It’s been awhile since my last post - a lot of life has happened in the interim, namely having a baby. Now that baby is three months old, I’ve spent a cumulative ~180 hours breastfeeding and another 139 hours pumping milk, which has given me a lot of time to wonder about… breastfeeding. Why do I have to pump every three hours while a cow only has to be milked twice a day? How long do other apes breastfeed their young for? Why do other apes not have visible breasts unless they’re breastfeeding? This post will answer those questions and more.

How did milk evolve?

You may have heard (or not) that milk derives from sweat - that sweat glands came first, and as newborn mammals licked their parents’ skin for bonding and hydration, the young of those with more nutrients in their sweat did better. After that, localizing this nutritious sweat into smaller areas that are easier for babies to feed from was advantageous, leading to teats. This is supported by the fact that monotremes (platypuses and echidnas) don’t have nipples, instead “sweating” milk diffusely out of their bellies, which the babies lap up. Makes sense. Except when you realize that a vast majority of mammals don’t actually sweat.

Sweating to cool down is a rare trait, exhibited only by humans, horses, and camels, who all developed the ability independently. Other animals have to pant, wallow, or find shade. This includes pigs; the phrase “sweating like a pig” apparently derives from pig iron, which would gather condensation on it as it cooled, rather than actual pigs. So if sweating is a recent development, how did milk arise?

It turns out that sweating to cool down is a recent development, but sweating to be smelly is older than mammals. Humans have two different types of sweat glands: eccrine glands all over the body, which excrete salt water, and apocrine glands in the armpits and groin, which excrete thicker, oily, smelly sweat. Eccrine glands derive from the older apocrine glands, and many mammals have this type of gland, which they use for scent signalling. A dog’s anal glands, which put informative pheromones onto their poo, are derived partly from apocrine glands and partly from sebaceous (skin oil) glands. Those glands, by the way, are why dogs are instantly terrified when walking into the vet: since vets routinely do stool samples, an unpleasant experience, the place reeks of dog pain and fear. These glands are also usually why dogs scoot on the carpet. Rather than trying to loosen bits of poo, which are uncommonly present because dogs’ anuses can turn inside out a little ways, keeping things contained, they’re trying to unclog their gland ducts.

Anyway, the earliest mammals back in the Triassic Period (252-201 million years ago) would have had apocrine glands, but laid eggs, like modern monotremes. Over time, as they invested more energy into caring for their young in their burrows, the young would have had the chance to lick their parents’ skin, selecting for more nutritious proto-milk, and the rest is history.

Why don’t female apes have visible breasts?

If you’ve ever been to a zoo that has primates, you’ve probably seen female chimps, gorillas, and orangutans, who appear to have flat, masculine chests. How do they breastfeed, and why do humans have breasts?

Apes grow breasts during pregnancy, the way a pregnant cat gets distended, low-hanging teats on her belly. Milk glands form under the skin, which go away when the animal is no longer lactating. Convenient.

Human women, on the other hand, have nonfunctional fatty tissue in the breast area at all times. During pregnancy, we also grow glandular tissue, making the breasts much larger during pregnancy and lactation than at other times. But in humans, breast size doesn’t correlate with milk production. You can have lots of fatty tissue making your breasts large, but not grow that much glandular tissue and therefore have low milk production, and vice versa, lots of glandular tissue and not much fatty tissue. So what’s the evolutionary purpose of the fatty tissue?

Well, for apes, the butt is a very important female secondary sexual characteristic. Since they’re quadrupedal most of the time, they often find themselves face-to-butt with the ape walking in front of them, and in tests, chimps have been able to identify which butt belongs to which face with great accuracy. When humans started walking upright, we stopped being able to easily see the butt in the course of normal social interaction. So the permanently engorged breasts took the keister’s place as the most important female secondary sexual characteristic. Some people hypothesize that the bosom is supposed to look like a false butt, but I think it’s more likely that large breasts simply signal reproductive fitness in a similar way as a round bottom does: if you have enough spare fat to have nice round squishy features, you’re more likely to be able to support the demanding caloric needs of pregnancy and nursing.

This also means that the “free the nipple” people are wrong, from an evolutionary standpoint. They argue that breasts are simply a tool for feeding infants, and should not be sexualized. But the fact that they persist outside of pregnancy and lactation means that evolutionarily, they must be objectively sexy.

Why two breasts? Most animals have them in rows.

This is likely due to great apes’ dextrous arms and small litter size. Other animals have to be able to nurse lots of young at the same time, and can’t move while nursing. Laying on the ground is a pretty vulnerable and inconvenient position to be in for hours every day, so when apes became able to carry their young around and hold them to the breast, that was an evolutionary advantage. We only have two arms, so two breasts makes sense; additional breasts would just waste calories and potentially get in the way. And we have only one or two babies at a time, so two breasts is enough to meet demand. Lemurs, which are more basal on the primate family tree, still have multiple rows of nipples and nurse their numerous young while laying on their side, which likely represents the ancestral primate condition.

Pediatricians recommend breastfeeding for a year. How long is “natural”?

Before birth control, that is, for a vast, vast majority of human history, it was common to be either pregnant or lactating all the time. This was necessary to keep up the population due to the high rate of infant mortality. If you could expect half of your children to die before sexual maturity, and maybe another one to simply fail to reproduce for whatever reason, you had to have six children just to achieve the replacement rate of two! Aren’t you glad you’re not a caveman?

