From Humanoids to Heptapods: The Evolution of Extraterrestrials in Science Fiction

Guest Blogger Russell Powell writes...

Troodon, a genus of Cretaceous therapods, with a Troodontid: their counterfactual, human-like descendants, imagined by Dale Russell.

Troodon, a genus of Cretaceous therapods, with a Troodontid: their counterfactual, human-like descendants, imagined by Dale Russell.

Science fiction has long been the stomping grounds of grand philosophical thought experiment—a vehicle for imagining radically different biological, social and technological worlds. By probing the boundaries of physical, morphological and psychological possibility, science fiction transcends cultural and anthropocentric biases to draw lessons about the nature of human existence and how it fits into the cosmic scheme. Contemplating the properties of intelligent life on other worlds has been a mainstay of this philosophical enterprise. Rarely, though, have these depictions broken free of our hominin and more broadly vertebrate-centered perspective on the history of animal life. In this post, I will argue that although recent depictions of extraterrestrial life in science fiction have been fruitfully informed by evolutionary theory and comparative animal cognition science, bio-essentialistic tendencies and unwarranted extrapolations from terrestrial life continue to distort popular conceptions of life on other worlds. The depiction of aliens in the recent film Arrival nicely illustrates both the theoretical advances that have been made on this front, as well as the obstacles that remain to imagining the nature of complex, and complexly cognitive, life on other worlds.

In short, I think that problems with the depiction of extraterrestrial life in science fiction are not due solely to inadequate consultation with evolutionary biological expertise, or to the ineluctable balancing of scientific plausibility with the pragmatic demands of literature and cinema; nor are they the result of straightforward limitations of the imagination. Rather, the main problem stems from our failure to distinguish contingent or accidental features of organisms from robustly replicable ones in what we might think of as complexly ‘bundled’ evolutionary outcomes. I will unpack this shortly. The key challenge is how we can properly parse contingent from robust aspects of evolutionary outcomes in the absence of an extraterrestrial data set. The most promising avenue for doing so, at present, is to examine patterns of contingency and convergence in the history of life on Earth. If this framing of the problem is correct, then the question of extraterrestrial life is intimately intertwined with the philosophical discussions of life’s history that fill the virtual pages of this esteemed blog.

The portrayal of intelligent extraterrestrial life in science fiction and in speculative philosophical imagination has changed dramatically over the last few centuries. As with all literature, the depiction of aliens and the way they interact with human beings has been shaped by prevailing cultural zeitgeists. Dominant social narratives of competition, cooperation, and degeneration have each shaped tales of extraterrestrial contact at different times and places in the history of science fiction. The development of evolutionary science in the 19th and 20th centuries, however, marked a particularly transformative turning point for thinking about life on other worlds. Prior to the advent of evolutionary theory, theorizing about extraterrestrials remained in the grip of pre-scientific cosmological and theological assumptions. In ancient and early modern literature, for instance, the intelligent inhabitants of other worlds were not conceived of as ‘alien’ beings at all, but rather as fundamentally familiar human-like beings and animals—apart from the introduction of very minimal anatomical or behavioral variations. Darwinian theory, with its focus on variation, natural selection, and adaptation to local environmental conditions, provided the first major theoretical framework for contemplating the nature of extraterrestrial life and its possibilia.

Extraterrestrial environments are likely to pose ecological design problems that do not exist on Earth, raising the specter that alien life forms will, to the extent they exist, be adapted to very different conditions and thus assume radically different forms from life as we know it on Earth. But selection to different environments is not the sole theoretical basis for the conclusion that extraterrestrial life will be truly alien. Even highly similar selective environments on Earth can, due to a series of historical contingencies such as the ordering of mutation and recombination events, the constraints of developmental patterning mechanisms, stochastic extinctions of lineages, etc., lead to dramatically different evolutionary outcomes. There is nothing in the structure of evolutionary theory which suggests that certain specific evolutionary outcomes are inevitable or even highly probable. Evolutionary theory describes mechanisms of blind variation, natural selection, and drift, but it does not specify laws of form or function. For instance, when natural selection is formulated in a general way—one that applies to all living populations of organisms that meet certain generic conditions (such as heritable variation causally related to differential reproduction)—it does not make any specific predictions about the kinds of life forms that will evolve. It can only make such predictions once all of the relevant ecological and developmental variables of a population are filled in (e.g., predation pressures, foraging problems, metabolic requirements, genetic mutations, developmental interconnections of the phenotype, etc.)—and, crucially, these details are not themselves predicted or covered by evolutionary theory.

