Evolutionary contingency and SETI revisited

Abstract

The well-known argument against the Search for ExtraTerrestrial Intelligence (SETI) due to George Gaylord Simpson is re-analyzed almost half a century later, in the light of our improved understanding of preconditions for the emergence of life and intelligence brought about by the ongoing “astrobiological revolution”. Simpson’s argument has been enormously influential, in particular in biological circles, and it arguably fueled the most serious opposition to SETI programmes and their funding. I argue that both proponents and opponents of Simpson’s argument have occasionally mispresented its core content. Proponents often oversimplify it as just another consequence of biological contingency, thus leaving their position open to general arguments limiting the scope of contingency in evolution (such as the recent argument of Geerat Vermeij based on selection effects in the fossil record). They also tend to neglect that the argument has been presented as essentially atemporal, while referring to entities and processes that are likely to change over time; this has become even less justifiable as our astrobiological knowledge increased in recent years. Opponents have failed to see that the weaknesses in Simpson’s position could be removed by restructuring of the argument; I suggest one way of such restructuring, envisioned long ago in the fictional context by Stanislaw Lem. While no firm consensus has emerged on the validity of Simpson’s argument so far, I suggest that, contrary to the original motivation, today it is less an anti-SETI argument, and more an astrobiological research programme. In this research programme, SETI could be generalized into a platform for testing some of the deepest assumptions about evolutionary continuity and the relative role of contingency versus convergence on unprecedented spatial and temporal scales.

Introduction: Wallace, Simpson, and the irresistible charm of evolution

The central importance of evolutionary biology for answering questions about existence and properties of life elsewhere has been recognized for quite some time. Alfred Russel Wallace understood it in 1903, and almost a century in advance set the stage for the enterprise we nowadays call astrobiology (Wallace 1903; see also Gould 1987; Dick 1996; Ćirković 2012). That in itself is rather uncontroversial. Unfortunately, the strategies used in bringing evolution to the problem of life in its general cosmic context have been quite controversial, to say the least. Perhaps the most authoritative of them all, George Gaylord Simpson’s 1964 paper “The nonprevalence of humanoids,” has not been critically scrutinized enough. In that study, the great palaeontologist argued that ubiquitous biological contingency prevents the emergence of any species capable of meaningful communication with us; thus, contingency makes SETI projects unlikely to succeed (Simpson 1964). While, as we shall see, he did not argue that this would make such projects unjustified in principle (after all, they could be construed as testing the importance of biological contingency; he called them “a gamble”), subsequent authors definitely argued for a stronger conclusion.

By 1964, Simpson had been hailed as one of the key figures in the neo-Darwinian Modern Synthesis, a successor to the “holy trinity” of Fischer, Haldane, and Wright. Coming from such a distinguished figure in the highly visible Science, soon after the dawn of the Space Age and the beginning of SETI, and during a time of fierce debates provoked by the ongoing Apollo Program and heightened Cold War tensions, the paper could not fail to be an “instant classic”. Almost half a century later, it continues to exercise a strong influence upon thinking about life and intelligence elsewhere. Even supporters of SETI, such as Paul C. W. Davies, call it “the most powerful and persuasive argument of all” (Davies 1995).

Simpson had been clearly inspired by Wallace’s vision. In 1903, more than four decades after publishing a treatise proclaiming biological evolution by natural selection, Alfred Russel Wallace, wrote Man’s Place in the Universe, indicatively subtitled A Study of the Results of Scientific Research in Relation to the Unity or Plurality of Worlds. In it, the co-discoverer of evolution sketched a truly vast, cosmic vision unifying terrestrial and cosmological views and offering a provocative counterpoint to the then prevailing naively optimistic views on all-pervading life (Heffernan 1978). In a famous paragraph, Wallace prefigured much of the subsequent scepticism towards the existence of extraterrestrial life and intelligence¹:

Lastly, I submit that the whole of the evidence I have here brought together leads to the conclusion that our earth is almost certainly the only inhabited planet in our solar system; and, further, that there is no inconceivability – no improbability even – in the conception that, in order to produce a world that should be precisely adapted in every detail for the orderly development of organic life culminating in man, such a vast and complex universe as that which we know exists around us, may have been absolutely required.

However, superficial similarities can be deceptive. Those who have hurried to make Wallace a precursor of modern “rare Earth” scepticism (Ward and Brownlee 2000; Conway Morris 2003) are making two different kinds of mistakes. The first is the claim that Wallace, having understood evolution by the action of natural selection, argued on that basis for the absence of intelligent life (“mind”) on other planets. In fact, Wallace thought that if evolution were the only force acting throughout the universe, it would be very difficult to argue for the uniqueness of the terrestrial biosphere or the human mind (which was, obviously, his foremost concern).

