I'm reading I. Persson's (great name) The retreat of reason while I mull over hedonic calculus-type things again. It's not an easy read but it's very rich and will take me a while to digest properly. But here's a taster (from p.223):
It is of survival value that thoughts directed at one's present hedonic sensations should motivate one and tend to monopolize one's attention, just as these sensations themselves do. Similarly, it is also of survival value that impressions of some activities in which one currently is engaged should attract one's attention; otherwise it would be much harder for immature individuals to expand their behavioural repertoire beyond instinctive patterns. However, it also enhances the prospect of survival if a power of mentally representing states of affairs beyond ones that are sensibly present—in particular possible future states—develops. But the more this power of envisaging possible futures is developed, the more of a drawback the P-bias will be, for it will prevent this representational power from having full impact on motivation. Thus, akrasia is the upshot of the conflict between two traits that each individually is of survival value. In better adjusted beings than actual human ones, the P-bias would be quite marked during the earlier stages of ontogenetic development, but would gradually loosen its grip as the capacity of mentally representing sensibly absent states of affairs expands and the power to estimate the probability of their materialization is refined.
P-bias is a bias towards the perceived (which P. helpfully distinguishes from N-bias, a bias to the temporally near, see pp.206-7)... Not sure about the grand diagnosis of akrasia, but it's certainly a suggestion worth taking seriously. Also interesting is that Aristotle does not figure at all in the index. Plato figures only twice (pp. 176 and 177 for his views on parts of the soul).