Sanskrit: mutation of "u" > "au" in names of descent.

[Michael Zwingli]

I have recently noticed this phenomenon in reading about the Kurukshetra War. For instance, at the time of the events of the Mahabharata, the descendants of Kuru were named Kauravas, and the daughter of the king Drupada was called Draupadi. Can anyone explain this morphological convention to me? Does this u > au shift only occur when the "u" is in the first syllable of the name in question?

 

[Dib]

This is a very common adjective-from-noun as well as noun-from-adjective building process. As adjectives from nouns, they are common in descendants, as you have correctly noticed, but they can also mean other kinds of relations. There are also instances of noun-from-nouns where the meaning remains the same (e.g. kutūhala - kautūhala). The noun-from-adjective derivation bears the meaning of "-ness", and the result is always neuter in gender.

The morphological process consists in replacing the first vowel by its vṛddhi variant, and if the original stem ends in a non-a vowel then replace that by its guṇa variant, and the final result is always a-stem.
kuru > kaurava (au/āv is the vṛddhi form and o/av is the guṇa form of u)

For the adjectives, the feminine is then formed by converting them into ī-stem. Hence, draupadī.

One more morphological curiosity is that in the plural for the "descendants" often the original stem is used, instead of the vṛddhi derivative, i.e. you may see kuravaḥ instead of kauravāḥ, kurūn for kauravān, etc.

 

[Au101]

Note that this doesn't just happen with u/ū, e.g. bhaimī- 'Bhīma's daughter i.e. Damayantī', dhārtarāṣṭra- 'a son of Dhṛtarāṣṭra, a Kuru (cf. nir-), esp. patr. of Duryodhana', and kaikeyī- 'the daughter of a prince of the Kekayas (one of the wives of Daśaratha and mother of Bharata)'

 

[Den falska sköldpaddan]

One more morphological curiosity is that in the plural for the "descendants" often the original stem is used, instead of the vṛddhi derivative, i.e. you may see kuravaḥ instead of kauravāḥ, kurūn for kauravān, etc.

Interesting. Is this for metrical reasons, forms like kuru, kaurava &c occuring in specific positions in the stanzas? [Edited typos.]

 

[Michael Zwingli]

Note that this doesn't just happen with u/ū, e.g. bhaimī- 'Bhīma's daughter i.e. Damayantī', dhārtarāṣṭra- 'a son of Dhṛtarāṣṭra, a Kuru (cf. nir-), esp. patr. of Duryodhana', and kaikeyī- 'the daughter of a prince of the Kekayas (one of the wives of Daśaratha and mother of Bharata)'

You perfectly anticipated my follow-up question!

 

[Michael Zwingli]

au/āv is the vṛddhi form and o/av is the guṇa form of u

Please, @Dib, one follow up question to this precise point. I have always thought that within the context of Sanskrit phonology, the "o" phoneme was simply a reanalysis of the diphthong "au", as is famously the usual explanation of the initial sound in ॐ, generally given as "om", but actually "aum". Am I thinking incorrectly with respect to this?

 

[Dib]

You are thinking perfectly correctly, but it's just a bit confusing to state it like that because "au" (and ai) have two different meanings in this formulation.

Classical Skt. o/e (standard IAST transliteration) = older diphthongs au/ai (as in Old Persian) -> These are called the guṇa grade.
Classical Skt. au/ai (standard IAST transliteration) = older long diphthongs āu/āi -> These are called the vṛddhi grade.

Thus these o-au (and e-ai) always remained distinct but their pronunciations changed over time. The old diphthong pronunciation of o became the new pronunciation of au.

Not sure if I am making sense, or making things further confused.

 

[Michael Zwingli]

You are thinking perfectly correctly, but it's just a bit confusing to state it like that because "au" (and ai) have two different meanings in this formulation.
[...]
Not sure if I am making sense, or making things further confused.

Thank you! Yes, you make sense...at least as much sense as possible to someone like myself, who has not systematically studied Sanskrit. I became quite interested in IE linguistics by means of my study of Latin. I have just begun with Ancient Greek, and am fascinated by the differing ways that the PIE roots and lemmas are expressed in the Greek as opposed to the Latin. If life should remain with me, I will someday learn Sanskrit, and then my "Classical" IE education will be well-rounded.

 

[ahvalj]

There is a treatment of this phenomenon in Burrow T · 2001 · The Sanskrit language on pages 199–201 (pages 205–207 of the pdf file). (And overall, a very good chapter about the prehistory of the Sanskrit word-formation in this book).

