Life

I stole this text from a paper I wrote nearly ten years ago. It is fun though, and worth reprinting as a blog. The paper itself is permanently unpublishable, because thought it was actually ACCEPTED FOR PUBLICATION in a very prestigious planetary science journal, the editors wanted me to do stuff with the model I wrote that are just plain impossible, given our current state of knowledge.

The Drake equation predicts the approximate number of intelligent, technological, civilizations in the galaxy at any particular time (Shklovski and Sagan, 1966);

(1) N=R_f f_g f_p n_e f_l f_i f_civ L_civ

when N=the number of technological civilizations, R_f =the rate of star formation, f_g =the fraction of stars able to sustain habitable planets, f_p =the fraction of such stars with planets, n_e =the number of habitable planets per solar system, f_l =the fraction of these planets that support life, f_i =the fraction of these planets on which intelligent species evolve, f_civ =the fraction of these species which create technological civilizations, and L_civ =the average lifetime of such a civilization.

These parameters touch upon diverse areas of human knowledge such as astronomy, biochemistry, evolutionary biology, and anthropology. A few are known to an order of magnitude. Others are outright guesses. Although of limited predictive value, the Drake equation it is an interesting heuristic tool. It allows us to organize ideas about how each factor might influence our likelihood of someday encountering another intelligent civilization.

Advances in astronomy, planetary science, and biochemistry have shed light on some parameters, including R, f_p, n_p, and perhaps even f_l. Some analyses are sobering, such as a computer modeling study by Michael Hart suggesting a star’s so called “habitable zone” may be very narrow (Hart, 1979; but also see Kasting et al., 1993), and Shiv Kumar’s theoretical study suggesting stable planetary systems may be uncommon (Kumar, 1974). Orbital eccentricities of recently discovered planetary systems seem to support Kumar’s prediction (Resio and Ford, 1996; Glanz, 1997).

Other areas of research have provoked a resurgence of optimism, such as the discovery of planetary systems around a growing number of nearby stars (i.e., Mayor and Queloz, 1995; also see Schilling, 1999). Discovery of complex communities of organisms associated with deep-sea hydrothermal vents, deep under the Earth’s crust, and in a growing variety of exotic microenvironments, has provoked speculation that suitable environments for life may be more abundant in the galaxy than once thought (Vogel, 1999). Additionally, there is growing geological evidence the Earth underwent at least two periods of global glaciation (Hoffman et al., 1998; see also Warwick and Howard-Williams, 2000). The survival of life on Earth through such an event suggests Hart’s definition of the habitable zone may be unduly narrow, and supports Kasting’s et al.’s less restrictive model for the habitable zone of a terrestrial planet (Kasting et al., 1993).

SETI searches are designed to detect intelligent life, and these findings naturally beg the question of what fraction of life-bearing planets might ultimately be expected to produce an intelligent species.

F_i, the fraction of life-bearing worlds that should ultimately be expected to develop intelligent species, remains one of the most nebulous parameters of the Drake equation. Values cited by various authors range from approximately 1 (Baugher 1985, Mentzel 1965), to .1 (Sagan 1973), to essentially zero (Simpson 1964, Mayr 1985).

Diversity of opinion is inevitable, considering that an intelligent species has evolved exactly once (of which we are aware), on the one planet that we know sustains life. The prevailing view among astronomers and physicists has been that f_i should equal approximately 1, since greater intelligence is (supposedly) unconditionally adaptive and confers greater possibilities to the organism. By contrast, evolutionary biologists have been pessimistic about the probability of intelligent species evolving on other earthlike planets (see Mayr, 1985). This view is based upon the notion that, given a replay of the Earth's history for the last few hundred million years, the odds against the evolution of anything like a human being are virtually nil.

While the first view is ill founded from a biological perspective, the latter view might prove to be unduly pessimistic.

