The Object of the Social Function

The question I want to raise here is this: for what entities can social phenomena be functional?

In biology or physiology these matters are relatively clear. Traits can be functional for species – of course there are huge debates in biology over what a species is, but the criteria make good sense in particular contexts. In physiology there is even less debate over say the function of, say, the heart for a human being. For organisms have very clear boundaries.

 Regarding social affairs things are not that clear. I think three entities for which social phenomena can be functional come into view:

1) society. But what is a society? Is it a state? This seems to be too minimal, as there is a lot going on beyind the state, and beyond that, there can be stateless societies, as in the mountainess areas of south east Asia or the cossacks that lived above the black sea, to name but a few examples.

Perhaps one could say that society is simply a number of people and the relations between them, but that also does not quite work. For would one say that, following Luhmann, society decreases in size when I cut my hair? One could reduce the definition to just the relations between these people. But then there are people with a double nationality, or citizens living abroad, who make the boarders of a society very fuzzy. Relations between people are moreover cultural to a large extent, and culture tends to float the boundaries of groups unless there is absolutely no contact with other "societies", but that is extremely rare. Moreover there are economic ties between what we call  ‘societies’ in that subsidiary firms settle abroad, or European law entering national European legal systems.

Let’s make this concrete. What does it mean when I say: lowering tariffs is functional for Dutch society? Or: the performance of the national soccer team is functional for German society. Or: trade is functional for world society. These appear to be empty phrases because they do not have a clear reference.

A way out of this problem is to define society as communication, but the costs are high. Elaborate and implausible notions of self-reference and self-reproduction must be introduced, which cannot be scrutinized here, but serve to undermine this approach. 

2) and 3) would be social actors and social mechanisms, and these look more promising.