The second mistake is assuming that Wallace’s reasoning is easily “transferable” from his tiny universe to the world of present-day cosmology. Cosmological background is, for Wallace, not a decoration—it is an important, active player. That is exactly why he was such an important precursor to astrobiology; erroneous conclusions reached through bad cosmological input cannot stay the same when the input changes. Therefore, we can hardly argue that Wallace’s claim for the Earth being unique in the universe was on the right track any more than, for example, we celebrate Thomas Wright’s conclusion that the Earth is located at the periphery of the Milky Way on the basis of morality (see, e.g., Belkora 2002). It is exactly the “non-transferable” nature of Wallace’s proto-astrobiology that makes his work and its lessons valuable for the present-day researcher. As we shall see, there is a parallel with Simpson’s treatment of the astronomical background of the study of cosmic life.

I wish to put one aspect of the record straight at the very beginning: while Simpson’s argument needs to be discussed on its own merit, some contamination from the wider debate about the justification of SETI projects is unavoidable. In the last half a century it has been very effectively used to argue against SETI projects and their funding, with a high degree of success (as documented by Dick 1996; Garber 1999). Simpson’s scepticism has been used to cut support for SETI searches, as well as to diminish the scientific status of astrobiology and SETI in many circles (cf. Crowe 1986; Ulvestad 2002; Chyba and Hand 2005; Davies 2011; Ledford 2012).

Since we have recently celebrated 50 years since the beginning of practical SETI projects (e.g., Penny 2011), and the half-centennial of Simpson’s paper will be soon upon us, it is worth reexamining Simpson’s arguments in some detail. To that end, after considering some philosophical and astronomical background in Sect. 2, I shall discuss the key argument from biological contingency usually cited to justify SETI scepticism in Sect. 3. In Sect. 4, I briefly consider some of the strategies that could be used—with hindsight, unavoidably—to undermine the substance of Simpson’s argument; a possible strengthening of the sceptical argument by Stanislaw Lem, in the fictional context, is presented in Sect. 5. Some concluding comments about the role played by the arguments from evolutionary contingency in the contemporary and near-future research are given in the final section.

Epistemology and astronomy: Simpson’s time and now

In the first part of his article, Simpson surveys the prospects of finding life elsewhere based on the then existing astronomical knowledge. Before everything else, he uses the opportunity to mock people like Joshua Lederberg or Melvin Calvin, who were at the time engaged in creating the new discipline, which would subsequently be called astrobiology (for historical reviews, see Dick 1996, 2006; Wolfe 2002; Strick 2004; Morange 2007), on the grounds that extra-terrestrial life had never been observed, and hence could not be the subject matter of a new science.

On this extremely constraining epistemological criteria, much of what has happened in the last half a century, even in biology, was not science! This has been aptly commented upon by Chyba and Hand (2005):

If exobiology (or astrobiology) were understood to mean solely the study of extraterrestrial life—which it is not—Simpson’s criticism would remain strictly true but might nevertheless seem bizarre to many astronomers or physicists. Astrophysicists, after all, spent decades studying and searching for black holes before accumulating today’s compelling evidence that they exist… The same can be said for the search for room-temperature superconductors, proton decay, violations of special relativity, or for that matter the Higgs boson.

Interestingly enough, all the examples listed by Chyba and Hand were either defined or at best became prominent after Simpson’s essay. One can add additional items to the list, more or less specialized: gravitational lenses (between 1924 and 1979), gravity waves, cosmological inflation, molecular assemblers, dark matter particles, global climate change (until recently), and, the most pertinent of all from the point of view of Simpson’s topic, extrasolar planets themselves (until 1995). Since the existence of extrasolar planets was purely a matter of theoretical faith and not empirical evidence in Simpson’s time, would it be justified for him to claim that “statements about extrasolar planets are not really about anything—or, at the very least, they are not science”?

Of course, as argued by Chyba and Hand, the real subject of astrobiology is cosmic life, not just extraterrestrial life (even disregarding that the notion of “extraterrestrial” is today hard to define clearly, since the Earth is not a closed-box system and the exchange of matter with its cosmic environment is evident). To the extent that such research deals with testable hypotheses and comparing their predictions with the existing evidence, there is no epistemological anomaly in astrobiology compared to other scientific fields. It is visible, inter alia, from the timely appearance of books such as Astrobiology of Earth (Gale 2009), or the many chapters, reviews, and research articles on apparently “terrestrial” astrobiological topics.

