In an email, Dr. Chris Heathwood of the University of Colorado - Boulder - suggested a potential problem with the desire fulfillment theory of value that I recommend here.
The specific context of this discussion was the "desire-satisfaction theory of well-being." In an article, "Desire Fulfillment Theory" (in G. Fletcher (ed.) The Routledge Handbook of the Philosophy of Well-Being, Routledge, 2016), Heathwood brought up Derek Parfit's objection that, if the desire-satisfaction theory were true, one person can benefit another by giving that person an addiction to a readily available drug.
I am about to make your life go better. I shall inject you with an addictive drug. From now on, you will wake each morning with an extremely strong desire to have another injection of this drug. … This is no cause for concern, since I shall give you ample supplies of this drug. Every morning, you will be able at once to fulfill this desire.
Against this, I employed Bernard Williams' theory of reasons - "A has a reason to φ iff A has some desire the satisfaction of which will be served by his φ-ing" - and argued that an agent would have a reason to choose to become an addict only if it served an existing desire.
Though claiming that this would provide an answer to Parfit, Heathwood suggested a couple of potential counter-examples to this theory. In one of them, he wrote:
Consider a toddler. He currently has no desire to be able to read. But we can be confident that in the future he will desire to be able read (if the future comes and he can’t read, while all of his classmates can, he will wish he could read; if the future comes and he can read, he will be glad that he can read). The theory that you are suggesting seems to imply that we would not benefit the toddler if we now do things to see to it that he will be able to read in the future. Am I right that your theory has this implication?
Dr. Heathwood is correct. However, I do not see this as a problem with the theory.
My answer to Heathwood is as follows:
In the case of the toddler, you wrote, "we can be confident that in the future he will desire to be able read."
In order to respond to your objection, I would like to shake this confidence a little and see what falls out.
Allegedly, in 210 BC, Qin Shi (the first emperor of the Qin dynasty in China) ordered that a number of scholars be rounded up and buried alive. This provides a case in which we have reason to ask whether the parents of those scholars provided them with a benefit by teaching them to read. It seems reasonable to say that they did not.
Similarly, the Khmer Rouge regime in Cambodia (Kampuchea) under Pol Pot engaged in a campaign of killing intellectuals. Reading is a primary cause of myopia (near-sightedness), and the Khmer Rouge used near-sightedness as a sign that one was a scholar, and would put those people to death. Here, too, we have reason to deny that these people obtained a benefit when their parents taught them to read.
We can add to this list any case in which the child dies before the child can obtain any benefit from learning to read. If one knows that a child only has a few months to live, then teaching the child to read - at least in the sense described above - is not a benefit. There may be some benefit in sitting with the child and reading, but that benefit comes from the current desires for companionship - which may also be fulfilled by playing a game or holding the child while watching her favorite movie.
This suggests whether teaching a toddler to read counts as a benefit depends entirely upon whether the toddler grows up to have desires that are fulfilled by reading. If the future desire fulfillment does not take place, then the present benefit does not exist.
I can see two different ways of describing this situation.
(a) We could say that the activity at time T provides a benefit at time T contingent upon its being put to use at a later time T + n to bring about the fulfillment of future desires. If no future desire fulfillment takes place, then the claim of "benefit" is revoked.
(b) Or we could say that the activity at time T does not provide the person with a benefit at time T; instead, the benefit comes at time T + n when the effects of that activity are put to use in the actual fulfillment of desires.
Of these two, (b) is technically more accurate. I find it odd to say that a benefit appears and is then lost – in fact, never existed – if not used.
However, in a community where an activity is almost certain to provide a benefit, is close enough to the truth for matters of public and social policy. We can loosely speak of such things as public education, immunizations, and the development of good eating habits as providing a benefit to the child in anticipation of the fact that it almost certainly will lead to the future fulfillment of desires. However, we can take the fact that this attribution is revoked if no future desire-fulfillment comes of it as reason to believe that the benefit is found in the future desire fulfillment, not the direct effects of the current activity.
In writing this response, I imagined another related example that would be much like the one Heathwood provided here which would be closer to the case of Parfit's addiction. In this alternative version one does not give the child an ability to read, but one gives the child a desire to read. Is this desire - to be fulfilled at a later date - any different from Parfit giving a person an addiction to a readily available drug?
I will examine this alternative in the next post.