Pain and Unintelligent Design

This is an earlier paper that I am re-posting because of the interest shown in “Evolution of Pain”.

Pain and Unintelligent Design

Pain is a very widespread biological adaptation. Pain receptors are everywhere in the animal world. Evidently pain serves the purposes of the genes—it enables survival. It is not just a by-product or holdover; it is specifically functional. To a first approximation we can say that pain serves the purpose of avoiding danger: it signals danger and it shapes behavior so as to avoid it. It hurts of course, and hurting is not good for the organism’s feeling of wellbeing: but that hurt is beneficial to the organism because it serves to keep it from injury and death. So the story goes: evolution equips us with the necessary evil of pain the better to enable our survival. We hurt in order to live.  If we didn’t hurt, we would die. People born without pain receptors are exceptionally prone to injury. So nature is not so cruel after all. Animals feel pain for their own good.

            But why is pain quite so bad? Why does it hurt so much? Is the degree of pain we observe really necessary for pain to perform its function? Suppose we [encountered alien creatures much like ourselves except that their pain threshold is much lower and their degree of pain much higher. If they stub their toe even slightly the pain is excruciating (equivalent to us having our toe hit hard with a hammer); their headaches are epic bouts of suffering; a mere graze has them screaming in agony. True, all this pain encourages them to be especially careful not to be injured, and it certainly aids their survival, but it all seems a bit excessive. Wouldn’t a lesser amount of pain serve the purpose just as well? And note that their extremes of pain are quite debilitating: they can’t go about their daily business with so much pain all the time.  If one of them stubs her toe she is laid off work for a week and confined to bed. Moreover, the pain tends to persist when the painful stimulus is removed: it hurts just as much after the graze has occurred. If these creatures were designed by some conscious being, we would say that the designer was an unintelligent designer. If the genes are the ones responsible, we would wonder what selective pressure could have allowed such extremes of pain. Their pain level is clearly surplus to requirements. But isn’t it much the same with us? I would be careful not to stub my toe even if I felt half the pain I feel now. The pain of a burn would make me avoid the flame even if it was much less fierce than it is now. And what precisely is the point of digestive pain or muscle pain? What do these things enable me to avoid? We get along quite well without pain receptors in the brain (or the hair, nails, and teeth enamel), so why not dispense with it for other organs too? Why does cancer cause so much pain? What good does that do? Why are we built to be susceptible to torture? Torture makes us do things against our wishes—it can be used coercively—so why build us to be susceptible to it? A warrior who can’t be tortured is a better warrior, surely. Why allow chronic pain that serves no discernible biological function? A more rational pain perception system would limit pain to those occasions on which it can serve its purpose of informing and avoiding, without overdoing it in the way it seems to. In a perfect world there would be no pain at all, just a perceptual system that alerts us non-painfully to danger; but granted that pain is a more effective deterrent, why not limit it to the real necessities? The negative side effects of severe pain surely outweigh its benefits. It seems like a case of unintelligent design.

            Yet pain evidently has a long and distinguished evolutionary history. It has been tried and tested over countless generations in millions of species. There is every reason to believe that pain receptors are as precisely calibrated as visual receptors. Just as the eye independently evolved in several lineages, so we can suppose that pain did (“convergent evolution”). It isn’t that pain only recently evolved in a single species and hasn’t yet worked out the kinks in its design (cf. bipedalism); pain is as old as flesh and bone. Plants don’t feel pain, but almost everything else does, above a certain level of biological complexity. There are no pain-free mammals. Can it be that mammalian pain is a kind of colossal biological blunder entailing much more suffering than is necessary for it to perform its function? So we have a puzzle—the puzzle of pain. On the one hand, the general level of pain seems excessive, with non-functional side effects; on the other hand, it is hard to believe that evolution would tolerate something so pointless. After all, pain uses energy, and evolution is miserly about energy. We can suppose that some organisms experience less pain than others (humans seem especially prone to it)—invertebrates less than vertebrates, say—so why not make all organisms function with a lower propensity for pain? Obviously, organisms can survive quite well without being quite so exquisitely sensitive to pain, so why not raise the threshold and reduce the intensity?

