The first weeks: The psychotic casserole

I've been a father for a couple weeks and already I could use a vacation. Friends, no matter what you've heard, parenthood is not all sunbeams and rainbows. There are also rainclouds.

Of course rainclouds have their silver linings. Tornados, on the other hand, do not have linings because no material can withstand the 300 mph winds. Parenthood this week was like a trailer park in Kansas.

Babies have only one way to communicate, and luckily only a few needs: a nap, a feeding, a burp, a diaper change, the Meaning of Life. The last one is tricky, because they need it NOW! You tell them "42," but they have no sense of humor. They do not accept that it's an ill-posed question. They will only cry. And cry. A lot.

The baby books say some amount of crying for no reason is normal. Expect an hour a day, they say. We get more than that, but an hour of pointless crying is still a lot. It's unpleasant, even when it's your own baby. It's frustrating, especially when it's your own baby. Why should this happen for no reason? It's wasted energy (13% more metabolic energy compared to not crying), which babies need--they lose weight after birth and then gain back ~15% of their body weight in a week. For me that would be 25 lbs! But the crying seems worse than pointless because of everyone else within earshot, whose nerves are getting shredded along with the baby's vocal chords. Is the wanton spreading of bad will really the best strategy for someone utterly helpless and dependent on others?

Being a scientist, I looked into this.

People have indeed studied infant crying, but not as much as you might think. Search for "colic and crying" on Web of Science and there's around 100 papers. For comparison, "black holes" has 12,000! Since both black holes and crying babies are singularities that suck in all nearby resources you would think scientists would focus on the ones in people's living rooms before looking 1,600 light years away--wouldn't you?

Anyway, before getting into pointless crying, let's mention the theories for regular (purposeful, consolable) crying. To my knowledge there are only two that are undisputed. One is common sense: that crying signals a discomfort or need, like hunger. The other is that crying is to maintain parental proximity, which is again straightforward, but perhaps less appreciated because the dangers of being left in the savannah, where we did most of our recent evolution (and where the baby is "expecting" to be), are more severe than in the nursery (SIDS is nothing compared to Sudden Infant Lion Food Syndrome). Also, being held is useful for thermo-regulation, which is under-developed in humans and other mammals where the young are carried. Crying that is "pointless" (or "excessive" or "inconsolable") is in relation to these two reasons. There is no apparent need (e.g. the baby refuses to feed) and it continues when the baby is held and soothed. There are other possibilities however.

There is a third theory for crying: that it signals vigor, to encourage parents to give their full support. In this case the purpose could be easily missed. In abundant societies--what we're used to--parents almost always give full support to all their children, but not so in harsher environments. With survival questionable, resources might be better used on a more vigorous child, or on the parents to get them through to a more prosperous time. Throughout most of history, even outright infanticide was a relatively common option. (Today there's plenty of food and adoption agencies, so infanticide is a heinous crime--even on a long transcontinental flight!) Less extreme, parents might just give less care to a baby that's less vigorous (cries less). Supporting this, a study of preterm twins found that the healthier twin received more attention from their mothers, particularly when crying, though without the mothers noticing the discrepancy. 

However, parents want to have vigorous children, but they don't want excessive crying. I think some amount of crying does signal vigor--it's more costly for sick children to cry, and some very sick children do not cry at all--but when the signal is perceived negatively (as irritating or frustrating) then I don't think it's likely to be entirely a signal for something unambiguously positive.

The conflict above suggests yet another theory for crying: that it's manipulation. In the parenting books, you will find no allowance for manipulation--in fact some explicitly state that the baby is incapable of it. (They likely mean conscious manipulation, but this is a false distinction, as consciousness is not required for the theory.)

In fact, we have good reason to expect some amount of manipulation, because of a theory put forth by Robert Trivers in the '70s called parent-offspring conflict theory. It begins with the observation that in sexually reproducing species, like humans, parents share half of their genes with each of their offspring, while offspring share all of their genes with themselves (obviously). Thus the interests of parents and offspring overlap but are not identical; some conflict is expected. And with conflict comes the possibility of manipulation. There is a fair amount of evidence supporting Trivers' theory in birds, insects, plants, and humans, as well as in pregnancy and in the genome itself, all of which is touched on here.

But if the crying really is pointless, then it can't be manipulation. There has to be some goal. Generally, babies that cry more receive more attention, but why continue crying when they're being held and offered food? It would make more sense to cry only when let down, say. However, all babies, even those with colic, have the same number of crying bouts; it's the duration that varies. The same is true of the traditional hunter-gatherer societies that have been studied (they typically have lower amounts of crying than in the West). Other universal features of crying is that it intensifies from birth until the sixth week and then decreases (the so-called n-shaped curve); and each day it peaks in the early evening.

