[Hindi/Urdu] Disappearing ka of patient and noun incorporation

bargolus

New Member
Danish and English - British
On (the now deceased) Peter Hook's webpage with lessons on the grammar of Hindi ("The Mellon Project"), the mentions two grammatical features.

In my limited understanding, the two features are as follows:

(1) "Disappearing ka of patient" where the ka is optional and you can either say

Bhaiye ne apni taakat ka istemaal kiya / भैये ने अपनी ताकत का इस्तेमाल किया / بھائی نے اپنی طاقت کا استعمال کیا

or

Bhaiye ne apni taakat istemaal ki / भैये ने अपनी ताकत इस्तेमाल की / بھائی نے اپنی طاقت استعمال کی

This stands in contrast to words where the ka is obligatory, e.g. X ka khayaal karna, where the ka is not optional (you can't say *X ko khayaal karna).

(2) Noun incorporation, where a noun is so thoroughly part of a verb that it stops being a noun on its own. For example, the kharch in kharch karna / 'to spend' creates gender agreement with the patient of the action rather than kharch in perfect tense:

Usne bahut paese kharch kiye / उसने बहुत पैसे खर्च किये / اسنے بہت پیسے خرچ کیے

'not' / nehin can be put in front of kharch rather than the verb:

Usne bahut paese nehin kharch kiye / उसने बहुत पैसे नहीं खर्च किये / اسنے بہت پیسے نہیں خرچ کیے

Passive ignores the noun:

bahut paese kharch kiye gaye / बहुत पैसे खर्च किये गए / بہت پیسے خرچ کیے گئے



Now Peter Hook listed a number of examples of both cases:

bayan kar, talaash kar, tyaag kar, chori kar, shuru kar, istemaal kar for (1)
bhent kar, dhaaran kar, sahan kar, bardaasht kar, arz kar, kharch kar, shuru kar for (2)

Given shuru kar appears on both lists - I wonder if the two things really refer to the same phenomenon or are there nouns that fit into pattern (1), but don't fit into pattern (2)?
 
  • Dib

    Senior Member
    Bengali (India)
    The way you have described the phenomenon, makes perfect sense. However, I am, myself, not a native or extensive enough user of Hindi/Urdu. So, I won't comment on the accuracy of the two lists in general, especially on the items that are not colloquially frequent in my milieu. However, "shuruu karnaa" is a common expression and it belongs only to type (2) around me. There is also "... kii shuru'aat karnaa", which, for me, takes the "kii" obligatorily. Not sure if this can belong to type (1). I'll leave that to the experts.
     

    aevynn

    Senior Member
    USA
    English, Hindustani
    Perhaps one way to reformulate the distinction that Hook's getting at is whether the incorporation of the noun into the predicate is optional or obligatory: it's optional for one group, and obligatory for another group.

    [I suppose acceptability judgments can vary from person to person, and what's obligatory for one person could be optional for another person; I'll throw some asterisks in front of things to mark the ones that sound off to me, and others can chime in to share if they disagree with my acceptability judgments :)]

    It seems like sometimes the noun incorporation is obligatory:

    usne kaam shuruu kiyaa​
    *usne kaam kaa shuruu kiyaa​
    usne bahut musiibateN bardaasht kiiN​
    *usne bahut musiibatoN kii bardaasht kii​
    usne bandar kaa ruup dhaaran kiyaa​
    *usne bandar ke ruup kaa dhaaran kiyaa​

    Sometimes, the incorporation is optional, like in the taakat sentence you have, or:
    usne paise chorii kiye​
    usne paisoN kii chorii kii​

    And, of course, there are nouns (like the shuruaat that @Dib points out) where incorporation is forbidden:

    *usne kaam shuruaat kiya​
    usne kaam kii shuruaat kii​

    It seems to me like the passivization that you mention is basically linked to the same phenomenon, in the sense that if a noun is incorporated into the predicate, passivization leaves it be:

    usne paise chorii kiye -> paise chorii kiye gaye (or: paise chorii hue)​
    usne paisoN kii chorii kii -> paisoN kii chorii kii gayii (or: paisoN kii chorii huii)​
    It's also true that, when the noun gets incorporated into the predicate, you can move the nahiiN before that incorporated noun, but I think this changes discourse pragmatics...? It feels to me like the "default" word order for "He didn't steal money" is still usne paise chorii nahiiN kiye, while usne paise nahiiN chorii kiye sounds more like the kind of thing someone might say if they were about to mention that this person stole something else besides money...?
     
