Belief

by Joel Marks

Published in Philosophy Now, Issue No. 70, November/December 2008, p. 39.

"JOOOOOOOOELLLLLLLLLL?!" The shrill call of my name made me jerk the phone from my ear. In that instant I thought, "Mom." I was flabbergasted: My mother had been dead for five years.

As the person on the wire continued to speak I had time to question myself. Could the death of my mother, even the sense that so many years had passed by, be only something I had dreamt the night before? Was it just that I had had no occasion to doubt it since waking up this morning, until this telephone call had jarred me back to my senses that my mother is still alive?

 I know there was a time in my life when I believed I could fly, for I remembered having done so. The image was of my body parallel with the ground, close to it but not touching it. My arms were crossed like a Cossack dancer's, and I was moving forward steadily, following the path of a sidewalk near my home. In my daily life I felt I had this power, although it wasn’t exactly clear to my child’s mind when I could exercise it. Then one day I realized it must have been something I had dreamt; so it was not a memory – or it was a memory, but of a recurring dream, not something that had actually happened. This was the kind of realization one experiences when, in the light of increasing knowledge, the belief in Santa Claus evaporates like dew at daybreak.

A few more words from the person on the telephone dispelled my current confusion. It was Svetlana, a new acquaintance. The way she had spoken my name had been her enthusiastic greeting, probably also prompted a bit by nervousness because of the novelty for her of speaking English without seeing the person she was speaking to. Now that I thought about it, her age was not far from my mother's when I was in college and would receive calls from her. That was how they would begin: "JOOOOOOOOELLLLLLLLLL?!"

The experience of Svetlana's call served as a reminder to me of the fragility of belief. For one second I had believed my mother was still alive. The belief was patently false, but I was taken in by it all the same. It is not an uncommon experience, is it? I’m sure you can empathize. Here is another example: How many times have I found myself gripping the chair when (for no particular reason) I have fallen into a momentary reverie of being in a plummeting airplane. I am feeling real fear. I believe I do believe at those moments that I am in an airplane. I am not asleep and dreaming; nor are my eyes even closed. It is just that there has been a shift of belief brought on by an image in my mind.

Yet, at other times, belief is recalcitrant. If somebody were to hold a gun to your head and demand that you believe in Santa Claus or he would shoot, could you do it? I doubt it. But that’s not because there is no Santa Claus; it’s because you don’t believe there is. If you were a Creationist, you would be just as much at a loss to conjure up a belief in evolution under the gun.

The startling revelation is that the entire world one inhabits is in some significant sense not the world that exists but the world one believes to exist. Everything that we know is first of all something that we believe, and in the end is that as well. In other words, what we know is, for all we know, something we only think we know. Our belief may be more or less justified, but even our deepest conviction is still a belief. And the hallmark of a belief, unlike a fact, is that it could be mistaken. That is the problem of skepticism: if beliefs are only buttressed by other beliefs, how can we know we have anything “right”? It is humbling, then, to realize that one's mind has a mind of its own.

But skepticism is, in the end, just a bugbear, for reasons that Wittgenstein explained in philosophy and sociobiologists have explained in science: We must be getting it all basically right or we couldn’t function -- we wouldn’t even be here. Indeed, for all the pleasure there is to be had from pondering the occasional lapse from perfection, such as mistaking Svetlana for Mom, the educated mind takes an even greater delight in understanding the inevitability of our exquisitely fine-tuned cognitive faculties. As others have pointed out: The question was never, “How could I have made such a mistake?” but, “How do we get it right so much of the time?” And now, amazingly, we know the answer: natural selection. What is more, the answer, now that we know it, seems totally obvious.

            Descartes’ intuitions were sound when he “forgave” the occasional illusions to which we are liable by pointing out that we also have the ability to disabuse ourselves of them (although he misattributed the source of that ability to the goodness of God rather than to the even more astonishing, because self-explanatory, mechanism of evolution). The late perception psychologist J.J. Gibson further developed this idea when he argued that illusions typically occur only under very limited or artificial circumstances, such as in the psychology laboratory, and are quickly remedied. Hence my swiftly figuring out that the person on the telephone was not my mother but Svetlana. 

Desire – Thirty Years Later

Published in Philosophy Now, Issue No. 93, November/December 2012, p. 44.

