This diagram, lifted from a lively paper by the astrobiologist Charles Lineweaver, is the tree of life on earth. The “root” at the center is the last common ancestor of all life. Toward the bottom left, you’ll find the genus “Homo”, to which you and I belong, at the end of a twig representing animals. The two neighboring twigs, ending in Zea (i.e. corn) and Coprimus, represent plants and fungi, our two closest relatives.
Professor Lineweaver offers this diagram as an antidote to the superstition that evolution has some tendency to converge on intelligence; his criterion is that we ought not say that evolution “converges” toward some feature unless we observe that feature arising independently in at least two or more twigs. By that same criterion, evolution has no tendency even to converge on heads, which (says Dr. Lineweaver) are likely to be prerequisite for anything like human intelligence.
If human-like intelligence is a fluke, then presumably the ability to build radio telescopes is also a fluke, which stands as a cautionary note for those who expect to communicate with extraterrestrial civilizations.
Now it’s certainly true that if you magnify that animal twig to display separate branches for birds, dolphins and carnivores, you can fool yourself into thinking that intelligence has emerged independently along a multitude of pathways. After all, the last common ancestor of birds and men had a very small brain; after 310 million years of independent evolution, birds and men both have much larger brains. The last common ancestor of dolphins and men had a small brain; after 85 million years of independent evolution, both dolphins and men have much larger brains. But Dr. Lineweaver observes that the “independence” is illusory. By the time we and the dolphins went our separate ways, we were already endowed with highly similar biochemical neural pathways and constraints that left us both with (in Dr. Lineweaver’s words) a finite number of highly evolved “toggle switches” that could be successfully tinkered with.
Moreover, even if evolution were biased toward intelligence, there’s no reason to suppose that intelligence would be the sort we can communicate with across interstellar distances:
About 600 million years ago, two kinds of metazoans, protostomes and deuterostomes, diverged from each other. Both evolved independently for ~600 million years and were very successful. Today there are about a million species of protostomes and about 600,000 species of deuterostomes (of which we are one). We consider ourselves to be the smartest deuterostome. The most intelligent protostome is the octopus. After 600 million years of independent evolution and despite their big brains, octopi do not seem to be on the verge of building radio telescopes.
Doctor Lineweaver closes with an endorsement of the Search for ExtraTerrestrial Intelligence, on the grounds that it’s cheap and he might be wrong. Do you agree with him?
I agree that it is common to think that something you like that is mostly paid for by others is ‘cheap’…
The primary value of searching for extraterrestrial life probably lies in the enjoyment of people who assume it’s being searched for because it might exist. The contemplation alone is worth something(maybe – it could also be bad, e.g. religion).
So if the search encourages the contemplation, and the search is cheaper than the value of that extra contemplation, it’s already worth it.
It is good to have the antidote to the belief that “evolution” is just that, a one-way process towards some higher goals. I see Darwinian evolution as a sort of random walk in the (infinite-dimensional) space of all possible DNA strings. Obviously, if the starting point is a single species with zero intelligence, the probability that some more intelligent species will evolve is greater than zero. In fact, given infinite time, it should be equal to one.
I won’t comment on Dr. Lineweaver’s reasoning because I haven’t read his paper, but I note that there is a species capable of building radio telescopes on the only planet where we found life. That is a small sample size, but the success rate is 100%.
@ Snorri:
It seems at least a little unfair to call something a significant observation when it would be ridiculous to observe the opposite. That is to say, you forgot to mention the “possible” many other ways one might communicate across vast distances. If extraterrestrial life did exist, and if there were, say, 1,000,000,000 technologies such that contact was possible if both species developed the same technology, then each species knowing of one such technology would not imply a high probability that they make contact (1 in a million in fact).
Actually I’m curious now – how significant can such an observation be? Clearly it’s sensitive to how complex the universe can possibly be. Unfortunately, I don’t know that either.
There is a common conception that Man is “top-of the-tree” in evolution terms, as though all life had just been leading to us. This is typified in the film “Night at the Museum” where a character encourages another to behave better than a monkey with the words: “who’s evolved?”. Of course, both Man and Monkey are equally evolved.
