Finding the metrical scansion of a Persian ghazal is not an easy task: many possibilities are offered in each line and hemistich. But several (voluntary !) limitations or regularities are always present , which can help us to find the (probable) solution:

1. All lines present the same metrical structure:

1.a. Lines are divided into two hemistiches.

1.b. All hemistiches show identical sequences, i.e., the same succession of long or short syllables.

1.c. Hemistiches can be divided into smaller (equal or different) parts (groups of long/short syllables ) called feet.

1.d. Only the final syllable of a foot can be either short or long, offering (to) the poet a certain flexibility.

1.e. Optional shorts can configured from a single letter with the addition of a neutral motion , a schwa sound that prolongs (moves) a consonant into a new syllable, inexistent in plain speech. This is relatively common in overlong syllables, those composed by three or more letters: i.e. the word Í±Åœ [nist], can be uttered as a single overlong syllable: [ni:st], or as two, long and short: [nis|te], by adding that extra motion.

This option is solved by the underlying metric pattern, which in turn is suggested (even enforced) by the extant lines of the ghazal.

These difficulties explain the necessity of a previous study of the ghazal ‒ in its shortest form, a thorough (?) glance ‒ of the whole poem. However one can always suspects that other possibilities, other metrical patterns, could be found, specially in very short poems.

Other difficulty, not well defined nor solved, lies in finding the inner divisions of a line into chained feet.