In 1982 I had my first “major” philosophical publication, a journal article entitled “A Theory of Emotion” (Philosophical Studies vol. 42, no. 2., pp. 227-42). My thesis was that the new cognitivist revolution in the study of emotion, associated at the time with the philosopher Robert C. Solomon, needed a supplement, namely, desire. (O. H. Green had reached the same conclusion independently.) Solomon, and even more explicitly, my target in the article, William Lyons, held that emotions are essentially a type of belief. This was a welcome change from the previously prevailing view of emotions as “brute feelings.” But I argued that this was not enough, for one could believe, say, that one was about to be mauled by a rabid dog, and yet not be in an emotional state unless one also possessed a desire not to be so mauled.

This insight had no doubt been prompted by my dabblings in Buddhism, for the Buddha preached that all suffering comes from desire. The Buddha’s recommendation was that we therefore cease to desire. I defended this thesis in an article on “Dispassion and the Ethical Life” in a volume I co-edited with Roger T. Ames on Emotions in Asian Thought (Albany: SUNY press, 1995). But to deflect the obvious objection that eliminating desire would be throwing out the baby with the bath water – since what would be the point of living at all if we desired nothing? – I analyzed the Buddha’s notion of desire as emotion, and emotion in turn as involving strong desiring.

Subsequently I saw an opening to the study of motivation, for it seemed natural to extend the belief/desire analysis to what moves us to action. And it is not only emotions that do this but, more generally, what might be called attitudes. I analyzed these as belief/desire sets, but now without the “strong desire” qualifier, since one need not feel deeply about something in order for it to produce behavior (or, for that matter, to be a “mental feeling”).

But now I came up against a distinction, first brought to my attention by Wayne Davis in an essay he wrote for my edited volume on The Ways of Desire (Chicago: Precedent, 1986). For it seems that “desire” is ambiguous between two quite distinct psychic phenomena. On the one hand desire is simply synonymous with motivation, so to say that one was moved by desire is just to say that one was motivated. On the other hand desire is a specific type of mental state, on a par with belief, such that a particular belief and a particular desire could jointly constitute a motivation (or a feeling). The mental-state desire would be desire proper or genuine desire, since the other type of desire is only another name for motivation.

An example of desire (proper) is wanting to go for a walk for its own sake. An example of motivational desire is wanting to go for a walk because you believe it will help you lose weight and you want (desire) to lose weight. But here again the latter desire (to lose weight) is ambiguous, since you might simply wish to lose weight or you might be motivated to lose weight by some further belief/desire set, such as that you desire to date someone and you believe s/he will only date you if you lose weight. And so on. The thesis I defended in another essay in that same volume – “The Difference between Motivation and Desire” -- was that, even though motivation as such is not the same as genuine desire, a genuine desire is always involved in motivation, simply because the regress must stop if there is to be any action at all.

I am no longer so sure about that last thesis. Bill Lycan, on behalf of his graduate seminar a few years ago, planted a seed of doubt in my mind. But even if we could be sure that “genuine desire” is an essential component of all of our motivation, we would still want an account of what it is. More specifically, it has always been a teaser to tease apart desire from belief. The best accounts I’ve seen, quite different from each other, are by Dennis Stampe (in my desire volume) and, more recently, Timothy Schroeder (in Three Faces of Desire from Oxford).

Despite my uncertainty about what I am even talking about, however, I remain a fan of desire. In fact, my interest in it has returned with a vengeance after a long hiatus. This time I am taken with desire’s role in values. In fact, I have quite given up on objective value as anything but a figment, and see all value as subjective – specifically, as a function of our desires.

I do still find room for more than one legitimate category of value, but, instead of objective and subjective values, there are intrinsic and extrinsic (or instrumental) values. The latter pair corresponds to intrinsic and extrinsic desires. So for example, to want to go for a walk for its own sake is to value walking intrinsically, while to want to do so for one’s health is to value walking instrumentally. What I no longer accept is that in addition to these there is such a thing as objective or inherent value, such that, for example, going for a walk might be “good in itself.” In a word, I no longer recognize the reality of value that is independent of desire.

Therefore I now consider desire to be the key to ethics, and so it becomes incumbent on me to try once again to figure out what the hell desire is. For starters I think I will pick up a fading offprint of an article from 1982 entitled, “A Theory of Emotion”!