Despite over a century-and-a-half’s worth of evolutionary science, the portrayal of extraterrestrials has been little more than a mildly naturalized version of the cross-cultural human tendency to postulate supernatural beings, or what cognitive scientists of religion call ‘minimally counterintuitive agents’ (MCAs): beings that are on the whole fundamentally familiar, but that have some properties that breach intuitive expectations for how entities in that familiar ontological domain look or behave (such as a bear that talks). Contemporary works in popular science fiction—from ET and Close Encounters of the Third Kind to the Star Trek and Star Wars universes, to (as we shall see) even more sophisticated films like Arrival—continue to be mired in the psychologically attractive but ultimately fallacious MCA paradigm: There but for the point of an ear and an interstellar spacecraft go I. Science fiction in the post-Darwinian world has widely (though not uniformly) continued to present intelligent extraterrestrials as distinctively humanoid in form. This notwithstanding reactive essays by eminent evolutionists in prominent venues with blunt titles like “The Nonprevalence of Humanoids” (by G.G. Simpson in Science in 1974).

A menagerie of fictional extra-terrestrial convergences with  Homo sapiens.

A menagerie of fictional extra-terrestrial convergences with Homo sapiens.

One reason for this is that thinking about extraterrestrial life, both in the academy and among the public at large, has been influenced more by physics than it has by evolutionary biology. And cosmologists, especially those who advocate the search for extraterrestrial life, are often readier to project Earth-like evolutionary outcomes onto other worlds. Astronomers deal with a universalizable and spatiotemporally invariant subject matter—stars, for instance, arise and evolve predictably in accordance with the laws of stellar evolution everywhere in the observable universe. Cosmologists may operate on (implicit?) assumption that evolutionary outcomes are similarly generic—that Earthly evolution is likely to be more or less representative of the types of outcomes we can expect to find on other living worlds. This mode of inference, useful in physics but problematic in biology (see below), may be partly responsible for the cosmic humanoid epidemic.

The cosmologists’ optimism has relied primarily on general statistical considerations (such as the Copernican or Mediocrity principles), or extrapolation from the familiar law-driven modes of inference and explanation that are endemic to the physical sciences—rather than any specifically biological justifications. And yet, as preeminent evolutionist Ernst Mayr complained in a letter to the journal Science, the success or failure of the search for extraterrestrial intelligence will be determined not by physical considerations but by distinctively biologically factors. And indeed, there are distinctively biological reasons to be suspicious of projecting Earth-like evolutionary outcomes onto alien histories of life. In other words, there are reasons to think that the evolutionary processes that direct the course of life are of a fundamentally different character than the processes that govern the evolution of physical systems. The fundamental difference, many biologists and philosophers of biology would contend, lies in the irreducible historicity of biology.

This deeply historical view of life is most forcefully articulated in the work of the late paleobiologist Stephen Jay Gould. In his book Wonderful Life: The Burgess Shale and the Nature of History, Gould famously poses an evolutionary thought experiment: imagine replaying the ‘tape of life’ from an early period in the history of animal evolution, and consider how its story would again unfurl. Would the replay result in a fundamentally alien biological world—a parade of forms, functions and relations that bear little resemblance to life as we know it? Or would a strikingly familiar set of designs, properties, and organizational patterns emerge? Gould argued in Wonderful Life, and throughout his career, that replaying the tape of life would result in radically different evolutionary outcomes, though he was never entirely clear about the grain of resolution at which this hypothesis applied. Due in part to Gould’s influence, philosophers of science have argued that there is a ‘nomological vacuum’ in biology—a striking absence of non-accidental, law-like generalizations at all but the most generic levels of description. Although Gould’s argument focused on animal morphology, his radical contingency thesis is particularly provocative for the implications it might have for the emergence of cognitively complex life. Is the evolution of mind a historical accident, unlikely to be repeated in hypothetical replays of the tape of life on Earth and actual replays on Earth-like planets elsewhere? Or is mind, in any or all of its variegated forms, likely to be an evolutionarily important feature of any living world?

