The Rise of Biological Non-Zero-Sumness (A Brief History of Organic Life)

A beehive is a collaborative enterprise on far more levels than first appears.

—Matt Ridley

You may be under the impression that you have a single set of genes, arrayed along chromosomes in the nucleus of each cell. A common misconception. The nucleus is just one of many little subcellular bodies called organelles. And one kind of organelle—the mitochondrion, which processes energy—has its own genes, passed down separately from the genes in your nucleus. Whereas your nuclear genes were drawn equally from your mother and father, all your mitochondrial genes came from your mother. And if you are male, you have no chance of passing them to future generations.

Why on earth would each cell have two sets of DNA? The answer, now generally acknowledged after decades of resistance, is this: Once upon a time, before there were multicelled organisms, the distant ancestor of your mitochondria was a free-living, self-sufficient cell, something like a simple bacterium; and the distant ancestor of the cells the mitochondria now inhabit—your nucleated cells—was also a free-living, self-sufficient cell. Then the two free-living cells merged; the mitochondrion specialized in processing energy, and the larger cell handled other matters, such as locomotion. The two lived happily ever after in blissful division of labor.

I’m such a romantic—always stressing mutual benefit. Some biologists would say that the story as I’ve rendered it glosses over ugly details. According to the Nobel laureate Renato Dulbecco, the once-autonomous mitochondria are now "subservient to the needs of the cells in which they reside." According to John Maynard Smith and Eors Szathmary, authors of The Major Transitions in Evolution, mitochondria are "encapsulated slaves," subject to ruthless "metabolic exploitation."

As we’ll see, one can argue with such cynical interpretations of the mitochondria’s plight, but for now the point is just that biologists actually talk this way—as if a mitochondrion, a dinky blob presumably lacking sentience, had a plight. What do they mean? In what sense could an organelle be a "slave" that gets "exploited"? You would think, to hear this kind of talk, that a mitochondrion, like a person, has interests that are either served or not served. Is there any sense in which that’s true?

Yes, in a Darwinian sense. In Darwinian terms, living things are "designed"—by natural selection—to get their genes into subsequent generations. To serve their "interests" is to aid this genetic proliferation. To frustrate their interests—to "exploit" them, for example—is to reduce their genetic legacy.

With this vocabulary in hand, we can apply game theory to biological evolution. When two organic entities can enhance each other’s prospects for survival and reproduction, they face a non-zero-sum situation; to the extent that their interests are at odds, the dynamic is zero-sum. In this light we will see that biological evolution, like cultural evolution, can be viewed as the ongoing elaboration of non-zero-sum dynamics. From alpha to omega, from the first primordial chromosome on up to the first human beings, natural selection has smiled on the expansion of non-zero-sumness.


How did life begin? Beats me. But from the beginning, one of its driving forces was nonzero-sum logic.

Consider the two genes that linked up to form the world’s first chromosome. (I’m using "chromosome" loosely, to mean any strand of genes, not just the neatly packaged strands that evolved in the kinds of cells that constitute plants and animals.) What were the genes’ functions? Again: I have no idea (except that they must have been able to replicate). But this much is clear: by virtue of now being in the same boat, they had a highly non-zero-sum relationship. By and large, what was conducive to the survival and replication of one was conducive to the survival and replication of the other. So whatever functions they did perform—building protective insulation, say—they were best advised to cooperate, to unite in pursuit of their common goal, dividing labor synergistically.

Of course, genes don’t take advice. And they don’t "cooperate" in the sense that people cooperate—size up the other players and decide that coordination is the most sensible option. In fact, once these two genes had hooked up, there were no options. The nature of both genes was now set; either they would by their nature cooperate to accomplish more than they could accomplish alone, or they would by their nature miss this opportunity. If the latter, then their lineage would stand a poorer chance of flourishing. Unless a genetic mutation soon brightened prospects for their progeny, this particular experiment in inter-gene cooperation might well fail. The two genes that invented the world’s first chromosome, lacking a legacy, would fall into the dustbin of organic history.

That’s life: most experiments fail. Still, maybe the pair of genes that invented the world’s second chromosome, or its third, or whatever, would interact more constructively. Prospects would be brighter for this lineage, which could be fruitful and multiply.

