Your glossing of the case of tomaatista is inconsistent.


		Your current lexical entries allow verbs and verb complements
		nouns as noun specifiers, etc (modulo semantic constraints
		specified by the matrix, so this isn't as bad as it might be).
		Also, it's more `standard' to have the lex-types say SPR <>
		on the dependents than to have the phrase structure rules require
		non-head daughters to be phrases, though your solution worked too.


		Case is for the next lab (and we won't need to mess with
		the phrase structure rules to handle it, aside from your
		bare np rules and their interaction with case).
		

		You don't want to constrain the head daughter  here to be
		a subtype of noun-lex, since that will rule out detless NPs
		with modifiers or complements.


		NPs in subject position don't have case agreement between
		the determiner and the noun?

	
		So are indefinite NPs completely barred from subject position?
		If not, how do you express indefiniteness on subject NPs?


		The discussion of the relationship of case to definiteness
		on bare NPs was a little hard to follow.  

Write up #4 (20): 15
	discussion of spurious strings and what might need to be done
	to rule them out

		This part of the lab didn't ask you to try to solve the
		overgeneration problem, but to discuss what aspects of the
		grammar you would need to incorporate in order to solve it.