The established literature on utility, seems deliberately to evade the single governing tendency of human actions, and instead allows itself to be distracted by the many subservient activities that incline statistically to the attainment of the over-riding end. The fact of the matter is that:
Statistically, humans gather resources with which to secure their reproduction.
This is axiomatic, and the phrasing is deliberate. It makes no presumption about temporal sequence, which plainly varies to a degree from individual to individual, and a little between the two sexes. No more does it rule out extended and indirect efforts to ensure that their children and grandchildren succeed in their endeavours, though these commitments are discounted due to fading degrees of relatedness as well as temporal distance, thus encouraging the concentration of scarce resources on nearer and dearer.
Composing this over-arching routine are a large number of sub-routines, these activities, as already noted, having a statistical property of inclining towards the achievement of the governing end; governing because if it fails, it exerts a statistical tendency to remove the sub-routines that led to failure. In other words, they are or will be selected against over time.
The apparent disconnection between the activities undertaken by human beings and the securing of reproduction is just that; in appearance only. Often, indeed, it is obvious that there is a plausible link, as when a person gathers resources enabling them to house a family and educate the children, but on occasion this may not be obvious at all, as when a person paraglides or climbs mountains or seeks to win a soft toy in a shooting gallery, or any of the multitudes of other things that seem either pointless in themselves or possibly even counterproductive because of wasted time or needless risk. It may even be the case that individual cases of behaviour are not adaptive (assuming that we could ever determine this), but in fact this is to be expected since the routines are selected for because they have a statistical bearing on the securing of reproduction (inclusive fitness).
Part of the impression of disconnection arises from the boundaries set around an action, say paragliding. In itself, it appears a needless risk, but if the context of this activity is expanded we would see that 'paragliding + meeting certain people + boasting both these activities" might well be statistically conducive to securing reproduction. Shooting for the soft toy is in itself useless, but seen as a part of "going out with friends + etc + etc" is likely to be reproductively functional.
Thus, and as so often in these cases, the boundaries of the categories employed are altogether crucial and may be used tendentiously to prejudice the conclusion to which we are driven. What after all, is an 'action'. Why say that 'sitting down' or 'crossing the road' are actions? Why not say that 'going to work (in which are combined many other 'actions') is an action'; or that 'fitting in with one's friends and contemporaries, of which climbing mountains or playing tiddly-winks may be a part', is an action? The whole life may be taken as an action, if you so choose, with all the fragments of behaviour that compose it being seen as contributing statistically to the outcome. Indeed, taken as a whole, human lives, all of them, leave no doubt that the tendency is for humans to act so as to secure their reproduction. In this trajectory we see individuals 'doing things' (arbitrarily defined) that obviously contribute to that, such as collecting resources, acquiring a mate, giving birth, but also many others that do not seem so, such as brushing crumbs off the breakfast table, choosing to wear grey not black socks, and going on holiday to Greece to study antiquities. The first of these cases no more proves the case than the latter disproves it. They are incidental to the principal observation, which is based on the lifetime tendencies of millions of individuals.
But we owe the critic some sort of comment on the fine texture of behaviour, which is often puzzling. The answer is four-fold:
1. We cannot see, always, the links that make an 'action' probabalistically conducive to securing reproduction.
2. The 'action' may well need to be understood as part of a larger constellation of behaviours that may constitute such an 'action'.
3. The propensity to behave has evolved against a background of contextual facts that will never be repeated, so we should not surprised if the system produces actions that are imprecisely focused on the securing of reproduction; some divagation is to be expected.
4. Evolution is itself a statistical phenomenon, and adaption will also be so.