The general consideration of objective theories of reasons is based on the relation to the sense that facts about the objects of desires gives us object-given reasons. However, the commitment to the view that all practical reasons are like this is a theoretical one, which is what is termed "objectivism" about reasons. This theory can be contrasted with the "subjectivist" one that regards appeal to desires or aims as involving facts about ourselves as being the entire truth of reasons. The difference is particularly one concerning what the "force" of reasons derives from. In articulating this distinction, Parfit takes Christine Korsgaard to be committed to a "subjectivist" view of reasons due to her general assumption that value is not contained in objects but in our relation to objects.
If part of the difference between subjectivist and objectivist views of reasons consists in the different conception of where the force of reasons resides, another part of the difference concerns which desires we have the stronger reason to fulfil. If one is committed to an objectivist view of reasons then the strength of commitment to an end depends on how worthwhile the end's achievement would be in relation to what is worth adopting as an end. By contrast, the subjectivist about ends may emphasise the strength of our preferences as what is decisive (I should stress here "may" since it is not clear that if Korsgaard is to be regarded as a "subjectivist" that this is what she would say).
Parfit is himself committed to an objectivist theory of reasons. However, he does nonetheless formulate a distinction between two types of response to reasons. On the one hand, there are reasons to want something to happen, and on the other there are reasons to try to make something happen. The two are distinct in the sense that reasons to try to make something happen appeal to voluntary considerations whilst reasons to want something to happen do not, in themselves, necessarily lead us to actually want it to happen (as we cannot easily make ourselves have desires).
Similarly, epistemic reasons to have particular beliefs can be provided by facts that are related to the truth of some belief but we cannot voluntarily respond to these reasons by, for example, just forcing ourselves to believe something to be true or entirely rejecting the grounds for something being true if we really see that they are grounds. Some of the basis of this point is not difficult to see since if part of what makes beliefs reliable is that they reflect circumstances in the world then the point is we cannot make the world be a certain way and if we could make our beliefs be any which way there would be no way of determining how any beliefs could be reliable.
But these points are not merely epistemic since desires are similarly not easily adaptable. We don't, as a matter of course, choose to prefer things (cheese rather than chocolate). This point leads Parfit into a long discussion of what he terms "state-given reasons" but that will have to be left to another posting given the complexity of his discussion of these types of "reason".
Separately from the discussion of "state-given reasons", however, is the account of hedonic reasons which are taken by Parfit not to be really reasons at all. Included here are "instinctive" responses such as are provided in certain kinds of organic conditions in which we do not respond to reasons (as when we prefer red to blue for example). Parfit also assumes that aesthetic states do not have reasons though this claim seems to partly depend on viewing such states purely sensationally. However, more importantly than this point is his distinction between hedonic and meta-hedonic desires where the latter concern our attitudes towards sensations. A hedonic state is one in which something is undergone whereas the meta-hedonic state is the reaction expressed to this state undergone.
When responding to Korsgaard's description of the "incorporation thesis" Parfit views Korsgaard as committed to the conception that we "create" value but Parfit views this as only true of meta-hedonic states and not of hedonic states. The reason to emphasise this is that present suffering is itself a bad for us given our experience of it not due to our present wish to be rid of it. "Since our meta-hedonic desires do not make their objects good or bad, the examples of pleasure and pain do not decisively, or even, I believe, strongly support the view that our other desires have such value-creating power. Though it is good to have sensations that we like, nothing is good merely because we want this thing." (55)
These points certainly seem right and suggest that Korsgaard's statement of the "incorporation thesis" has, in these cases, been sloppy but they do not cut against any possible statement of it. What they do show is that the desire itself is not made good if it relates only to a sensation merely by wanting to have it. But Korsgaard should never have exposed herself to that argument, which doesn't touch on what she is really meaning to get at in relation to the general claim about how it can be that there is "construction" of "value".
The real point of the distinction about hedonic and meta-hedonic states, however, consists in the fact that if the former are not responses to reasons, the latter are. We have reasons to adopt these meta-hedonic attitudes but not to respond to sensations in the way we do.