The second factor arguing against this possibility
is empirical. We now have compelling neurobiological
evidence that subjective values of some kind are
represented in the brains of monkeys ( Dorris and
Glimcher, 2004 ; Sugrue et al ., 2004 , Louie and Glimcher,
2006 ; Padoa-Schioppa and Assad, 2008). In those
experiments and others like them it has been demonstrated
that the subjective values of individual
options, and not choice probabilities, are represented
by neuronal firing rates.
For these two reasons we can consider choiceprobability-only
based systems as empirically and
theoretically falsified. Of course, it may well be that
groups of neurons (or the local ensemble connections
of those neurons) do explicitly represent choice probabilities.
Some evidence suggests that this may be the
case in posterior parietal cortex, but we now have sufficient
evidence to conclude that the representation
of subjective values – or something much like them –
occurs within the central nervous systems of primates.