Joel Marks is Professor Emeritus of Philosophy at the University of New Haven and a Bioethics Center Scholar at Yale University. He continues the tale of desire in his trilogy: Ethics without Morals: In Defense of Amorality (Routledge, 2013), It's Just a Feeling: The Philosophy of Desirism (CreateSpace, 2013), and Bad Faith: A Philosophical Memoir (CreateSpace, 2013).

Pons Asinorum

Copyright © 2002 by Joel Marks
Originally published in Philosophy Now magazine, no. 35, March/April 2002, page 48

Three travelers seek lodging for the night. They come upon a pension that charges 10 euros per person. It turns out that there is only one room available, but they don't mind sharing; so they pay the clerk 30 euros. When the proprietor returns, however, she decides that the guests should be given a discount for having to bunch up, so she summons the bellhop and hands him 5 euros to refund to them. Not being a completely honest fellow, the bellhop pockets two euros; this conveniently leaves one euro to be returned to each guest. Therefore each guest has now paid nine euros, for a total of 27 euros. But 27 plus the two in the bellhop's pocket = 29. What happened to the thirtieth euro?

When I first heard this puzzle, I was bedazzled. It seemed so simple; yet no matter how I turned it over in my mind, I could not come up with a solution. I even entertained the hypothesis that I must be dreaming, or under the influence of Descartes' evil daemon, "who has directed his entire effort to misleading me, [for] how do I know that I am not deceived every time I add two and three or count the sides of a square or perform an even simpler operation, if such can be imagined?" (Meditation One).

Soon, however, I came up with this surprising conclusion: There is no thirtieth euro! The travelers ended up paying 27 euros. The proprietor had 25, and the bellhop kept two. That's it. And yet ... I still could not shake from my head the notion that there was a missing euro. So it occurred to me that the puzzle could be conceived as a kind of illusion -- a calculative illusion, we might call it. An analogy can be drawn to a visual illusion, like the bent-stick-in-water, which is not really bent, but, even when one is fully knowledgeable of its straight shape, continues to appear bent at the waterline (due to the refraction of light). Just so, I now knew there was no thirtieth euro, but I couldn't dispel the mental impression that there was.

Finally I was able to dispel even the illusion. This came about precisely because of its refractoriness. I could not rid my mind of that thirtieth euro; there had to be a way to account for it. And so there is: For at the end, the proprietor has 25 euros, the bellhop two, and the guests three. Voila: 30 euros! So NOW the puzzle became: Why had there seemed to be a puzzle in the first place? Indeed, for some of my more logically adept friends and colleagues, there had been no puzzle about the 30th euro, and they were only puzzled about what was puzzling me. I can still experience a kind of Gestalt switching (as when viewing the picture of a vase and two facial profiles) between my puzzlement and my lack thereof. What makes for the difference?

The answer I have come up with is that this "puzzle" arises from a simple "mental mishearing": Where the situation at the end is that the guests have paid 27 euros, one might inattentively "hear" this as their now possessing 27 euros. Then indeed there would be a mystery (for the bellhop only possesses two, so where's the thirtieth?). But in fact at the end the guests only retain three euros of the original 30.

I have therefore passed through three stages: (1) puzzlement (indeed, astonishment), (2) knowledge, but with remaining unease or residual illusion, and (3) "total enlightenment" or "wisdom," with no puzzle or illusion extant (and even understanding why there had been puzzlement in the first place). The progression is instructive: From time to time life throws us for a loop, and, indeed, philosophy is in the very business of questioning fundamental assumptions. But sometimes, as with the three lodgers puzzle, we eventually discover a way to buttress our original conception of things; Wittgenstein considered philosophy itself to be one big faux-puzzle maker, which it was his calling to foil. However, the history of thought -- not to mention, the narratives of our individual lives -- is surely rife with cases of a new conception's replacing the old after some initial shock, such as the discoveries of pi, the stellar nature of the Milky Way, the absence of an ethereal medium, radioactivity, the expansion of the universe, the incompleteness of arithmetic, and so many others. So the truly philosophical task may be to discern which are the real and which the ersatz puzzles.

Which, for example, is the Anthropic Cosmological Principle? It seems that the various physical constants of our universe are exquisitely fine-tuned for the coming into being of ... us! The odds of this having come about by chance are said to be infinitesimal; ergo, we have empirical evidence of some (vast) intelligence and purposiveness (God?) pre-existing the universe. Is this a genuine problem for the secular mind?