Natural selection does not inevitably favor increasing intelligence. For any given organism, greater intelligence is usually maladaptive, since it carries a high cost in terms of energy and increased developmental time. This is one case of a widespread phenomenon is known as stabilizing selection ( ). Since the benefits of increased intelligence are balanced against the disadvantages of higher energy cost and slower development, intelligence remains stable over long periods of time. As an example, the human brain uses up to one sixth of the resting body’s energy budget (Smith, 1984), a fraction that would be unacceptable to organisms with different life histories. Large, warm-blooded animals with relatively high intelligence have slower developmental times than simpler organisms and are often more vulnerable to extinction.

Two influential evolutionary biologists, George Gaylord Simpson and Ernst Mayr, have argued that extraterrestrial intelligence is unlikely because the odds against the evolution of anything like a human being are virtually nil. This argument does not preclude the existence of intelligent species on extrasolar planets, since we do not expect an intelligent extraterrestrial to resemble a human being. There is no need to recapitulate the evolution of our own species to produce intelligence. Even on a planet identical to Earth such an event would be inconceivably improbable. It is more appropriate to ask; given a planet like Earth, how probable is the eventual evolution of at least one species with the human grade of intellect?

Studies of animal cognition suggest that the human mind differs from those of other animals largely in matters of degree (Wilson, 1975; Budianski, 1998, Butterworth, 1999). In the macro-evolutionary sense, human intelligence can be viewed as the endpoint of a long random walk that ultimately produced one species with an exceedingly complex nervous system. With the increasing diversity of species on Earth, the most complex animals on Earth have become more complex for the last 540 million years This does not imply that evolution is an inevitable march to complexity, and most terrestrial life forms exhibit simple behavior.

The evolution of increasingly intelligent forms is not unique to our own lineage of animals. The most recent common ancestor of humans and cephalopods (octopi, squids, and their relatives) was comparable to a flatworm, yet cephalopod nervous systems share much in common with our own, including image-forming eyes and reliance upon learned behavior. There may, in fact, be an infinite number of evolutionary pathways that ultimately produce intelligent organisms.

Periods of rapid speciation foster the diversification many different lineages of organisms. Within each lineage, species spread along axes of behavioral and morphological complexity determined by their development and body plan. (see Valentine and Campbell, 1975; Valentine, 1992). These adaptive radiations depend upon novel adaptations, or "evolutionary innovations". Evolutionary innovations enable the temporary escape from interspecific competition, and allow many new species to proliferate and replace existing ones.

For example, the evolution of a chitinous exoskeleton and jointed appendages allowed one (or perhaps several) marine annelid-like organism to diversify into the vast array of arthropod forms, each with a different ecological niche. This, in turn, facilitated the invasion of land by arthropods, and the spectacular adaptive radiations of the insects. Each novel adaptation is built upon previous ones, however, and they constrain the future evolution of a lineage as well as facilitate it. Returning to insects as an example, the geometry of the chitinous exoskeleton and the design of the insect respiratory system are excellent means by which to supply a small organism with mechanical support and oxygen. While fostering their enormous success, both adaptations constrain insects to a relatively small body size. This and certain other features of insect development constrain insect nervous systems to simple but efficient structures that make due with a limited number of cells. Social bees have existed on Earth for perhaps 40 million years, since the Miocene. In this time, they have evolved a wide range of problem-solving behaviors, interesting architecture, and complex societies. It is doubtful, however, that any of them will ever build a radio telescope (but not impossible, as I will discuss later). This because insect nervous systems are constrained to a mere 7000 neurons or so, compared to the fifteen billion neurons of the human nervous system.

Evolutionary constraint is critical to the evolution of intellect, which is the product of multiple sequential adaptive radiations. Each radiation is built upon an evolutionary innovation. Some facilitate greater neurological integration and behavioral complexity (i.e., the vertebrate neural crest, a tissue that gives rise to our brains and a wide range of other specialized tissues). Others may be only peripherally related to intelligence (i.e., the tetrapod limb). Each novel adaptation has the potential to stymie the evolution of greater intellect much as the arthropod exoskeleton ultimately stymied the evolution of intelligent insects.