The Dinosaurs ruled for 140 million years without evolving intelligence (although what traces would be left of their radio telescopes now?). The protostomes seem to have done it for 600 million years. There is no requirement to develop technology to be succesful in evolutionary terms.
On the other hand, as Snorri Godhi says, we only have one sample of a world with life, and it has developed a technological species. That is 100% record. Whilst technology is not required for survival against nature’s teeth and claws, it does provide one succesful strategy, at least in the short term. So perhaps it is very likely that eventually technology will occur on suitable worlds. After all, the Earth may have a few more billion years to go.
We don’t know yet if technology is a good or bad evolutionary strategy. We may wipe ourselves out within a million years or so. Alternatively, after a million or so years it could prove to be hugely succesful, with entire solar systems colonised. We have only had technology for a few thousand years so far. There is an argument that if technology is succesful and reasonably likely, it is inevitable that ET would already have reached us, so where are they? Maybe they are all listening to their radio telescopes, and no-one is sending.
The cost is relatively modest, and it has had spin-offs in creating networks of volunteers allowing processing to be done on their computers (SETI@home). This network could have great value, of itself or as an example. The enjoyment and inspiration mentioned by mcp are also of value, and in my opinion it is well worth the cost.
I also think that the search for near-Earth objects which have the potential to detroy us is under-funded.
mcp: So far, our technology has detected only one way in which information is spread accross vast distances – the electromagnetic spectrum. It is at present our only option. It also seems very likely that other species in other parts of the universe will also have experience of this. They may also have discovered tachyon or gravity wave transmitters, or some other thing entirely beyond our technology. From their point of view, it might make no sense at all to transmit electromagnetic signals, but it seems very likely they would be aware of the possibility.
Snorri: There is also a species capable of speaking English on the only planet where we found life. Does that significantly increase your confidence that there are other English-speaking species out there?
Harold:
It also seems very likely that other species in other parts of the universe will also have experience of this.
Does the same reasoning convince you that it’s very likely that other species on earth have some experience of this?
Steve, absolutely; all species that can see have experience of the electromagnetic spectrum. Should one of these species wish to communicate with us, and were technological, they might use this method. On Earth, it is perhaps more likely they would use sound, using noises such as woof or meow.
Regarding the English thing, I am sure there are other reasons why this is a red herring, but here’s one. On Earth, there are thousands of languages. There is no particular reason to think that English would be the one selected by aliens. However, we expect that they would have a language, but we may be wrong. The key is language, rather than the specific one. An interresting thought, if English had always been the only language on Earth, I wonder if we would expect aliens to speak English?
Maybe I should make it clear that my last paragraph was a bit of a joke. To the extent that there was a point to it, it was this: if you are going to count the fraction of living species capable of interplanetary communication, then you should also count the number of living species per planet. On the only planet where the 2nd number is greater than zero to our knowledge, the 1st number is the reciprocal of the 2nd.
As for the English-language thing, language is a matter of convention, as Steve reminded us recently.
More on the English language thing, not that I want to get too bogged down with this. As a counter to my last argument, you can say there are thousands of non technological species, as well as thousands of languages. There is no particular reason to think the aliens would select technology as their strategy. So why is the existence of technological life on Earth supportive of the likelihood of technology evolving elsewhere, whereas the existence of English is not? We have one world with a huge variety of life and languages. We have 2 hypotheses. Given the lifetime of a world:
1) Technology is likely to evolve as a strategy for survival.
2) English is likely to evolve as a strategy for survival.
You could say we have the same evidence for both, in that the one world where we have evidence exhibits both technology and English. However, whilst we can conceive of a mechanism whereby 1 is a sensible hypothesis, I can think of no reason why 2 is. Technology provides a unique and distinct method for surviving. It can exist among the variety of other strategies, and provide this unique benefit. English does not provide a unique benefit, but exactly the same benefit as other languages. Therefore hypotheses 2 makes very little sense in the first place. I am sure this can be put much better than this.
