Invention and Necessity
An evolutionary theory of design
As we look back on human evolutionary history, we can’t help but notice a paradox. The capacity to manipulate our environment to provide safe and comfortable lives was always desirable, and in retrospect, we know it was never more than a few hundred generations away. That power would also have been clearly beneficial in an evolutionary sense. Any society that began to acquire it should have gained an advantage in competition for resources. Given that all it took for these innovations to come about was a little thinking and tinkering, it’s hard to understand why our ancestors had to go so long without them.
This paradox is a restatement of the central problem that the field of cultural evolution emerged to solve. If you’re familiar with cultural evolution or the history of technology more broadly, it might seem like a naive statement, and of course it is. But this naive framing captures a problem that most treatments of cultural evolution have failed to adequately conceptualize.
Invention’s Mother
Jared Diamond’s controversial classic Guns, Germs, and Steel contains one of the earliest popularizations of cultural evolution. Diamond frames his discussion of the problem and its apparent solution through the proverb “necessity is the mother of invention.” Under the traditional view of innovation,
inventions supposedly arise when a society has an unfulfilled need: some technology is widely recognized to be unsatisfactory or limiting. Would-be inventors, motivated by the prospect of money or fame, perceive the need and try to meet it. Some inventor finally comes up with a solution superior to the existing, unsatisfactory technology. Society adopts the solution if it is compatible with the society’s values and other technologies.
Since unfulfilled needs were never in short supply, the bottleneck in the flow of inventions must have been the infrequent occurrence of genius inventors. But Diamond quickly steps in to correct this misconception. In fact, he explains, most technologies were developed in spite of the aspirations of their inventors. They were products of idle curiosity, or the attempt to solve a distinct, unrelated problem. Only after they existed, and someone else found a problem they happened to be able to solve, did they become necessary. Diamond tells us that this inverted proverb, “invention is often the mother of necessity,” captures the “usual roles of invention and need” (emphasis added).
As an alternative to the genius-inventor story, Diamond outlines a preliminary sketch of what would become the cumulative cultural evolution or “collective brain” model. Most cultural change occurs through playful tinkering and trial and error. Complex solutions emerge through the gradual accumulation of simple modifications. Once they exist, they may be copied and spread, but even this is not guaranteed. Technology that seems clearly superior to us can be ignored or even abandoned. Humans, in Diamond’s depiction, are better modeled as nodes in a network of gradually accumulating, evolving ideas than as passive users receiving and transmitting the insights gleaned by a few Great Man inventors.
For the record, I think that idea is not just correct but fundamental to a true understanding of human history. It poses an elegant solution to the paradox I posed above. It does so in a way that describes many finer details of the history of technology, including the Industrial Revolution and collective forgetting in isolated, low-population societies. Its key insight, that culture is a direct, unbroken extension of Darwinian evolution, finally broke the grueling deadlock between evolutionary psychology and constructivist anthropology.
Dethroning the Designer
With that insight in mind, all sorts of productive analogies fall into place. Tools developed to study the phylogeny of living organisms can be fruitfully applied to everything from songs and folktales to trumpets and looms. Many of the things about culture that feel counterintuitive to us as individuals are commonplace in evolutionary biology. Of course most innovations are repurposed for uses beyond their original conception. Of course most small changes are made for arbitrary reasons. How else could genes design anything?
The next book to popularize cumulative cultural evolution, Joe Henrich’s Secret of Our Success, provides a set of contrasting case studies that illustrate the primacy of this “design without a designer” in human culture.
Chapter 3 includes a discussion of the Burke and Wills expedition. These men, British and Australian, came from an empire that had launched many successful expeditions through difficult and unfamiliar landscapes. Their society had a well-developed scientific culture and a long tradition of innovation. And yet, even given the chance to learn directly from Aboriginal foragers how to find an abundant edible plant, nardoo, they starved to death. They failed to copy the details of the processing steps necessary to detoxify the plant, so no matter how much they ate, it didn’t nourish them. Even for men from a culture already delivering history-altering innovations at a rapid pace, dire necessity failed to birth invention.
