Good Grief

A new paper in Personality and Social Psychology Review is out, arguing that grief serves an adaptive function. Thus, the title. Today’s post discusses the argument in the paper.

(First, yes, I’ve been gone for a while. I was trapped near the inner circle of fault. But now I’m back.)

The paper, by Winegard et al., opens with the following vignette:

A bereaved wife every weekend walks one mile to place flowers on her deceased husband’s cemetery stone. Neither rain nor snow prevents her from making this trip, one she has been making for 2 years. However poignant the scene, and however high our temptation to exclude it from the cold logic of scientific scrutiny, it presents researchers with a perplexing puzzle that demands reflection. The deceased husband, despite all of his widow’s solicitude, cannot return to repay his wife’s devotion. Why waste time, energy, effort, resources—why, in other words, grieve for a social bond that can no longer compensate such dedication?

Grief... what is it good for?

Grief… what is it good for?

This does seem to be a good question. I mused about this a little in a discussion of love and broken hearts a year ago. The emotional pain, and everything that goes along with it, does seem puzzling. Why cry over spilt milk?

Winegard et al. locate the puzzle in costs. This is a key point given the argument that they want to make. As you can see from the opening passage, their idea is that the “time, energy, effort, resources” are being wasted on the dead, who, they correctly point out, are notorious for failing to reciprocate. (I except here of course the Dead Men of Dunharrow, who really came through in a pinch.) Time and resources are being wasted in the sense that they could be used more productively. Walking to the grave in this example, then, carries an opportunity cost, which is the next best thing that one could do with a given resource. (I’m gratified that these authors lean so heavily on the notion of opportunity costs, which I have recently written about in a quite different context.) It’s clear that the puzzle that Winegard et al. have in mind has to do with the very large opportunity costs being paid by those who grieve.

Their explanation is that bearing these costs acts as a signal. Drawing on Costly Signaling Theory (CST), they argue that paying these costs sends signals to other people regarding one’s value as a social partner. Recall from CST that for a signal to convey information, the cost must depend on the underlying quality being signaled. Crucially, the size of the cost must depend on a property of the organism doing the signaling; in the usual example of a peacock’s tail, the cost of a bit tail is marginally lower for higher quality organisms. The authors write:

We suggest that grief functions like these (and other) hard-to-fake signals because it is costly and conveys information about the underlying traits of the griever. Humans’ prolonged grief response may act as an honest signal of prosocial proclivities, most importantly, of the proclivity to form strong, non-calculated bonds.

Their claim, then, turns on the idea that grief will be less costly for people with greater “prosocial proclivities.” (As a complete aside, this is more or less what the anonymous yet obviously insightful commenter Aliera suggested in my “broken heart” post, writing: “…perhaps extreme reactions to unrequited love or rejection (in the form of creative endeavors, passionate manifestos, devotion-displays) might serve as signals of one’s ability and willingness to commit to a romantic partner in general. These signals, then, are actually – and unknowingly – directed toward new potential mates who might now consider the individual attractive as a long-term mate based on the quality, costliness, and honesty of the display.”) In any case, returning to the paper, the argument rests on the idea that less prosocial people have better things to do with their time than more prosocial people (my underlining, their italics):

A relatively low commitment social strategy, one that consists of cheating and manipulating others, may constitute a viable social strategy… If so, intense grief would cost those who pursue such a strategy more relative to those who are inclined to form strong bonds because their time, energy, and resources would be better spent searching for and exploiting less costly opportunities.

I find myself puzzled when I take this claim in juxtaposition with the opening vignette. The story about the woman turns on the idea that she could be doing something else with her time and energy. That, indeed, is supposed to be the root of the puzzle: that she is paying substantial opportunity costs by visiting the grave. These costs are supposed to animate the issue in the first place: Why are people paying such huge costs, in the form of all the things that they’re not doing because they’re grieving? It seems clear that the woman in question has more than just the two options of either grieving on the one hand or exploiting others on the other. People have many things they might be doing at any given moment besides those two activities. In short, it seems from the opening vignette that the authors not only concede but require that it be true that grieving carries very big opportunity costs, even if one is a prosocial sort of person. Yet their argument also requires that the opportunity costs of grieving people to be small, at least relative to non-grieving people.

So, while the logic of the opportunity cost argument turns on the idea that non-grieving people have bigger opportunity costs than grieving people, I see no particular reason to believe that the benefits of searching for and exploiting others – the thing that non-grievers are up to – carries especially greater benefits than other activities that either they or grievers might do.

