It is an interesting fact about many of our most important choices, such as the choice of what kind of education to pursue, whether and whom to marry, and whether to have children – for short, life choices – that they transform us in ways we can’t fully anticipate, so that the person who lives with the consequences of the choice won’t be quite the same as the person who makes the choice. Recently, L.A. Paul has argued in a stimulating paper that the existence of such transformative experiences causes serious trouble for rational decision-making. I’ll grant here that her argument is more or less successful to the extent that the phenomenal quality of our experiences is central to the value of a choice or preference among options.
Alas, I think that precisely when it comes to the life choices that are associated with transformative experiences, phenomenal quality is relatively insignificant. After all, there is more to prudential value than positive experiences. Consequently, we make our life choices not on grounds of what we expect it to be like to lead a certain kind of life, but rather on the grounds of how it will shape our relationships, roles, and identities, in short, the story of our life, and have good reason to do so. As I prefer to put it, life choices are not experience-regarding, but primarily story-regarding, and should be such. Insofar as we can reliably compare the value of or form preferences between relationships and activities, and estimate their likelihood given our choices, we can after all make approximately rational self-interested life choices. Even if Mary can't know what it will be like for her to have a child, she can ask herself whether having a child would best promote connecting her to something larger than herself or best express her love and commitment to her partner, and use the answers as a basis for rational choice.
Here is an argument that is suggested by Paul’s paper:
- Rational choice is a matter of (approximately) maximizing expected utility, which is calculated on the basis of expected value of outcomes and their probability given possible actions.
- When it comes to prudentially rational choice, what centrally determines the value of an outcome is what it would be like for the agent if an outcome were realized.
- “When a decision involves an outcome that is epistemically transformative for the decision-maker, she cannot rationally assign a value to the outcome until she has experienced the outcome.” (p. 12)
- “[S]ince having one’s own child is unlike any other human experience, before [a person] has had the experience of seeing and touching her newborn child, not only does she not know what it is like to have her child, she cannot know. … [H]aving a child is epistemically transformative.” (p. 8)
- So, it is not possible to rationally assign an expected value to the outcome of having or not having a child. (3, 4)
- Hence, it is not possible to make a prudentially rational choice about whether or not to have a child, since a central piece of information is missing. (1, 2, 5)
I say this argument is suggested by Paul’s paper. She doesn’t quite make it, since she is very explicit that the conclusion is that “you cannot rationally choose to have a child based on what you think it will be like to have a child” (p. 20) – that is, she doesn’t say there is no way to make a rational choice. But it is clear that she thinks that making the decision on a phenomenal basis (what it is like to have a child) is a) the ordinary, common sense way of making it, and b) the rationally preferable way of making it, as long as you don’t take into account moral or, say, environmental considerations, or instrumental reasons like the need for an heir or having more hands at the farm, because the value of the phenomenal outcome is likely to “swamp” the value of non-phenomenal outcomes. (For detail of her actual argument, see the appendix below.) So the popular press is not entirely wrong in claiming she argues "there is no rational way to decide to have children—or not to have them".
Alas, this pessimistic conclusion rests on premise 2, which is false, especially when it comes to life choices. What motivates the premise is either the thought that what is objectively valuable for us are above all or exclusively positive experiences, or that what we subjectively care about are above all or exclusively positive experiences. Since neither of these claims is true, our at least partial ignorance of what outcomes that involve transformative experiences are like for us doesn’t threaten the possibility of making approximately rational self-interested life choices (which is not to say that it is easy to make them).
The standard way of arguing against the notion that prudential choices are or should be exclusively experience-regarding, as I will say, is by appeal to thought experiments like Nozick’s Experience Machine. As I have emphasized, the relevant comparison is between a qualitatively identical outcome in the Experience Machine and in reality. It is illustrative, I believe, that it is precisely when it comes to outcomes of life choices that the intuition against plugging in is strongest. Would I be indifferent between actually going to see Gravity or eating a nice meal, and having the perfect illusion of doing so? Yes. Would I be indifferent between actually having and interacting with my two children or doing my job, and having the perfect illusion thereof? Certainly not. What matters to me about being with my children and doing my job is neither exclusively nor centrally the quality of associated experiences.
So how do we, and should we, go about making life choices? On Paul’s picture, recall, the way people ordinarily decide whether to have a child is asking themselves how they would feel if they had one, or more broadly what it would be like for them if they had a child. I am, first of all, highly skeptical of whether the empirical claim is true. It doesn’t even remotely match my own experience of making the choice to have children, nor does it reflect what conversations with friends suggest.
