Re: [tied] Re: The role of analogy, alliteration and sandhi in coun

--- In cybalist@yahoogroups.com, Richard wrote:

>--- In cybalist@yahoogroups.com, "alex" <alxmoeller@...> wrote:
>> Richard Wordingham wrote:
>> > --- In cybalist@yahoogroups.com,
>> > "alex" <alxmoeller@...> wrote:
>> >> The phonetic change of "kW" to
>> > "p" cann be
>> >> just one way I think:
>> >>
>> >> -the lost of velar and
>> > consonating of the frontal "w" to a
>> > clean labial,
>> >> thus kW > W > b/p
>> > An intermediate stage [w] (or [W])
>> > is quite unlikely. From Latin the
>> > route may be [kw] (cluster) > [kW]
>> >> [p], but an intermediate step may
>> > be unnecessary.
>> > Richard.
>>
>>
>> Richard, I try to let the theory on a side and the practic aspect

on another

>> side. If theoreticaly any change can be made since there won't be

any

>> problem to _write_ "s" > "a" the paractic aspect should be the one

who say

>> something about the possiblity of a such change and the

explanation should

>> be searchd in real life not on the paper.
>>
>> The lost of the velar "k" or "g" should simplify the things. Why?

Because we

>> do know of the easines of W > v or w > U or w > b (in face b,p and

further

>> f, v beeing alophones of the same phonem; that is: b, p, f, v are

all

>> variants of consonantic "w" at different levels). This way we can

have

>> practicaly the change. Mentaining the velars there is no way to

change of

>> "k" or "g" to p" simply it does not work, the difference is too

big for

>> making it practicaly.
>> Thus, there remains these two ways:
>> -analogicaly due "sprachgefühl" as Torsten mentioned, with living

example as

>> in Rom. "piatrã/kiatrã" (stone), antic Ulpiana/Ulkiana,

Lykos/Lupos ( for

>> that see again Rom. lukii/lupii "the wolves"
>> -lost of velar and from that point the things are easy how I

supposed

>> before.
>>
>> Practicaly it seems to me the easiest way to get it. That doesn't

mean the

>> easiest way is always the true way:-)
>> What would say Oçam here?
>
>Entites are not to be multiplied without reason?
>
>If we considered the simpler case of kW > p, notice that both the
>start and endpoint are voiceless stops. [W] is not a stop; [w] is
>voiced to boot. One could imagine a more gradual shift from [k_w]
>to [k_p_w] (labialised true labiovelar - i.e. co-articulated and
>still with lip rounding) to [k_p] (plain true labiovelar) to p.
>However, if the primary mechanism is children not learning the
>language properly, one does not need all these intermediate steps.
>
>I am not unbaised enough to believe that [kw] and [p] sound similar -
> I must take that on trust. If true, then even their merger could
>occur in one fell swoop.
>
>A problem with kW > W > p, assuming you meant the voiceless labio-
>velar approximant, is that the only relevant developments from it I
>can think of are [W] > [w], [W] > [f] and [W] > [p\] (bilabial
>fricative). If you mean kW > w > p, well, w > p in one step is
>unusual, and would be inconsistent with gW > w > b !
>
>Richard.

Alex, as Richard says, the problem is in several features of the phones.
Let's list the differences:

1. kW: stop, voiceless, (bi)labial+velar
2ai. p: stop, voiceless, (bi)labial(-velar) [1 diff.]
2aii. k: stop, voiceless, (-bilabial)+velar [1 diff.]
2b. f or F: fricative, voiceless, (bi)labi(odent)al(-velar) [2 diffs.]
2c. w: approximant, voiced, (bi)labial(-velar) [3 diffs.!]

1. gW: stop, voiced, (bi)labial+velar
2ai. b: stop, voiced, (bi)labial(-velar) [1 diff.]
2aii. g: stop, voiced, (-bilabial)+velar [1 diff.]
2b. v or B: fricative, voiced, (bi)labi(odent)al(-velar) [2 diffs.]
2c. w: approximant, voiced, (bi)labial(-velar) [2 diffs. !]

Let's (for our convenience) imagine that the number of differences is
somehow related to the amount of energy necessary to change the sound :-),
and the higher the energy is, the lower the probability of that change on
the one side is, or the longer time necessary for that change is, as it
needs more stages or steps to gradually do the changes, then it is clear
from the list above that {kW>p} and {kW>k} are equally probable and your
{kW>w} is the least probable option, and would need much more energy or time
to procede.

This doesn't mean that kW>w is impossible at all. The velar component might
change e.g. to a glottal one, hence a glottal stop or fricative and then
disappear: kW > either 'W > hW > HW > w or, also possible via velar set
itself, kW > xW > XW (where H is a voiced glottal fricative and X is a
voiced velar fricative). Notice the number of stages, show how improbable
(although possible) that would be (or take a long time to gradually happen).

So, just as Richard says, it is not only unusual to take place in a single
step, but also improbable and, as far as I can remember, unattested.

Petusek