Scansion for urdU Poetry

Disclaimer

This page is still in its formative stages and most of the information is not necessarily complete in itself ... I present a semi-formal, though non-historical, approach to urdU prosody - but it is by no means an expert's thesis. All this information is culled mostly from my memory and the meagre notes from having read Agha Sadiq's treatise on urdU prosody "nikAt-e fann" a long time ago.

The basis for generating the meters as employed in urdU poetry is available from the urdU Prosody (Basics) page.

taqtI` (scansion)

Merely knowing all the buhUr (meters) employed in urdU poetry is not sufficient to determine if a given line is in [some] baehr (meter) or not. A good portion of metrical scansion relies on the shortening/strengthening of syllables to fit a given meter. Broadly, the following sections outline the heuristics that encode most of the rules of taqtI` as employed in urdU prosody. Obviously, these are not an expert's comments - and are intended only to provide the spirit of taqtI`. I've outlined as many exceptions as I am aware of at each stage.

This section is taken directly from one of my posts on ALUP with some modifications/augmentation.

Terminal Long Vowels

In words that end in long vowels (whether nasalized or not), the terminal long vowel is a candidate for vowel reduction.
eg. kahI.n can be scanned as kahi.n, mujHE can be scanned as mujHe etc..

Exceptions to this rule:

• Words like kyA that start with a consonant cluster (in which the initial consonant is already considered subsumed; see section "Miscellaneous Cases").
• In words that end in a nasalized "A" (eg. yahA.n, kahA.n etc.) this reduction is best avoided.
• A final "A" cannot be reduced in the preposition tA and the negative particle nA.
• In fArsI words, a final nasalized "I" cannot be reduced.
• The "E" in bE (without) cannot be reduced.
  

However, I *personally* try to AVOID this terminal reduction in the following additional cases because it "sounds" bad to me. This is only *my* preference, aehl-e `ilm may well disagree :)

• Most Nouns (both substantiative and adjective).
• Various usages of "hindI" verbs eg. samjHE
• All infinitive verbs eg. samajHnA, karnA, pakaRnA
  
• Numerals eg. dO (2), saO (100).
• (Most of) Address and emphatic cases. eg. sAqiyA!, sunO!, aE ... !
  

Note: izAfat (-E) and conjunction (-O) are effectively terminal long vowels of the preceeding term and hence are subsumed by this rule i.e. they can either be scanned as long or short syllables. But, composition (tarkIb) is a special case altogether - and resulting words (eg. gulistAn) usually have fixed aOzan (metrical weights).

Non-Terminal Long Vowels

By default, non-terminal long vowels cannot undergo reduction! But there are exceptions, ofcourse:

1. aOr (and): Only when used as a conjunction can the "aO" vowel be reduced ie. aOr (vazn LS) can be scanned as aor (vazn L).
Thus, "ham aOr tum" can be scanned as "ham aor tum".

However, when the word is used in the sense of 'more, different' this reduction CANNOT be allowed.
Thus, the aOr in "uskA tarz aOr haE" cannot be reduced to aor.


2. For certain words like tabAh, rAh, ShAh, mAh, ChAh, gunAh, siyAh, nigAh, etc. when used standalone, the "A" vowel can be reduced to another short vowel.
eg. gunAh can be scanned as guneh. This is also applicable when the word is used, as is, in a compound. eg. "nigah-e yAr".

Modified/derived forms of these words can only use the long "A" vowel. eg. siyAhI etc.

Exception: This "A" reduction is not applicable to gavAh and panAh.


3. The "I" vowel in the word A'InaH can be reduced. (Also see section "The Impeceptible H", and "Miscellaneous Cases" below)


4. For mErA, tErA (and related words like mErI, mErE etc.) the "E" vowel can be reduced. Thus, mErA can be scanned as merA etc. (Most people consider merA, terE etc. as entirely different words altogether - preferring to pronounce them as mirA, tirE etc..)

Liaisons

When a word begins with a *pure* vowel, it is possible to attempt a liaison with the preceeding word if the latter ends in a consonant.
eg. "ham uskE" (vazn LLL) can be scanned as "ha muskE" (vazn SLL), "ab AgE" can be scanned "a bAgE" etc.

Note that this is not possible if the second word begins with the `ayn (glottal stop) since, technically, there is a glottal stop (a consonant) guarding the beginning of the word.
Thus "jab `uzr" CANNOT be scanned as "ja buzr".

(I personally feel that forcing too many liasions in a single line to fit the demands of meter is poor practice - and usually ends up sounding not so good.)

The Imperceptible H

In cases of words that use the "H" symbol in the persian script merely as a placeholder to indicate a final short vowel, the "H" can be strenghted and adds on to the vazn. eg. A'InaH (vazn LLS) can be scanned as A'Inah (LLL), zalzalaH (vazn LSS) can be scanned as zalzalah (LSL) etc. This is, however, NOT applicable to the relative pronoun keH and the negation naH.

Note, ofcourse, when these words are used in izAfat or conjunction this rule hardly matters :)

Special Note: The opposite of this rule CANNOT be applied for words in which the "h" is an actual sound (and not merely a placeholder). Thus kOh (vazn LS) remains kOh and CANNOT be scanned as kOH (vazn L).

Miscellaneous Cases

a. tarah: This can scan either as tarah (vazn SL) or tarh (vazn LS). However, when the word is used to mean "foundation" it can scan only as tarh (vazn SL).

b. gulistA.n: This is allowed to scan as gulistA.n (vazn SLL) or gulsitA.n (vazn LSL).

c. A'InaH: As already mentioned in the sections above, this word can be scanned variously as, A'InaH (vazn LLS), A'Inah (vazn LLL), A'inaH (vazn LSS), and A'inah (vazn LSL).

d. Words that start with consonant clusters (eg. pyAr, prEm, kyA etc.) consider the initial consonant cluster as subsumed into a single consonant.

e. Classical urdU scansion allows Khizr to be scanned as Khizar. I personally, refrain from this in my usages.

f. Classical urdU scans the sa.Nsk.rt word brAhmaN as barhaman (vazn LSL) or barehman (vazn SLL). I personally scan it only as brAhmaN (vazn LSL) in my own verses.

g. gulistA.n: This scans either as gulistA.n (vazn SLL) or gulsitA.n (vazn LSL). Again, I prefer only the SLL form.

h. Few cases of repeated consonants can be considered as words with a single consonant. eg. nazzAraH (= nazAraH), naShShA (= naShA), ummId (= umId), baChChaH (= baChaH), hadd (= had), rabb (= rab), haqq (= haq)

i. Finally, for most fArsI nouns that have a terminal nasal vowel, opinion is divided on whether such words can be used with the nasal component being converted into a bonafide consonant. i.e. Whether the use of words gulistAn, parEShAn etc. is legal; and if so, under what conditions. This discussion is outside of the scope of this document.