Most estimates of the likelihood of receiving signals from ET are similar to the Drake equation:
N = R*.fp.ne.fℓ.fi.fc.L
where:
N = the number of civilizations in our galaxy with which communication might be possible;
R* = the average rate of star formation per year in our galaxy
fp = the fraction of those stars that have planets
ne = the average number of planets that can potentially support life per star that has planets
fℓ = the fraction of the above that actually go on to develop life at some point
fi = the fraction of the above that actually go on to develop intelligent life
fc = the fraction of civilizations that develop a technology that releases detectable signs of their existence into space
L = the length of time such civilizations release detectable signals into space.
I do not think that science is able to determine many of these with any degree of certainty, although we are finding out more as time goes on. The existence of intelligent life on Earth (if it exists) is evidence that all terms are above zero. That we evolved intelligent life somewhere about half way from the origin of life and the end of life (depending on what happens to the sun) must give us some information about the likely value of fi.
I don’t disagree with his reasoning about evolution — strongly agree, in fact — but I think he’s erecting a straw man in re SETI. You don’t have to believe that there is any sort of convergent evolutionary tendency to produce intelligence to think that extraterrestrial life might exist. You just have to believe that there is a definable probability that any given ecosystem could over a set amount of time produce an intelligence, and then goose the odds a bit by believing that there are many (hopefully billions) planets capable of supporting life. These beliefs are, to be sure, so far largely untested, but neither have they been definitively falsified.
Doesn’t the expected cost-benefit analysis of SETI depend very heavily on what your conditional expectation is that intelligent life, if it exists, is friendly? It seems that several reasonably intelligent species or sub-species of human evolved, and the reason all but one are extinct may very well be that our ancestors wiped the others out. I’m not sure if our own experience should make us more or less confident about the wisdom of SETI.
As I said in response the last time you raised this, I think there is at least some evidence that we could communicate with another intelligence insofar as we’ve managed at least some form of communication with every other species we’ve met so far that has anything even resembling what we call intelligence.
You’re also conflating the notion of extraterrestrial life, with which there’s no a priori reason to suppose it has communication, and extraterrestrial civilizations which I think presupposes communication. If you’re talking about the latter then yes I think we can figure out how to communicate with them.
Ryan said: “Doesn’t the expected cost-benefit analysis of SETI depend very heavily on what your conditional expectation is that intelligent life, if it exists, is friendly?”
SETI is passive eavesdropping. I would agree with you if a program of actively signaling our presence were under consideration.
Natural selection has no goal, of course, but intelligence has obvious survival and reproductive advantages to complex life forms, so it is not unreasonable to expect it will develop elsewhere.
Neil,
Fair point, but presumably a large reason why you’d want to locate other civilizations is in order to make contact.
Ryan,
If we did overhear evidence of alien intelligence, we sure would need to think hard and long about letting them know we are here. But even if we remain passive listeners, it would be stupendously important. For one thing, it would end the last bastion of human exceptionalism.
Intelligent life on earth may be an exception, though I would not bet on it. Every scientific discovery has disproved some notion or other that the earth is unique, from the fall of the geocentric theory on. When I was young, I remember learned scientists pontificating their opinions that the sun’s planetary system was unique. Now we know that planetary systems are the rule, not the exception.
Neil:
Intelligent life on earth may be an exception, though I would not bet on it. Every scientific discovery has disproved some notion or other that the earth is unique, from the fall of the geocentric theory on.
Well, if this is a good argument, it should apply more generally. Let’s try it:
English spoken on earth may be an exception, though I would not bet on it. Every scientific discovery has disproved some notion or other that the earth is unique, from the fall of the geocentric theory on.
Do you think this is a good argument for expecting to find extraterrestrial English speakers?
Steve,
Of course not. There are good reasons for thinking that English is peculiar, but not so for life. There is no particular combination of elements nor physical or chemical property that would distinguish earth. In fact, life on earth is made of three of the most common elements in the universe–hydrogen, oxygen and carbon. Since life on earth is just chemistry, there is no reason to think that chemistry on earth would be different than on planets in habitable zones around other stars. Life is determined by the operation of scientific laws that exist throughout the universe, language is not.
To put a point on it, if I were to travel to a far away country about which I knew nothing, I would be not be surprised to find that they did not speak English there. I would be very surprised to find that the laws of nature operated differently there.
