THIS PAPER PROPOSES SOME INSIGHTS INTO THE PROBLEMS FACED BY SOMEONE ABOUT TO EMBARK ON A GENERATIVE GRAMMAR OF AZERBAIJANI. IT IS NOT A SYSTEMATIC ANALYSIS, BUT A SAMPLING OF IDEAS PERTINENT TO SUCH AN ANALYSIS. THE PROBLEMS DEALT WITH ARE (1) GRAMMATICAL DEPTH IN AZERBAIJANI BASED ON VICTOR YNGVE'S DEFINITION OF "DEPTH"--A NUMBER OF SENTENCES OF DIFFERENT LENGTHS ARE ANALYZED AS TO DEPTH AND NUMBER OF TRANSFORMATIONS NEEDED FOR THEIR GENERATION, (2) GRAMMATICALNESS IN AZERBAIJANI, BASED ON HOUSEHOLDER'S DEFINITIONS OF GRAMMATICALNESS AND SAPORTA'S "DEGREES OF UNGRAMMATICALNESS"--AZERBAIJANI USES BOUND MORPHS TO A HIGH DEGREE, AND THESE ARE FIXED IN THEIR ORDER, BUT THESE SO-CALLED WORDS, CONTAINING MANY MORPHS, CAN BE MOVED AROUND IN THE SENTENCE FAIRLY FREELY, WITH A"GRADATION OF CONSISTENCY", (3) THE IDIOM GRAMMAR OF AZERBAIJANI--THE IDIOM IS DEFINED, AFTER HOCKETT, AS "SOMETHING NOT REGULARLY GENERABLE FROM SOME PRODUCTIVE FORMULA OF THE GRAMMAR...." THEY MUST BE LISTED AND GLOSSED IN THE LEXICON AND ARE CONTAINED UNANALYZABLE IN THE LAST LINE OF THE SENTENCE GRAMMAR, THEREFORE BELONGING TO THE MORPHOLOGICAL GRAMMAR. THE ANALYSIS INCLUDES A TWELVE-PAGE HANDOUT. (KL)