Hindi, Urdu: light, heavy, super-heavy syllables in relation to stress

MonsieurGonzalito

Senior Member
Castellano de Argentina
Friends,

What are the proper Hindi/Urdu words to refer to light, heavy, and super-heavy syllables, in the context of syllabication to determine the stress of a word?

For example, the page here gives al algorithm in order to determine where the stress falls in some words.

The idea is that:

  1. a consonant followerd by short a i u is considered a light syllable
  2. any other vowel, or a consonant followed by it, is a heavy syllable
  3. syllables starting and ending with a consonant, are super-heavy syllables
Leaving aside for a moment the obvious fact that this categorization doesn't cover all possibilities, the idea is that the heaviest syllable is stressed, and in words where there is more than one syllable having the most weight, the one before the last should be preferred.

For example:

- ghaRii (watch) gha-Rii light-heavy, hence: ghaRii
- billii (cat) bil-lii superheavy-heavy, hence: billii
- kaalaa (black) kaa-laa heavy-heavy, hence kaalaa

So my question is: what would be the vernacular names for these light, heavy, super-heavy terms?

Thanks in advance.
 
  • MonsieurGonzalito

    Senior Member
    Castellano de Argentina
    Wow, so the whole mentioned construct is rubbish?

    In the Urdu Lughat sound clip for ghaRii, much as I try, I can't really say that one syllable is pronounced with more stress than the other.
    Urdu Lughat
     

    aevynn

    Senior Member
    USA
    English, Hindustani
    Stress is just not very important -- nothing like English or Spanish. As that article you link to points out, there isn't even agreement that stress is an actual feature of Urdu-Hindi. The rule given in that article also disagrees with the rule on Wikipedia (or at least, what I can understand of that rule, since the rule on Wikipedia isn't very clearly stated...).

    But, putting stress aside, there are words for "syllable" [1] weight in the context of poetry.

    In the Hindi poetic tradition, I think "syllables" come in two varieties: laghu (optional consonant plus a/i/u) and guru [or diirgh] (anything else, ie, optional consonant plus aa/ii/uu/e/ai/o/au, or optional consonant plus any vowel plus consonant). I think a schwa that's orthographically present in Devanagari transcription but typically silenced in the modern language is counted as a laghu.

    In the Urdu poetic tradition, I think "syllables" come in two varieties: kotaah (optional consonant plus a/i/u) and buland (optional consonant plus long vowel or optional consonant plus a/i/u plus consonant). I believe that optional consonant plus long vowel plus consonant is typically counted as a buland followed by a kotaah. You can look through Frances Pritchett's book about more details about this if you like.

    ---

    [1]: I say "syllable" in quotes because neither of these poetic "syllables" quite corresponds to a phonological syllable... In fact, the phonological syllable, the Hindi poetic "syllable," and the Urdu poetic "syllable" are all distinct from each other!
     
    Last edited:

    MonsieurGonzalito

    Senior Member
    Castellano de Argentina
    What a loving book, that Urdu Meter one!

    But I guess we can put that Urdu clasification aside, since it pertains to scansion rather than being operative in determining stress.

    So much complexity and contradictory views even on where the stress lies in a common word, begs one last question (in the understanding that stress is not that important):

    Which of these 3 is true?

    1) words have an agreed-upon stress among most speakers (even though the rules to determine it are complex or poorly stated)
    2) words have an agreed-upon stress, for which some non-systematic regularities can be found, but which is ultimately idiomatic.
    3) words don't have an agreed-upon stress at all
     

    aevynn

    Senior Member
    USA
    English, Hindustani
    Which of these 3 is true?

    1) words have an agreed-upon stress among most speakers (even though the rules to determine it are complex or poorly stated)
    2) words have an agreed-upon stress, for which some non-systematic regularities can be found, but which is ultimately idiomatic.
    3) words don't have an agreed-upon stress at all

    There probably is stress. In fact, there are probably even minimal pairs somewhat akin to Spanish esta vs está or English permit (noun) vs permit (verb). For instance, using green to denote the stressed syllable, I think we have:

    (A) mujhe dikhaa would be "Show it to me" (imperative of dikhaanaa).
    (B) mujhe dikhaa would be "I saw [it]" (perfective participle of dikhnaa).

    The rules I've seen (either the article you shared or whatever Wikipedia might be saying) seem too simplistic. The above dikhaa examples show that the syllable weight structure is just not enough information to completely determine stress; one at the very least needs to know something about morphology as well. But I suspect there probably is something pretty systematic going on. I think I manage to figure out how to pronounce a new word I encounter in writing without having access to any stress-related phonological information.

