From Old Turkic to Turkish

[ancalimon]

I was wondering. Are there examples of P sound changing to M sound in Turkish?

Like for example EB changing into EV.

[hjernespiser]

Uh.... what? Dunno about P to M, but I have examples from Tuvan of B/P to V or visa versa. I'll get to it later....

[H. İhsan Erkoç]

Actually one example of a p -> v sound change just came to my mind. According to Zeki Velidi Togan, the Turkic title Yabghu derived from the root Yap- meaning "to cover" (as in the word Yaprak, "leaf") with the suffix -ghu, meaning "(one) who covers", ie "one who protects the people/land/etc". In all the written examples, we have the word with a b instead of p, so the form Yabghu was already around when the Turks starting using writing. This Yabghu later became Yavghu during the Qarakhanid and Seljuk periods. Thus, in Old and Middle Turkic, p could easily revolve to b of which we know many examples of b's turning to m's (such as Bän -> Män, Biŋ -> Miŋ, Bäŋgü -> Mäŋgü, etc).

[massaget]

What Ihsan writes is right, there are many examples for b becomes m in several turk languages. And p can esaily become from b, so the change is possible, but not a direct short term change.

[hjernespiser]

B to M is found at the beginning of words. P to V is found in the middle of words. That's why P to M elicited a big "huh?" from me.

[hjernespiser]

Feb 5, 2012 1:40:02 GMT 3 ancalimon said:

I was wondering. Are there examples of P sound changing to M sound in Turkish?

Like for example EB changing into EV.

Phones, Phonemes, and Allophones

Phones are just sounds. The International Phonetic Alphabet attempts to assign single glyphs to each humanly possible phone. Phones are usually written between two brackets []. Proboards won't really display it right because the formatter uses brackets! So I will put empty brackets before the phone to indicate that the letter is meant to represent a phone.

Phonemes are abstract constructs representing the most basic contrasting sounds of a particular language. These are written between slashes like /p/.

Allophones are the little collections of phones that are the possible sounds of a phoneme. In English the phoneme /p/ can be pronounced like []p or []p^h and the meaning of the word would not be changed (there is no contrast). Japanese has no contrast between []r and []l. []R and []l are allophones for /r/, which is why it sounds funny when a Japanese person tries to pronounce a word containing []r or []l in a language that does have such contrast. In Tuvan, /p/ and /p^h/ are separate phonemes because the meaning changes between words like par and p^har (NOT /b/ and /p/). See en.wikipedia.org/wiki/Tuvan_language#Consonants

For Turkic languages, it is also important to talk about archiphonemes. These are basically the same as phonemes but their sound is underspecified. These are usually written with capital letters with or without slashes. For example the plural -lAr can be -lar or -ler. The /A/ is an archiphoneme and the allophones for it are and [e]. It represents sounds that are low vowels with back harmony but we don't yet know if it is a front or back vowel (underspecified). Archiphonemes are still contrastive like you couldn't substitute /I/ for /A/ like -lIr without the meaning changing.

It is important to understand these concepts when examining the sounds of a language and how those sounds change. More later...

[hjernespiser]

According to "A Grammar of Old Turkic" by Marcel Erdal, Old Turkic had the following consonant phoneme system:

/p/, /v/, /m/
/t/, /d/, /n/, /s/, /z/, /l/
/č/, /y/, /ñ/, /š/
/k/, /g/, /ŋ/, /r/

(Wikipedia has it, basically the same, as:
labial: p, v (β), m;
dental: t, d (δ), n;
palatal: č, y, ń;
velar: k (q, χ), g (γ), ŋ;
sibilant: s, š, z;
liquid: r, l.)

Keep in mind that these are abstract constructs and not the phone/sound that would have actually been spoken...

Erdal writes:
There is a []b (the phone) at the onset; within words, []b appears only in late texts: Onset []b could be equally assigned to /v/ as to /p/.

And:
In the word onset []b could, as an alternative to being an allophone of /v/, be assigned to the phoneme /p/.