As far as the astronomical background of the problem is concerned, the state of our observational knowledge of extrasolar planets as possible habitats for life in Simpson’s time and our own time can serve as a useful gauge of how much have things have changed since the onset of the astrobiological revolution. At the time of Simpson’s article, there was no firm evidence for any extrasolar planetary system; Van de Kamp’s results on Barnard’s Star, quoted by Simpson in a footnote, had not yet been debunked as a result of instrumental error, but most astronomers had been quite sceptical towards them. It was still possible to claim, albeit harder and harder as cosmogonical theory progressed, that the Solar System is unique planetary system—and hence a unique abode of life—in the Milky Way. Simpson did not take that road; he admits that probably there are many planetary systems in the Galaxy (and in other galaxies), but seriously underestimates the capacity of observational astronomers to detect them:

It is not impossible that our descendants may some day make pertinent direct observations on other planetary systems, but that is far beyond our present capabilities or any reasonable extrapolation from them. (p. 770)… In the observable universe the lowest recent estimate for earthlike planets by a competent astronomer is, as far as I know, that of Shapley… who considers 100 million a highly conservative figure. (p. 772)

“Our descendants”, it turned out, were some of the still living astronomers, at most “our children”. Since 1995, we are witnessing an explosion in our knowledge of extrasolar planets; more than 860 have been found by February 2013 (for a continuously updated resource, see http://exoplanet.eu/). And it is not just the sheer quantity: our overall knowledge and understanding of various aspects of extrasolar planetary systems has been increasing exponentially (e.g., Mason 2008, already outdated in many respects). For example, the atmospheres of extrasolar planets have become a standard research topic since the first chemical data on exoplanet atmospheres were obtained in 2002 (e.g., Seager and Deming 2010; Kaltenegger and Sasselov 2010). More to the point, papers with titles such as “Detecting life-bearing extrasolar planets with space telescopes” appear in the world’s premier research journals today without raising eyebrows (Beckwith 2008; for various detailed aspects of habitability, see also Turnbull et al. 2006; Valencia et al. 2007). Even putative extragalactic planets discovered by pixel lensing can be mentioned as an achievement universally considered impossible just a decade ago (Ingrosso et al. 2009, 2011).

All this is not to fault Simpson for his overcautious or failed predictions in this area; most astronomers have been taken by surprise by the pace of discovery as well. However, it is useful to put into perspective the incredible change that has occurred over such a short time—particularly in the light of the optimism of early exobiologists about the number of extrasolar planets in the Galaxy, which seems to have been vindicated thus far. Even a rather conservative approach such as the one of Lineweaver and Grether (2003) affirms that:

[t]he hypothesis that ~100 % of stars have planets is consistent with both the observed exoplanet data that probe only the high-mass, close-orbiting exoplanets and with the observed frequency of circumstellar disks in both single and binary stars. The observed fractions… are lower limits that are consistent with a true fraction of stars with planets, f_t, in the range 0.25 ≤ f_t ≤ 1. If the fraction of Sun-like stars that possess planets is representative of all stars, our result means that out of the ~300 billion stars in our Galaxy, there are between ~75 and ~300 billion planetary systems.

Put together with much better insight into some of the other relevant astrobiological variables, like the age distribution of habitable planets (Lineweaver 2001; Lineweaver et al. 2004), all this means that the astronomical basis for “science without the subject matter” has surged to the point where—in sharp contrast to Simpson’s time—very specific questions can be posed and answered.

The central argument revisited

Simpson’s key evolutionary argument is usually construed as follows: due to the ubiquitous biological contingency, evolution is as opportunistic and unpredictable on Earth as anywhere else in space. Homo sapiens is the result of a 3 + billion-year-long chain of contingent evolutionary events, which is extremely unlikely to be repeated anywhere else with a sufficient degree of similarity to us (“humanoids” of the paper’s title) to be able to meaningfully communicate with us. Hence, our SETI efforts are wasteful and futile.² This is an oversimplification that needs to be dispelled.

What are the “humanoids” of the paper’s title? Simpson was not naïve—in contrast, for instance, to the usual SF portrayal of aliens in his time, and often even today—to claim that those must be bipeds with human-like morphology. (Ironically, he left such naïveté to some of the latter-day convergentists; see Conway Morris 2003; I shall return to this point later.) On the contrary, he writes³:

A humanoid, in science-fiction terminology adapted to the present also somewhat fanciful subject, is a natural, living organism with intelligence comparable to man’s in quantity and quality, hence with the possibility of rational communication with us. Its anatomy and indeed its means of communication are not defined as identical with ours.

In retrospect, the term “humanoid” was perhaps not a fortunate choice, since from the very outset a confusion reigned as to whether it is meant in the morphological sense (as the term is used in the vernacular⁴ or in robotics; e.g., Harada et al. 2010) or in the more complex sense explicated by Simpson. His surprisingly meek disclaimer about being “not defined as identical with ours” [emphasis M. M. Ć.] could not but fuel the confusion further, since it created the impression that the humanoids under discussion were likely to be similar to us in both of those two respects (morphology and the means of communication).