It is in this blind, stumbling way that genes come to "pursue strategies" without thinking about the goal. Over time, natural selection preserves those genes that just happen to do a good job of playing the game.

And when genes are on the same chromosome, cooperation is generally the name of the game. The maturation of a frog is a construction project of such intricate harmony that by comparison the building of a modern skyscraper seems crude and chaotic. Ditto—only more so—for a bear. To say that more and more complex organisms have evolved over time—as they have—is to say that genes have over time gotten involved in more vast and elaborate non-zero-sum interactions. From bacteria to people, biological evolution moved in fundamentally the same direction that cultural evolution has moved in.

Is the story of life really so simple? Just more and more non-zero-sumness piling up? No. For one thing, there are a few fairly daunting evolutionary thresholds that had to get crossed somewhere between a well-designed bacterium and Homo sapiens. But, conveniently, crossing these thresholds tends to depend on harnessing non-zero-sum logic, a task at which natural selection manifestly excels.


One of the greatest thresholds was the coming of the "eukaryotic" cell. For much of the early history of life, the most complex being was the prokaryotic cell, which persists today in such forms as bacteria. Prokaryotes are a bit slovenly. For example, their DNA is bloblike, constituting a chromosome only in the loose sense. The eukaryotic cell, which went on to be the building block for plants and animals, is more tidy and bureaucratic. Its DNA is arrayed neatly along distinct chromosomes and housed in a nucleus, whence it issues commands that are shuttled forth. The eukaryote has much division of labor, thanks to its many organelles. These include the mitochondria, mentioned above; and, in plants, the green bodies called chloroplasts, which handle photosynthesis and which, like the mitochondria, descended from a free-living ancestor that fatefully merged with another cell.

Back when many biologists doubted the autonomous origins of mitochondria and chloroplasts, the foremost proponent of the scenario was the biologist Lynn Margulis. Margulis contends that various other organelles, too, had free-living ancestors. The nucleus itself may even be an example. Most mainstream biologists doubt that Margulis is right, but then again, not so long ago they were doubting her story about mitochondria and chloroplasts.

Why all the mainstream skepticism? Margulis, for one, believes that biologists (who tend to be male) have a bias in favor of competition, and against cooperation, as the formative force in evolution. She might plausibly use the scientists quoted above, talking about "exploitation" and "subservience," as Exhibit A in her indictment. Though they accept her thesis that mitochondria came about by merger, they still insist that mitochondria are brutally subjugated by dominant partners. Or, to put their claim in the language of biology: they accept that two distinct entities came together through "symbiosis"—which just means "living together"—but insist that the symbiosis is parasitic, not "mutualistic." Or, to put the claim in the language of game theory: yes, it all started with a relationship between two sovereign beings, but the central dynamic has been zero-sum, not non-zero-sum.

Are these stereotypically male biologists right? Not demonstrably. Consider Dulbecco’s contention that the "needs" of mitochondria are "subservient" to the needs of the larger eukaryotic cell. It rests entirely on his observation that the actions of the mitochondria are governed mostly by genes in the larger cell’s nucleus; the big shots in the nucleus give the orders, and the mitochondrion obediently follows them—the very definition of servitude.

It’s true that many instructions governing mitochondria issue from the nucleus. Indeed, as other biologists have noted, some genes that were initially in the mitochondrion, and controlled it from there, seem to have migrated to the nucleus, where they exercise remote control. But, as some of those biologists have also noted, this transfer may have been favored by natural selection because it raised the efficiency of the overall cell. If so, then the transfer brightened prospects for the DNA remaining in the mitochondrion as much as for the DNA in the nucleus—since the overall cell is, after all, the boat in which both kinds of DNA find themselves.

Dulbecco is anthropomorphizing mitochondria. Human beings like autonomy, and often resist control. But there’s no evidence that mitochondria have a strong opinion about autonomy one way or the other. The only sense in which they can be said to have "needs" or "interests" is in a Darwinian sense. And in Darwinian terms, they are just as interested in the efficiency of the larger cell as is the nuclear DNA. For both of them, the larger cell is home. The two are locked into a highly non-zero-sum relationship.