    That being said (and to re-iterate an earlier point), stress in Urdu-Hindi is somewhere between subtle and unimportant. In most situations one can shift the stress without changing the meaning --- and probably also without raising any eyebrows. I've found myself unable to make up my mind about which syllable of ghaRii I think is stressed for over an hour now, presumably because neither sounds decidedly "weird." (I'm currently on team -Rii, but I'm sure I'll change my mind soon after I hit "Post Reply"). Also, with the dikhaa examples above, I'm pretty confident about the stress placement in (A), but I'm less confident about (B) (eg, maybe it would be better to say that neither syllable is stressed in (B)...?). It's just very hard for me to introspect about stress in Hindi-Urdu. In contrast, I never have any trouble figuring out which syllable is stressed in an English word I know.

    TLDR; I suspect option (1) is the correct one, but this is likely a question that requires further study by phoneticians who have access to acoustic machinery and are willing to collect large samples of phonetic data (as opposed to being a question that requires further introspection, which personally doesn't feel very reliable to me on this issue).
     

    Pokeflute

    Senior Member
    English - American
    I will caveat this by saying (like @aevynn said) stress is not a very salient feature of Hindi. Native speakers (my family in this case haha) often have a very difficult time hearing stress. It's not something they really think about unless I specifically prompt them for it (unlike in English or Spanish, where word stress is very important).

    That said, below are a few personal observations I've found related to stress (based on probing my family - from Mumbai if it's relevant). In general the mora rule on Wikipedia (that @MonsieurGonzalito listed) is 90% right. It covers the vast majority of cases, with few exceptions.

    One major exception, however, is that usually the "stem" of a word carries the stress.

    For example, verbs stress syllables in the root. So "dikhaa" (show) and "dikhaa" (seen) have different stresses because they have different roots. That said, there are some conjugations where you evaluate stress on the whole word, not just the stem, but this varies by speaker. My family does not stress verb endings (they'll say "karegaa") but I've noticed that many speakers say "karegaa" with the second syllable stressed.

    This also occurs for nouns. "jaadoogar" stresses the first syllable because the root word is "jaadoo". Similarly "asliiyat", not "*asliiyat" or "*asliiyat"

    Note that, for declensions, stress does change depending on the ending. So "qismat" vs. "qismateN" and "laRkii" vs. "laRkiyaaN".

    I'm not 100% about compound words, but I believe that each compound is stressed separately with one stress being more pronounced (e.g. "gatiisiimaa", "asthiivisarjan").

    A second exception is that voiced aspirated sounds (I think) are heavier. So "dhaniiyaa" and "ghaRii." (From a technical perspective I think this makes sense as "voiced aspirates" don't really exist, but that's a different discussion)
    (EDIT: See responses below, it's more complicated for voiced aspirates)

    Another exception is allophonic changes. In general, Hindi stress is based on underlying forms, not realized forms. So a word like "pariivaar" is stressed on the final syllable (even though you might expect "*pariivaar). This is because the underlying form is "parivaar", and the "i" is lengthened due to the following /v/. (Note that "i" lengthening before /v/ isn't universal, but "dhaniyaa" above also illustrates the same point for speakers that lengthen "i" before /j/).

    Another thing is nasalization. I believe a nasal vowel gets an extra mora (for example "aaNkheN") but I'm not 100% sure (is "juRvaaN" stressed on the first or second syllable - I can't tell).
     
    Last edited:

    littlepond

    Senior Member
    Hindi
    A second exception is that voiced aspirated sounds (I think) are heavier. So "dhaniiyaa" and "ghaRii"

    Yet, wouldn't you say it's "raa" that is stressed in "raadhaa"?

    Another thing is nasalization. I believe a nasal vowel gets an extra mora (for example "aaNkheN") but I'm not 100% sure (is "juRvaaN" stressed on the first or second syllable - I can't tell).

    I think "juR" is the stressed part.

    On the whole, I agree with both @aevynn jii and @Pokeflute jii that stress is not a salient feature of Hindi, and in fact for many words I have to think hard where the stress is, and sometimes I am still not sure.
     

    Pokeflute

    Senior Member
    English - American
    Yet, wouldn't you say it's "raa" that is stressed in "raadhaa"?

    You're right. Same with "dhamaakaa". Back to the drawing board I suppose...

    think "juR" is the stressed part.
    That throws a wrench into that theory too (as you'd expect "*juRvaaN").

    I suppose the conclusion is "kind of accurate but there are things that complicate it (nasalization, voiced aspirates, etc.)"
     
    Top