And:
[]f (the sound) is an allophone of both /p/ and /v/ appearing before /š/: yafšur- < yap-ïš-ur- 'to stick or fasten something onto something else'.

The point being that a phone could be an allophone of more than one phoneme.

I found this in Erdal that might be directly related to the question:
In referring to runiform texts, scholars have often assumed that the voiced labial consonant in inscriptional Turkic is []b wherever it appears, also between vowels. They give 'äb', for instance, where I write äv also when transcribing runiform texts, or 'yabïz' and not yavïz for 'bad'. It was V. Thomsen who chose this rendering, presumably in view of the principle that the runiform characters b¹ and b² should be transcribed the same way wherever they appear; he was followed e.g. by Tekin. I find myself in agreement with Clauson and Zieme in this matter. Since there is no runiform character for []v (the sound), there is no solid basis for Thomsen's assumption, as the users of the alphabet had no choice but to use the b letters; nor was any additional letter needed for any relevant phonemic distinction.

So what you are seeing is one limitation of alphabets where they don't always reflect the actual spoken sound. It is possible there was never an EB that changed into EV. It was always EV. But there are different examples with different phonological "conditions" (I'm not using the right term) from EV...

[hjernespiser]

Feb 5, 2012 17:06:27 GMT 3 H. İhsan Erkoç said:

Thus, in Old and Middle Turkic, p could easily revolve to b of which we know many examples of b's turning to m's (such as Bän -> Män, Biŋ -> Miŋ, Bäŋgü -> Mäŋgü, etc).

The onset b became m in the presence of nasals, n and ñ and ŋ (or something like that, don't have my book in front of me now).

The word Tuva itself, with the intervocalic -v- is pronounced differently amongst different groups of Tuvans. There's Tuva, Tuba, and Tofa.

There's also the interesting middle-of-the-word archiphoneme /B/ in Tuvan, which becomes (surface variants) b, p, v, or m. It's used in the negation suffix /-BA/, amongst others. You'll also find this archiphoneme in certain western dialects where 'men' becomes attached to the verb rather than separate like in standard Tuvan. Some consonant assimilation occurs.

Men bilbes men becomes men bilbespen.

[H. İhsan Erkoç]

I see. Yeah that is a commonly-seen case in many Turkic languages (such as in Kazakh where they say Men Qazaqpın).

[hjernespiser]

ooOOo it's online: books.google.com/books?id=l4yGbSQcrBoC&pg=PA67&lpg=PA67&dq=old+turkic+p+%3E+v&source=bl&ots=cqkBQ66XTi&sig=SPv4zgmbYBtkF5R2ZmE-zsHpd4k&hl=en&sa=X&ei=L0YwT-y-CciMiALovLWrAw&ved=0CDwQ6AEwBA#v=onepage&q=old%20turkic%20p%20%3E%20v&f=false

[H. İhsan Erkoç]

Oh cool!

[hjernespiser]

Some other example from Tuvan of a p changing into intervocalic -v- due to consonant assimilation.

Khap means a sack
Khep means clothes.

His sack and his clothes become khavy and khevi respectively.

[hjernespiser]

Heh, my son keeps reminding us that he wants to watch the movie "The Pirates of the Carivean".

[ancalimon]

Feb 7, 2012 0:37:23 GMT 3 hjernespiser said:

Some other example from Tuvan of a p changing into intervocalic -v- due to consonant assimilation.

Khap means a sack
Khep means clothes.

His sack and his clothes become khavy and khevi respectively.

Thanks for reminding me. I have another question.

You mentioned khap and khep. I guess khep became >khiv>giy in Turkish. (to wear) and giysi (clothes).

What I wonder is since khap is "a sack which contains things", khep are "clothes which contain our bodies", kap means "to catch~capture something (and contain it)" and kavra means "to grasp and hold something" and kap kacak means "kitchen utensils";, ....
Would it be wrong to think that the KHA sound originally (in Proto-Turkic) could have meant something related to "catching~capturing and containing inside something that has an ... inside"?

[massaget]

I dont think that words develop that directly how you think they do.