The jaundiced view of Maynard Smith and Szathmary also has anthropomorphic overtones, but of a different sort. To back up their view of mitochondria as slaves, they theorize that long ago, at the beginning of the symbiosis, nuclear cells kept mitochondria around "as humans keep pigs: for controlled exploitation." That is: they would let the mitochondria reproduce in captivity, then eat a few, then let them reproduce some more. In this scenario, the cell’s current handling of mitochondria—inserting "tapping proteins" to extract energy, rather than eating the whole mitochondrion—is just a higher-tech version of the original enslavement; it is a "more elaborate metabolic exploitation."

But, even granting that mitochondria started out as the subcellular equivalent of pigs, would this really be exploitation? Don’t get me wrong. I’d rather be a person than a pig, and I do believe that pigs get the raw end of the pig-person relationship. But that’s because when I think about human benefit—and, in a way, even when I think about pig benefit—I think about happiness; and in both species (presumably) a certain amount of freedom furthers happiness. But when we talk about nuclei and organelles, we’re talking only about Darwinian benefit, about genetic proliferation. And in Darwinian terms, domesticated species do very well, thank you. There is today a lot more pig DNA around than its undomesticated kin, wild boar DNA. In that sense—in the Darwinian sense— getting eaten is the best thing that ever happened to pork. Analogously, the "controlled exploitation" of those proto-organelles may well have boosted their legacy, in which case it wasn’t exploitation. Certainly they have lots of descendants today—billions in every sizable animal on earth.

In harping on intra-cellular non-zero-sumness, I don’t mean to say that the interests of organelle DNA and nuclear DNA entirely coincide. Though the two spend most of their life in the same boat, they do take separate boats to the next generation. Since a mitochondrion’s DNA is passed down only via mothers, its Darwinian interests might be served by biasing reproduction in favor of females, so that daughters were the norm and sons the exception. Even if this sexual imbalance cut down a bit on the reproductive success of the overall organism, the trade-off could still be worthwhile from the mitochondrion’s point of view. But the nucleus would take a different view, since it gains nothing by a surplus of daughters.

Hypothetical as this logic may sound, it has actually found incarnation—in plants at least. In various plant species, mitochondria have genes that cause the (male) pollen to abort, biasing reproduction in favor of (female) seeds. That this works against the nuclear DNA’s interests is evident in the countermeasures it takes. In some cases, nuclear "restorer" genes have evolved to neutralize the bias by boosting the supply of pollen.

This is a reminder that non-zero-sum relationships almost always have their natural tension, their purely zero-sum dimension. It is also a reminder that biologists such as Dulbecco are not wrong to say that tension can exist between mitochondria and nuclei. But they are sometimes wrong in what they see as evidence of tension, and in leaping to the conclusion that tension is pervasive, when in fact it is a small part of the overall picture.’

At the very beginning of the mitochondrion-nucleus relationship, to be sure, zero-sumness was probably the main story line. Apparently the proto-mitochondrion first got inside the cell in an act of exploitation gone awry. Either the big cell tried to eat the little cell and failed to digest it, or the little cell invaded the big cell and failed to kill it. But these ignoble origins of a non-zero-sum relationship shouldn’t surprise us. The general theme of this topic is that non-zero-sumness tends, via both biological and cultural evolution, to emerge. And it has often emerged among entities—villages, cities, states— whose relationship had once been overwhelmingly zero-sum. Even if the relationship between mitochondria and the larger cell was initially full of strife and bitter recrimination, today they are locked into an essentially non-zero-sum game, and both play the game well, to their mutual benefit.

The underlying reason that non-zero-sum games wind up being played well is the same in biological evolution as in cultural evolution. Whether you are a bunch of genes or a bunch of memes, if you’re all in the same boat you’ll tend to perish unless you are conducive to productive coordination. For genes, the boat tends to be a cell or a multicelled organism or occasionally, as we’ll see shortly, a looser grouping, such as a family; for memes, the boat is often a larger social group—a village, a chiefdom, a state, a religious denomination, Boy Scouts of America, whatever. Genetic evolution thus tends to create smoothly integrated organisms, and cultural evolution tends to create smoothly integrated groups of organisms.


In crossing the next great threshold after the eukaryotic cell—the chasm between single cells and multicellular life—natural selection again followed non-zero-sum logic, but this time the logic had a different foundation.