            Compare pleasure. Pleasure, like pain, is motivational, prompting organisms to engage not avoid. Food and sex are the obvious examples (defecation too, according to Freud). But the extremes of pleasure are never so intense as the extremes of pain: pain is really motivational, while pleasure can be taken or left. No one would rather die than forfeit an orgasm, but pain can make you want to die. Why the asymmetry? Pleasure motivates effectively enough without going sky-high, while excruciating pain is always moments away. Why not regulate pain to match pleasure? There is no need to make eating berries sheer ecstasy in order to get animals to eat berries, so why make being burnt sheer agony in order to get animals to avoid being burnt? Our pleasure system seems designed sensibly, moderately, non-hyperbolically, while our pain system goes way over the top. And yet that would make it biologically anomalous, a kind of freak accident. It’s like having grotesquely enlarged eyes when smaller eyes will do. Pleasure is a good thing biologically, but there is no need to overdo it; pain is also a good thing biologically (not otherwise), but there is no need to overdo it.

            I think this is a genuine puzzle with no obvious solution. How do we reconcile the efficiency and parsimony of evolution with the apparent extravagance of pain, as it currently exists? However, I can think of a possible resolution of the puzzle, which finds in pain a unique biological function, or one that is uniquely imperative. By way of analogy consider the following imaginary scenario. The local children have a predilection for playing over by the railway tracks, which feature a live electrical line guaranteed to cause death in anyone who touches it. There have been a number of fatalities recently and the parents are up in arms. There seems no way to prevent the children from straying over there—being grounded or conventionally punished is not enough of a deterrent. The no-nonsense headmaster of the local school comes up with an extreme idea: any child caught in the vicinity of the railway tracks will be given twenty lashes! This is certainly cruel and unusual punishment, but the dangers it is meant to deter are so extreme that the community decides it is the only way to save the children’s lives. In fact, several children, perhaps skeptical of the headmaster’s threats, have already received this extreme punishment, and as a result they sure as hell aren’t going over to the railway tracks any time soon. An outsider unfamiliar with the situation might suspect a sadistic headmaster and hysterical parents, but in fact this is the only way to prevent fatalities, as experience has shown. Someone might object: “Surely twenty lashes is too much! What about reducing it to ten or even five?” The answer given is that this is just too risky, given the very real dangers faced by the children; in fact, twenty lashes is the minimum that will ensure the desired result (child psychologists have studied it, etc.). Here we might reasonably conclude that the apparently excessive punishment is justified given the facts of the case—death by electrocution versus twenty lashes. The attractions of the railway tracks are simply that strong! We might compare it to talking out an insurance policy: if the results of a catastrophic storm are severe enough we may be willing to part with a lot of money to purchase an insurance policy. It may seem irrational to purchase the policy given its steep price and the improbability of a severe storm, but actually it makes sense because of the seriousness of the storm if it happens. Now suppose that the consequences of injury for an organism are severe indeed—maiming followed by certain death. There are no doctors to patch you up, just brutal nature to bring you down. A broken forelimb can and will result in certain death. It is then imperative to avoid breaking that forelimb, so if you feel it under dangerous stress you had better relieve that stress immediately. Just in case the animal doesn’t get the message the genes have taken out an insurance policy: make the pain so severe that the animal will always avoid the threatening stimulus. Strictly speaking, the severe pain is unnecessary to ensure the desired outcome, but just in case the genes ramp it up to excruciating levels. This is like the home insurer who thinks he should buy the policy just in case there is a storm; otherwise he might be ruined. Similarly, the genes take no chances and deliver a jolt of pain guaranteed to get the animal’s attention. It isn’t like the case of pleasure because not getting some particular pleasure will not automatically result in death, but being wounded generally will. That is, if injury and death are tightly correlated it makes sense to install pain receptors that operate to the max. No lazily leaving your hand in the flame as you snooze and suffering only mild discomfort: rather, deliver a jolt of pain guaranteed to make you withdraw your hand ASAP. Call this the insurance policy theory of pain: don’t take any chances where bodily injury is concerned–insure you are covered in case of catastrophe.[1] If it hurts like hell, so be it—better to groan than to die. So the underlying reason for the excessiveness of pain is that biological entities are very prone to death from injury, even slight injury. If you could die from a mere graze, your genes would see to it that a graze really stings, so that you avoid grazes at all costs. Death spells non-survival for the genes, so they had better do everything in their power to keep their host organism from dying on them. The result is organisms that feel pain easily and intensely. If it turned out that those alien organisms I mentioned that suffer extreme levels of pain were also very prone to death from minor injury, we would begin to understand why things hurt so bad for them. In our own case, according to the insurance policy theory, evolution has designed our pain perception system to carefully track our risks in a perilous world. It isn’t just poor design and mindless stupidity that have made us so susceptible to pain in extreme forms; this is just the optimum way to keep as alive as bearers of those precious genes (in their eyes anyway). We inherit our pain receptors from our ancestors, and they lived in a far more dangerous world, in which even minor injuries can have fatal consequences. Those catastrophic storms came more often then.