Steve: English vs. Radio, et al…
Please do not conflate a social construct with the laws of
physics and chemistry. Or or you trying to tell us that
non-English speaking aliens will also use an unrecognizably
different math? (I.e.: I don’t mean base 10, but the rules
themselves)
Neil: I would be not be surprised to find that they did not speak English there. I would be very surprised to find that the laws of nature operated differently there.
Ron: Please do not conflate a social construct with the laws of
physics and chemistry.
But you have both ignored the question of whether the laws of nature are any more likely to generate intelligence than to generate English. Here on earth, they’ve done both. Elsewhere, you think they’re more likely to do one than the other. What’s the difference?
what we have here is an infinite number of monkeys!
i feel strongly that the probability of another monkey somewhere in the universe that is typing these exact sentences is 1!
it is the curse of existence — the probability that said monkey is too far away to drop me a note about their ncaa basketball finals is also 1.
“Doctor Lineweaver closes with an endorsement of the Search for ExtraTerrestrial Intelligence, on the grounds that it’s cheap and he might be wrong. Do you agree with him?”
my conclusion: cheap in this case better be free if he wants my support. i agree that he might be wrong.
@Neil would you be surprised if they had quantum computers?
Although there is no evidence, I think the law of natural selection is universal, like physical law. Although natural selection is blind and does not have a goal such as intelligence (or sharp teeth, ability to run fast, etc), intelligence is adaptive and increases survival and reproductive success. The fact that the majority of earth’s life forms are archaea and bacteria does not in any way refute that.
While intelligence is not inevitable, I see no reason why the circumstances that led to its emergence on earth would not operate in suitable locales elsewhere. In other words, I do not think intelligence is some sort of miracle or fluke–it came about for a reason. Now an alien speaking English, that would be a miracle. But, then again, maybe not–perhaps they’ve been watching our 1950s TV shows!
Dave,
I would be surprised if they could keep more than a half dozen qubits coherent.
Have you ever considered the reason that we have not made contact with any other worlds is that the ability to survive once you have developed nuclear weapons, or some other way to destroy your planet, means that most all worlds are only capable of something like radio transmission for a few hundred years?
I know, for a fact, that a planet about 50 light years away has been sending us radio signals for about 400 years. The recognized that our planet had living creatures on it. Unfortunately, they blew themselves up about 700 years ago and the last radio transmission passed Earth about 1350.
Just a thought.
This doesn’t seem like the most useful way of looking at it, to me. He doesn’t have to be “wrong” for the concept that evolution will produce intelligence to be likely true, since it doesn’t contradict what he’s saying.
For the purposes of that statement, we don’t care how many branches of an evolutionary tree lead to intelligence, we just need to care about the probability that, given enough time, an evolutionary tree will produce intelligence on *any* branch. Given that the only one we’ve got did in fact do so, but that we wouldn’t be here asking the question if it hadn’t, I don’t think we can use the existence or shape of this particular evolutionary tree as evidence one way or the other on that statement.
Convince me otherwise. Tell me why this tree we’re looking at provides evidence either that evolution is likely to produce intelligence, or that it isn’t?
“But you have both ignored the question of whether the laws of nature
are any more likely to generate intelligence than to generate English”
We have ignored it because it’s self-evident. A dependent
probability is always lower than its depended-on probability.
E.g.: What’s the probability that I’ll throw three sevens in a
row followed by three eights in a row. You don’t have to know
much at all about math, statistics, or probabilities to know it’s
less likely than my simply throwing three sevens in a row.
Similarly, the chance of a race developing intelligence and
speaking English is lower than the chance of a race developing
intelligence. I also suggest that the chance of generating
English without any intelligence is self-obviously essentially
nil.
Ron: Point taken. I should have said: “But you have ignored the question of whether the laws of nature are *substantially* more likely to generate intelligence than to generate English.”
Cos: If many branches led to intelligence, we could conclude that intelligence is something other than a mere fluke. In the absence of those many branches, we cannot draw that conclusion. So (for starters) that’s one reason why I think this tree should lower our assessment that intelligence is a likely outcome of evolution.
But separate branches HAVE developed intelligence. Not only mammals, where it is most extravagant, but also Aves, especially Corvidae, and there is increasing evidence of intelligence in Chephalapoda. Intelligence is a matter of degree. The honey bee’s dance is a form of intelligence.