Further, suppose that it were true that, generally, how much one grieves depends on what else one might be doing with one’s time, such that people who grieve have less they might be doing, and so are bearing lower opportunity costs by grieving. In that case, unless one grants that “searching for and exploiting others” is an especially valuable way to spend one’s time, then grieving will wind up, just like other costly signals, signaling the underlying quality that keeps the signal honest: that one doesn’t have big opportunity costs. This property – not having much else valuable that one might be doing – seems like a puzzling quality to signal, but I suppose it’s possible.

In short, if the asymmetry in opportunity costs for grievers versus non-grievers doesn’t hold, then the rest of the argument in the paper doesn’t hold. I hope I am not misunderstood. I think grieving is indeed mysterious, as my Love post implied. And of course I think the evolutionary point of view will help to clarify matters.

For my part, I’m skeptical in general of explanations that turn on the notion of Types, to use the language of game theory. It seems perfectly plausible to me that many people might form very deep attachments to particular friends, kin, and lovers, yet be very exploitative in other relationships. I have little doubt that people who viciously exploit strangers nonetheless grieve when their parents die, limiting the information that is conveyed by observations of grieving. The fact that people can vary their degree of exploitation versus prosociality over time makes me very skeptical that grief and related emotions have to do with signaling one’s Type. Nothing, as a logical matter, prevents someone from grieving at time one from being exploitative at time two.

Not that I have a much better idea. Tooby and Cosmides propose a simulation view and a recalibrational view, both of which seem plausible:

Paradoxically, grief provoked by death may be a byproduct of mechanisms designed to take imagined situations as input: it may be intense so that, if triggered by imagination in advance, it is properly deterrent. Alternatively-or additionally-grief may be intense in order to recalibrate weightings in the decision rules that governed choices prior to the death. If your child died because you made an incorrect choice (and given the absence of a controlled study with alternative realities, a bad outcome always raises the probability that you made an incorrect choice), then experiencing grief will recalibrate you for subsequent choices. Death may involve guilt, grief, and depression because of the problem of recalibration of weights on courses of action.


Winegard, B. M., Reynolds, T., Baumesiter, R. F., Winegard, B., & Maner, J. K. (in press). Grief Functions as an Honest Indicator of Commitment. Personality and Social Psychology Review.

10. February 2014 by kurzbanepblog
Categories: Blog | 19 comments

Comments (19)

  1. I don’t see the logic of the recalibration view. Do other animals which might need to recalibrate their behavior act similarly? How does returning to a grave each day or depression help one recalibrate behavior in general? They seem like awkward design features.

    • I have no information about other animals, but depression in humans enables deep rumination useful for problem-solving (Thinking, Fast and Slow, D Kahneman). In grief the problem is how to continue living after loss, which will usually require some recalibration in behaviour.

      • There’s very little good evidence to suggest that depression helps people make better decisions, as I wrote about here:

        Further, making better decisions again seems to be an odd bedfellow for other behavioral characteristics, like suicidal thoughts, a lack of self-care, and other such features of depression. The same goes for recalibration.

        • Melanie Klein advanced the risky idea of the manic-depressive cycles which we all experience. The depressive movement involves a reflective disposition or mood which benefits the person and is the most mature adaptation to events.

          • Many people have advanced the idea that depression helps people make better decisions. There just hasn’t been good evidence to accompany it yet.

          • Depression as a clinical disease is probably not the same thing as Klein’s depressive phase. However, it seems intuitive that decision making during the depressive phase is different from decisions during the manic phase which by definition are made in haste.

          • To the extent that these decisions were important, we should have expected that evolution designed mechanisms to make them quickly and efficiently. Just because “thinking is slow”, whatever that is supposed to mean, we have no guarantee that it’s better.

          • That is correct, the actual procedures utilized during an evaluation are relevant to the validity of the decision. As I understand Klein, human emotional states cycle between a manic phase and a depressive phase. The depressive phase ‘may’ be more reflective which usually implies a somewhat more comprehensive and thoughtful procedure but obviously things can be overlooked, relevant procedures may not be implemented, and certain influences may be more determining or controlling than the person can understand. The manic phase, being more impulsive and automatic, ‘could’ be more effective because certain habits (genetic or learned) are repeated in a pattern that is usually effective. Sometimes, using tried and true pathways of response produces the useful results. The depressive thought process may interfere with automatic habits. But, in each case the procedures involved whether automatic or reflective should be analyzed in terms of the relevance for validity and outcome. As I understand grief, there is a loss of role because of the absence of the deceased; there is a loss of identity because that relationship is no longer actual. So, it just seems to me that certain habitual patterns can no longer be active requiring reflection, namely, a reconstitution of identity and a search for a new role. The visitations and the memory of the past person could lead to a recalibration of interpersonal relationships – not necessarily to repeat the old patterns with a new person but to integrate the past and the future.

          • My sadness is that whenever someone explains something to me by citing Melanie Klein, I feel I end up knowing less about it….

