Peter Clark wrote:

>        Well, as if I didn't already have enough projects, my tired brain
thought of
>yet another idea: a collection or dictionary, if you will, of sound
>The thought occured to me as I considered the changes that occured in
>Proto-Enamyn that resulted in Enamyn today. Earlier this year, I did some
>research on historical linguistics, but didn't find much in the way of
>information on possible sound changes.

Probably because, in some language family somewhere, _any_  change can be
shown to have taken place.  Aargh. (Well, almost any.... And it's true that
some are more likely than others, but it can be a judgment call.)

Nevertheless, this is a very interesting idea, and well worth doing.  I've
never encountered such a listing, at least not one covering many language
families; it could even be publishable, certainly web-worthy-- with the
caveat that it perforce will be highly schematic and cannot cover every
exceptional case (of which there are many, everywhere, alas).  The data for
individual families are probably available (in multi-volume works!!)-- and
to summarize it all might result in a work of considerable magnitude, but
don't let that discourage you ~ us .....;-).

Some other considerations that might be useful:
_Does a set of sound-changes result in systematic (phonemic change) or
merely phonetic-- i.e. Gmc. retains the systematic 3-way IE contrast *vl.
stop - *vd. stop - *vd.asp.stop, though changed phonetically /fric - vl.
stop - vd. stop/

_Mergers_ can be total/unconditioned (e.g. Austronesian *p,b > Oceanic *f)
or partial/conditioned (e.g. AN *b,d,g > Malayic *p,t,k in word-final; or
the equally common (many families) *k,g > *tS,dZ before i or front V

_Definitions_ of common terms that affect sound change: assimilation,
dissimilation; syncope, apocope, haplology, epenthesis, metathesis, etc.
etc. -- maybe more appropriate to a FAQ-type list.

_Subgrouping_ (I mistyped "sobgrouping"...calling Dr Freud!!) What, if
anything, determines whether one language is "more closely related" to
another? What, if anything, does this say about how a family broke up?

_Semantic change_ though a rather vague and knotty area, important
nonetheless in determining whether forms are indeed cognate-- how  much
variation in meaning is permissible across languages???? (I was taken to
task for merely suggesting that an isolated Sulawesi word meaning "banana
blossom" (an obscene red pouch-like/human heart-like thing) _might_ be
cognate with an isolated Polynesian word meaning "quiver, for arrows",
admittedly a stretch.)

(I suspect some of these topics are ranging further afield than you intend)

Incidentally, specific topics like the Great Vowel Shift would be of great
interest, I think.  IMHO it's one of the strangest, messiest things that
ever happened to a language....
(sudden ObConlang, whee:  IMO could be LPM (Lusongi Pila-Mi 'according to my
thought' in Kash e-mail......)

>        For instance, here's a small dictionary that combines Grimm's law
with some
>of the sound changes between Germanic and High German (plus a little
>dialectal variation thrown in):
>b > p (PIE > Germ.)
>b_h > b (PIE > Germ.)

(snip)  One thing to emphasize is that reconstructed "sounds" only represent
sets of correspondences in attested languages, and any implied
proto-phonetic value is often a "best guess" based on the modern reflexes.
Obviously /m/ everywhere in a group of langs. can safely be ascribed to *m,
but more complicated sets could just as well be represented with wing-dings
or numerals (and sometimes are in preliminary work, as Ventris did in his
decipherment of Linear B).  There is on-going argument about the exact
nature of the IE stop system.  And what sort of proto-"sound" produced (in
modern AN languages) the set of correspondences / r l d g y dZ s D /-- at
one time it was simply called the "RLD law", then *r2 (abstractions, and
based on limited data), then *g-acute (implying palato-velar stop), now *j
(originally a typing convenience, but implying a palatal stop or affricate).

>However, all of this would be too much for me to research, so here's where
>would need some help in gathering data. What say you all? Is this a project
>that people would find useful and helpful? I'm willing to compile the data
>into presentation form, but I need a wide data sample, preferably from
>different language families.

Yes to all questions. A thought: perhaps ultimately publishable as a
"Special/Supplementary Edition" of our putative "Conlang Journal".......????
With appropriate drum-beating, it could even bring in a little cash....

I can contribute Austronesian/Malayo-Polynesian in any amount of detail.
Perhaps Dravidian-- there is a table in Burrow & Emeneau's Drav. Etymol.
Dictionary, which I have, but suspect it needs updating.