Petran,
theoretically D2 blockers do very well getting rid of the positive sx's, and what is left once treated is the negative sx's. Of course the sedation from many meds can easily look like negative sx's as well. Whopper already mentioned anticholinergic effects (which can ultimately cause a delirium but at lower doses just a cognitive cloudiness). 5HT2A antagonism is what was theorized to help negative sx's, which was particular to most atypical/2nd gen antipsychotics. There is pretty good research that those with classical schizophrenia (schneiderian) have cognitive deficits, including (but not limited to) inhibition and acclimation to stimuli (such as measured with pre-pulse inhibition). In the area of PPI a lot of the research shows antipsychotics correct PPI, though, and since antipsychotics don't seem to help negative sx's this doesn't seem to be the sole explanation/component for negative sx's.
Selectivity is a huge problem. We're not bad at hitting receptors, but not the specific brain regions. We talk about the mesolimbic area being related to positive psychotic sx's (paranoia, hallucinations), and mesocortical being related to negative sx's. Conceptually why you would consider limbic encephalitis in the ddx of a new onset psychosis case.
Here's an article I found quickly (review) on neurocog deficits in schizophrenia.
http://www.ncbi.nlm.nih.gov/pubmed/21312407
This is being written about a lot more, as the real frontier being focused on is cognitive rehab for schizophrenics, with a lot of possible meds that Could help (as usual, all theoretical), including galantamine (and other cholineesterase inhibitors) and namenda. Glutamate antagonists are
talked about as maybe the next big thing too, but it could all be hype.