  2. Thanks for flagging up this paper Rob. You are right that “if the asymmetry in opportunity costs for grievers versus non-grievers doesn’t hold, then the rest of the argument in the paper doesn’t hold” – and the asymmetry indeed doesn’t hold.

    In fact, this was pointed out to the authors in the question session after they presented this work at HBES last year. They didn’t have a good response then. It’s disappointing that the PSPR reviewers didn’t pick up on this hole in the argument.

  3. I think the theory could work if it were paired with the tooby/cosmides model. The more one values an individual, the more that individual’s imagined death will function as a deterrent, and the more that individual’s actual death will be felt as grief. Thus, grief could function as a signal for how intensely one valued the deceased, and — to the extent that one is more likely to have valued the deceased if one is a prosocial person — one’s prosocial tendencies. Such a signal would be hard to fake, because the less one valued the individual, the more maladaptive it would have been to have used their imagined death as a deterrent.

  4. The grieving process is like a recalibration of one’s social role and identity towards the person who has died and towards others who are in the network of related individuals and the general public as potential new relationships. For example, the repetition of grave visiting reviews the relation to the deceased and restructures the relationship to the relevant others, and it also structures your own role towards the public because your social role has shifted and is evaluated by others which is modified by the grieving behaviors, and by the cognitive restructuring of your own identity. Although there is a difference between the social status of a widow and widower, or the parent/sibling of a deceased child, the aftermath of the death of as person includes a shifting of the social hierarchy and the anticipated future relationships. But, these possible relationships require an adjustment of one’s personal identity which has incorporated and integrated the deep memories of the other which provides an openness towards the future.

  5. What really, is the opportunity cost of a weekly, two mile walk–or of other manifestations of grief? If the widow’s walk was done for exercise instead of out of mournful duty, the widow would be applauded for taking care of herself! Is there a presumption here that she neglecting her contribution to the well being of her offspring while she’s walking (that is, grieving?) What if the solicitousness of her practice in fact inoculates her from other costs that may accrue when a spouse dies, and that this inoculation actually benefits her fitness? If you want me to be interested, show me that grieving behavior is correlated with decreased fitness…or anything….

    • I like the idea of grief serving “to recalibrate the behavior of social others who might invest in you” because “events that induce grief … might tend to leave the griever in a weakened social position.” But how does that fit with the grief that a parent may feel upon the loss of a young child? No-one was dependent on that child (in fact the child was using up resources), no weakening of social position occurs but there is undoubtedly massive grief.

  6. Pingback: Hacia una taxonomía de las emociones (XI) » Yo Evoluciono

  7. I have a different theory, that some emotions including grief serve an altruistic function, namely to flag up events of fitness relevance, for others to learn from. If I bash my thumb with a hammer, I yell “ouch” to say “Pay attention! If you hit yourself with a hammer, you will pay a fitness cost.” I yell with joy to communicate “Pay attention to the sequence of events I caused because they resulted in a fitness gain.” Grief then is an extreme message saying to others “Pay attention to the mistakes I made (perhaps inattention or selecting a weak partner) that resulted in a massive loss.” Under this theory, grief lasts a long time to make sure that the communication is received by others, and perhaps duration is part of the encoding by which severity of fitness cost is communicated. If we evolved in tight knit family groups, then the cost of altruistic signals may be outweighed by the benefits to relations of learning the fitness lessons we bring to their attention.

    • Grief is a behavioral sequence that signals that ‘I have control of my feelings of anger and fear.’ Grief is a period of time following a loss (and I would not limit my understanding of loss only to someone’s death, but also to loss of role, identity, or self which may be precisely what is at risk following someone’s death) which contains the memories of the past and the anguish of the missing person in the future. But, most importantly, it signals that there will not be a contagion of the feeling that could result in either retaliation or self-destruction. Therefore, grief involves symbolic acts and the presence of others who are more distant. Those who have experienced the loss are watched to determine where their reactions will lead as well as to providing symbols that reflect the deepness of the feelings, e.g. the pipes, the crack of a gun-salute, or a celebration. If there is a fitness lesson, it may be in the valuation of the person who has left which reinforces the self-worth of those who are close. But, unfortunately, there are many cases where one’s self experiences a loss similar to the loss of the other person, sometimes these persons do not overcome the feeling of loss and the grieving process may take over a year or even persist for longer. I think that we need to explain death by our feelings of grief which may not be experienced and may be denied which lessens our value of life. I wonder now whether the experience of loss is related to stunningly high rates of accidents and suicides, and deaths by murder and during wars or genocides. Although grief is a behavioral sequence that communicates stability to others, it is also a psychological experience of images and emotions.

Skip to toolbar