Steve:
If many branches led to intelligence, we could conclude that intelligence is something other than a mere fluke. In the absence of those many branches, we cannot draw that conclusion.
This is a good, tight formulation of the Lineweaver thesis. But the fact is, intelligence IS a mere fluke: every step in Darwinian evolution is a mere fluke.
Also, that we cannot draw a conclusion does not imply that we can draw the opposite conclusion.
The English thing again! I am sure this is a red herring. If we were to speculate that English were to exist in the universe, we would need a new term in the Drake equation, call it fe – the fraction of intelligent species that go on to speak English. We therefore know that the number speaking English is less than the number of intelligent species, as Ron pointed out. The fact that English exists means this fraction is above zero. Can we hypothesise that languages may possibly converge on English, even if given given fairly specific pre-conditions? No. Therefore fe is very very small. Can we hypothesise that life could converge on intelligence, given some specific pre-conditions? Yes. Therefore fi may be quite large. There is a possible mechanism fore making it large.
I agree with the article, that the idea that life will always converge on intelligence is unlikely. Intelligence is one mountain in a range. Each peak is a succesful strategy for survival. It is by no means certain that life will ascend this particular peak, but if it stumbles upon the base of the mountain, then it is likely to ascend. We don’t know how broad that base is, but it exists and provides a ramp to the top. English is like a spike – there is no ramp to the top. You have to stumble upon the very point to get there, which is vanishingly unlikely.
Those who subscribe to the convergence hypothesis assume that the base is very broad, and all other peaks are subsidiary peaks on this mountain. I see no reason to assume this.
That is just not true. Astronomical evidence of the last 30 years has been almost entirely against the likelihood of finding Earth-like planets elsewhere.
Roger, why do you say that? 30 years ago we knew of no extra solar planets, now we have confirmed hundreds. Most are gas giants, but these are the only ones we would have been able to detect. What is this astronomical evidence of which you speak?
30 years ago it was suspected that Earth-like planets could orbit those stars. No we know that it is impossible.
@Roger:
“30 years ago it was suspected that Earth-like planets could orbit those stars. No we know that it is impossible.”
I must have missed that. Source?
Jumping in very late here, but human language is a function of the fact that we evolved in a gaseous atmosphere (actually, we started in a liquid), and it was a survival advantage to be able to detect waves in that atmosphere. After animals evolved the ability to here, we next evolved the ability to make sound, and finally the ability to communicate via language. If we had evolved in a different environment, would we have evolved language in the same way?
While most water-dwelling species can detect sound in a limited way, I believe the only water species that communicate via sound are those that evolved the ability on land: whales and porpoises.
I could envision another planet with intelligent life, where the mode of communication is so different from ours that we would not recognize it as communication.
@Harold, it’s quite possible that a species could have the ability to detect electromagnetic radiation via a sense like sight, but have no way to artificially generate it. A species that evolved under water might not have experimented with electricity and magnetism, because doing so would be very challenging under water, or any other conductor.
Al V.,
You do not watch the Discovery Channel, obviously. There is a number of fish that make sounds. Croakers and grunt fish come to mind, and the toadfish. Pompano and sea catfish make a sound. I do not know if the sounds are used to communicate though, probably not.
Oh yes, and lobsters clack their claws together to attract females.
>> Cos: If many branches led to intelligence, we could conclude that intelligence is something other than a mere fluke. In the absence of those many branches, we cannot draw that conclusion. So (for starters) that’s one reason why I think this tree should lower our assessment that intelligence is a likely outcome of evolution. <<
You're making a big, unsupported leap there. Whatever we could conclude if many branches led to intelligence, let's talk about what we can('t) conclude given what we see: An evolutionary tree that led to intelligence – but which happens to be the only evolutionary tree we know about. Is it a "mere fluke"? How can we tell?
On the level of probabilistic reasoning you're using, without explaining what evolutionary mechanism supports one conclusion or another, it's just as likely that perhaps every evolutionary tree is likely to create one branch with intelligent life eventually, but the presence of that branch then makes it unlikely that other branches will develop intelligence. Or, it may be that evolution eventually leads to intelligence on *all* long-surviving branches, but the one we're observing hasn't been around long enough yet. Or, perhaps it is a fluke.
