An important question to ask is what I mean when I speak of "higher" desires. It would be totally undue to simply assume that there are such things, particularly in a discussion with an intelligent and rigorous determinist. More commonly, should we not assume that the perception of "grades" of value or joy are themselves mere anthropomorphic hypostasizations of immediate pleasures into/onto mediating objects? Hierarchies of more/less successful disavowals and sublimations of our graspings for food, sex, and power? Aren't all 'hierarchies' of goodness in the world ultimately an illusion crafted by a brain which has tricked itself into thinking that its levels of meaning correspond to any objective reality?
Let's pull a "Husserl" here and forget about the objective world for a moment. We know as a basic fact that, objectively substantial or not, each of us prefers some goods to other goods; and more often than not these preferences have wide similarities across large populations. Consider the effectiveness of beer advertisements: very often, a beer ad will place the subject male in a dillemma between a beer and a woman. A beer that a man chooses over a woman must be a VERY good beer. Of course, such ads delight with their irony (because the ad-makers know that the better-than-a-woman beer will never fail to surprise). Even if a beer ad has the man choose the woman (to distinguish itself from other, more predictable beer ads), the dillemma itself is sufficient to inflate the value of the beer.
The point is that we take for granted that a beer is inferior to a woman. Why? A Fruedian would speak of the "overvaluation" that occurs when we take the pleasure of genital climax and associate it, in various and diffuse ways, with the proximate signs the woman carries about herself--the rest of her body, her fashion tastes, even her politics.
Yet there is something disingenuous about this idea. The critique against it takes the same shape as the critique against the reduction of all desires to "happiness". Consider the following thought experiment: an amatuer writer has two options placed before him by a brilliant scientist: he can either (1) take a pill that will make him believe he is a brilliant and well-loved writer, no matter what, in spite of the fact that he is not (thereby granting him happiness), or (2) not take the pill, and work towards becoming a brilliant and well-loved writer. The results of such an experiment, idiosynchratic exceptions notwithstanding, yield that the first option is revolting. It is not merely the happiness of being an accomplished writer the writer wants; he truly wants to write and write well.
The point is that our desires are what they are; they are not something else in disguise. When a man wants the company of a woman, it is not just a sublimation of genital lust to the level of a relationship--it is a genuine desire for companionship.
Now, real, authentic comanionship, few would disagree, is almost universally more desired than a beer--unless the subject in question is a far-gone acoholic, and such people cannot be representative.
But why is companionship valued more? I would suggest two virtually indisputable reasons. First, the goods that result from 'companionship' are experienced on a level on which they can coexist with deprivation on several other levels, and yet still be preferred to the converse scenario in which companionship is lacking but the competing goods are met. This is what is first meant by its being "higher."
The second reason is that the goods that result from 'comanionship' are experienced as being more enduring, and experienced in time. Companions are found and established across time, and they are received as embodiments of history, not only of past memory and present attention, but future expectation as well. Other goods--such as the buzz from a beer--are strictly "now" goods; the memory of a past buzz has no real desireable effect when there is none present now. This is to say nothing of the desolation resulting from a genital sexual release removed totally from the context of companionship.
Standing back now and surveying the wide expanse of all the sorts of things people may decide they want, we can see that the grades of desires are, in fact, real. They are real, not only in the drastically differing subjective experiences of their benefits, but in the immanent demand that the benefits not be experienced apart from their sources. In other words, the amateur writer does not want to merely 'believe' that he is great; he wants to become great; and neither is he willing to substitute more easily-gained pleasures of sex or food for his goal of becoming a great writer.
This brings up another important point: 'higher' desires are naturally more difficult to attain with any certainty. A relationship requires more devotion, more skill, and more perseverance than the enjoyment of a hamburger. It does not necessarily follow that whichever desires demand the most effort are thereby higher; a drug addict may have to brave fearsome obstacles over long periods of time in order to get his fix. However, the converse is never true; there is no art, no friendship, and no romance that is had cheaply.