Given that we are privy to but a single history of life, how can we even begin to approach let alone adjudicate these matters? Perhaps the strongest biological support for the cosmologist’s optimism may be found in studies of ‘convergent evolution’—the independent origination of similar biological forms and functions. Classic examples of convergence include the evolution of camera-type eyes in vertebrates, cephalopod mollusks and jumping spiders (lineages which diverged ~1 billion years ago); the dolphinoid form of porpoises (mammals) and ichthyosaurs (Mesozoic marine reptiles) (whose last common ancestor lived ~300 million years ago); and the saber-toothed lion morphology that evolved independently in placental and marsupial mammals (clades which split in the mid-Jurassic). On its face, the logic of the evidence is straightforward and compelling: convergent evolution is tantamount to natural experimental replication in the history of life, and to the extent that evolutionary replication is ubiquitous, this would seem to cast doubt on the radical contingency thesis. If this is right, then we can look to convergence for the sorts of properties that might be projected onto alien histories of life.

Now, I have argued in a series of publications that this sweeping evidential reading of convergent evolution is problematic, for reasons I will not go into here. Nevertheless, I do believe that there are some cases of convergence that indicate the evolutionary robustness of the repetitions observed, and that these, in turn, could support specific, contentful laws of form—laws that tell us, at some level of description, what extraterrestrial organisms might be like. The problem for popular science fiction is that there is no evidence of robust evolutionary convergence on humanoid forms or even anything approaching the anatomical parameters of the vertebrate clade. The sample size is big (including ‘chance setups’ throughout the entire history of life), yet these outcomes are singular. This was essentially the macroevolutionary lesson that Gould read off of the fossil record: that had the ancestor to all modern vertebrates gone extinct in the Cambrian (and there was no good functional reason why it should not have), vertebrates and all of its nested groups would never have arisen again, and the future history of life would have assumed a markedly different shape. The same is true for other large-scale animal groups. There may, however, be more generic functional properties that are evolutionarily robust in a way that would support their projection onto alien worlds. Indeed, mind may be such a robust property. Regardless, the key is to distinguish features of complex organismic outcomes that are robust in this way from organismic parameters that are essentially contingent—and it is this distinction that some of the best recent science fiction fails to draw.

In order to determine whether, for example, complex forms of cognition are evolutionarily robust, we need to resolve their phylogeny, or distribution in the tree of life. Exciting work in comparative (animal) cognition has begun to illuminate the complex cognitive mechanisms that may be at work in very distantly related groups of animals, not only within vertebrates but also between vertebrates and cephalopod mollusks and arthropods (especially insects and arachnids)—groups whose last common ancestor lived in the base of the Cambrian and in all probability lacked a centralized brain. Cephalopods—which in extant forms include octopuses, squid, cuttlefish and nautilids (and in extinct forms ammonites and belemnites), have become the poster boys of convergent cognition. Cephalopods are rightfully considered intelligent aliens on Earth—complex problem solvers with convergently evolved brains, the anatomically bizarre and functionally impenetrable architecture of which is the stuff of science fiction. Although less widely popularized, work on insect and arachnid cognition is also producing surprising and in some cases astounding results that are forcing us to rethink the nature of cognition and its neurological correlates. If these findings are borne out, they may suggest that mind is evolutionarily robust, at least given the origins of animals: even if vertebrates had gone extinct in the Cambrian, mind would still have evolved on Earth. This is a much more telling finding, from the standpoint of thinking about extraterrestrial intelligence, then, say, the fact that different mammal clades (such as primates, cetaceans, and elephants) have independently evolved high levels of intelligence—since it is not clear that any of these latter outcomes would have occurred had a certain subgroup of vertebrates (Mammalia), let alone vertebrates themselves, never arisen.