Many single-celled forms of life reproduce without sex—clonally. A cell just splits in two, creating a carbon copy of itself. One consequence is that adjacent cells often have exactly the same Darwinian interest, a fact that makes their merger into a single multicelled organism more feasible. If this makes immediate sense to you, congratulations. If it doesn’t, don’t feel bad; evolutionary biologists didn’t quite fathom the logic until the 1960s, more than a century after On the Origin of Species, when William Hamilton authored the theory of kin selection.

What is in an organic entity’s "Darwinian interest," remember, is what aids the transmission of its genetic information. So if a bunch of cells have exactly the same genetic information, their Darwinian interests are by definition identical. Suppose, for example, that two cells face starvation, but cell A can somehow save cell B by committing suicide. The net effect of cell A’s "sacrifice" is to raise the chances that its own genetic information will reach the next generation—since, after all, its own genetic information is being carried by cell B. By Darwinian math, a cell has as large a stake in a clone’s welfare as in its own welfare.

Of course, cells don’t do math, or conscious calculation of any other sort. So how might a cell come to pursue its Darwinian interest in this circuitous manner—via "kin-selected altruism"? Imagine a mutant gene that just happens to incline a cell to make some sacrifice on behalf of nearby cells. After a few rounds of cell division, a cell containing this gene is surrounded by cells that also contain the gene. So when the occasion arises for this "altruistic" gene to spring into action on behalf of an imperiled nearby cell, the gene is actually rescuing a copy of itself. Sure, the gene stands a chance of perishing in the course of such heroic acts (that’s why they call them "altruistic"); but if, on balance, more copies of the gene are saved than lost, the gene can spread through the population (that’s why they put "altruistic" in quotes). This gene will outreproduce an alternative, risk-averse gene that would stand idly by while nearby copies of itself perished en masse.

Consider a real-life example of altruism at the cellular level: the cellular slime mold. The slime mold is reminiscent of that old TV commercial about whether Certs is a candy mint or a breath mint. (Both—"it’s two, two, two mints in one.") Is a slime mold a society of cells or a single multicelled organism? Both—or, if you prefer, neither; it sits at the boundary between society and organism, never making a firm commitment. Its cells spend lots of time on their own, scooting along the forest floor looking for nutrients, occasionally reproducing by splitting into two. But when food grows scarce, the first cells to feel the shortage emit a kind of alarm call in the form of a chemical called acrasin. Other cells respond to the call, and a transformation ensues. The cells bunch up together, form a tiny slug, and start crawling as one.

Finally, having reached a propitious spot, they set about to create a new generation of slime mold cells. The slug stands up on end andturns into a "fruiting body" that features a sharp division of cellular labor. While some cells—depending on where they happened to be in the sluggish aggregation—become bricks in a sturdy stalk, other cells—again, depending on where they find themselves—become spores, designated for reproduction. The spores rise to the top, for widespread dispersal, and are launched into space to carry the slime mold legacy into posterity. The stalk cells, having spent their last full measure of devotion, now die. They have "sacrificed" their own reproductive prospects for those of their neighbors. But the "sacrifice" isn’t real. The stalk cells stand a very good chance of being genetically identical to their next-door neighbors, so they have a strong Darwinian stake in the dispersion of the spores.

Once again, two organic entities—the cells that turned into bricks and the cells that turned into spores—cooperate because they are in the same boat. But in this case the "same boat" doesn’t mean "the same physical vehicle," as it does for genes that share a chromosome. After all, for much of the cells’ life cycle they are on their own, free to reproduce autonomously; that they "choose" to cooperate in constructing a common vehicle suggests some prior non-zero-sum logic, some sense in which they were already in the same boat. And they were. They were in the same boat by virtue of having a common Darwinian interest—getting the same set of genetic information sent to the next generation.

The slime mold’s aggregation seems so magical that back in the 1930s, the German biologist who first filmed it described it in Bergsonesque terms—as the result of an immaterial "vital force." But in fact the aggregation is orchestrated concretely—by the acrasin that cells emit when it is time to aggregate.

This sending of signals is of course fairly standard procedure in non-zero-sum relationships. In the case of humans the signals may be light waves conveying gestures, or sound waves that carry words, or patterns of ink, or electronic blips. In the case of intra-cellular communication, the signals are often proteins. For example, the nucleus governs the cell’s energy processing by having proteins sent to the mitochondrion. But at all levels of organization, when entities coordinate their behavior to mutual benefit, information tends to be processed.