The use of the lengthened grade in nominal derivation is actually not confined to Indo-Iranic: you may find examples of it in other branches (e. g. Lithuanian vìlkė “she-wolf” and várna “crow” are derived from vil̃kas “wolf” and var̃nas “raven”, the prosody suggesting the former presence of long vowels in the feminines: *u̯īlkīṣ and *u̯ōrnā), though of course nowhere remotely as often as in Sanskrit.

In Germanic, we find this pattern in *đaǥaz “day” → *-đōǥaz “pertaining to a day”, e. g. in the Gothic fidurdogs “which lasted four days” and ahtaudogs “which lasted eight days”.

 

[Au101]

Thank you! Yes, you make sense...at least as much sense as possible to someone like myself, who has not systematically studied Sanskrit. I became quite interested in IE linguistics by means of my study of Latin. I have just begun with Ancient Greek, and am fascinated by the differing ways that the PIE roots and lemmas are expressed in the Greek as opposed to the Latin. If life should remain with me, I will someday learn Sanskrit, and then my "Classical" IE education will be well-rounded.

If you're familiar with Ancient Greek, have you encountered the use of the augment to make the past tense? (I don't know any Greek really but Wikipedia gives the example κελεύω (keleúō) "I order" but ἐκέλευον (ekéleuon) "I ordered".) Well in Sanskrit, the augment is अ a and when added to a root that begins with a vowel it also causes a change to the vṛddhi grade, e.g. √iṣ-, present stem iccha-, 3rd person singular imperfect a- + iccha- + -t = aicchat.

(We would generally expect a + i to give e (from original *ai, see Dib's post above). Classical Sanskrit ai (as transliterated in IAST) is from original *āi of course, again see Dib's post. I'm not perfectly sure but I think Classical Sanskrit ai is usually pronounced as āi despite the misleading (albeit cleaner) transliteration).

 

[ahvalj]

If you're familiar with Ancient Greek, have you encountered the use of the augment to make the past tense? (I don't know any Greek really but Wikipedia gives the example κελεύω (keleúō) "I order" but ἐκέλευον (ekéleuon) "I ordered".) Well in Sanskrit, the augment is अ a and when added to a root that begins with a vowel it also causes a change to the vṛddhi grade, e.g. √iṣ-, present stem iccha-, 3rd person singular imperfect a- + iccha- + -t = aicchat.

(We would generally expect a + i to give e (from original *ai, see Dib's post above). Classical Sanskrit ai (as transliterated in IAST) is from original *āi of course, again see Dib's post. I'm not perfectly sure but I think Classical Sanskrit ai is usually pronounced as āi despite the misleading (albeit cleaner) transliteration).

The traditional explanation of this behavior is that the augment was originally a separate particle, which became attached to the verbal forms during the history of separate languages. What looks like a lengthened grade may therefore rather reflect contraction: a + a, a + i, a + u, a + r̥ (so that the initial ai and au still weren't diphthongs in Vedic imperfects and aorists but two adjacent vowels in hiatus).

 

[Au101]

The traditional explanation of this behavior is that the augment was originally a separate particle, which became attached to the verbal forms during the history of separate languages. What looks like a lengthened grade may therefore rather reflect contraction: a + a, a + i, a + u, a + r̥ (so that the initial ai and au still weren't diphthongs in Vedic imperfects and aorists but two adjacent vowels in hiatus).

I didn't know that, but I am very interested to learn it

 

[desi4life]

(We would generally expect a + i to give e (from original *ai, see Dib's post above). Classical Sanskrit ai (as transliterated in IAST) is from original *āi of course, again see Dib's post. I'm not perfectly sure but I think Classical Sanskrit ai is usually pronounced as āi despite the misleading (albeit cleaner) transliteration).

According to Macdonell’s “A Vedic Grammar for Students” (1916), “The diphthongs ai and au are at the present pronounced as ăi and ău and were so pronounced even at the time of the Prātiśākhyas. But that they etymologically represent āi and āu is shown by their Sandhi.”

So they are likely still pronounced ăi and ău unless a corrective has been introduced in recent times to the etymological āi and āu.

 

[Au101]

According to Macdonell’s “A Vedic Grammar for Students” (1916), “The diphthongs ai and au are at the present pronounced as ăi and ău and were so pronounced even at the time of the Prātiśākhyas. But that they etymologically represent āi and āu is shown by their Sandhi.”

So they are likely still pronounced ăi and ău unless a corrective has been introduced in recent times to the etymological āi and āu.

You're absolutely right, thank you.

