Since the sixties functional analysis is in retreat in the social sciences, but not so in biology. From its very beginning, sociology has struggled with the question what a functional explanation is, and whether it applies also to social affairs.
The first question one has to deal with is whether functional explanantion is just a form of causal explanation, or whether it stands alone. For the purposes of this Logi, I assume with Anthony Giddens (1996; original 1976) that all functional explanantions can be rendered as "functional effects", which are therefore simply types of causal explanation. This equation is certainly not accepted by all functionalist sociologists, most notably not by Niklas Luhmann, but that would be a different discussion.
If we say that functional explanantion is a form or type of causal explanation, what then makes it different? According to Giddens, there are two options:
1) By way of assuming needs of a social system. A phenomenon is functional when it contributes to the survival of a more inclusive system within which it exists, and thislarger system then has "needs" which are defined as this "survival value" (87). In the world of biology this can be defined simply as the life of an organism, but in social affairs we must always ascribe needs rather arbitrarily to a social system (93). In other words, there are no independent system needs in the social world, therefore this agument does not work. I have no problems here.
2) Instead of supposing indpendent social needs, one can also argue for survival value by way of defining functional prerequisites. The problem Giddens has with this argument is that these prerequisites are either a) a tautology or, if they go beyond that, b) no longer prerequisites.
a) In the world of living systems, one can distinguish anatomy as the structure, from the operation of that structure as its function, Giddens argues. But in the social world one cannot make this distinction - i.e. one cannot observe social structure the way one can observe a dead body as the anatomical structure apart from its operation in social affairs. In the social world the structure only is insofar as it is functioning. For that reason, the description of the prerequisites for the survival of a social system often amounts to simply a definition of that social system, like having similar cognitive orientations and the like (94). In other words, the prerequisites for its functioning are the same as the prerequisites for its structure.
I do not see how this argument works. Even if structure and function would be the same in social systems, I do not see why this would prevent defining a social system by way of the prerequisites of its existence.
b) Some functionalists in sociology argue by way of factors that enhance the adaptive capacity of a social system, say a society. For Giddens they either restate the definition of a system, and if they go beyond that, they no longer talk about prerequisites, but about an adaptive advantage over other societies (95). Indeed, this no longer is a prerequisite, but this argument hinges on the validity of 2 a), which is doubtful.
This famous 1976 article is invoked by many as the one that drove the nails into the coffin of sociological functionalism, but I do not quite see how it has done so.