My point is that given merely the shape of this tree, and the level of reasoning you're using, we have no evidence whatsoever about the relative likelihood of these and other possibilities (I certainly haven't thought up an exhaustive list). So we can conclude nothing, and that means we can't conclude that it is a fluke, nor can we conclude that it isn't a fluke.
Given that the Drake equation
has been mentioned here, you
might like to see this recent
xkcd strip:
http://xkcd.com/718/
A few heavy logs to throw onto this smoldering fire…
a) On Steve’s amusing “intelligence vs. English” question, there is a logical case to make that intelligence, as a general property, is adaptive – that it leads to greater success in survival. This hypothesis may be wrong, but the case can certainly be made. Perhaps the most important goal of SETI is to test this hypothesis.
One can similarly make the case that language, as a general property, is adaptive, since it allows multiple intelligent organisms to coordinate their actions. So it’s reasonable to hypothesize that intelligent organisms will evolve the ability to communicate with each other, and that those that do not will succumb to those that do (e.g., our ability to communicate with each other so that we can build boats and other implements seems to have given us an edge over the octopus empire). However, I am not aware of any logical argument that English is adaptive over other forms of audible language, let alone that audible language is adaptive over other forms of communication. Audible language is just one way to communicate, after all. So “English” and “intelligence” are fundamentally different.
b) I haven’t read Charlie’s paper, so the following argument may have been addressed already: The diversity on the “tree of life” is mostly the diversity of the adaptation of single-celled organisms to the dizzying diversity of environmental conditions on the Earth. These “simple” organisms have figured out how to survive in conditions that seem to us very extreme (e.g., boiling acid) and how to eke out a living “eating” and “breathing” materials that we wouldn’t think of as “food” and “air”. But many of these environments and strategies are not conducive to the development of multi-cellular organisms, and hence will never lead to intelligence. Therefore, comparing a branch that doesn’t lead to intelligence to our branch is not the apples-to-apples comparison that is implied.
Put another way, the diversity of the tree tells us that there are many ways to live an un-intelligent life. That does not mean that it is not adaptive to evolve multicellular communities and eventually to build radio telescopes in those environments in which it is conducive to do so. It doesn’t even tell us that it is unlikely that such adaptations will inevitably occur, given enough time. It only tells us that there are many more ways to survive without such adaptations than with such adaptations.
To put it in terms for an audience of economists: There is a dizzying diversity of ways for people to make a living on the Earth, while there are many fewer ways to become a billionaire (most of which are only possible in certain types of cultural environments). These facts do not mean that it is not adaptive to become a billionaire. Only that it is not as easy to get there as it is to work at Walmart.
I might extend this analogy a bit in a different direction to point out that, just as intelligent beings may destroy themselves with H-bombs, billionaires have many extravagant paths to self-destruction that are not open to a greeter at Walmart. Yet, we do not consider this an argument that wealth is not advantageous for survival.
c) The impressive and persuasive argument that intelligence will be rare in the cosmos comes NOT from the diversity of the genomic record, no matter how pretty that tree looks, but from the geological record (efforts to put a “clock” on the genomic record can also generate a persuasive argument). That’s because it is the element of TIME that is necessary to give us a sense of the LIKELIHOOD that a particular adaptation might occur, which is the real essence of this discussion.
When we consider the history of life through time, we see that while microbial, solitary organisms emerged perhaps as early 3800 million years ago, multi-cellular organisms didn’t evolve until around 600 – 500 million years ago. “Intelligence” didn’t evolve for another couple of hundred million years (depending on one’s definition) and then did not lead to tool-using intelligence for quite a while longer. Intelligence capable of building radio telescopes is very recently on the scene, of course. So the strong impression one has from the history of life is that while “simple” life originates easily and survives tenaciously, complex life is a more difficult and hence less likely trick, and intelligence particularly so. These observations still do not tell us that it is not adaptive to have the ability to build radio telescopes, but it is strong evidence that this ability will not be a common feature of planets that have life.