Apparently not. Here is a homely analogy. Suppose you hit a golf ball into the air and it comes down in a dark forest. Well, no mystery there: Where it came down is where it came down. If we want to explain why it landed where it did, we would naturally look to physical laws and conditions. Now change the point of view: Pick a particular point hidden in the deep woods and challenge somebody to strike that precise location with the ball. We would expect only a Tiger Woods to attempt the feat, but even he would probably find it impossible.

Just so, the "fine-tuning" of nature that resulted in us may seem unlikely to the point of impossibility (sans an act of intentional design or creation), but the refutation of this "mystery" is that we are just "looking at things through the wrong end of the telescope": We pose the "problem" from the vantage of the end point, whereas causality works from the beginning, and then, whatever happens, happens. Thus, the "problem" needs no solution because it is not really a problem.

Yet there are others who see a deeper riddle posed by the constants of nature, and who consequently disparage the formulation above as the "Weak Anthropic Principle," or "WAP." Is there a Strong Anthropic Principle constituting a real puzzle? (Or would one just be a SAP to think so?) You will have to consider that for yourself outside the confines of this column.

Car Seats and the Absurd

Copyright © 2002 by Joel Marks
Originally published in Philosophy Now magazine, no. 38, October/November 2002, page 51

The extra minute you take to secure your child into her car seat could be just what it takes to bring your whole family into the path of a Mack truck half an hour down the road.

But that is obvious. It is the cruel, rueful, and ironic face of the contingency of existence. And of course it can work the other way around: Had you not taken the extra minute to secure your child into her car seat, you might have driven right into the path of a Mack truck. What does this tell us? Only, one might suppose, that we do not know the future. It doesn't change the fact that the only rational way to conduct one's affairs is to consider the odds: Children in automobile accidents are more likely to survive if they are strapped into a car seat. Therefore it is rational, not to mention morally obligatory, to do this for your child, even though it is within the realm of possibility that there will be a freak coincidence of circumstances, which converts your caring action into a contributing cause of the very catastrophe you were attempting to avert.

Only ... further reflection leads me to make a more bizarre inference. Put aside for the moment our epistemological situation and consider the metaphysics. Do you grant the following? Most accidents where there is a child passenger and an adult who has been responsible enough to purchase a car seat and secure the child into it, will not be due to some such aggravating factor as the driver drunkenly weaving in and out of traffic or drag racing or the like. Rather, the scenario will more likely be one of encountering some other car which has such a driver, or of the first driver's doing something foolishly spontaneous, like miscalculating when the light was going to change, OR of his being momentarily distracted, as by the family dog wagging his tail in the driver's face at a bend in the road, etc. In sum, I assume that the typical accident involving a child in a car seat occurs because the car was in the wrong place at the wrong time. Accidents are the thing of a moment, and moments are conditioned as delicately as a house of cards.

But if that is so, then do we not arrive at a rather startling conclusion, namely, that it is not the freak coincidence, but in fact the norm that accidents involving a child secured into a car seat would not have happened at all if the child had not been secured into the car seat? The logic of my argument is that everything else would have remained the same ... ceteris paribus, to use a logician's term. And I think that is a reasonable assumption in most cases. For instance, your not taking an extra minute with the car seat (because you were rushed, say) would not in any way affect whether the driver of the Mack truck takes another drink, or runs the stop light, etc. So that truck would still be at the very spot it would otherwise have been had you taken the extra minute. Except that because you didn't, there would be no accident: Your car and the Mack truck would pass through the same space but at different times.

In other words, although your alternative behavior would indeed affect the whole universe given enough time, the vast majority of the universe would remain the same in the short term. It is like the ripples in a pond after you plunk the pebble in: They will eventually reach the far shore and make the frog croak, but at first a nearby fish will not even notice anything has happened. Just so, the fate of the Mack truck and its driver, and of all who would be affected by them in turn into the indefinitely far future, would not begin to alter until later, after the moment at which the accident would have occurred. Up until then, all else with the truck and driver would be identical, so the accident won't occur provided you are careless about the car seat.

Singing the praises of car seats because your child's life has just been saved by one seems, therefore, as odd as extolling the virtues of kidnappers because your child has just been released by one. It is understandable, of course; there is a certain psycho-logic to it since your relief makes you feel grateful. But in strictly logical terms ... it ain't, is it?

