Below is a response, by W. Ford Dootlittle and Drew Inkpen to my commentary on the Gaia Hypothesis, published here on January 4th. It’s followed by a brief counter-commentary by yours truly.
We (I’m joined by my former postdoc Drew Inkpen, now a member of the faculty at Mt. Allison University and co-author with me on a Proceedings of the National Academy paper explaining some of the tenets of our expansion of Darwinian thinking) have several things to say in response to Massimo’s critique. Most of this has already been published in academic journals, which do not in general publish rebuttals of blogs, and neither of us runs a blog. So we thank Massimo for putting this on his.
The point of the Aeon article was to sketch out for a more general public our attempts to broaden Darwin’s theory, as it is now generally promulgated by philosophers of biology and biologists themselves, so that the Gaia Hypothesis would not be so automatically dismissed (as Massimo does again here) as “impossible in theory.” Even those who accept multilevel selection theory, which Massimo does and which figures large as the Aeon article develops, would not see natural selection as applying to entities that don’t reproduce as unified. Species might (they speciate) but ecosystems (of which Gaia is possibly the largest) don’t, and that’s really what Massimo and most of his colleagues don’t like about Gaia. And because she doesn’t reproduce, it’s hard to imagine that there’s a population of Gaia-like entities in which natural selection could operate. That’s the second thing Massimo and colleagues don’t like. We’d agree, which is why we attempt in the Aeon essay, and elsewhere, to rethink natural selection.
Another of Massimo’s issues is around “selfishness” as a useful, albeit metaphorical and somewhat dated, way to describe evolution. We’d continue to defend this, but applying it (as the Aeon article does) at many levels of the biological hierarchy. All that we (or Dawkins) meant by “selfish” was that (at whatever level), natural selection as usually construed selects for changes that result in an entity’s having more progeny. In this sense, there can also be selfish organisms, selfish populations or selfish species. Some object to the anthropomorphizing implications of the S-word, but when they get over that, most see the meaning.
Massimo next recounts Lewontin’s Recipe (as does the Aeon article) and again infers that it implies the existence of populations (populations of reproducing entities, we’d add). This “Where’s the population?” question is one of the oldest objections to Lovelock’s view, and dealt with in the Aeon article and elsewhere. And, independently of some of our arguments, there’s more and more evidence of habitable planets other than our own and more and more evidence that it’s fairly easy to get life going. So a population of independently arising individuals, in which differential survival can in principle be assessed, likely does exist. We are left with the reproduction issue, of course, which we admit is the big one.
Actually, reproduction is not required to get some sort of selective sorting going (think of the sizing of stones on a storm-wracked beach) but it does seem to be required for the sort of iterated natural selection that leads to complex adaptations like the vertebrate eye. If the first mutation is rare and that selected rare mutant does not differentially reproduce, there won’t be enough of a population to expect a second mutation, let alone all the many many more sequential mutations that are necessary to produce an eye.
There are several ways around this, though, and the Aeon article describes one though Massimo seems not to like it. We (along with Darwin) proposed that all Life on Earth descends from some single common ancestral species or cell (LUCA, the last universal common ancestor). Since many of the core enzymes shared by all living things already have well-defined and conserved enzymatic machinery, it’s most reasonable to suppose that LUCA was already a fairly advanced species or cell. Perhaps we could not even readily identify it as “primitive” if we were to use a time machine and bring it back to the lab. And it defies imagination that LUCA was the sole descendant of its slightly more primitive ancestor, which was the sole descendant of its, and so forth back to the origin of life. There must have been a population of wannabe LUCAs contemporaneous with LUCA, in spite of Massimo’s insistence that “we don’t have any empirical evidence for multiple LUCAs”. It’s most unparsimonious on the basis of all the evidence we do have and the phylogenetic theory we use to interpret it to assume otherwise. Yes, we’d agree with Massimo’s claim that “even if there were multiple LUCA’s we don’t know that all but one of them went extinct because of competition,” but if competitive natural selection operates now, why not back then?
Massimo also points out that “LUCAs are not biospheres, unless we assume that a given clade […] living in the same environment thereby counts as a separate biosphere,” and that we’d only assume that if we were trying to “artificially rescue Gaia.” We’re not sure what artificially rescuing Gaia would mean: if we need to modify the theory of Gaia in order for it to be scientifically tractable, why not do so? Of course, if this pushes Gaia so far as to be unrecognizable, then we could call what we’re doing by a different name. But keeping the name has the benefits (i) of clearly showcasing, in a popular venue, the relation between what we’re up to and the work of Earth System Scientists, like Tim Lenton, and (ii) of helping people who think about Gaia to think about it more critically.
We’d agree, further on in Massimo’s critique, that we cast Gaia as a persister, and “persistence is not the same thing as reproduction.” Indeed not, but that’s sort of the point on which we have been insisting and the philosopher Frédéric Bouchard has been insisting for decades before. We could rework the theory of evolution by natural selection so it included differential persistence and differential replication/reproduction as mechanisms or outcomes, if we wanted. The Aeon article and other of our more scholarly papers show how. Massimo doesn’t want to do that, and he does point out what was for us once a sticking point — that evolution by persistence is just like the sorting of stones on a beach, a largely non-reiteratable one-off sort of process. Turns out it’s actually not: the arguments about LUCA’s descendants evolving by “clade selection” in the Aeon article, our previously articles in Trends in Ecology and Evolution, Philosophy of Science and the Journal of Theoretical Biology, as well as Rich Lenski’s ongoing experiments involving serial transfer over 60,000 generations of independent clones of E. coli all point in this direction, if we are willing to go there. Massimo isn’t, though.