Maybe inconsolable crying is caused by discomfort that the parents aren't aware of, or can't do anything about? Digestive problems is the usual hypothesis--indeed "colic" derives from the Greek word for colon. However, studies suggest that in, at most, 5% of cases is colic (using the "Wessel" definition: >3 hours of crying, for >3 days/week, for >3 weeks) caused by some organic problem, usually cow's milk protein intolerance, but also acid reflux or indigestion. There is no apparent difference between breast fed and formula fed babies. Recently, an intriguing study did find that probiotic supplements for breast-fed babies significantly reduced the time (by 2/3) of crying compared to simethicone (the active ingredient in over-the-counter anti-gas medication, which in other studies worked no better than placebo). This suggests that digestive issues do contribute to colic, at least in some cases, though it should be mentioned that the study was not blind, so parents--who also made the crying assessment--knew which treatment they were getting. Still, even babies with this treatment cried on average for an hour a day. Also, studies (e.g. this one) find that cortisol levels--a sign of pain and stress--are not elevated in babies with inconsolable crying, further arguing against gas pains or other discomfort.

The universal patterns of crying (the n-shaped curve, the evening clustering) suggests that inconsolable crying might be a feature of neurological development. Chimpanzees, which are our closest evolutionary relative, have similar patterns in their crying, however their crying is always consolable. Perhaps human babies' relative immaturity at birth (neoteny) has made this development more fragile? The behavior change at 3 months, when inconsolable crying largely stops, is correlated with physiological changes in the brain (increased cortical-subcortical neural connections, an increase in cortical control of subcortical activity), and changes in the neurons themselves (myelination). Other authors (Jenni) have argued that the patterns of crying may be due to missteps while establishing the alignment of the circadian rhythms. Colicky babies seem to sleep about 2 hours less, for instance, and their nighttime sleeping is disturbed even after controlling for the additional crying (Ref. here).

It's possible that inconsolable crying is merely a side-effect of development and neoteny. However, I find this explanation suspect--inconsolable crying causes too much emotion in parents (and others). It's the leading complaint to pediatricians. It's also the number one triggering cause for abuse and infanticide. Plus, as mentioned it's energetically costly at a time when energy is at a premium (presumably this is one reason babies sleep so much). The above changes are of course relevant, but they're proximate causes; I'm interested in ultimate causes. Lastly, I think the n-shaped curve similarity with chimpanzee cries--which are consolable--suggests the details in inconsolable crying in humans is probably not highly optimized. My guess is that one "knob" related to crying was adjusted (perhaps the off knob), which may or may not have been a side effect (e.g. of neoteny). It then remained that way, or underwent only minor modification, because of the ultimate causes considered here.

With that in mind, I will now turn to my own idea, which I call the hazing hypothesis. I propose that inconsolable crying may be inconsolable and irritating on purpose, to make the parents value the child more afterwards, using the same logic as fraternities that haze. The logic there, in turn, is the "effort heuristic," which is the tendency to attribute more value to outcomes that required more effort. Whether or not this heuristic, or heuristics in general, are adaptive shortcuts, or outright flaws is debatable and interesting (here is an interesting discussion on the rationality of heuristics in general), but I won't get into it here. 

I put forth this hypothesis because it seems plausible and yet to my knowledge has not been considered. Note that it's an extension of the manipulative signal hypothesis. Anyway, one implication of the hazing hypothesis is that after the hazing period parents will value their child more if they cry more. So here's a testable hypothesis: Fathers present during the 6 week hazing period and who subsequently leave the family (e.g. through divorce) will devote more resources (time, money) to offspring that cried more excessively. I would think the hazing would be targeted more at the fathers since they have invested less into the offspring than the mother and have more to gain by leaving. (I chose divorced fathers because their resource expenditures would be easiest to quantify.) Plus, the generally difficult and painful delivery is already a form of hazing for the mother. Another hypothesis might be that infants cry more with a male caregiver present, assuming the strategy is responsive to the environment (weak evidence against this here). Of course, it's possible that the crying itself drives the father away, or worse, and certainly compared to a fraternity where potential pledges are abundant this would be much worse for the baby.

But perhaps it's not worse for the baby's genes? As Richard Dawkins has famously pointed out, evolution acts on genes, not organisms. And the baby shares half its genes with its siblings. If the parents can't handle the hazing then perhaps they don't have what it takes to raise the baby anyway, and it would be better for the genes to have the parents focus on those genes in the baby's current or future siblings. 

Though I predict the hazing to be targeted mainly at the father, I think the mother will experience an effect as well. Here we can hope to find some evidence already in the literature, since research on parenting more commonly focuses on the mother. Studies have compared children with and without colic, and their mothers, after the colic period was over. They found no significant differences in personality between the two groups of children, but they found that mothers of children with colic have more separation anxiety. This is consistent with these mothers valuing their child more, though of course it doesn't imply it.

Here's a final possibility: Maybe inconsolable crying is meant to make us anxious. Anxiety is generally seen as a negative quality--and in an extreme form it is--but in moderate amounts, in the right situation, it is almost certainly adaptive. A baby is a fragile beast. Perhaps you, and people around you (which in tribal days were largely your extended family), should be anxious. You are not on vacation. In fact you're still in Kansas.

---

Note: One review article (by Joseph Soltis) was particularly helpful in my research. It included responses from a dozen or so researchers and then another response from the reviewer. It's around 50 pages total, so I have left some things out, for instance the role of the cry tonal frequency.