            This puts the extremes of romantic suffering in a new light. It is understandable from a biological point of view why romantic rejection would feel bad, but why so bad? Why, in some cases, does it lead to suicide? Why is romantic suffering so uniquely awful?[2] After all, there are other people out there who could serve as the vehicle of your genes—too many fish in the sea, etc. The reason is that we must be hyper-motivated in the case of romantic love because that’s the only way the genes can perpetuate themselves. Sexual attraction must be extreme, and that means that the pain of sexual rejection must be extreme too. Persistence is of the essence. If people felt pretty indifferent about it, it wouldn’t get done; and where would the genes be then? They would be stuck in a body without any means of escape into future generations. Therefore they ensure that the penalty for sexual and romantic rejection is lots of emotional pain; that way people will try to avoid it.  It is the same with separation: the reason lovers find separation so painful is that the genes have built them to stay together during the time of maximum reproductive potential. It may seem excessive—it is excessive—but it works as an insurance policy against reproductive failure. People don’t need to suffer that much from romantic rejection and separation, but making them suffer as they do is insurance against the catastrophe of non-reproduction. It is crucial biologically for reproduction to occur, so the genes make sure that whatever interferes with that causes a lot of suffering. This is why there is a great deal of pleasure in love, but also a great deal of pain–more than seems strictly necessary to get the job done. The pain involved in the loss of children is similar: it acts as a deterrent to neglecting one’s children and thus terminating the genetic line. Emotional excess functions as an insurance policy about a biologically crucial event. Extreme pain is thus not so much maladaptive as hyper-adaptive: it works to ensure that appropriate steps are taken when the going gets tough, no matter how awful for the sufferer. It may be, then, that the amount of pain an animal suffers is precisely the right amount all things considered, even though it seems surplus to requirements (and nasty in itself). So at least the insurance policy theory maintains, and it must be admitted that accusing evolution of gratuitous pain production would be uncharitable to evolution.

            To the sufferer pain seems excessive, a gratuitous infliction, far beyond what is necessary to promote survival; but from the point of view of the genes it is simply an effective way to optimize performance in the game of survival. It may hurt us a lot, but it does them a favor. It keeps us on our toes. Still, it is puzzling that it hurts quite as much as it does.[3]

Colin McGinn

[1] We can compare the insurance policy theory of excessive pain to the arms race theory of excessive biological weaponry: they may seem pointless and counterproductive but they result from the inner logic of evolution as a mindless process driven by gene wars. Biological exaggeration can occur when the genes are fighting for survival and are not too concerned about the welfare of their hosts.

[2] Romeo and Juliet are the obvious example, but the case of Marianne Dashwood in Jane Austen’s Sense and Sensibilityis a study in romantic suffering—so extreme, so pointless.

[3] In this paper I simply assume the gene-centered view of evolution and biology, with ample use of associated metaphor. I intend no biological reductionism, just biological realism.

Share
12 replies
  1. Pain
    Pain says:

    Thanks for reposting this article. I would have added a link at the top of the Evolution of Pain to indicate that there is more to the argument. Evolution of Pain is Part 2 of your case.

    Few random thoughts on the topic:
    1. Pain is an umbrella term to include various types of experiences that people categorize as pain. Common cold is a similar umbrella term that combines the similar symptoms from multitude of the viruses. When we generalize the reasons for pain, we may be losing some subtlety as there are multitudes of pain.

    The related consideration is that the pain might have evolved to deal with real emergencies but then got adapted for less emergent cases. It works but could be optimized. The body might find reuse of the existing blunt tool better than developing a toolset of 200 different responses that all have to be optimally evolved while engaging in survival and pro-creation. In my layman’s view, the species usually narrow down the environment variables rather than increasing adaptation strategies.