The “biggest problem” that Massimo sees is also the lowest blow, we think. He says, and we endorse this, that we should not “make up scientific hypotheses because they are politically useful.” Of course not. And the Aeon piece was not an example of doing so, nor is our other work. At the end of the piece, Ford pointed out that our acceptance of a modified version of Gaia theory may have some benefits, including: providing some credibility to the idea of conceiving of nature as a unified whole and healing an unproductive rift between Darwinians and Earth System Scientists. That is to say, if worthy of acceptance on the grounds we provided (and more as needed), the theory might have such benefits. We see little difference between this and a biochemist pointing out that their work would, if true, cohere well with the goal of alleviating disease. Ford did not claim that a fit with certain political goals should be seen as a reason in favor of accepting our theories. That would indeed be problematic. Massimo may have, falsely it turns out, taken this paragraph to suggest that we develop the theories we do because they fit with a certain political ideology, but that would (i) be untrue of the in fact genesis of this theory and (ii) irrelevant anyway. Scientists develop (and reject) theories for all kinds of reasons, and it’s the theories themselves that need to be scrutinized.
W. Ford Doolittle
S. Andrew Inkpen
I wish to thank Ford and Andrew for taking the time to respond to my critique. However, my first counterpoint is that I did not “automatically dismiss” the Gaia Hypothesis. On the contrary, I wrote a commentary on it years ago, as well as the more recent one to which Ford and Andrew are responding. It is a questionable rhetorical move to label one’s critics as dismissive while in fact they advance arguments against your position.
Similarly, it isn’t a question — as Ford and Andrew repeatedly put it — of me or other critics of Gaia “liking” or not certain ideas. It is a question of agreeing or disagreeing on the basis of evidence and arguments.
Ford and Andrew’s first substantial point in their rebuttal concerns the population problem: if Gaia is not part of a population (of reproducing entities) then there can be no natural selection. Surprisingly, they claim that there is increasing evidence of such a population, because we are discovering more and more extra-solar planets. This is entirely irrelevant and misleading. First off, they would have to be Earth-like planets, that is, planets characterized by a biosphere. And there is no evidence of that, yet. But even if there were, it is more than a bit strange to claim that these other planets “compete” with Gaia / Earth for common resources, which is what is necessary for natural selection to get going. There is no shared competitive environment to speak of, and therefore no possibility for natural selection.
Next, Ford and Andrew introduce a red herring, when they tell us that their theory doesn’t really need reproduction, sorting will do. But their own example of the sorting of pebbles by size on a beach (caused by the continuous action of waves) shows why this response is wholly insufficient. Systems subject to sorting do not evolve, reproduction and inheritance are needed for that. Accordingly, no geologist speaks of the evolution of beach pebbles.
Immediately after we come back to the issue of LUCA, the Last Universal Common Ancestor. Ford and Andrew reject my objection that we have no empirical evidence for the existence, at one time, of multiple LUCAs, saying that “it’s most reasonable to suppose” that did exist. But science doesn’t advance by reasonable suppositions alone. That’s metaphysics. And it is dangerous when scientists gingerly slip into what some have labelled a “post-empirical” approach to the validation of hypotheses.
But even if we follow Ford and Andrew there, they seem to be confusing multiple LUCAs with multiple ecosystems, which is what they really need to get their hypothesis going. LUCAs were organisms, which competed within the same broad environment (Earth), for similar resources. This is standard evolution by natural selection, no need to invoke Gaia.
And yet Ford and Andrew insist in that mistake even in their rebuttal, writing that they are puzzled by my charge that they are trying to artificially rescue Gaia. They claim that there is nothing wrong with modifying a hypothesis so that it becomes scientifically tractable. But this is not what they are doing at all, because: (i) they are not just modifying a hypothesis, they are changing the meaning of scientific concepts (competing organisms become competing ecosystems) to fit their hypothesis, and (ii) they are not making Gaia more “tractable” because their modification does not make contact with the empirical. At all.
Next, Ford and Andrew claim that I don’t want to expand the theory of evolution to include mere persistence and sorting. But it’s not just me, most working biologists also resist that move, for the reasons I explained in my original response, as well as those laid out by Richard Lewontin in his famous treatment of natural selection. Ford and Andrew bring in Richard Lenski’s famous long term experiments concerning the evolution of bacteria in the lab as if somehow such experiments lent credibility to their approach. I am very familiar with Lenski’s work, and I have even visited his lab years ago. Those experiments fall squarely within the standard framework of evolution by natural selection, including mutation, variation, and reproduction. They have nothing to do with what Ford and Andrew are talking about.
Finally, they object to my comment about the political / ideological aspect of Gaia. I stand by my original remarks. I don’t know whether their interest in Gaia is ideologically motivated or not, but Ford’s comment at the end of the original Aeon article is troubling because Gaia has a long history of such mingling between science and politics, as Ford surely knows. It seems to me, therefore, a bit disingenuous to claim that his remark about the possible relevance of Gaia to environmental policies is simply analogous to a biochemist pointing out that his work may be used to cure diseases. Most articles in biochemistry, incidentally, do not end with that sort of statement.
Regardless of other considerations, the crucial question is: if Gaia is a scientific hypotheses, can it be tested empirically? The answer is a resounding no.