This definitional issue requires a bit of philosophical reflection. After all, ellipses are not defined as identical with circles, nor are arthropods defined as identical with insects, but it is plain that a genetic—in the ontological sense—relation exists between these sets of objects. The set of humans of the Homo sapiens species is a subset of all humanoids in both the morphological sense and in the sense of means of communication (let me call the latter biosemiotic sense). Hypothetical intelligent arthropods, like Orson Card’s Formics, would not belong to the humanoid category in the morphological sense, but would—since it is, after all, the major point of the plot in the whole series of novels—belong to the humanoid category in the biosemiotic sense. I shall show below an example in which Simpson himself confused the two, and most SETI sceptics in the decades since have cleverly exploited this ambiguity about the meaning of “humanoids”.

Simpson has not been adequately credited for formulating one of the first rare-Earth arguments: that is, a probabilistic argument suggesting that, while simple microbial life is probably ubiquitous throughout the Galaxy, complex biospheres, like the terrestrial one, are very rare due to the exceptional combination of many distinct requirements. While this is usually associated with the turn-of-the-millennium Rare Earth (Ward and Brownlee 2000), argument against SETI from independent small probabilities, it was clearly present in Simpson’s writing decades earlier:⁵

There are four successive probabilities to be judged: the probability that suitable planets do exist; the probability that life has arisen on them; the probability that such life has evolved in a predictable way; and the probability that such evolution would lead eventually to humanoids. … the first probability is fair, the second far lower but appreciable, the third exceedingly small, and the fourth almost negligible. … The product of these probabilities, each a fraction, is probably not significantly greater than zero.

This is the “real” argument of Simpson, although it is not usually perceived as such; it is the probabilities #3 and #4 in the quote above that contain what is usually labelled as the argument from evolutionary biology (or the argument from evolutionary contingency). But, before I pass to this core argumentation, one needs to perceive that the whole of Simpson’s discourse is a clear prefiguration of the “rare-Earth equation” of Ward and Brownlee. In contrast to these latter-day rare-Earthists, Simpson does not distinguish complex metazoans as a critical phase in evolution. Instead, we have something which Simpson calls “the probability that such life has evolved in a predictable way” without further explicating it. It does not seem to play an important role in his core argument (to be discussed below), since it anyway presupposes that evolution is unpredictable. And why would predictability of evolution be an issue in debates about SETI at all?

Thus, it remains unclear what purpose the requirement of predictability serves in Simpson’s argument. As I see it, there are two different ways to interpret Simpson’s probability #3: (1) as rejection of stronger forms of convergentism (e.g., as espoused recently by Vermeij or Conway Morris), and (2) as a claim about macroevolutionary processes. Strong convergentism would, in this particular context mean that, for example, upon detecting any extraterrestrial intelligent being, even indirectly, we would be entitled to conclude that it resembles ourselves. Rejection of this extreme position is irrelevant to SETI, since SETI deals with the search for and detection of extraterrestrial intelligent beings without prejudicing anything about their morphology, unless we revert to the morphological meaning of “humanoids”. Simpson himself has been ambiguous in this respect, but we wish to try to make the overall argument as forceful as possible. So we are left with interpretation (2), which might be supported by the fact that Simpson previously promoted what he called “quantum evolution”, a strongly contingent macroevolutionary mode.⁶ As an example of the macroevolutionary interpretation, imagine—in a thought experiment—a planet virtually identical to the Earth, but populated only with prokaryotes.⁷ Obviously, such a biosphere would be completely in accordance with Darwinian evolution, since, after all, about three-quarters of the history of the terrestrial biosphere was spent in such a stage. Equally obviously, this prokaryote paradise would not be a viable SETI target. However, one is entitled to ask how exactly is such a stipulation about the unpredictability of evolution different from simply lowering his probability #4, namely claiming that the evolution of humanoids (now in the true biosemiotic sense) is still lower than hitherto assumed? Whether evolution is predictable or not—even if we completely understand the abiotic environment—is, of course, important for a student of evolution, but does not seem a pressing concern for those searching for potential SETI targets at present or in the near future (Fig. 1).