Conversely, in Chapter 7, he describes an indigenous society in South America, the Tukanoans, who have developed a 7-step processing routine to remove cyanide from their staple crop, cassava. They didn’t develop that routine using biochemical assays, and they didn’t stick to it to avoid the slow, chronic debilitation caused by consuming low doses of cyanide. It came about through the blind operation of natural selection. They maintained it, per Henrich, simply because that was the way things were done in their society. The cultural justifications are never spelled out explicitly, but whatever they are, the Tukanoans follow them faithfully. Any individuals or communities who deviate face the business end of natural selection, eliminated without ever understanding how they went wrong.
I chose these two examples because they have become staple case studies for introducing newcomers to cultural evolution. Together, they paint a picture Henrich and others aren’t shy about stating explicitly. In a world as complex as ours, humans are simple creatures. The ultimate causal mechanisms underlying nearly any system we engage with are obscured to us. We rely entirely on cultural selection to furnish us with viable responses to every threat to our survival. It is a far “smarter” designer than we are.
By emphasizing the faith-based nature of the Tukanoan processing routine and the subtle but inexorable natural selection that enforces adherence to it, Henrich implies a very specific analogy with genetic evolution. In this picture, each human inherits a bundle of culture in the same way we inherit a bundle of genes. Then that dual cultural/genetic inheritance is expressed throughout our lives and tested against our environment. Natural selection removes the combinations that fail to satisfice. It is taken for granted that the selection environment of culture, as it is for genes, is the external world. Yes, there might be intermediate processes analogous to selfish genes, biased transmission, and recombination. But these are just details. The main show is cultural adaptation to environmental conditions.
In that framing, what humans think about their actions has no more inherent meaning than the sequence of nucleotides that codes for a certain protein. Most changes in our way of thinking are equivalent to neutral “drift.” We might experience ourselves as proactive and independent problem-solvers, creators, and designers, but our relationship to our cultural output is ultimately identical to our relationship to our genes. Darwinian evolution was forged in a war of ideas against Intelligent Design and now that we’ve unleashed it on culture, it has once again vanquished its old foe.
This conclusion makes a big impression. It has the same existential down-a-peg-knocking effect as the Copernican or Darwinian revolutions. It’s easy to understand why this angle has become the spearhead of the push to popularize cultural evolution. But there’s a glaring problem with it, an obvious discrepancy that lay audiences seize on immediately and which gestures at a significant imprecision in the theory itself. The problem, of course, is that design is obviously real. We do it all the time. In order to bear the explanatory weight we’d like to place on it, cultural evolution can’t be an alternative to design. It needs to fully and convincingly explain it.
The alternative is that people will either dismiss cultural evolution as plainly inconsistent with everything they know about human culture, or misunderstand it as a separate process operating alongside design. In Tim Harford’s gloss of Henrich’s nardoo and cassava stories for the BBC, for instance, he concludes that
Of course cultural evolution can take us only so far. Now we have the scientific method to tell us that, yes, we really do need to leave the cassava for two days, but, no, the volcano does not care about the goats.
In other words, blind evolution operated where people were still superstitious and ignorant, but don’t worry, we don’t have to rely on that brute force trial-and-error anymore. Science and engineering now light a clear path.
As long as cultural evolution leads with this kind of story, we will keep begetting this same misunderstanding. Rather than holding cultural selection up as an alternative to intentional design, we need to show how cultural evolution brought about the Tukanoan cassava processing routine AND the Manhattan Project and everything in between. And that means understanding that cultural selection is not blind.
Genes In Our Eye
Genes were until very recently the only game in town when it came to evolution. Biologists have spent decades arguing about whether they are the “unit of selection.” But even where selection at higher levels of abstraction has been entertained, genes represent the type case. We learned what evolution is by studying how these sequences of nucleotides change over time. We may never have noticed cultural evolution at all if we hadn’t first developed an intimate familiarity with the genetic system. But the weaknesses of the analogy have hindered our understanding of cultural evolution on its own terms.
Some biologists have gone so far as to question the applicability of evolution to the cultural system. They’ve come to mistake certain aspects of genetic evolution—the existence of discrete “replicators,” consistently defined boundaries between generations of “individuals,” and unbiased replication of inherited material—for definitional properties of evolution in general. The truth is that genetic evolution is merely a specific manifestation of the process, with very particular features that are by no means universal. Studying cultural evolution as it actually is, with special attention to the ways it is disanalogous to genetic evolution, will broaden our concept of evolution in general.