Nonetheless, it is still true that it is rational (and, again, surely also ethical, even morally obligatory) to strap the child in. That is because the epistemology of the human condition leaves us with no rational option for deciding what to do other than relying on known, general probabilities. And in this case they presumably tell us that in otherwise matched populations, the one employing car seats will suffer fewer casualties. You simply cannot outwit Mother Nature on this one.

I conclude that ... life is absurd. (Although it is perhaps also absurd to employ logical argument to arrive at such a conclusion. But then ... life is absurd!) For the summation of the above is that it is rational to use a car seat for the safety of your child, even though on any actual occasion when the car seat shows its effectiveness for that purpose, it has likely also occasioned the risk to which your child has been exposed. In short, the car seat (in any given case but not in general) brings about the need for itself. It sounds like a marketer's dream ... or a metaphysical wizard's "perpetual justification engine" ... or the answer to a theologian's prayers for a Necessary Being ... but it is really a kind of joke, akin to: "Why am I hitting myself on the head with a hammer? Because it feels so good when I stop!" Also, this realization seems to have no practical import, and yet it changes everything, like a Gestalt shift (as from the contour of a vase to two facial profiles).

The Car Seat Paradox REDUX

August 5, 2018

Note: The following essay is a much expanded consideration of the above puzzle from 2002. At that time I concluded that life is absurd. I still think life is absurd, but at least I am now able to offer a detailed explanation of why it is (nevertheless) rational to use car seats.

Note: Despite the puzzle, I reiterate in the strongest terms that I believe (for the reasons given) it would be irrational not to use a car seat, and I encourage everybody to use a car seat when conveying a child.

We atheists or agnostics and even some thoughtful or compassionate believers know why it is ridiculous for the sole survivor(s) of an airplane crash to thank God: What loving, all-powerful and omniscient God would stand by, not to mention cause, a horrific event like this? Only an extreme egotism would suggest our meriting such special regard (God has a plan for me!) as to be spared the terrible fate that befell everyone else. Why not instead curse a deity who is so cruel and capricious, or at least callous?

            It may come as a surprise, however, that it is also ridiculous to thank God (literally or simply as an expression of emotion) that we strapped our child into a car seat when she has been spared injury or death in an automobile accident. (And let us suppose everyone else involved was also spared.) It certainly came as a surprise to me when I had this thought many years ago, and I have struggled to make sense of it ever since.

            Here is the basic idea. An accident while driving is typically a matter of bad timing (whatever else it may also involve, such as carelessness or bad luck). The smash-up occurred only because you entered the intersection at the exact moment a drunk coming down the cross street ran the light. The dog chose to jump into the front seat at just the moment you entered the bend. The deer leapt into the road just as you were passing by the thicket that had hidden her from view. You turned your attention from the road to your Google map just as the car next to you started drifting into your lane because the driver was texting his boss. And so on.

            Meanwhile, using a car seat takes a few moments. Here are instructions from a YouTube video:

To buckle the child you’re going to want to start with your harness straps nice and loose. Then you’re going to put the child’s arm through the hole. Make sure the shoulder strap is over their shoulder, and buckle between the legs. Do the same thing on the other side. … And now buckle the chest clip. But importantly keep the chest clip low. If you move it up to the right place right now, as you tighten your strap it’s going to get caught under the child’s throat, and that would not be comfortable. Now I’m going to take hold of the shoulder straps anywhere above the chest clip. I’m going to pinch them and pull firmly upward. See how I gathered all the slack out of the legs, out of the stomach and up to the shoulders? If I need to I can slide the chest clip down a little bit at this point. Now I’m going to take the tail at the bottom of the seat and I’m going to pull firmly. The I’m going to check. I’m going to pull upward again on the shoulder strap, checking that no slack comes up towards the shoulders. I’m going to put a finger at the collar bone and pull it away from the child’s body. One finger should fit. But if you can do a two-finger salute like this, that is too loose. So I have a little bit left to pull out from the tail, and now when I check – again, pull upwards – there’s no slack that came up. … Next I’m going to move the chest clip up so that the top of it is at the top of the armpits. I like to call it the tickle clip to remind you to run your fingers across the top and tickle the child’s armpits.           

So what first occurred to me was that an accident in which a child is saved by a car seat might very well not even have occurred if the driver had not used a car seat. Why not, then, curse God (or your partner or the manufacturers of car seats or your own conscientiousness or just your unlucky stars) for inducing you to spend so much time making sure your child was properly strapped into the car seat, since this served only to place your car in the wrong place at the wrong time?