Fig. 1

Two versions of Simpson’s argument against SETI: the original, proto-rare-Earth argument shown as the product of probabilities on the left-hand side, and the vernacular contingency version in the shaded area on the right-hand side. Popular accounts often identify Simpson’s argument with just the latter version. Both versions use an ambiguous definition of humanoids, shifting between morphological and biosemiotic interpretations. The proto-rare-Earth version contains an unnecessary step (“the probability that such life has evolved in a predictable way”; see text)

Since we now know (and need not conjecture) that there are many possible abodes of life in the Galaxy (probability #1), and since Simpson himself did not find biogenesis improbable under the right conditions (#2), we are left with his #4, which is the “real” probability of evolving potential SETI targets. The claim that probability #4 is extremely small is based on the contingent nature of biological evolution, made famous in Wonderful Life (Gould 1989). Simpson summarizes this core sceptical argument as⁸:

The assumption… that once life gets started anywhere, humanoids will eventually appear is plainly false. The chance of duplicating man on any other planet is the same as the chance that the planet and its organisms have had a history identical in all essentials with that of the earth through some billions of years. [emphasis by M.M.Ć.]

This is what is elsewhere claimed as the “opportunistic” and “contingent” character of biological evolution. Notice, however, that here Simpson (consciously or not) plays on the ambiguity of the definition of humanoids: we have “humanoids” in the first sentence of the quote, and then talk about “man” in the second. Even allowing for the possibility of an accidental slip of pen, it is important to understand that the usual argument from biological contingency is ambiguous in its original form.

Let us try to formulate the argument in syllogistic form (while keeping in mind that Simpson did not formulate it in this manner), using the conventional concept of evolutionary morphological space and giving it, for the purpose of clarity, a specific label.

Simpson’s contingency argument

• 1. Homo sapiens is an intelligent, communication-capable species currently existing in the Galaxy.

• 2. The part of morphological space representing humanoids in the morphological sense is of negligibly small measure within the entire morphological space.

• 3. The rate of independent evolution of intelligent, communication-capable species on habitable planets within a sufficiently large ensemble of such planets is proportional to the size of that part of morphological space representing such species.

• 4. The part of morphological space representing intelligent, communication-capable species is congruent with that part of morphological space representing humanoids.

• 5. The probability of Homo sapiens detecting another intelligent, communication-capable species within the ensemble of habitable planets in the Galaxy is negligible.

Much is uncontroversial here. Premise #1 is obviously true. On the other hand, #3 is a reasonable mathematical consequence of our conventional understanding of stochastic events and our concept of the biological morphospace. It is similar to the phase-space interpretation of the famous Boltzmann formula $S = k\ln W$ relating entropy and statistical weight (“probability” in a loose sense). The “size” of any part of the morphological space can be understood as its n-dimensional volume. Note also that for the conclusion #5 to follow it is not necessary that proportionality is a linear relation; we can use weaker versions of #3 in which the rate of evolution of intelligent species is simply any increasing function of the size of the corresponding chunk of morphospace.

Premise #2 is nothing less than Copernicanism applied to biological evolution. The late epoch of evolution of hominids on Earth clearly testifies to its truth; Simpson and other evolutionary biologists sceptical towards life and intelligence on other worlds list many additional arguments for this premise. When it is stated that out of “billions and billions” of species having evolved on Earth during the last 3.8 Gyr of our planet’s history only a single one has evolved intelligence (Mayr 1993), it is just restatement of premise #2.⁹ But we need to be aware of the selection effects accompanying any such general conclusion about the (incomplete) past record (Vermeij 2006). If the emergence of a trait like intelligence (noogenesis) prevents its independent re-emergence at later epochs, we obviously cannot infer much about the general frequency of noogenesis from a single data point of our past record.

Thus we come to the premise #4, which is the prime suspect, and not only because its formulation is by necessity a complex one. Gould, in his criticism of Simpson’s argument and its usage by Tipler and other SETI-pessimists, invoked exactly the distinction between the general and particular claim as the underlying logical fallacy.¹⁰ The particular claim is the supposed low relative frequency (= rarity) of humanoids; the general claim is the supposed low relative frequency of any kind of intelligence. Gould and many proponents of SETI (e.g., MacGowan and Ordway 1966) deny the general claim.

This would amount to the alternative hypothesis that the part of the morphospace corresponding to intelligent observers capable of detection via SETI programmes and activities is significantly larger than the part corresponding to humanoids. These alternatives are shown in Fig. 2 as options (A) and (B). Simpson’s argument corresponds to the case (A), in which the darkened part of the morphospace representing intelligent beings is extremely small, and Gould’s rejoinder to the case (B), in which the same part is of moderate size. Of course, since morphospace is many-dimensional, with perhaps a very large number of dimensions, it is impossible to graphically represent it except symbolically. We might be sure that by far the largest part of it corresponds to simple lifeforms, i.e., microorganisms. Now, the question on which Simpson’s argument hinges is whether reality corresponds to case (A) or case (B).