The most obvious disanalogy is that cultural evolution takes place within human minds. This imposes considerable selection pressures on the kinds of things that can be inherited. Ideas are only passed on if they are simple and memorable enough to be expressed, learned, and retained. But more importantly, we are not passive receivers of the culture around us. In order to be inherited, an idea needs to be subjectively appealing to us. And after we’ve received it, we actively modify it for our own ends. This insight, called "subjective selection," is the subject of a new paper by cultural evolutionist Manvir Singh. It’s one of those notions that, like evolution itself, feels entirely obvious in hindsight, but which no one had quite managed to express before.
Subjective selection isn’t a bias in the otherwise blind transmission of cultural variants. Subjectivity is the native domain of culture. The vast majority of cultural variants have no direct relevance to our survival and reproduction. Nonetheless, they are under furious selection simply to keep our attention. That insight opens a fascinating frontier in the study of cultural evolution in “nonadaptive” or “maladaptive” belief systems, from shamanism and con artists to clickbait and self-deception.
It also reconciles cultural evolution with design. What is design, if not subjective selection acting on possible solutions to a consciously framed problem? The dichotomy between design and evolution was only ever apparent. Cultural evolution was acting within the minds of the Tukanoans as they decided how to handle their cassava crop and it was acting within the minds of the Manhattan Project scientists trying to weaponize nuclear fission and it is acting within my mind now as I draft and edit this piece. As the “collective brain” hypothesis argues, we are all nodes in a network of cultural variation. But we are not each just a single node. We contain multitudes. Generations of replication, mutation, and selection can take place within a single person’s mind in one day.
The Nature of Necessity
Of course, that makes the question I outlined at the beginning all the more paradoxical. If cultural evolution can produce a nigh-unlimited number of new variants every day, it shouldn’t have taken long at all for our ancestors to hit on innovations akin to the world-changing technology of the last few centuries. The path (and the outcome) would probably have been a lot stranger, with a lot more “Darwin Award” candidates along the way as wackier combinations were tried and found to be unviable. But it didn’t happen that way either. Why not?
Fortunately, subjective selection solves the problem even more decisively than it raises it. Unlike mutation or sexual recombination, the cultural variants tested against the external environment aren’t random at all. They’re not just “biased” by certain elements in the human mind. They’re the outputs of a completely distinct evolutionary process. A cultural variant can only come about when a niche for it exists within the landscape of subjective selection.
This is why Diamond found necessity to be such a poor mother of invention. He was look for need in the wrong place. In order to design an innovation that meets a need, it’s not enough to recognize the need. You have to start with some concept of what might address it. Then you need to have access to pieces that can be assembled to make that concept a reality. And, most importantly, you have to have some way to test each iteration of your design against external reality. None of those steps can be taken for granted. Every single one has to be acquired as an evolved cultural variant.
Subjective selection actually provides a fairly direct synthesis of Diamond’s opposing hypotheses. The traditional view is right to posit that, in order for an invention to come into being, we need a consciously framed problem, some incentive to solve it, inventors with a repertoire of relevant components and the ability to assemble them, and a society prepared to adopt it. But as in the “collective brain” view, there’s no reason that more than a few of those elements need to coexist within a single individual. As long as they’re all present in a single network, they can find their way together.
Ultimately, I think the clash between evolution and design was always about ego. Design happens within our heads, and we consciously create each iteration using the discernment of ideas we collectively identify with as our selves. It’s natural that we’re proudly possessive of our creations. There’s a sense of loss in accepting that we are also the vehicle of this mindless natural process operating across millenia. Fortunately, I don’t think these perspectives have to remain at odds.
Individual human creators are still irreplaceable. We just need a little change of emphasis in what we value about them. Personal hyper-intelligence and creativity are key ingredients in someone who can make a massive contribution to our collective cultural evolution. But what makes someone truly valuable is the ability to engage with the whole multi-generational network of human culture. Maybe our vision of the genius will evolve to lionize that particular kind of virtuosity.