But of course there is an obvious reply. On some other occasion you might with equal likelihood, and for the same reason of bad timing, have ended up in a different accident if you had not used a car seat. And this time – what is even worse – your child would not have been protected and so been more likely to be injured or killed.

Well, OK: This sounds like a good reason for people to use car seats. This is what makes it rational, perhaps even morally obligatory, to use them. Nevertheless, I find something peculiar about the situation. For one thing, it is not clear what bearing the rationality of using car seats has on your emotional reaction to your child being saved by your having used a car seat. While it is true (or I will assume) that a society in which people regularly use car seats has lower casualty figures for children in moving vehicles, it still seems, by my reasoning above, that on any particular occasion when a child is saved by a car seat, it might well or even usually have been better if a car seat had not been used. After all, your motivation for using a car seat is not public spirit. You are not first and foremost trying to make society safer (as your motive might be if, say, you sent your offspring off to war); you are trying to make your child safer. So how could it be rational to be happy you had used a car seat on the particular occasion when it would have been better if you hadn’t, just because (as if by a statistical hand) widespread use of car seats is beneficial to society?

I don’t buy, by the way, that you yourself would have been in a different accident with your child had you not used a car seat and hence avoided this accident. That’s just superstition. Most people do not get into an accident when driving with their children. So you would be mighty unlucky if you not only got into an accident while using a car seat but also would have gotten into one (on this or a different occasion) if you had not used a car seat on this occasion.

Now, it is old news that a rational action can lead to an undesired outcome. Rationality is what we rely on in practical affairs precisely in the absence of certainty. What is rational to do is what is the most likely to achieve our ends under the circumstances; but this implies that sometimes things won’t turn out as we want them to even when we behave rationally. It is irrational to refuse to fly just because, in an exceptional case, an airplane will crash; but if most flights crashed, it would be better not to have boarded an airplane most of the time, and hence it would not be rational to fly for routine purposes. What creates the air of paradox in the present case is that it is better not to have used a car seat in most of the cases when the car seat does exactly what we want it to. This is the rule, not the exception. How, then, could it be rational to use a car seat?

            The answer, I now think, goes like this. What we want and expect a car seat to do is protect our child in an accident. This is surely rational because car seats have (I presume) been amply demonstrated to reduce the likelihood of injuries to children in accidents. What is not rational, however, is something different, namely, to use car seats in order to prevent accidents. That is not rational because there is no reason to think that using car seats is more likely to prevent accidents than not using them. In particular, as we have seen, the timing argument works equally well either way. Ergo Q.E.D.: It is rational to use a car seat to protect your child in case of an accident, even though if an accident occurs, it might well or even usually have been better had you not used one.

Here’s another way to think about the kind of situation I am talking about. There are actually two main ways that using a car seat could help to protect your child. Only one of them is if you are in an accident. An even better way is if the mere passage of time it takes for you to strap that wiggly body into that complex array of straps causes you to miss out on being in an accident in the first place. This means that when you are expending time and effort in this way you are contributing to one of the following consequences (although you don’t know which): (1) You will narrowly miss being in accident; (2) You will be in an accident but the car seat plays no further role (the child will be hurt or unhurt as much with as without the car seat); (3) You will be in an accident and the car seat makes it worse (I will omit the gruesome details); (4) You will be in an accident and the car seat works as advertised, and intended and hoped, to spare your child (greater) injury. All of these are highly unlikely.

Much more likely is that your using a car seat makes no difference whatever: You won’t be in an accident when driving with your child whether you use a car seat or not. So why bother using a car seat? What is more, of the four scenarios wherein your use of the car seat does make a difference, three of the four cause an accident while only one (1) prevents it, and of those three, one (3) even makes things worse in the case of the accident. So it looks like using a car seat actually makes things worse!

But that last calculation is cheating, since the likelihood of (1) equals the combined likelihood of (2)-(4): Your using a car seat is just as likely to prevent as to cause an accident. So those cancel out. Furthermore, among (2)-(4), one (3) makes things worse, one (4) makes things better, and one (2) is neutral; so these too would seem to cancel out. However, this still leaves the question: What is now left to tip the balance toward the rationality of using a car seat? The answer is clear: the greater likelihood of (4) than (3). A car seat is much more likely to help than to harm in an accident.