Fig. 2

A schematic representation of the alternative ways of looking at the evolution of intelligence. In case A, intelligent beings, morphologically humanoid or non-humanoid, occupy a negligible part of the evolutionary morphospace, making the emergence of intelligence elsewhere highly unlikely (even the size of that part is here greatly exaggerated for graphic purposes—it is best to think about it in terms of something that is 10⁻¹⁰⁰ or similar astronomically small fraction of the total). In case B, intelligence of various kinds is represented by a sizeable chunk of the morphospace, making parallel evolution in various habitable locales likely. (Courtesy of S. Popović)

As an initial consideration in favour of (B), it is worth thinking to what extent conceivable intelligent beings incapable of communication as per Simpson’s argument would still be detectable by our—actual and potential—means. After all, communication is, of necessity, a much stronger requirement than mere detection. Ants, for example, can hardly communicate with humans; and yet, as their appearance at picnics amply testifies, they are perfectly capable of detecting human activities. As a long and rich tradition in SETI studies witnesses (e.g., Dyson 1960; Sagan and Walker 1966; Freitas 1985; Harris 1986, 2002; Tilgner and Heinrichsen 1998; Annis 1999; Arnold 2005; Ćirković and Bradbury 2006; Bradbury et al. 2011), it is entirely plausible that we shall detect traces of activities of extraterrestrial civilizations much sooner than we shall be able to communicate with them; the latter could turn out to be impossible without impairing prospects for the former. Such detections are still, obviously, part of the SETI enterprise; and yet, Simpson’s criticism does not seem to apply to them. Obviously, Simpson, like many in the SETI community itself, took the “Galactic Club” (Bracewell 1975) scenario of intelligible and intentional communication of advanced Galactic civilization too seriously and linked all SETI activities with this particular, rather improbable, scenario.

Whether the case (A) or the case (B) are better description depends, among other things, on how high the acquiring of intelligence moves the given lifeform towards the peak of fitness landscape. If intelligence is very advantageous, then we could be right in speculating that there are many different evolutionary trajectories leading to such an important trait. Another factor might whether the adaptive platform from which intelligence is built (say, something like an organised central nervous system with plasticity and learning) can be constructed through many evolutionary pathways.

Let me sketch one of the specific ways in which premise #4 of Simpson’s contingency argument might be undermined. Encephalization quotient (EQ: the ratio of cerebral mass to the body mass of an organism; Hofman 1982) is often used as a metric that gives some pointers to the physical description of intelligence (/self-awareness/tool-making capacity/whatever vague property is necessary for SETI targets). EQ has generally increased since the Cambrian explosion, which might be regarded as an instance of the general increase of complexity in that period. However, its rate of increase was faster until 251 Myr BP, i.e., until the famous end-Permian mass extinction episode, the most destructive mass extinction in the Phanerozoic eon. Afterwards, it continued to increase, to reach the historical maximum in Homo sapiens, but at a slower pace.

Nowadays, no serious palaeontologist would argue that the end-Permian catastrophe, “when life nearly died”, to quote the title of Michael Benton’s popular book, was necessary in the historical sense.¹¹ So, it is possible to argue that the rate of EQ increase during the Palaeozoic could have brought about the emergence of intelligent observers much earlier if the end-Permian cataclysm did not happen. This statement ought to be distinguished from naïve convergentism: a view that, without the end-Permian cataclysm, Homo sapiens would have evolved sooner. Such a view is implicit in the more extreme writings of Conway Morris (2003), who holds a particularly strong view on the inevitability of humanoids in the vernacular, morphological sense: “Rerun the tape of life as often as you like, and the end result will be much the same.” (He even takes seriously the “Star Trek” idea that distant planets could be populated by “smiling bipeds.”¹²)

Such strong convergentism is in itself a refutation of Simpson’s argument, because according to Conway Morris, if a planet is habitable at all, it is likely to eventually produce intelligent bipeds, i.e., humanoids in both senses. But the position of Conway Morris is arguably an extreme one. The question, therefore, is to what extent a weaker convergentist assumption would undermine Simpson’s argument as well. A minimal assumption in this sense would be the one on the adaptive value of intelligence. If we allow for the positive adaptive value of intelligence in “replaying the tape” thought experiment on Earth, there is certainly no reason to believe that intelligent observers which would have evolved without the end-Permian extinction would have been mammals or remotely similar to humanoids. The particular peak in the fitness landscape corresponding to a human level of intelligence and communication capacity—or so we fancy in our narcissism—would have been reached, in the fullness of evolutionary time, by the vector of evolution from another direction.

