstein, and S. Peters (eds.), Studies in Comparative Germanic
Syntax II. Amsterdam: Kluwer. 140-179.

Kroch, Anthony. 1994. Morphosyntactic Variation. K. Beals et al. (eds.),
Papers from the 30th Regional Meeting of the Chicago Linguistics
Society. Vol. 2: Parasession on the Variation and Linguistic Theory.
180-201. Also available at
http://www.ling.upenn.edu/~kroch/online-papers.html

Kroch, Anthony, and Ann Taylor. In press. Verb movement in Old and
Middle English: Dialect variation and language contact. To appear in Ans
van Kemenade and Nigel Vincent (eds), Inflection and syntax in language
change. Cambridge University Press. (due out this year). Also available
at http://www.ling.upenn.edu/~kroch/online-papers.html or in Penn
Working Papers in Linguistics. 1: 45-68 (earlier version).

Lightfoot, David. 1991. How to set parameters. Cambridge: MIT Press. Ch.
6.

Lightfoot, David. 1997. Catastrophic change and learning theory. Lingua.
100: 171-192. The references contain a number of interesting citations
of work on the nature of s-shaped curves in language change.

Lightfoot, David. In press. Shifting triggers and diachronic reanalyses.
To appear in Ans van Kemenade and Nigel Vincent (eds), Inflection and
syntax in language change. Cambridge University Press. (due out this
year)

Need and Schiller. 1994. An autolexical account of variation. K. Beals
et al. (eds.), Papers from the 30th Regional Meeting of the Chicago
Linguistics Society. Vol. 2: Parasession on the Variation and Linguistic
Theory. 218-231.

Ogura, Mieko. 1993. The development of periphrastic _do_ in English: A
case of lexical diffusion in syntax. Diachronica. 10(1): 51-85.

Stein, Dieter. 1990. The semantics of syntactic change: Aspects of the
evolution of ‘do’ in English. Berlin: de Gruyter

Tottie, Gunnel. 1991. Lexical diffusion in syntactic change: frequency
as a determinant of linguistic conservatism in the development of
negation in English. Dieter Kastovsky (ed), Historical English Syntax.
Berlin: de Gruyter. 439-468.

Traugott, Elizabeth C. 1992. Syntax. Cambridge History of the English
Language I: The beginnings to 1066. pp. 168-289.

Weinrich, Uriel, Willam Labov, and Marvin Herzog. 1968. Empirical
foundations for a theory of language change. In Winfred P. Lehmann and
Yakov Malkiel (eds.), Directions for historical linguistics. Austin: UT
Press. 95-188.