We are not home free yet. The percentage of cases in which there is an accident involving a child in a car seat is still very small. Why, then, go to all the expense of purchasing a car seat and the trouble of using it? Is this just another capitalist scheme to scare us into buying something we don’t need? No. The standard analysis of risk provides the solution: We are concerned not only about the probability of an event, but also the nature and magnitude of the event. It is very unlikely that your house will burn down while you and your family are in it; but the magnitude of such a loss counsels the relatively minor expense and inconvenience of installing smoke detectors and testing them every week. Just so, the death or injury of your child in an automobile accident is very unlikely, but it would be a harm of such magnitude that you are wise to purchase and use a car seat.

So we have managed to dispel any suspicion that it is irrational to use a car seat. But this has not been my concern in the first place. No, the wrinkle that wrinkles my brow is that it is rational to use a car seat. Why does this perplex me? Because, as I keep saying, a car seat works as advertised in an accident just in case its use was (more often than not) responsible for causing the accident. And that’s not all: It remains rational (and probably also obligatory from a moral point of view) to use car seats despite its turning out to be the case that one’s happiness at having used a car seat in the case of an accident where it worked as advertised is misplaced. This to me has it all over Sisyphus in the absurdity department.

So I doubt that the logical explanation of the rationality of using car seats will penetrate so deeply into the psyches of even most of us who understand it to change our feelings (now that we have been bitten by the bug of paradox). Speaking for myself, if I am ever in an accident where a child is saved by my having used a car seat, I am sure I will thank my lucky stars that I used it. I liken this phenomenon to visual illusions, which will often persist even after we come to understand they are illusions. For example, the parallel lines in the Müller-Lyer will likely forever appear of unequal length, no matter how often we measure them with a ruler.

            Of course there is a legitimate source of joy after the accident we have been discussing, namely, that the child has not been hurt or hurt badly or killed. Even if you could kick yourself for having used a car seat on this occasion (which would be irrational – just as well rail against your partner for kissing you goodbye before you got into the car), since you did use one and ended up in an accident it is wonderful that she was not hurt. Thank God!

            But let me finally reiterate that I do not conclude that using a car seat is irrational. (If this article gains popular currency, I know that it will be misread on a thousand occasions by those who merely skim it.) Quite the contrary: I believe it would be irrational not to use one, and I encourage everybody in the strongest terms to use a car seat when conveying a child. But it is precisely this that creates the sense of puzzlement, namely, that it is rational to use a car seat despite its having the feature I have been describing. So I do not dispute that it is rational to use a car seat, but I marvel that it is.

Note to analytic philosophers: Our discipline is riddled with tantalizing thought experiments that have challenged both common sense and deeply held theories. There is the Gettier Problem, the prisoner’s dilemma, Mary the visual neuroscientist, Nozick’s experience machine, Parfit’s split brain speculations, the Chinese Room, the Nonidentity Problem, the Knobe Effect, and so on. The car seat paradox (so to speak?) has the feel of a perfect thought experiment to me. Unfortunately I have not been able to come up with any great issue it might speak to, so this could be yet another illusion generated by the rationality of using car seats. However, I am sure that the actual genesis of many of our favorite thought experiments was not from wrestling with a philosophical issue but just from having a puzzle suddenly occur to somebody. So I invite my colleagues to find some application(s) of this puzzle-in-search-of-an-issue that would ultimately earn it a place in the pantheon of Great Gedanken Experiments. Could this quirk in the rationality of using car seats be, as it were, the next precession of the orbit of Mercury that will change the universe of knowledge? My delusion of grandeur doth entertain the prospect.

            Alternatively, you are invited to argue or demonstrate that there is no quirk to begin with, for example, by coming up with a counterexample. That would be an act (other than using a car seat) that is rational despite the fact that, most of the time it achieves its purpose, it would be better if it had not been done, and yet does not strike us as odd for that fact. I have not been able to come up with one, nor (by my lights anyway) have my interlocutors; yet I certainly cannot rule out that there is a whole class of such acts. But even in that case, it may yet be possible to salvage something of value if the existence of this feature of some rational acts (viz., that, most of the time they achieve their purpose, it would be better if they had not been done) is felt to alter our conception of rationality in an interesting way, even if for no further reason than that it had never before been noticed.

Many thanks to Thomas Pölzler and Mitchell Silver for very helpful assistance in unraveling this puzzle (if not my puzzlement).