If, on the other hand, intelligence is only weakly advantageous or neutral, or—in the extreme pessimist view—maladaptive, it will be ephemeral (on an evolutionary timescale) on Earth and its chunk of the general morphospace is negligible. We can think of many ways that intelligence could be advantageous. On the other hand, there is one very disturbing family of scenarios for the negative adaptive value of intelligence—those of anthropogenic existential risks, where humans destroy themselves or permanently curb their creative potential. Simpson was fully aware of it; he wrote¹³:

Even if, as I believe, any close approximation of Homo sapiens elsewhere in the accessible universe is effectively ruled out, the question is not quite closed. Manlike intelligence is, after all, a marvelous adaptation, especially in its breadth.… There is, to be sure, another serious hitch here. Man may be going to use one wild aspect of his intelligence to wipe himself out. I do not believe that will occur, but no realist can now deny it as a possibility. If it did happen, the adaptiveness of human intelligence would have been short-lived indeed, and the argument from its apparent broad adaptiveness would be negatived.

Unfortunately, today we have much more to worry about than in Simpson’s time. In addition to the ever-present threat of nuclear war and subsequent nuclear winter, at the beginning of the third millennium we are facing such large risks as anthropogenic global warming or the misuse of biotechnology (Bostrom and Ćirković 2008). Obviously, if any of the adverse outcomes are realized, it will be due to our intelligence, one way or another, so an impartial observer would be forced to assign a large negative value at least to our kind of intelligence. Here, it is easy to notice how tightly astrobiological thinking is tied to future studies—roughly, optimism towards the future of intelligence on the Earth tallies well with case (B), which in turn would imply a high relative frequency of emergence of intelligent observers and justification for SETI. Conversely, pessimism with respect to intelligence on Earth could be associated with the validity of premise #1 in Simpson’s argument and negligible chances for SETI success. It is clear that much further research is necessary before we can empirically discriminate between cases (A) and (B). In the meantime, it is very difficult, to say the least, to use Simpson’s contingency argument as a justification for cessation of practical SETI research.

Fiasco argument

There is another version of the argument from evolutionary contingency that is not based on the rate of evolution of intelligent species per se, but on our capacity for detecting radically different forms of intelligence and their capacity for communication—and the duration of both. I shall call it the Fiasco argument, since in the last novel of Stanislaw Lem, Fiasco, a version of this argument has been advanced.¹⁴ It can be schematically represented in a similar manner.

Fiasco argument:

• 1′. Homo sapiens is an intelligent, communication-capable species currently existing in the Galaxy.

• 2′. The part of morphological space representing humans is of negligibly small measure within the entire morphological space.

• 3′. The rate of passage of intelligent species through the part of morphological space corresponding to intelligent species detectable by humans (those within the “window of contact”) is proportional to the size of that part of morphological space.

• 4′. The part of morphological space representing intelligent species detectable by humans (those within the “window of contact”) is not much larger in comparison to that part of morphological space representing humans.

• 5′. The probability of Homo sapiens detecting another intelligent, communication-capable species in the Galaxy is negligible.

This is still closely related to Simpson’s contingency argument (we note that the premises #1′ and #2′ are very similar to #1 and #2 above), but it has some specific advantages. First, the Fiasco argument explicitly takes into account the fact that intelligent species themselves evolve. Therefore, this argument avoids the “timeless” nature of Simpson’s reasoning. It provides a missing extinction term, which might balance the emergence-of-intelligence term; see the discussion of possible negative adaptive value of intelligence in Sect. 3 above. The “window of contact” is defined in an operational way, without prejudicating upon the deep definitional issues of “what counts as an observer?” type. Lem introduces it in one of his long discoursive passages in the novel as a result of fictional theoretical work on the general structure and evolution of civilizations¹⁵:

The commencement of the main road was the early technological age, which was short-lived, allowing no offshoots for the thousand of years between the rise of mechanical tools and the advent of the informational.…

The turning point in the main road was the moment when the engineering ability of the Intelligent Beings matched the life-creating potential of Nature. It was not possible to predict the further career of any individual civilization; this followed from the very nature of the crossroads. A certain percentage of the civilizations would stay on the main road—by putting the lid on an attainable but unrealized auto-evolution. …

The conservatives on the main road would be silent: that was obvious.

For the biotically nonconservative there were many solutions. Decisions to autoevolve, once made, were generally irreversible. Hence the great diversity among the older psychozoics. Ortega, Nielssen, and Tomic introduced the concept of a “window of contact.” This was the interval of time in which Intelligent Beings had already reached a high level of applied science but had not yet undertaken to change the natural Intelligence given them—what would correspond to the human brain. The “window of contact” was, cosmically, a moment. From the resinous torch to the oil lamp 16,000 years passed; from that lamp to the laser, it was a hundred years. The information needed to make the torch-to-laser step was on the order of the information needed to go from the discovery of the genetic code to the code’s implementation in a post-atomic industry. Increases in knowledge were, in the “window of contact” phase, exponential—and, toward the phase’s end, hyperbolic… Outside the window, for civilizations either immature or too mature, silence reigned. The immature lacked power to communicate, while those too mature closed themselves off—or else formed groups that communicated with one another by means faster than light.

