Examined Life

Disclaimer. I regret to report that I no longer espouse the argument hereunder set forth, if one could call it that. I wrote this long rant at night after a drunk conversation with a friend. I re-read it the day after, thought it through, and concluded that the argument does not work structurally, and the drug-children analogy does not hold. The Frankfurtian distinction, in fact, exerts little or no explanatory power applied to this specific case, so I may need to discard that as a conceptual tool. I have two reasons for nevertheless posting this in its original state : (i) because I want to revisit this topic, and when I do, I want to track how my view has changed; and (ii) because I think this piece captures, or vaguely hints at, a couple of fundamental intuitions which I want to salvage and develop. These are: (a) that I believe it is epistemically rational to exclude post-childbirth testimonials (which is not contrary to what Paul claims); and (b) more importantly, that factoring in post-childbirth satisfaction may undermine rational autonomy. These will, hopefully, be examined in a follow-up essay in due course.

What is the difference between having a child and taking drugs? I know what you’re thinking: what the hell kind of a question is that? You might even think the question should be not how they differ, but whether they have anything in common at all. Nevertheless, I intend to argue that having a child is, in one important respect, identical to — or at least structurally analogous to — taking drugs.

I have no particular desire to have children. When the topic is brought up at dinner parties, I hear this same response every time: “I used to think exactly like you, but once I had a child, I couldn’t imagine life without one. I could kick my former self for being so foolish.” I believe this completely. Not only do I trust that these people are entirely sincere, I am fairly confident that if I were to have a child, I would feel exactly the same way. And yet, I still don’t want one. Let me explain why.

In her paper “What You Can’t Expect When You’re Expecting,” philosopher L. A. Paul argues that it is impossible to make a rational decision in prior to and about whether to have children.¹ In the standard model of rational decision-making, a rational choice is, roughly speaking, a matter of calculating the expected value of available alternatives and selecting the one with the higher expected value. This squares with our everyday intuitions. Cheryl Strayed, for instance, in an online column, advises a man who is uncertain about fatherhood to ‘imagine life with and without children, and then choose whichever feels more compelling’, which is basically telling him to choose the alternative with the highest expected value.² Paul’s claim is that this approach simply doesn’t work when it comes to the decision to have children. Having a child is what she calls an epistemically transformative experience. Having a child changes you so fundamentally that the person you become after having a child is effectively a different person from the one who made the decision. Since the person who has not yet had a child cannot genuinely anticipate the expected value of childbirth for the person who has undergone this epistemic transformation — just as it is impossible for someone who has never tasted a banana to imagine, let alone know, what it tastes like, it is impossible — it is impossible, according to Paul, to make a genuinely rational decision about parenthood in advance. Of course, one can decide to have or not have children on social, religious, or moral grounds, but a rational decision in the traditional sense is, she argues, out of reach.

My own thinking begins from the same observation as Paul’s, but the structure of my argument and the conclusion I end up with are very different. I agree with Paul that having a child is an epistemically transformative event. But what I take that “epistemic transformation” to mean is slightly different, and I would put it this way: having a child effectively undermines the ability for reflective endorsement based on second-order desires. And the mechanism, I should add, is biological.

Let me first explain what I mean by “second-order desire.” In his paper “Freedom of the Will and the Concept of a Person,” philosopher Harry Frankfurt distinguishes between first-order and second-order desires.³ First-order desires are the most immediate desires we experience: wanting a bowl of ice cream or a glass of whisky. A second-order desire is a meta-desire — a desire about a desire. If you want ice cream but decide to resist because you’re trying to lose weight, and you order a chicken salad instead, you have, based on your second-order desire, reflectively endorsed a particular first-order desire.

All else being equal, human beings want to maximize pleasure and minimize pain. We also know that drugs deliver a level of pleasure that is categorically different from anything else we have experienced. By syllogism, it should follow, then, that everyone should do drugs — and yet only a handful of people in fact do. The picture, though, becomes very different when we limit our observation to those who have already started, in that very few people are able to stop. Frankfurt’s framework of first- and second-order desires is useful in understanding this phenomenon. We all have a first-order desire to maximize pleasure, but we also have a second-order desire not to become enslaved to pleasure, and it is through this second-order desire that we reflectively endorse the decision not to touch drugs. The problem arises for those who have already started. Addiction, as a biochemical phenomenon, neutralizes the capacity to autonomously select first-order desires based on second-order reflection. The agent is trapped in the biochemically-induced, unending pursuit of pleasure.

It is in this respect, I want to argue, that having a child is structurally analogous to taking drugs. I am not saying that the two are qualitatively or morally similar — the former is a profound blessing, the latter something I would rather not contemplate. But both make reflective second-order reasoning difficult, and in both cases the mechanism is biological. This is where the similarity lies. Just as taking drugs neutralizes the second-order desire to evaluate whether one should desire drugs at all, having a child neutralizes the second-order desire to evaluate whether one should desire (or should have desired, in retrospect) parenthood in the first place.

Again, I do not have a strong desire to have children. And yet, when I’m with my nephew, I feel like I could throw myself in front of a truck to save him. I mean, he isn’t even my child — imagine how I’d feel if it was my own child. I can just picture myself in pure delight, and looking back and cringing at the version of me writing this piece, ranting about epistemic transformation, second-order desires, and whatnot. Because having a child is an epistemically transformative experience, and that experience fundamentally reorders one’s value hierarchy in a way that creates a structural bias toward endorsing the first-order desire for parenthood. (Whether this is an appropriate characterization of what’s happening is something I need to think through further. For now, this is more a sketch than a proper argument.)

There is biological evidence for this as well. A substantial body of research points to the significant psychological and neurobiological changes that new parents undergo. In other words, once we have children, we are biologically hard-wired to assign the highest possible priority to parenthood.

At this point it is worth being clear about the scope of the argument — such as it is. The point is not that having children is better than not having them, nor is it the reverse. What I want to show is that the so-called “testimony of the converted father” — “I used to feel just like you, but once I had a child, I couldn’t understand why I ever thought otherwise” — carries no rational persuasive force. The reason is that from the moment one becomes a parent, one is already subject to a structural bias that endorses a particular first-order desire. Of course, one might object that just as the post-parenthood self cannot be epistemically neutral, neither can the pre-parenthood self, since they have never had the experience of having a child. But if we consider neutrality not as a binary but as a matter of degree, it is reasonable to hold that the self who has not yet undergone the biological rewiring of parenthood occupies a relatively more epistemically neutral position than the self who already has. The problem with this objection becomes obvious when you apply it to drugs: “You’re making a decision about whether to have drugs without ever having tried them, so you’re not really neutral, either.”

So the next time someone at a dinner party tells me to “just have one and you’ll understand”, I think I’ll show them this piece and ask: “have you ever tried drugs?”

1. Paul, L. A. (2015). What you can’t expect when you’re expecting. Res Philosophica, 92(2), 149–170. ↩︎
2. Strayed, C. (2011, April 21). Dear Sugar, the Rumpus advice column #71: The ghost ship that didn’t carry us. The Rumpus. https://therumpus.net/2011/04/21/dear-sugar-the-rumpus-advice-column-71-the-ghost-ship-that-didnt-carry-us/ ↩︎
3. Frankfurt, H. (2018). Freedom of the Will and the Concept of a Person. In Agency And Responsiblity (pp. 77-91). Routledge. ↩︎