Think back to the E. coli bacterium featured in the last topic, the one that builds a propulsive tail to escape low-carbon environs. This deft maneuver is actually a cooperative venture between two kinds of genes on the bacterial DNA: genes responsible for ensing the carbon shortage, and genes responsible for building the tail. Such cooperation naturally calls for communication between the sensing genes and the building genes. And communication there is, involving a "symbol," the cyclic AMP molecule that, when carbon grows scarce, is sent to genes that initiate the tail-building.

As we saw in part I, communicating—breaching the "information barrier"—is just one of two things that generally have to happen for positive sums to ensue; there is also the "trust" barrier to overcome; the threat of "cheating" must be dampened. At least, that’s the case in non-zero-sum games among human beings. But presumably "trust" isn’t a big problem among cells. Right?

Wrong—in a certain sense, at least. Biological evolution, like cultural evolution, creates opportunities for cheater—pirates and scoundrels who would parasitically subvert the greater good if left to their own devices. So natural selection designs technologies of "trust"—anti-cheating mechanisms.

For example, we’ve seen that the cells in your body get along famously in large part because they are genetically identical. But suppose that, while your cells are dividing after birth, a mutation happens. A new, genetically distinct type of cell is born. Rather than focus on serving the needs of the larger organism, it replicates itself manically. By the time you are old enough to reproduce, there are so many of these mutant cells that they stand a much better than average chance of getting their DNA into the next generation.

In theory, such a cell would be favored by natural election—at least, in the short run. But in real life, this sort of parasitism couldn’t happen. The reason is that back when you were very, very, very young, your "germ line" was "sequestered." That is, the cells that will form your egg or sperm were put aside for safekeeping; try as some mutant skin cell might, it will never get into the next generation, no matter how prolific it is.

Why do animals thus seclude their germ lines? In the view of some biologists, it is precisely to avoid this sort of parasitic mutiny; germ-line sequestration is an anti-cheating device—the functional equivalent of a technology of trust. It presumably evolved as organisms that lacked such a technology died out, their coherence compromised by rampant cheating—rather as human cultures are extinguished if they fail to discourage the sorts of parasitism to which humans are prone.


When our own lineage crossed the blurry line on which the slime mold sits—the line between society and organism—the result probably didn’t look much like a slime mold. (For one thing, our incipiently multicellular ancestor probably floated rather than crawled, since it seems to have evolved in the sea. Ever wonder why there’s so much salt in your body?) Still, the evolutionary logic behind the harmony among our cells is the same as the logic behind the harmony among slime mold cells: the non-zero-sum logic of kin selection, empowered by the fact that the cells in our bodies are clones of one another. This clonal relationship is the reason that your kin cells issue not the faintest protest while drying up and blowing away. So far as they are concerned, the safety of your sperm or egg cells, as the case may be, is a cause worth dying for.

This logic has encouraged harmony not just within multicellular organisms but among them. The reason is that kin selection can operate, in diluted form, among organic entitities that aren’t genetically identical. Full human siblings, for example, while not haring all of their genes, do have lots of genes in common. So some altruism makes Darwinian sense among siblings—it is just a more measured altruism than among clones. You might take a chance on behalf of a sibling—even run into a burning building to save him—but you don’t normally act as if his welfare were just as important as your own. The reason is that you are related to siblings by a factor of 50 percent, not 100 percent. In general, according to Darwinian theory, when kinselected altruism evolves, the degree of altruism will roughly reflect the degree of relatedness.

For example: Ants, thanks to a strange form of reproduction, can be related to siblings by 75 percent, and some biologists think this is the reason for their extraordinary cooperation. Most ants even—like our skin cells—forgo reproduction entirely, and instead perform some mundane task that makes life easier for their kin, such as hanging bloated from the ceiling, waiting to be tapped as a food source in time of need.

We don’t know how these bloated, dangling ants feel. Resentful that their genes forced them into this humiliating position? Delighted that they can aid cherished kin? Or do they feel anything at all? But we do know what feelings accompany kin-selected altruism in humans, and one of them is love. Indeed, kin election is the original source of love. What was almost surely the first form of love in our lineage was a mother’s love of offspring, and it is because the off-spring assuredly shares the mother’s genes that the genes bestowing love, and the altruistic behavior accompanying love, could flourish. So kin selection, in addition to impelling life through key organizational thresholds, carried it across a threshold of no mean spiritual significance as well.