The film Arrival has clearly been influenced by findings in comparative cognition. It depicts intelligent extraterrestrials called ‘heptapods’ (Greek, ‘seven foot’) whose visit to Earth is motivationally opaque, with the communication of their intentions stymied by a linguistic barrier that (in frustratingly tropey and scientifically implausible fashion) only one reluctant human is bold, clever and creative enough to break. Apart from a single negligible modification—namely, that heptapods have seven appendages, rather than eight like octopods (octopuses) or ten like decapods (squid and cuttlefish)—heptapods are, anatomically speaking, unmistakable cephalopods. They boast several traits in the cluster of ‘apomorphies’ that delineate the cephalopod clade: they are equipped with sucker-coated tentacles, ink sacs, jet-stream propulsion, and camera-type eyes. Oh, they also fly interstellar crafts, see into the future, and inexplicably squander a great deal of metabolic resources writing complex messages to humans in their own biosynthesized ink. This alien rendering may not be a huge improvement over the MCA supernaturalism paradigm discussed above—there but for an extra tentacle and interstellar ship go Cephalopoda—but it deserves kudos for picking out genuinely alien intelligence among us as a model for the evolution of extraterrestrial intelligence, and thereby declining to provide another vector for the cosmic humanoid epidemic.

Some ink-based heptapod communication from  Arrival.

Some ink-based heptapod communication from Arrival.

For all its improvements, however, Arrival makes the very same evolutionary theoretical mistake that has been made by countless iterations of the humanoid epidemic. Namely, it falls prey to a particular species of biological essentialism: the tendency to bundle contingent, accidental traits with evolutionarily robust ones and to assume they all come as part of a single, law-like package. As I noted earlier, I think the humanoid epidemic is not due solely to promiscuous anthropomorphism or a sheer lack of imagination. It stems in part from the progressivist notion that the evolutionary process leads inexorably toward higher intelligence—and that the trait of higher intelligence is nonaccidentally connected to a series of other traits (such as bipedalism) to which it is tethered by an underlying essence. Let’s call this the ‘humanoid-intelligence bundle.’ In the case of Arrival, the recognition that cephalopods are intelligent aliens on Earth prompts a novel bundling, namely the ‘cephalopoid-intelligence bundle.’ The crucial mistake in each of these cases of essentialistic bundling is to lump evolutionarily robust features, such as complex brains and cognition and eyes, with accidental features of animal body plans, such as hominin-style bipedalism, sucker-lined tentacles, or the jointed exoskeletons of arthropods. There is simply no reason to think that the distinctive features of animal body plans are connected in any law-like way to complex forms of intelligence, such that extraterrestrial evolution will gravitate toward one or another of these essential bundles. In essence, then, what the humanoid epidemic and the incipient cephalopoid outbreak in science fiction depict is not the plausible result of genuine extraterrestrial histories of life, but rather alternative histories of life as we know it on Earth—a most worthwhile, but very different (and in many ways much easier) philosophical pursuit. Avoiding essentialized bundles increases the imaginative load substantially. But it also opens up a new space for scientific artistic expression—one that transcends the parameters of terrestrial animal body plans to combine traits in novel and surprising ways, while retaining the truly law-like features of life that we can expect to arise on any living, thinking world.


Russell Powell is Assistant Professor of Philosophy at Boston University. He works on conceptual, methodological and ethical problems in biological and biomedical science. His research in the philosophy of biology is focused on philosophical dimensions of macroevolution, as well as problems in human evolution such as in relation to morality, religion and culture. He is currently working on a book manuscript, under contract with MIT Press (as part of the Vienna Series in Theoretical Biology), tentatively titled “Contingency and Convergence in the History of Life: Why Complex Bodies and Minds Are Common but Technological Species Are Rare.” Powell is also a co-organizer of and contributor to a recent issue of the Journal of the Royal Society Interface Focus titled “Convergent Minds: The Evolution of Cognitive Complexity in Nature.” A second major thread of Powell’s research involves applications of work in philosophy of biology to bioethics and political philosophy. This approach is exemplified by his recent book “Evolution, Culture and Moral Progress: A Biocultural Theory of Moral Change,” co-written with Allen Buchanan, which is forthcoming with Oxford University Press.