    Last edited:

    Pokeflute

    Member
    English - American
    Teach Yourself Hindi mentions this process with talaash karnaa

    Step 1: "Ki" is obligatory (main uski talaash kartaa hoon)
    Step 2: "Ki"is removed (main usko talaash kartaa hoon)
    Step 3: It becomes its own independent verb (main usko talaashta hoon).

    It's very rare for a verb to move from one step to the next, because Hindi is a language where verbs form a "closed class" (i.e. they don't readily adopt new members, unlike English).

    Also verbs may be at different stages for different speakers - this is one of those areas where grammar can differ even between educated native speakers (though I think most would consider step 3 for "talaash" to be ungrammatical/non-standard).

    A similar process exists in other languages with close class verbs. For example, Japanese (where "suru" is the equivalent of Hindi "karnaa") follows a similar pattern.
     
    Last edited:
    • Thank you!
    Reactions: Dib

    aevynn

    Senior Member
    USA
    English, Hindustani
    Looking at this three-step diachronic process leads me to an observation and a question:

    It seems like, when English verbs are loaned into Hindi-Urdu, they're typically loaned in such a way that incorporation of the loan into the predicate is obligatory. For example, consider pik karnaa in the sense of "to pick someone/something up":

    maiNne eyarporT se use pik kiyaa​
    *maiNne eyarporT se uskaa pik kiyaa (*uskii pik kii)​

    It's as if there is no pik in isolation, only the verb pik karnaa, but this verb still behaves like a complex predicate since the two elements can be pulled apart (eg, ... pik nahiiN kiyaa). This seems like a sort of general phenomenon with loans of this form (like the transitive liiv karnaa in the sense of "to leave" that just showed up in this thread!). Looking at it purely from a synchronic perspective, these predicates are a bit like dhaaran karnaa in that there's no dhaaran in isolation, there's only the complex predicate dhaaran karnaa (at least, I'm having trouble formulating an acceptable sentence that uses dhaaran purely as a noun that's unattached to karnaa).

    But there is a difference from an etymological perspective, in that the dhaaran of dhaaran karnaa derives from a noun while the pik of pik karnaa derives from a verb. Just the fact that pik comes from a verb makes anything like *uskaa pik kiyaa feel very unlikely. Which leads me to my question:

    Are there Hindi-Urdu predicates of the form X karnaa where X is etymologically a verb not from English?

    If there are examples with X deriving from a verb in one of Hindi-Urdu's "traditional" sources of loans (Persian, Arabic, Sanskrit), perhaps the resulting predicates behave like the English ones do today. Otherwise, perhaps Hindi-Urdu's interactions with English and its abundance of verbs is introducing a new process which effectively skips step 1 that @Pokeflute describes above.

    [Footnote: I suspect that what I've described above may not be identical to what seems to happens in Japanese with these kinds of loans...? For example, the English verb "fry" has been loaned into Japanese as furai. When I googled a bit, it seems like both of the following are possible in the sense of "to fry X" where X is some food item:

    Xをフライする​
    X-o furai suru​
    Xのフライをする​
    X-no furai-o suru​

    In the former, X is marked as a direct object and furai runs up directly against suru, whereas in the latter, X is marked gentive and furai is marked as a direct object. So, it seems as though furai, despite deriving from an English verb, has full status as a noun in Japanese, which may or may not be incorporated into a suru complex predicate...? But my Japanese is *far* from fluent, so don't take this footnote too seriously...]
     
    • Thank you!
    Reactions: Dib

    aevynn

    Senior Member
    USA
    English, Hindustani
    Sorry about the double post, but I also just realized that this rephrase by @littlepond is actually shows yet another way to see just how "thoroughly the noun becomes a part of the verb" :)

    Note that "usne bohat musiibatoN ko bardaasht kiyaa" also exists.

    It's sort of related to this comment:

    For example, the kharch in kharch karna / 'to spend' creates gender agreement with the patient of the action rather than kharch in perfect tense:

    Notice that bardaasht, when used as a standalone noun, is typically feminine (eg, ye merii bardaasht se baahar hai). However, in the above sentence, the verb agreement on kiyaa is masculine. Presumably this is because the noun bardaasht is incorporated into the predicate and the predicate cannot treat this incorporated noun as it's subject; speaking loosely, the predicate has to try to find another unmarked noun outside of the predicate to be its subject, and since it can't, it defaults to the masculine singular ending.

    In other words, it's not just that the verb inflection doesn't agree with bardaasht when there's another unmarked agent/patient for it to agree with, it's that it cannot agree with bardaasht, even when there's no unmarked agent/patient:

    *usne bahut musiibatoN ko bardaasht kii.
     