Kin selection isn’t the only Darwinian dynamic that binds organisms into webs of non-zero-sumness. There is also reciprocal altruism, which can evolve among non-kin. A vampire bat, on returning from a nightly blood-sucking expedition empty-handed, may accept a donation of regurgitated blood from a close friend—and will return the favor on some future night when fortunes are reversed. Both bats benefit in the long run. Of course, they aren’t smart enough to recognize this win-win dynamic. Still, it is non-zero-sum logic that natural selection followed in programming bats to behave as if they did understand such things.

The same thing happens in other species: chimpanzees, dolphins, even people.Generosity, gratitude, and outrage over ingratitude are genetically based impulses that can steer us into mutually profitable relationships and away from unprofitable ones. There is also a genetically based impulse that sometimes keeps us in profitable relationships—affection, an affection that, in the deepest friendships, can approach outright love. If kin selection can be credited with inventing affection, reciprocal altruism gets credit for extending affection beyond kin. And what gets credit for both feats is non-zero-sumness, the logic shared by these two evolutionary dynamics. Non-zero-sumness, in addition to being the reason that organic complexity exists, and the reason communication got invented, is the reason there is love.

The evolution of reciprocal altruism, along with the prior evolution of kin-directed altruism, imbued human societies with structure long before there were corporations (or even markets) or prime minister (or even Big Men). Biological evolution pushed us a ways up the ladder of social complexity before cultural evolution acquired much momentum. And biological evolution was in turn pushed, you might say, by non-zero-sum logic.

And so it had been from the beginning, since genes on the same chromosome first cooperated. Natural selection is a blind process, but in its fumbling, bumbling way, it has wound up, time and again, happening upon and harnessing the logic of game theory. Just as cultural evolution would follow that logic from hunter-gatherer band to global civilization, genetic evolution had followed it from tiny, isolated strands of DNA to hunter-gatherer bands. This logic organized genes into little primitive cells, little primitive cells into complex, eukaryotic cells, cells into organisms, organisms into societies.

Just how strong was the impetus behind all this? How likely were the amazing feats of non-zero-sum logic? How likely was the biological evolution of complexity, of richer and richer non-zero-sumness? This is a tough question. Certainly, crossing the various thresholds described in this topic entailed solving more knotty technical problems than I’ve bothered to enumerate. Still, there is reason to believe, as we’ll see in the next topic, that the answer is "very likely." In the meanwhile, let’s briefly marvel at natural election’s unconscious ingenuity, its ability to pursue varied paths to the same end: cooperative integration among organic entities.

But let’s not overdo it. The discovery of the role that symbiosis—and mutualism—played in the evolution of eukaryotic cells has led to some dubious rhapsody about the harmony of nature. One author titled a topic on the symbiotic origin of complex cells, "Is Nature Motherly?" Well, if by "motherly" you mean "gentle," the answer is no. The harmony created by biological evolution thrives on dissonance. It is because natural selection pits fiercely competing models against one another that models exploiting non-zero-sum logic become the norm.

Consider the bottle-nosed dolphin. Male dolphins, programmed to play win-win games, team up with other male dolphins to form a coalition. The coalition then abducts a female dolphin from a competing coalition, forcibly detaining her and taking turns having sex with her. Coalitions even play non-zero-sum games with other coalitions. Coalition A helps coalition B steal a female from coalition C today, and coalition B returns the favor tomorrow.

This is quite a display of cooperation—cooperation on top of cooperation—and is yet another tribute to natural selection’s genius. But it isn’t a tribute to natural selection’s weetness, and it isn’t cause to get all mushy about Mother Nature. In biological evolution, as in cultural evolution, Kant’s "unsocial sociability" is a recurring theme

Indeed, "unsociality"—that’s a polite way of putting it, actually—is central to natural selection: organisms vie for finite resources, and the loser slip from the pages of history. And the "slipping" assumes such forms as starvation, disease, and getting devoured. Natural selection creates by discarding, and it doesn’t discard gently. Tennyson, in a poem that was a favorite of Darwin’s wife, Emma, wrote of Mother Nature: "So careful of the type she seems, / So careless of the single life." In biological as in cultural evolution, breathtaking creations come at a horrible cost.

Next post:

Previous post: