Causes of language change
1. Introduction1
Why do languages change? Given its crucial nature, historical linguists have been concerned with
this question over the last two centuries; answers provided are sometimes quite fanciful, and
discussions of the causes of language change often start with a list of imaginative theories, the most
popular being breathing efforts in mountain environment as a possible cause for the first sound shift
in Germanic. Even without reviewing such proposals, current theories of causation in language
change are quite disparate, and, depending on the perspective from which they are seen, may also
look rather unlikely. Ultimately, one’s views on the causes of change are inextricably connected
with one’s general assumptions on language and on the real object of linguistic research.
2. Inter-generational transmission
Let us start with the apparently common place observation that languages display a wide margin of
synchronic variation. As uncontroversial as this statement may sound, it cannot help us much if we
assume, following the by now almost anecdotal quote from Chomsky (1965: 3), stating that
“[l]inguistic theory is concerned primarily with an ideal speaker-listener, in a completely
homogeneous speech-community, who knows its language perfectly”. Since, as noted by
Weinreich, Labov and Herzog (1968: 188), “[a]ll change necessarily involves heterogeneity and
variation,” such a view of language clearly rules out any possible study of language change, simply
because it leaves no possibility for change to happen.
Admittedly, since 1965 generative linguistics has tried to come to terms with the undeniable fact
that languages do change, and has focused on inter-generational language transmission as the locus
for change. Following this approach, language change corresponds to a different parameter setting
by the new generation as a result of reanalysis. According to I. Roberts (2007: 230), the issue of
causation in language change can be formulated as follows: “if the trigger experience of one
generation permits members of that generation to set parameter pk to value vi, why is the trigger
experience produced by that generation insufficient to cause the next generation to set pk to vi?.” In
the same vein, Lightfoot (2003: 505) claims that “[i]f one has a theory of language and a theory of
acquisition, it is quite unclear what a theory of change is supposed to be a theory of.”
The idea that the main cause of change, at least as far as so-called internal causes are concerned,2
lies in imperfect language transmission from one generation to the next is not new: as shown in
Weinreich, Labov and Herzog (1968) similar views were held by Herman Paul in the 19th century.
Similar to modern generativists, Paul, too, indicated the competence of individual speakers as the
proper object of linguistic research.
In spite of various implementations, the “child based theory” (cf. Croft 2000: 44) leaves some
basic questions unanswered, that is, in the first place: how do children independently come up with
the same reanalysis at exactly the same time (cf. Hock 1992: 229)? and, second, why does this
happen in certain precise moments, while preceding generations of children have apparently done
quite well setting parameters the same way as their parents did? In other words, the second question
shows that the child based theory does not account for the fact that not only languages may change,
but also that they may exhibit no changes over remarkably long periods of time.
Critics of the child based theory have often pointed out that children do in fact make deviations
and overgeneralizations in their L1 acquisition, but these are not of the type that generates language
change (cf. Hock 1992: 229; Aitchison 2003: 738). Besides, recurrent deviations and
overgeneralizations tend to be abandoned at a certain age, and this process repeats itself over
generations. In fact, to radicalize the argument, following the child based theory one might expect that features of baby talk go into language change, which is patently not the case (see the discussion
in Chambers, this volume). Moreover, proponents of the child based theory belonging to any school
of thought, whether generativists or structuralists or neogrammarians, have never really tackled the
serious problem that there is no positive evidence, in terms of real data from field research, for
language change to happen between generations, as pointed out in Aitchison (2003: 739).
1. Introduction1
Why do languages change? Given its crucial nature, historical linguists have been concerned with
this question over the last two centuries; answers provided are sometimes quite fanciful, and
discussions of the causes of language change often start with a list of imaginative theories, the most
popular being breathing efforts in mountain environment as a possible cause for the first sound shift
in Germanic. Even without reviewing such proposals, current theories of causation in language
change are quite disparate, and, depending on the perspective from which they are seen, may also
look rather unlikely. Ultimately, one’s views on the causes of change are inextricably connected
with one’s general assumptions on language and on the real object of linguistic research.
2. Inter-generational transmission
Let us start with the apparently common place observation that languages display a wide margin of
synchronic variation. As uncontroversial as this statement may sound, it cannot help us much if we
assume, following the by now almost anecdotal quote from Chomsky (1965: 3), stating that
“[l]inguistic theory is concerned primarily with an ideal speaker-listener, in a completely
homogeneous speech-community, who knows its language perfectly”. Since, as noted by
Weinreich, Labov and Herzog (1968: 188), “[a]ll change necessarily involves heterogeneity and
variation,” such a view of language clearly rules out any possible study of language change, simply
because it leaves no possibility for change to happen.
Admittedly, since 1965 generative linguistics has tried to come to terms with the undeniable fact
that languages do change, and has focused on inter-generational language transmission as the locus
for change. Following this approach, language change corresponds to a different parameter setting
by the new generation as a result of reanalysis. According to I. Roberts (2007: 230), the issue of
causation in language change can be formulated as follows: “if the trigger experience of one
generation permits members of that generation to set parameter pk to value vi, why is the trigger
experience produced by that generation insufficient to cause the next generation to set pk to vi?.” In
the same vein, Lightfoot (2003: 505) claims that “[i]f one has a theory of language and a theory of
acquisition, it is quite unclear what a theory of change is supposed to be a theory of.”
The idea that the main cause of change, at least as far as so-called internal causes are concerned,2
lies in imperfect language transmission from one generation to the next is not new: as shown in
Weinreich, Labov and Herzog (1968) similar views were held by Herman Paul in the 19th century.
Similar to modern generativists, Paul, too, indicated the competence of individual speakers as the
proper object of linguistic research.
In spite of various implementations, the “child based theory” (cf. Croft 2000: 44) leaves some
basic questions unanswered, that is, in the first place: how do children independently come up with
the same reanalysis at exactly the same time (cf. Hock 1992: 229)? and, second, why does this
happen in certain precise moments, while preceding generations of children have apparently done
quite well setting parameters the same way as their parents did? In other words, the second question
shows that the child based theory does not account for the fact that not only languages may change,
but also that they may exhibit no changes over remarkably long periods of time.
Critics of the child based theory have often pointed out that children do in fact make deviations
and overgeneralizations in their L1 acquisition, but these are not of the type that generates language
change (cf. Hock 1992: 229; Aitchison 2003: 738). Besides, recurrent deviations and
overgeneralizations tend to be abandoned at a certain age, and this process repeats itself over
generations. In fact, to radicalize the argument, following the child based theory one might expect that features of baby talk go into language change, which is patently not the case (see the discussion
in Chambers, this volume). Moreover, proponents of the child based theory belonging to any school
of thought, whether generativists or structuralists or neogrammarians, have never really tackled the
serious problem that there is no positive evidence, in terms of real data from field research, for
language change to happen between generations, as pointed out in Aitchison (2003: 739).
No comments:
Post a Comment