Sanskrit: Tracking lost sounds

[tarkshya]

Wikipedia page on difference between Vedic and classical Sanskrt mentions that...

"Vedic Sanskrit had a voiceless bilabial fricative (/ɸ/, called upadhmānīya) and a voiceless velar fricative (/x/, called jihvāmūlīya)—which used to occur when the breath visarga (अः) appeared before voiceless labial and velar consonants respectively. Both of them were lost in Classical Sanskrit to give way to the simple visarga."

Can somebody point out some words from Vedic Sanskrt which had these sounds /ɸ/ and /x/ in them? /ɸ/ is approximately same as English f, as in fan. /x/ is normal خ sound of Arabic/Farsi.

 

[Dib]

Wikipedia page on difference between Vedic and classical Sanskrt mentions that...

"Vedic Sanskrit had a voiceless bilabial fricative (/ɸ/, called upadhmānīya) and a voiceless velar fricative (/x/, called jihvāmūlīya)—which used to occur when the breath visarga (अः) appeared before voiceless labial and velar consonants respectively. Both of them were lost in Classical Sanskrit to give way to the simple visarga."

Can somebody point out some words from Vedic Sanskrt which had these sounds /ɸ/ and /x/ in them? /ɸ/ is approximately same as English f, as in fan. /x/ is normal خ sound of Arabic/Farsi.

Mostly they would have occurred in external sandhi situations, as visarga inside a word is not that common. I could find only two instances in the Rgveda where the upadhmānīya might have occurred inside a word: पु॒रःप्र॑स्रवणाः (8.100.9) and maybe अ॒न्तः॒पेय॑म्‌ (10.107.9). I found no instance of word internal jihvāmūlīya in RV. There are many examples from external sandhi though, e.g. दू॒तः क॒विः (10.110.1), इन्द्रः॒ किल॑ (10.111.3), etc.

 

[Au101]

Well, as the link explains, upadhmānīya and jihvāmūlīya were allophones of visarga. I don't know how much you know about the visarga, I've forgotten most of it myself, but the visarga is what you get when you have an -s at the end of the word. It does occur elsewhere, too, but I'm afraid these days I'm sketchy on the details. You may have talked about how to pronounce it though. In isolation, in modern pronunciation, it is pronounced as an with a slight echo of the preceding vowel. Thus, इः is pronounced [ihǐ] and उः is pronounced [uhǔ]. Originally, though, I think it really was nothing more or less than a simple, word final . So रामः would have been pronounced [raːmah], as opposed to [raːmahǎ]. However, before ś it is (and was originally) pronounced (and sometimes actually written as) ś, before ṣ it is (and was) pronounced (and sometimes written as) ṣ and before s it is and was pronounced s and may be written as s. Thus both राजानः संनिपातिताः and राजानस्संनिपातिताः are acceptable and they are both pronounced as the latter is written. As you know, it undergoes many many saṃdhi variations as well, so जनाः च becomes जनाश्च and पुत्रः अत्र becomes पुत्रोऽत्र, but in these cases the pronunciation is plainly reflected in the way it is written.

Now to get into the business of really answering your question. These days, visarga is left unchanged before k & kh and p & ph. Thus, प्रशशंसुः कुतूहलात् and देवेभ्यः प्रहस्य and word-internally दुःख. It's pronunciation is usually as though it were at the end of a sentence or otherwise isolated. However, originally, in Vedic Sanskrit, before k & kh, visarga was pronounced /x/ and before p & ph, visarga was pronounced /ɸ/. They even had their own characters, which can be seen in this pdf:

http://www.sanskritweb.net/cakram/chandas-encoding.pdf

The characters you are looking for are called jihvāmūlīyaḥ (F131) and upadhmānīyaḥ (F136), just scroll down till the bottom of page 7. There were, I believe, in fact a number of ways of writing them and another can be clearly seen in this sample (I don't know why the character isn't shown in the encoding pdf, it is in the font and has codepoint F132):

http://www.sanskritweb.net/cakram/vedic.pdf

See how अ॒ग्निः पूर्वे॑भि॒रृषि॑भि॒रीड्यो॒is written here.

So any word that ends in visarga would had final /x/ before k & kh and any word that ends in visarga would had final /ɸ/ before p & ph. So, as you can see, (without accents) अग्निः पूर्वेभिः would have been pronounced something like [ʌgn̪iɸ puːrʋebʱih]. I suspect दुःख was presumably pronounced [d̪uxkʰʌ] - although I'm not sure if the word even existed back then.

 

[Dib]

Thanks for the thorough explanation as always, Au101.

Just a very tiny addition: while the normal "h" was voiced (ghoSha-dhvani), the visarga was supposedly always voiceless, which can be also guessed at from the environments, it occurs in.

 

[tarkshya]

Thank you Au101 for the very informative post, as usual.

I was wondering if Vedic sanskrit had /x/ and /ɸ/ as full consonants, with some words beginning with these sounds. Now I know they were very specialized and rarely used sounds. Since they were simply the allophones of visarga, they couldn't have appeared in the beginning of a word.

Also, is it possible that rather than adding new sounds before k , kh, p & ph, they simply used to modify the sound of following consonant. So could दुःख have been pronounced [d̪uxʌ] rathar than [d̪uxkʰʌ] ?

 

[tarkshya]

Thanks Dib. I always used to wonder where is the voiceless h in Sanskrit, i.e. Voiceless glottal fricative. Now I found it. It is our good old visarga.

Thanks for the thorough explanation as always, Au101.

Just a very tiny addition: while the normal "h" was voiced (ghoSha-dhvani), the visarga was supposedly always voiceless, which can be also guessed at from the environments, it occurs in.