But the important thing is what actually makes a difference to the prudential value of the outcome of either having or not having children. Here are some of the non-experiential things you choose when you decide to have a child. You choose to be a father or mother, and potentially a grandfather or grandmother, a link in the chain of generations from the unknown past to the uncertain future. You choose to take on certain obligations and responsibilities towards the child, whatever he or she will be like. You choose to bring into the world and nurture a creature who is for a long time dependent on you. You choose to bind yourself to your partner in a new way: you will always have something distinctively in common, someone who shares some of the unique traits of both of you, both by biological transmission and (typically) by way of the example you give and the way you bring her up. This may be an expression of a deep kind of love. In short, you choose identities, roles, commitments, and relationships. (Including being a childless woman, for example.) These are outcomes that you can know in advance, regardless of how the actual experience transforms or would transform you.
And these outcomes have value for people independently of the associated experiences. Why? That’s not a simple question. I believe that it has to do with the fact that we are not just subjects of experience but also active agents shaping our lives, so that certain exercises of our agency are good for us independently of the experiences they involve. Subjectively, some people simply have preferences for having certain roles, relationships, or activities, regardless of or in addition to what they are like for them. They may feel their lives are wasted or rudderless without them, or take pride in doing them well. When it comes to objective value, there are many competing non-experiential theories. A perfectionist might say that the roles and relationships involved in having children are good for you to the extent that they enable the development and exercise of your essential human capacities. (In that case, you are better of not having children when doing so would hinder such development and exercise.) An objective list theorist would ask which kind of life is likely to involve higher levels of achievement, personal relationships, and self-respect (Fletcher 2013), or whatever is on the list.
My own view is that the non-experiential value of being a parent and a co-parent depends on whether it makes your life more (objectively) meaningful, which in turn is, roughly speaking, a matter of whether it continues your story in the direction of making a lasting positive difference in the world in a way that builds on your past efforts and makes use of your distinctive abilities. On this view, whether it is prudentially rational for you to have children depends centrally on what else you could be doing with respect to realizing some objective value (whether moral, alethic, or aesthetic), as well as what you have done up to now and are able to do. To be sure, it also depends in part on whether children would make you happy, but that is not a paramount value that would “swamp” others, but a relatively minor consideration, unless there is some good reason to think it would make you either extremely happy or unhappy.
So I prefer to say that the self-interested choice of whether to have a child, or go to college, for another example, should rationally be centrally story-regarding. Insofar as I’m abstracting from moral considerations, I should ask myself whether the story of my life is likely to be better if I make one choice rather than the other. This is by no means an easy calculation – nobody says that making even approximately rational life choices is simple. But if it is difficult, it is not because you can’t predict your experiences, but because it’s hard to predict exactly how your relationships and projects are transformed.
So: I reject the conclusion of the argument suggested by Paul’s paper, since what it is like to have a child is not a central prudential consideration when it comes to deciding to have a child. Contrary to Paul, I don’t believe common sense makes this error either, though the issue would need to be settled by proper empirical research. The result is a kind of a dilemma for Paul’s thesis about the implications of transformative experience. On the one hand, when it comes to ordinary choices for which the phenomenal quality of outcomes is paramount, we are in a fairly good position to know what the phenomenal quality will be (for example, I’ll get better food at Café Paradiso than at McDonald’s down the road). On the other hand, when it comes to life choices, the phenomenal quality of outcomes is indeed hard to tell, but it is not among the most important considerations either. We don’t have to set aside our preferences between lives to make a rational choice, since (and as long as) those preferences are above all for relationships, identitities, and activities rather than experiences.
Is Paul committed to the pessimistic conclusion of the argument I claim is suggested by her paper? Perhaps not, but she comes close when she emphasizes the importance of phenomenal quality in making personal choices, and then denies it can be the rational basis when it comes to transformative experiences. For example, she says that “it seems appropriate to frame the decision [of whether to have children] in terms of making a personal choice, one that carefully weighs the value of one’s future experiences” (p. 2), and even more strongly that
not only is the phenomenal outcome what it’s like to have your own child a relevant outcome of your choice, it’s an outcome whose value might swamp the other outcomes. In other words, even if other outcomes are relevant, the value of the phenomenal outcome, when it occurs, might be so positive or so negative that none of the values of the other relevant outcomes matter. (p. 17)
Paul considers the possibility that other outcomes might outweigh the phenomenal ones, but dismisses it with the claim that “What is much more likely is that the value of what it is like to have the child will swamp the other outcomes.” (p. 17n28) So it is clear that she doesn’t just think that one potential basis for making a prudentially rational choice is missing when it comes to transformative experiences, but that the very possibility of such choice is severely threatened, because phenomenal quality is of paramount importance. To be sure, Paul does grant that someone can decide rationally, as long as she “ignore[s] what she personally thinks about whether she wants to have a child” (p. 20), but given what she says about the importance of what it’s like to have a child, I read this as gesturing towards making the choice on the basis of something other than one’s personal preferences or intrinsic interests, such as the kind of considerations that motivated some pre-modern parents, or moral reasons. At the very least, the claim about the phenomenal value of an outcome swamping non-phenomenal values is in severe tension with the claim that it is possible to make a rational choice on the basis of non-phenomenal values.
(Thanks to David Killoren for comments on a draft of this blog.)