We can imagine the evolution of this high-complexity sector of the overall “astrobiological landscape” (Ćirković 2012) as the intelligent species emerging, evolving at different rates, entering the window of contact, and subsequently at some point exiting it; we can speculate what happens next, but we cannot really know it. In any case, they move beyond the domain of the operational definition—beyond being viable SETI targets.

Therefore, we may prefer the Fiasco argument over the original version of Simpson’s contingency argument, since the former does not constrain astrobiological complexity per se. Also, the Fiasco argument explicitly takes into account further evolution of intelligent species after they acquite intelligence as a trait, without prejudicing the mechanism of such further evolution. Finally, it cannot be easily linked to either pessimism or optimism towards the future of intelligence on Earth. It is easy to see that this is so: if civilizations can leave the window of contact without being destroyed (by either external or internal agencies), there is no obvious reason why they could not continue to increase in complexity.

Conclusions

No reassessment of Simpson’s argumentation could be complete without taking into account its “practical” element; near the end of the article, he writes¹⁶:

We are now spending billions of dollars a year and an enormously disproportionate part of our badly needed engineering and scientific manpower on space programs. The prospective discovery of extraterrestrial life is advanced as one of the major reasons, or excuses, for this. … We can, indeed, learn more about possible extraterrestrial life by studying the systematics and evolution of earthly organisms.

This has provoked some expected comments (Dick 1996, and elsewhere) that the whole point of the anti-SETI exercise is a quarrel about funding, and the feeling of inferiority among classical biologists—as contrasted with their molecular counterparts—in the new, space age. While there may be some truth in it, in retrospect we perceive history showing Simpson to be clearly wrong as far as motivation for space programmes is concerned. Both SETI and space programmes have been dying a slow death for decades, but the decline of public and funding support for SETI has been much quicker, especially after the NASA HRMS cancellation in 1993. While the astrobiological revolution since 1995 has somewhat rekindled the faded interest in astronautics and space research, it is mainly limited to planetary probes, at quite modest costs in comparison to manned missions. On the other hand, SETI projects are, by general consensus of both supporters and sceptics, among the cheapest ways of doing empirical science at all, and Simpson’s linkage between SETI and space programmes nowadays sounds contrived and hollow.

Since SETI is cheap, there is no powerful practical argument for killing it. Even if we consider the spectrum of outcomes, we might wish to retain it, since if the “SETI gamble” pays off, we learn an enormous amount; but even if it does not pay off, we learn something important, along the lines of Simpson’s discussion (e.g., about evolutionary contingency vs. convergence). Thus, on one key issue, Simpson was right on target: biological systematics and evolutionary theory on Earth are incredibly important and indeed unavoidable parts of astrobiology and SETI studies. This simple truth has been slowly diffusing to the forefront of research and in recent years has been quite openly embraced (e.g., Fry 2000; Chela-Flores 2003; Rospars 2011; Ćirković 2012; Dick 2012). Whether the 1964 article played an encouraging or inhibiting role in the process is open to debate: the forcefulness of Simpson’s attack on the then newly born exobiology and SETI, as well as the gusto with which his paper has been used (and misused) by staunch opponents of SETI like Tipler, Mayr, or Basalla, could have persuaded many researchers that it is just a cover for irrational and visceral opposition to a new enterprise, motivated by funding jealousy and interdisciplinary misunderstanding. It is still a question for the historians of science to resolve. On the other hand, the resistance on the part of the mainstream SETI community to consider the evolutionary context and implications of their own work (as manifested, for example, by the fact that the major review of the field, as late as 2001, neither cited Simpson’s article nor contained any discussion of the relevance of evolutionary theory for SETI; Tarter 2001) is clearly deplorable.

All in all, while both Simpson’s and the Fiasco argument are extremely important and thought-provoking, they are hardly as decisive as the critics have repeatedly presented them for the last five decades—rather, they are wake-up calls for more serious work on both the theoretical and practical sides of SETI. While it is nonsense to reject or even downplay the importance of evolutionary biology for SETI studies, it is also simplistic nonsense to claim, half a century after Simpson’s attempt, that evolutionary biology proves SETI to be misguided and unfounded. The road to the real SETI—that is, small-scale, cheap, diligent, both observational and theoretical, imaginative and innovative—has never been more open.