Thought Experiments on Desire

Continuing my commentaries on In Praise of Desire by Nomy Arpaly and Timothy Schroeder, their section on what desires are and what desires are not is filled with "thought experiments."

They refer to Galen Strawson's weather watcher thought experiment, where we are to imagine a group of plants that have a desire for sunshine and light rain but an aversion to heavy rain and dark clouds. The point of this experiment is to show that it is possible to have desires/aversions without having a disposition to act.

I have written quite a bit about another thought experiment - Radioman - who has a disposition to turn on radios even though he does not value doing so nor does he value any of the consequences for turning on radios. This is meant to discredit the idea that we can reduce desires to dispositions to act since Radioman has a disposition to act but no desire.

There are examples of sports fans who want their team to win even though they do nothing to cause their team to win, and alien races who care about justice or truth even though they have no capacity to feel emotions.

I dislike these types of thought experiments. Language is an invention. It exists to serve a function in our everyday lives. We do not know what to say about these elaborate cases because we have not had a need to invent a language for talking about them.

When a philosopher asks us to imagine plants with a desire for sunshine and light rain even though they cannot move, I am wondering . . . are they asking us how we would say that they had a desire using our current language, or whether we are trying to imagine a possible language that uses the term "desire" to talk about these beings. This question of whether we can imagine such a case never actually specifies an option. I worry that "imagining a universe in which the term 'desire' is used in the way described" is often getting in the way of "imagine us applying our actual term 'desire' in this bizarre set of circumstances for which we have never needed a language."

This is a relatively recent frustration that I have been giving thought to. People familiar with my writings will note that I have sometimes used these types of stories. I am now worried that it is easy to misuse them.

In using them - the physicist who asks students to imagine frictionless surfaces and massless strings can certainly use this type of thought experiment for pedagogical purposes. That's legitimate. That is how I have attempted to use my own "thought experiment" imagining a single person Alph with a single desire - to gather stones (or "that I gather stones" as held by Alph). So, that is legitimate.

Sometimes, a thought experiment yields clear results. I have used one involving a species that evolved to have a genetically-based fear of spiders. They get transported to another planet that does not have spiders. Eventually, they forget all about spiders - but the biological basis for their aversion remains. Do they still have an aversion to spiders even though they do not know what a spider is?

I argue that the answer is "yes" grounded on the fact that if we were to drop a large spider in the middle of a group of these beings, they would scream and run away. Which means, you do not have to have a concept of "spider" to have an aversion to spiders.

However, I would have to say that if anything more than a very small minority objects to the use of a term in a particular way, we have stepped out of the realm of "what the term means in English" to "What should we use the term to mean in English" - which meaning should we adopt as opposed to which meaning are we actually using.

It is a legitimate response to a "thought experiment" that gives some counter-intuitive result to say, "Okay, fine. Let's change the definition of the term so that it works this way rather than that." One can defend this use of the term on the grounds of ease of use or ability to focus attention on the distinctions that actually matter.

In contrast, a reference to "habit" as a counter-example to the thesis that desires are dispositions to act (since habits are dispositions to act but are not desires) is not a thought experiment. It applies to something that is real and known. Our invented terms already consider this possibility, and we can ask reasonable questions about what, in our language, we seem to have already decided.

I have also argued against Timothy Schroeder's theory of desire that we need to allow for the possibility that an agent can have a desire that p (that it is important to Agent that p be the case), without this desire having the power to influence the strength of other desires through the reward process. Here, I am not using a some strange science-fiction thought experiment to suggest a possibility we can only imagine. I am suggesting the possibility of something we might actually find in our empirical research - a reason for an agent to realize p that does not have the power to alter the strength of other desires and aversions (that does not have the power to serve as rewards or punishments).

In short, I tend to think that thought experiments have a more restricted use than many philosophers seem to think. As a result, I do not feel a strong urge to respond to arguments that, as far as I can determine, represent the weak form of thought experiment that is not particularly useful.