Anyway, in modern hunter-gatherer societies like the San, this is still the case. Yet, they only have children once every three to four years. This is because their lifestyle requires them to walk something like 1500 miles per year, and children younger than that can’t keep up on their own and must be carried. Carrying two closely-spaced children at the same time is prohibitive, so it’s unlikely both would survive, selecting for more distantly spaced children. The way they keep from having children more often is through constant breastfeeding, as the babies are strapped to the mother throughout the day as she forages, and through food availability. The hormone prolactin suppresses ovulation, making it more difficult to conceive while lactating. And being too thin also lowers the chance of conception, so the caloric demands of nursing an infant while living in an environment with seasonal food availability also suppress fertility.

All this to say, it’s likely that hunter-gatherers breastfed their babies, supplementing with solid food, for two to three years. This feels like a long time until you consider the fact that orangutans breastfeed for three to six years, only having another infant every four to seven years. They represent the most extreme case, but chimps and gorillas also have longer periods of breastfeeding than humans do. Usually, whatever apes do, humans do but more so. But in this case, humans are less ape-like than the other great apes. What gives?

Well, evolutionarily, it’s best to have children as frequently as you can. Orangutans only have them so infrequently because having them more frequently would decrease the infants’ chance of survival, likely for similar reasons that hunter-gatherer humans wait four years between children: because it takes that long for the child to be independent enough to keep up. Orangutans and other apes live in complex societies (while orangs are solitary these days due to low population, it’s thought that their typical condition is to be social) and forage in complex ways that require a long time to master. Humans are the same way, but we’ve developed technological and social tricks to enable us to overcome some of these limitations and have children more frequently than would be expected for our developmental schedule. First, baby food: cooking and tools allow us to produce higher-quality, easier-to-digest soft mush than simply by pre-chewing it. This is a double benefit, since it both allows babies to wean earlier, letting the mother get pregnant again sooner, but also takes some of the caloric load off of the mother. In other great apes, the mother provides 100% of the calories the infant needs for years on end; in humans, she’s only on the hook for a few months until the baby can start to supplement with prepared baby foods, spreading the cost of raising the child over more individuals and reducing risk. This is also why it’s hypothesized that humans undergo menopause, yet live decades longer, something no other ape does. Some studies have found that in premodern societies, for every decade a maternal grandmother lives past fifty, two additional grandchildren survive.

Premodern agricultural humans, who show some physiological adaptations to this new lifestyle compared to the hunter-gatherers that came before, such as shorter jaws with crowded teeth and smaller brains that allowed them to live shoulder to shoulder and work nine to five, would also have been able to wean babies earlier, having children approximately every two to three years. The reason for this is twofold: first, agriculture doesn’t require everyone in the tribe to walk great distances; there are tasks nursing mothers can do that are sedentary, allowing them to watch an infant in a playpen or similar rather than physically carrying them at all times. And second, when child labor isn’t illegal, children are a net economic benefit to the family. They can output more useful work than they consume in calories, incentivizing farmers to have as many children as possible.

For industrialized humans, the situation has changed yet again: babies are a net economic drain, since childcare is expensive and child labor is frowned upon, so people are incentivized not to have children. Furthermore, technology has improved to the point where infant mortality is mostly not a concern, so the replacement rate doesn’t require as many pregnancies to achieve. So modern humans tend to have only a couple children, closely spaced in order to reap the benefits of economies of scale (daycare for two often costs less than double daycare for one, baby clothes can be reused, and so on). And to have babies closely spaced, you have to stop breastfeeding early. But breastfeeding has significant health benefits for the child, in terms of setting up their immune and digestive systems for success. So a year is a good balance between physical health of the child and economic health of the family.

All this to say, the “natural” length of breastfeeding varies depending on the situation we find ourselves in and has generally decreased over time as technology has improved our ability to care for young babies and control reproduction in other ways.

Why do cows only have to be milked twice a day?

Cows are the result of around eight millennia of artificial selection centering around their udders. (They were first domesticated for meat and leather around ten millennia ago, but eight thousand years ago is when we find the first evidence of cheese making, a necessary step to digest milk products for the then-lactose-intolerant humans. Lactose tolerance only became widespread around three thousand years ago, after a long history of eating cheese and yogurt.) The wild ancestor of cows, known as aurochs, would’ve had the regular two rows of nipples along the underside, rather than a large udder. However, they did have a rudimentary milk storage system, called “cisterns”, or expanded portions of the ducts, that allowed some of the stored milk to come out much more quickly than the milk kept inside the alveoli where it’s produced, increasing the efficiency of the first few gulps during a feeding. Humans selected for animals that could produce more milk and store more of it for longer periods. In all lactating animals, milk production level is governed by demand: the more milk is removed, the more the body will produce. Cows are the same way. If you milk a cow more often than twice a day, the animal will produce more milk in total than it would otherwise, and indeed factory farms have cows hooked up to milking machines near-constantly, in order to maximize output. But in pre-industrial dairy farming, since milking every cow in a herd takes quite a long time, it’s prohibitive to milk more often than twice a day.

Conclusion

I hope this post has answered some of the questions you were wondering about and some that you weren’t when it comes to milk and breastfeeding. It’s one of the core things that define Mammalia, so it seems worthwhile to know about!