    We as humans are relatively unique as we have moved across the variety of the environments where we need a general approach strategy that is good enough but not perfect. This also gives a buffer for us when we enter the new domain to not fail and be forced to retreat to the prior habitat.

    2. There might be a timing difference. All the pain adoptions to get you to procreate are passed to the next generation. All the pain adaptations that you have after procreation die with you. Again calling upon the layman’s authority, the modern disease of cancer might be one of these cases. It just doesn’t matter in the procreation game. There is no need for body to optimize for this use case.

    Reply
  2. Paul Reinicke
    Paul Reinicke says:

    I agree as you state in the last sentence pain can be puzzlingly superfluous. Who knows, perhaps humans and AI working together will someday find a way to engineer it out of the human genome (reduce it down to a Goldilocks level). And that gives me another thought. People are so concerned about AI taking over, but what about the possibility that humans and AI working together might someday bioengineer a much more intelligent species than our own, and *it* might take over?

    Reply
    • Colin McGinn
      Colin McGinn says:

      I actually think it’s quite likely that another quasi-human species, descended from us, will take over, and that will not be a bad thing. We have so many flaws, harmful to ourselves and others, that a retooling might be in order.

      Reply
      • Paul Reinicke
        Paul Reinicke says:

        I hate to say it, but you might be right about that (all of what you said). And that gives me another thought. Imagine if scientists were able to find a way to bioengineer into the human chromosome that deep concern for ecological preservation and protection found in individuals like arne naess, rachel carson, aldo leopold, and yours truly. Could programing into us a deep concern for saving the planet and a willingness to make sacrifices be a solution? It’s an interesting thought at least.

        Reply
  3. Giulio Katis
    Giulio Katis says:

    I enjoyed the more specific question of why is there the apparent asymmetry between pain and pleasure (taking it as granted that sensation can act as a motivation for subsequent deliberate action). Could it be that something stronger than asymmetry is involved. Perhaps the only type of deliberate (not spontaneous) action that sensation can directly motivate is the action to try to stop something. Are there examples of deliberate pleasure seeking actions? Or are they really take the form of stopping some form of pain (physical, emotional etc – including the unpleasant feeling of unfulfilled desire or wish, fear of being alone, or guilt etc)? (If this were true, then all acts of genuine love that weren’t at root minimising some form of pain would have to be spontaneous.)

    If this is true, why would it be? Perhaps the following fact is relevant: in mathematical terms, minimising something to zero is a well defined optimisation problem, but unconstrained maximisation has no solution (it would be disastrous from an evolutionary perspective).

    Reply
    • Colin McGinn
      Colin McGinn says:

      It’s an interesting idea, but it is implausible that all pleasure-seeking is just pain-avoidance. Couldn’t an organism avoid all pain but never experience real pleasure? The two states are not identical–though they may be generally correlated.

      Reply
      • Giulio Katis
        Giulio Katis says:

        The hypothesis was not that all pleasure is pain avoidance. (It assumes both pleasure and pain exist as states of experience.) The hypothesis was regarding the nature of the algorithm that directly motivates deliberate (as opposed to spontaneous) action. If the algorithm sought to increase pleasure, what would atop it from continually increasing pleasure? Some more complicated mechanism could be developed, involving some new control variable to reduce the pleasure seeking urge (making pleasure a nonlinear function of this control variable). Maybe this happens some of the time – it probably does. But perhaps evolution found another simpler and more robust algorithm that has become more dominant. Develop some form of pain associated with the goal (eg the pain of unfulfilled desire), and simply (let evolution evolve algorithms to) take action to minimise that to zero. Put more simply, minimising something to zero is an effective control for an algorithm that needs to stop or reach an equilibrium state, and pain is in essence that which we wish to minimise to zero.

        Reply
        • Colin McGinn
          Colin McGinn says:

          But it is an interesting point that pleasure is strongly correlated with pain avoidance, so we can imagine a form of hedonism that recommends pain avoidance instead of pleasure seeking.

          Reply
    • Colin McGinn
      Colin McGinn says:

      That’s why it’s interesting: it can be hard to tell the difference. If we ask whether pain could be just a decrease in pleasure, the answer seems more obvious–it is more than that.

      Reply

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.