Neogrammarian vs. other sound change

John Hewson jhewson at mun.ca
Mon Jan 28 04:05:14 UTC 2008


I am happy to see this discussion, on problems that many of us have been
mulling over for decades. I long ago came to the conclusion that 
sound change is a multi-faceted phenomenon, and that we need some kind of 
descriptive machinery that would make it possible to classify the 
different types.

The true Neogrammarian kind of change is change in the system; it is not 
atomistic. When I learned (half a century ago) the sound changes from 
Latin to Modern French, we were lucky enough to be taught the early 
changes as systemic, and the rest then as atomistic. It was only later 
under the influence of Haudricourt & Juilland (1970) that I began to think 
about interpreting some of the later changes as systemic changes, and 
reported on the subsequent research at the 5th ICHL at Galway in Ireland 
in April 1981 in a paper entitled "Shifting Systems: Evidence for Systemic 
Change in French Historical Phonology" (Papers of the 5th ICHL:117-122)

The early changes from Latin to Romance are very simple: (1) short vowels 
are laxed, and (2) all length distinctions are lost.

Latin had 10 vowels: 5 cardinal positions, short and long.
Early Romance had 7 vowels; 7 cardinal positions.
All of that can be described atomistically, but it is quite complex, and a 
huge waste of time: 	i:>i:>i
 			i>e>e
 			e:>e:>e
 			e>3>3
 			a:>a>a
 			a>a>a
 			etc, etc

Vowels in open syllables then developed phonetic length and later 
underwent Romance diphthongization, which varies quite considerably from 
language to language, but can always be described in systemic terms.

The next major shift for French was /u/ to /y/, a shift which is also 
found in Ancient Greek and elsewhere; the usual explanation is the lack of 
space for back vowels forcing the fronting of the high vowel. In such 
circumstances, this is a systemic shift, it is not atomistic, as it 
appears to be. The systemic consequence of this shift, in French, was the 
creation of two other front rounded vowels, high mid and low mid: more 
systemic shifting.

The next major move is the shift of the diphthong /oi/ to /w3/, and later 
/wa/. This was taught to us completely atomistically: we were required to 
learn the sequence oi>oe>o3>u3>w3>, which I discovered is largely 
nonsense. What really happened was that the two new front rounded vowels 
were made out of two of the existing diphthongs, and to re-establish 
balance in the system of diphthongs, /oi/ was shifted from a closing 
diphthong to an opening diphthong /w3/ (details in my Workbook for 
Historical Romance Linguistics pp 38-40), the change taking place very 
rapidly.

So it is essential to distinguish systemic change from atomistic change 
which also exists, in many curious ways, which may in fact require many
separate categories

It is also important to realize that we are all particularists or 
generalists by temperament. I am aware that I am a generalist, and 
Malkiel, for example, was a particularist, who given twenty minutes could 
talk for an hour on endless detail (which in no way detracts from my 
admiration for him and his work). I have found it very profitable to work 
with particularists, who are pleased to get the big picture from me, and 
and reciprocate by filling it in with pointillistic detail!

(Malkiel, by the way was at the Galway conference. His only comment on my 
paper was to point out that my generalizations on the vowel shifts from 
Latin to Romance did not apply to the North African dialects of Latin! The 
Particularist responding to the Generalist: 98-99% yes, 1-2% no; there's 
always the exception that proves the rule...)

Among the atomistic changes there has to be a category of spelling 
pronunciations. Since Caxton set up his printing press in Westminster Hall 
in 1472 (?), the way that English has been spelled has had a vast 
influence on pronunciation. Isn't it true that without our archaic 
spelling the /h-/ of modern English would have been long gone? And now 
it's pronounced even where it was never pronounced even in Early MnE: "an 
humble and contrite heart" from the Prayer Book of 1549. Latin *humile(m)* 
gave French *humble*, borrowed into English during the great period of 
bilingualism (1350-1450). Since the h- was not pronounced in Latin it was 
never pronounced in French, and there are still little pockets of dialects 
(in the New World at any rate) where it is still not pronounced in 
English. But for most speakers of English today the word begins with /h-/. 
Isn't this a classic example of reversed sound change? In that case it is 
both systemic (pronounce your aitches -- haitches in Ireland), and in the 
case of *humble*, atomistic: this particular item never had an /h-/ in the 
older forms of English: it never had an /h-/ that could be reversed.

For a simple case of atomistic reversed sound change, consider the popular 
pronunciation of /t/ in *often* (the /t/ was lost in the 16th century). By 
contrast, the /t/ in *soften* is NOT pronounced: for this word the 16th 
century sound change is not reversed.

 	John Hewson
 	Memorial University of Newfoundland



On Sat, 26 Jan 2008, Sally Thomason wrote:
>
> But I don't think Brian's approach will work to solve the problem
> Andrew raises.  Years ago, when for some reason I read Yakov
> Malkiel's 1976 Language paper on a minor sound change in Spanish
> ("for some reason" because I am not a Romance specialist and Malkiel
> is not easy to read), I realized that the whole question of phonetic
> triggers of sound change is unanswered, for some or possibly even
> most changes.  Malkiel's example was a case of monophthongization
> that started in a diminutive suffix, where -ie- was replaced by -i-.
> He showed that the change happened because of analogy -- ordinary
> morphological analogy -- to two other diminutive suffixes that had a
> monophthong -i-; the monophthongization change then spread to a
> couple of inflectional verb suffixes.  Then it stopped: it didn't
> become a regular sound change.  But clearly it *could* have done so:
> monophthongization of diphthongs is common enough in regular sound
> change, and the only reason we *know* that this minor Spanish sound
> change had an analogic trigger is that it never progressed beyond a
> handful of suffixes.  If it had progressed to completion,
> monophthongizing all -ie- diphthongs to -i-, we would view it as a
> nice regular Neogrammarian sound change (and probably even Malkiel
> wouldn't have discovered the analogic source).
>
> In other words, even with phonetically reasonable sound changes, I
> don't think it's safe to assume that "phonetic factors are paramount"
> at the onset of the change.  As we all know, many phonetically
> natural changes fail to occur; we don't know why the ones that
> do occur happen.  It's possible that nonphonetic factors are
> frequently part of the trigger -- analogy in Malkiel's example,
> social factors in some other instances, who knows what in still
> others.
>
> The Rhenish Fan shows that what looks like a coherent set of
> perfectly regular changes didn't happen all at the same time, or even
> in chunks in all words at the same time, in the westernmost region of
> German-speaking territory.  Suppose the Rhenish Fan didn't exist, and
> we had a completely regular (bundle of) isogloss(es) separating the
> region where the High German Consonant Shift occurred from the region
> where it didn't: wouldn't we view that as a quite ordinary example of
> Neogrammarian sound change?  I think we would.  So I think it's very
> difficult, and maybe impossible (given that we generally have zero
> information about the earliest stage of a regular sound change) to
> draw a sharp line between Neogrammarian sound change and other kinds
> of sound change.  In some cases we can certainly point to
> non-Neogrammarian change processes in sounds, of course.  But
> excluding those processes doesn't, it seems to me, give us a reliable
> criterion for identifying Neogrammarian sound changes before they've
> completed their run and turned out to be regular.
>
>  -- Sally Thomason
>     Univ. of Michigan
> _______________________________________________
> Histling-l mailing list
> Histling-l at mailman.rice.edu
> https://mailman.rice.edu/mailman/listinfo/histling-l
>

_______________________________________________
Histling-l mailing list
Histling-l at mailman.rice.edu
https://mailman.rice.edu/mailman/listinfo/histling-l



More information about the Histling-l mailing list