On Earth, there are a limited number of independent lineages, which compete for ecological niches in the biosphere. Each lineage has a different body plan. By number, and possibly by numbers of species, most terrestrial organisms are prokaryotes such as bacteria, cyanobacteria, and archaeans. Our own lineage, the eukaryotes, is one branch in the early adaptive radiation of terrestrial life. Since there are no truly multicellular prokaryotes, we can guess that eukaryogenesis carries with it features necessary for the later evolution of multicellular life. Had another group of organisms displaced the eukaryotes early in their evolution, multicellular life might never have evolved on Earth.

Among multicellular lineages, there are many protists, the fungi, plants, and animals. Fungi and plants have adaptations that, although tremendously useful, prohibit their eventual evolution of intelligence (in both cases, their cell walls obviate the need for a central nervous system, since they restrict them to an immobile lifestyle). There may have been as many as 100 animal phyla in the past, of which approximately 25 remain. Most of these major phyla arose at a certain critical period approximately .7 billlion years ago to .4billion years ago, and subsequent diversifications of animal life have been built upon these plans. Likewise, the evolution of classes within phyla and families within classes has slowed considerably, so that new species almost inevitably arise from existing body plans that arose 540 million years ago.

Given these major body plans, the biosphere cycles through an enormous number of species. Conservatively, there are perhaps 10 million species on the planet right now. This number represents a small fraction of the total species that have ever existed. Although speciation and extinction rates vary considerably among taxa, species tend to last between 1 and 10 million years before going extinct.

In 2000, I came up with an intentionally oversimplified model to produce a guess as to this figure. I basically used taxonomic trees from biology textbooks as my data, and based my guess on the negative binomial distribution. This is the same distribution that tells you how many times you are likely to be able to flip a coin till it lands on tails. I basically assumed that each branching event was like flipping a coin for a while, trying to get a string of ten or fifteen heads, and no tails (not a fair coin, the odds of a lineage loosing the ability to evolve intelligence were higher than a bump in that direction, but there was a neutral state as well-it was a three headed coin-where landing on tails meant that a whole branch on the tree of life had to stop flipping), each flip, however, led to a branchpoint, so that there were multiple chains of coins flipping at any given time, but the descendants of a lineage that came up tails were always there, competing with the others, but fated to never evolve intelligence. I used a lot of C++.

I plugged a lot of data from guesswork biology into the thing, and my figure was .01. Just my guess. It is interesting to note that, although just under one percent of planets actually produce an intelligent species some time in their histories, we should not expect one percent of planets to actually harbor them. The proportion of habitable worlds sustaining an intelligent species at any given time should be much smaller. This is because any given species survives only a small fraction of the planet’s history. The average lifespan for a terrestrial mammal species is 1.5 million years (Savage 1988). Once established, however, intelligent species would tend to give rise to other intelligent species, increasing the duration a planet harbors intelligent life. If successful, a single intelligent species might give rise to an entire family of intelligent forms. The lifespan of a typical terrestrial vertebrate species is approximately 23 million years (Savage 1988). This is still a very short time compared to the lifespan of an entire planet, but long enough to suggest intelligent species from different solar systems might have time to come into contact by some means (perhaps with biological descendants of the species that initiated the communication).

Even if intelligence is short lived, the proliferation of intelligent species might have major implications for the evolution of technological civilization. Although the first intelligent species on a planet may have limited aptitude for technology due to behavioral and ecological traits inherited from its ancestors, subsequent species might overcome these obstacles. This is exactly the process that produced our own species. Although we are the lone intelligent species (of which we are aware) on Earth at present, our species is the product of an intense but short-lived radiation of African Homonids. This radiation quickly (in geological time) gave rise to Homo sapiens, a species which has subsequently developed sophisticated technology and transformed the biosphere.

Our species may be unusual in this regard. Once intelligence has originated in one species, the proliferation of intelligent species may continue for a long time before a technological civilization arises. Technological species that originate from established families of intelligent life forms would share the world with tens or hundreds of fellow thinking beings.