Re: [tied] Re: Daci

----- Original Message -----

From: tgpedersen

To: cybalist@yahoogroups.com

Sent: Tuesday, March 19, 2002 1:13 PM

Subject: [tied] Re: Daci

 

[Piotr:] >> ... The palatalisation of velars before front vowels or weak-vowel deletion are examples of commonplace changes that can play havoc with the morphology. People let them happen nonetheless and don't seem to worry about the consequences -- even a royal decree against sloppy pronunciation would not help :)

 

[Torsten:] > You are merely restating your position. And Danish has restored <sj> - <sk>, even without royal decree. Worry about consequences may lead to restoring paradigms.

 

[Piotr:] Correct me if I'm wrong, but what we have here is a puristic fashion imposed "from above" with the help of the educational system, not the usual kind of spontaneous change "from below". Such determined attempts occasionally succeed (cf. the English spelling-pronunciation of [h-] in French loans, originally hypercorrect) but play a very marginal role in the evolution of languages. Whatever motivated it, it was not the desire of the man in the street to prevent the corruption of grammatical paradigms (the replacement was not morphologically conditioned). Also, it certainly did not _prevent_ the palatalisation of old /sk/ (which had happened earlier, unresisted) but merely imposed "restored" <sk> as a substitute for the already palatalised sequence.

 

----------

 

[Torsten:] > ... I provided an example of the disappearance of grammatical categories being determined by their usefulness, whereupon you restate your position.

 

[Piotr:] You presented another circular argument: what was entirely lost was allegedly "useless" and what was recycled was "useful". The common/neuter distinction did not survive with anything like its original distribution. It was only its formal markers that were employed in a completely new function (countable/uncountable) in some dialects. You could say that they _became_ useful after the fact (since a new use was found for them), but not that their survival was determined by their inherent usefulness (since their original function was lost).

 

----------

 

[Torsten:] > You don't seem to have much confidence in the speakers of the language, do you? But given the mechanical nature of language wear and tear, try to provide an estimate of when eg. Polish will have lost its case endings to become as analytical as Bulgarian?

 

[Piotr:] There are many different attractors in linguistic evolution, which is why we cannot predict its course. I used the word "mechanical" in the sense "independent of human intentions" or "beyond the control of rational agents", which certainly does not mean "predictable". The typological change from inflected to analytical is not inevitable within any given period of time. Polish has been an inflected language ever since PIE times despite having lost many of the original inflections through phonological attrition (e.g. *-os, *-us, *-is, *-om, *-im, *-um, *-m > zero) and having merged or replaced many others. How long it will stay that way is anybody's guess. Most Modern Polish words are at least one syllable shorter than their ancestral forms of a thousand years ago, and the lost syllable is in most cases inflectional. Up to a certain point in the past pre-Polish and pre-Bulgarian had the same history, then they were separated geographically, which made them diverge and evolve in different ways -- partly because of different areal environments, partly due to random drift (which is always an important component of evolutionary processes).

 

----------

 

[Piotr:] >> Ever heard of entropy?
 
[Torsten:] > Yes, in the context of physics.

 

[Piotr:] Phonological naturalness is in a great part a matter of the physics of articulation. The tendency to maximise the "ease of articulation" is a manifestation of growing entropy.

 

----------

 

[Piotr:] >> ... Do you think the French changed [akwa], [augustus], [faktus] and [wi:ginti:] to [o], [u], [fE] and [vE~] through some kind of conscious language engineering (or other motivated behaviour) rather than mechanical lenition?

 

[Torsten:] > Mechanical? Rather sociological. For one reason or another, what used to be considerered sloppy pronunciation may acquire high status.

 

[Piotr:] The social propagation of sound change in progress does have a sociological dimension (being conditioned by factors such as prestige, age, gender and the social context of communication). But the mechanism that initiates change is intralinguistic, hence the directional and systematic character of sound change (for example, morphemes get shorter rather than longer, and changes reveal regular patterns of phonological conditioning rather than happening at random). We find the same familiar kinds of regular processes (lenitions, assimilations) occurring in language after language. This recurrence can't be due to any kind of social convention and is more like a natural phenomenon.

 

Piotr