FWP:
Well, here in {164,9-13} we have a verse-set that really is a verse-set! Arshi presents it as one, and also it feels like one in every way. Nazm treats it as one, and so do almost all the other commentators as well. Moreover, every single verse in it contains phir , which means that enough of a context develops so that in the case of this verse-set phir can basically be read as 'again', without much need to invoke its alternative meaning of 'then'.
To my surprise, Nazm's objection to its 'un-ghazal-like' language is echoed by several more commentators, although less passionately and at less length. I join Bekhud Mohani in telling them they should lighten up. Not only are there plenty of counterexamples to their narrow, restrictive claims about classical ghazal language, but also, who are they to say what kind of terms the ghazal can and can't assimilate? Who's to tell Ghalib (or Shakespeare, that powerful coiner of dozens of neologisms) what words he can or can't use? In a two-line verse, after all, 'a fresh word is equal to a theme'. And Nazm himself knows this perfectly well: he has cited this very quotation in his discussion of {17,2}. His literary instincts seem to be pulled in several directions at once; his commentary on different verses expresses attitudes that are sometimes strikingly self-contradictory.
In the present verse, do both lines describe the same situation-- that the beloved is coquettishly holding 'court', and judging her lovers guilty of all kinds of offenses? Or do the lines refer to two different situations? After all, as Shadan observes, in Ghalib's day colonial courts came in three varieties, so that 'court' might be a general term, used for the coquettish, arbitrary, dictatorial salon of the beloved, while the rush at the 'criminal court' [faujdārī] might refer to the cases of mad lovers doing crazy things and getting arrested in a whole different context (of actual criminal behavior).
Or, thanks to the versatility of the iẓāfat construction, the 'courthouse of coquetry' could even be the courthouse in which coquetry will be brought to trial, as Shadan proposes-- and since coquetry murders the lovers, this might well be an affair for the criminal court.
It's true, as Nazm sneers, that not much is done with the word bāzār . But then, this is the introductory verse of a verse-set, so it's entitled to be judged and enjoyed along with its companions. And (Nazm to the contrary) their technical, law-court language is indeed manipulated very cleverly and wittily.
Note for grammar fans: It's not possible in this context
to distinguish khulā hai , 'has opened' (present perfect),
from khulā huʾā hai , 'is in a state of having become
open' (past participle with the huʾā colloquially omitted). But surely it doesn't make much difference here.
Nazm:
In this verse, bāzār is an extremely uninspired/'cold' [ṭhanḍā] word. (178)
[Commenting on verses 9-13:] In this verse-set, ʿadālat and faujdārī and sarishtah-dārī and savāl denā and muqaddamah and rūbakārī -- all these terms are, to this day, rejected by the eloquent. The reason for the aversion is that these are not terms made by expert knowers of the language [ahl-e zabāñ]. Although by necessity everyone is forced to say them, as yet they are not properly established, and the Urdu language has not accepted them. And even if you consider them to have entry into the language, then in those special meanings they are all Indian words [lafz̤-e hindī]-- it will not be correct to bring them into a Persian construction [like the iẓāfat].... The objection has always been raised against this verse of Atish's
kisī kī maḥram-e āb-e ravāñ vuh yād āʾī
ḥabāb kā jo barābar koʾī ḥabāb āyā
[someone's delicate/'running-water' bodice came to mind
when a bubble came next to another bubble]
-- that is, although the word 'bodice' [maḥram] is not Indian [hindī], with the meaning of 'bodice' [angiyā] it is Indian, so why did he use it with a Persian iẓāfat ? [Further discussion and more examples along similar lines.] (178-79)
== Nazm page 178; Nazm page 179