I apologise, I was away from my books. Thanks for the correction.

I don't know about any modern corrections. Whitney seems to concur. The only thing of note is that the illustrious Wikipedia gives the pronunciation of ai as /ɑj/ (with short a-element) and au as /ɑw/ (also with short a-element). What's interesting about that though is the quality, rather than the length, of the first vowel. The noble Wikipedia gives the pronunciation of a as /ɐ/ (and I tend to agree with this), whereas it is widely recognised that Sanskrit ā is not purely a lengthened version of a but actually also has a slightly different quality to it, which Wikipedia (you've guessed it) transcribes as long /ɑː/

 

[desi4life]

You're absolutely right, thank you.

I apologise, I was away from my books. Thanks for the correction.

I don't know about any modern corrections. Whitney seems to concur. The only thing of note is that the illustrious Wikipedia gives the pronunciation of ai as /ɑj/ (with short a-element) and au as /ɑw/ (also with short a-element). What's interesting about that though is the quality, rather than the length, of the first vowel. The noble Wikipedia gives the pronunciation of a as /ɐ/ (and I tend to agree with this), whereas it is widely recognised that Sanskrit ā is not purely a lengthened version of a but actually also has a slightly different quality to it, which Wikipedia (you've guessed it) transcribes as long /ɑː/

The “illustrious” and “noble” Wikipedia 😆 also jumps to conclusions. For example: “The vowels e and o were actually realized in Vedic as diphthongs ai and au, but they became pure monophthongs in later Sanskrit, such as daivá- > devá-and áika->ekā-. However, the dipthongal behaviour still resurfaces in sandhi. The vowels ai and au were correspondingly realized in Vedic as long diphthongs āi and āu, but they became correspondingly short in Classical Sanskrit: dyā́us > dyáus.”

In reality, it seems we don’t really have any way of knowing whether the original etymological pronunciations of o, au, e, and ai were found in early Vedic, or whether they are pre-Vedic. Their late Vedic (Prātiśākhya) pronunciations appear to be the same as Classical Sanskrit.

 

[ahvalj]

The origin of this type is discussed in the above book by Burrow, but I'd explain it in Latin terms.

Many Proto-Indo-European nominals showed vowel alternations in the declension, in particular the nominative singular often possessed a lengthened grade: a pattern that survives (although modified) in the Latin pēs — pedis.

With time, this alternation tended to disappear, with either form generalized across the paradigm: for example rēx — *reges (from Proto-Indo-European *Hrēgʲs — *Hregʲ-) was replaced with rēx — rēgis. The relational adjective from this noun inherited the long vowel: rēgius.

From pēs — pedis, however, the attested adjective is -pedius (e. g. acūpedius).

Imagine that both variants of relational adjectives once existed, that is *regios/*pedios and *rēgios/*pēdios. If by some chance the noun generalized the full grade (**rex, **pes) while the adjectives still survived in both variants, we would get the situation described by Burrow for earlier Indic, where both full- and lengthened grade- derivatives are sometimes attested. At the next stage, the pattern **rex — rēgius started to expand, which has resulted in the system so characteristic of Sanskrit.

 

[ahvalj]

The traditional explanation of this behavior is that the augment was originally a separate particle, which became attached to the verbal forms during the history of separate languages. What looks like a lengthened grade may therefore rather reflect contraction: a + a, a + i, a + u, a + r̥ (so that the initial ai and au still weren't diphthongs in Vedic imperfects and aorists but two adjacent vowels in hiatus).

I think an alternative explanation is also possible.

The root imperfects originally possessed the same alternation of the full grade in the singular with the zero grade in the dual and plural as the root present, like for example in the Proto-Indo-European *hₑei̯t ‘he went’ : *hₑite ‘you went’ or *hₑedt ‘he ate’ : *hₑdte ‘you ate’. The augmented forms were, consequently, *hₑehₑei̯t : *hₑehₑite and *hₑehₑedt : *hₑehₑdte.

After the Indo-Iranic coloring *e>a and the fall of the laryngeals, the regular pre-Vedic phonetic outcome of this paradigm in the augmented forms must have looked like *a.ai̯t : *a.ita and *a.att : ātta, and, after contractions, *āi̯t : *ai̯ta and *ātt : *ātta — that is with the alternation vṛddhi : guṇa in some roots and with vṛddhi across the paradigm in others.

This second type was apparently generalized, as simpler, for all verbs with initial vowels, which produced the non-alternating *āi̯t : *āi̯ta and hence the attested Vedic ait : aita.