    • Agree
    Reactions: Dib

    Dib

    Senior Member
    Bengali (India)
    If there are examples with X deriving from a verb in one of Hindi-Urdu's "traditional" sources of loans (Persian, Arabic, Sanskrit), perhaps the resulting predicates behave like the English ones do today. Otherwise, perhaps Hindi-Urdu's interactions with English and its abundance of verbs is introducing a new process which effectively skips step 1 that @Pokeflute describes above.
    I'll be surprised if some of the Arabic and Persian loans did not directly start in stage 2, like was kharch really ever in stage 1?

    The other thing is that Arabic and Sanskrit verbs are heavily inflected, which often makes the verb root/stem rather obscure if viewed from the logic of Hindi/Urdu grammar. So, it is probably rather difficult to zero in on a verb-form to borrow. Therefore, the verbal nouns (prayog, isti'maal) and passive participles (prayukt, musta'mal) are prefered for both these languages.

    English is a lightly inflected language, the verb stem is a free form, often having the same form as the verbal noun. All this helps in borrowing English verb stems in the same way as the nominal forms from other languages.

    The status of Persian is somewhat in-between, I believe. The verb is as much inflected as in Hindi-Urdu, but also structurally rather similar to it. This allows isolation of the verb stem to an extent, which allows us to form simplex verbs like guzar-naa, farmaa-naa, etc. from the Persian present tense stems. There are also Persian verbal nouns with the same form as the past tense stem, and they are available for borrowing in Hindi-Urdu as compound verbs, e.g. bardaasht karnaa.
     

    aevynn

    Senior Member
    USA
    English, Hindustani
    I'll be surprised if some of the Arabic and Persian loans did not directly start in stage 2, like was kharch really ever in stage 1?
    I wondered this too, and then figured that probably Snell would have had some historical data to back up a proposal like this. But I just looked into it further, and the passage that I was able to find where Snell discusses this doesn't actually seem to propose a diachronic process:

    A few verbs of the second type can drop their possessive का/की/के, thus becoming first-type verbs: तलाश करना 'to search for.' Just occasionally, a base noun even sprouts a -ना ending and thus generates a verb: किसी को तलाशना 'to search for someone.'
    Snell's "second-type verbs" are of the form X kaa noun karnaa, and "first-type verbs" are adjective karnaa. It's not clear from this passage, at least, that there is a diachronic transition from optional noun incorporation to obligatory noun incorporation, and this does leave open the possibility that kharch has always had obligatory noun incorporation.

    It would be fun to have a historical corpus to test these kinds of hypotheses...

    The other thing is that Arabic and Sanskrit verbs are heavily inflected, which often makes the verb root/stem rather obscure if viewed from the logic of Hindi/Urdu grammar. So, it is probably rather difficult to zero in on a verb-form to borrow.
    This is a great observation, thanks for pointing this out! :) This is probably precisely why it is hard to find loans involving verbs from Sanskrit and Arabic. Both have word-internal vowel mutations, and Sanskrit also has its consonant sandhi, and I can definitely see how these kinds of transformations would effectively obscure verbal roots from the point of view of Hindi-Urdu grammar. In contrast, verb conjugations in Persian and English typically just tack morphemes on at the end, and it makes sense that this would make the verb stem easier to identify from the point of view of Hindi-Urdu grammar.

    I hope that some of these English loans do end up forming simple verbs like guzarnaa and farmaanaa. My family does occasionally say sentences involving hypothetical verbs like siTnaa or iiTnaa, but mostly for humorous effect; maybe one day, it won't just be for humor! :)
     
    • Agree
    Reactions: Dib

    Dib

    Senior Member
    Bengali (India)
    I hope that some of these English loans do end up forming simple verbs like guzarnaa and farmaanaa. My family does occasionally say sentences involving hypothetical verbs like siTnaa or iiTnaa, but mostly for humorous effect; maybe one day, it won't just be for humor! :)
    The same humorous phenomenon exists in my Hindi-speaking friend circle as well. But it (still) seems to be just a fringe phenomenon, both within my circle and in the wider society. On the other hand, I don't think, this phenomenon exists at all in my Bengali-speaking circle. I wonder whether the fact that half of my Hindi-speaking circle is composed of non-native speakers, who probably don't have older family members affecting their language-choices makes a difference. In my mind, the friend who uses this construction most heavily (always humorously, though) is a native Telugu-speaker.
     
    < Previous | Next >
    Top