The Individual Family
Until relatively recently, the concept of an atom was that it... well, was indivisible. It was the smallest unit. Yet things didn't work out quite right that way, and we ended up discovering that there were smaller interacting bits.
Likewise, we have lots of models of human behavior - all of which seem to fit... sometimes. It's here that modern neuroscience might point to a solution. There are lots of processes - perhaps a preponderance - that are acting subconsciously in our minds. Purely sociobiological solutions might discount entirely the influence of the conscious mind, but that's another error of oversimplification. Yet social sciences concentrate (nearly entirely) upon the conscious levels.
What if they're both right?
Consider that the organism... well, me... is not just my conscious mind. That there are nearly independent sub-processes that work in concert with my conscious mind. (Note - not in a subject/master relationship; more like a collaboration.) The subconscious data influences my conscious mind, my conscious mind influences the subconscious areas... and we've got a dialectic process that is "Steve". Throw in the rest of biology (hormones, physical structure, etc), and you've suddenly got a fairly complex triumvate that work independently, but work together and mutually influence each other.
Trying to model all three levels as a single "person" and talk about behavior at that level alone is like talking about the behavior of "families" without acknowleding that there are indivdiuals in that family.
This hypothesis would explain both the strengths and weaknesses of various threads of the social sciences, biology, and medical regarding human behavior.