Unobtainable goals are common. Many life goals, such as being the best artist (ever) and writing the best novel (ever), are unobtainable. Epistemically the goal of being certain of the truth is also unobtainable. One reaction to the existence of such unobtainable goals might be to try to get rid of them. Certainly we can’t accomplish an unobtainable goal (except in delusions), and so they might appear to lead to frustration at the exclusion of satisfaction. But a closer look reveals that unobtainable goals might be very desirable indeed, perhaps even necessary.
Consider someone who has only obtainable goals. This means that after some period of time they will have accomplished all of them. At that point then what should they do? If they were robots perhaps they would just sit down and starve to death. But of course people are not robots, in the absence of goals we acquire new ones. And this is in fact problematic. Let us suppose that a particular person will always be guided by three goals, and that when they accomplish one of them a new goal will present itself to them. These new goals we can assume are fairly random, relating to whatever the person was interested in when their previous goal was accomplished. Thus such a person will be pulled in one direction and then the next, in a kind of random walk over the space of all possible goals. This is to their disadvantage, because it is likely that they will not have suitable skills with which to accomplish their new goals, meaning that it will take more time and effort to accomplish those goals then someone with the proper skills would require. A better strategy in terms of accomplishing goals is to have goals that are related to each other, so that the work that goes into accomplishing one of them aids in accomplishing the others. Of course in reality new goals aren’t acquired totally randomly, social pressures certainly play a big role. But that does not necessarily improve matters, because often society can push for goals that are counterproductive or in other ways detrimental to the individual. In any case, were we to be in a position where we could choose to always have a consistent kind of goals versus a relatively random selection, consistent goals would obviously be preferable.
One way to have a consistent set of goals is to pursue unachievable goals, since such overriding goals increase the probability of acquiring achievable goals that are similar to them. For example, the unobtainable goal of writing the best novel leads to many obtainable goals regarding writing various things in preparation for writing the best novel; the unobtainable goal of being certain of the truth leads to many obtainable goals involving gathering information. Thus while pursing an unobtainable goal we would achieve many smaller goals along the way, many of which build on each other. But unobtainable goals aren’t the only way to enjoy these benefits, at least in principle. Having goals centered around continual improvement, such as continually improving as an artist, writing better novels, and getting closer to the truth, might very well have basically the same effect. But continual improvement has some hidden defects. The major difficulty is that the idea of improvement by itself doesn’t include the idea of an endpoint or target. Without such a target in mind the person who aims at continual improvement will likely simply convince themselves that whatever they happen to be doing is improvement (because that is the easiest way to satisfy the goal). If we have no fixed standards then acquiring a worse writing style may very well be thought of as satisfying the goal of continually improving as a writer, so long as it is accompanied by a change of attitude which values what was previously thought as bad writing over good writing. Thus, like the person who picks up goals at random, continual improvement can wander all over the map, “wasting” the past effort put into improvement, because the skills developed towards what was previously thought of as good may no longer be useful when the standard for good changes. Now obviously this can be “fixed” by adding in some endpoint which all improvement is improvement towards. But this is simply to have an unobtainable goal.
So unobtainable goals may be better than they first look. But what about the obvious objection, that an unobtainable goal can never be reached, by definition, and thus that it cannot bring us satisfaction? All our satisfaction would then come from obtainable goals, and so while an unobtainable goal or two might be useful to guide which obtainable goals we put ourselves towards it would make sense then to deprioritize it. However, this reasoning turns on a misunderstanding of satisfaction. It is true that we get satisfaction from reaching our goals, but there is also satisfaction to be had in our efforts to achieve them. Thus it is possible to get more satisfaction from an unobtainable goal than an obtainable one. We can only work so much at the obtainable goal, and so we can only get a fixed amount of satisfaction, coming from both the work we put towards achieving it and the accomplishment itself. In contrast we can always work more towards the unobtainable goal, and thus there is no fixed upper bound to the satisfaction we can derive from it. And so, contrary to the earlier analysis, it may make sense then to deprioritize our obtainable goals and to put more emphasis on the unobtainable ones.