Problems with major transitions: Maureen O’Malley & Russell Powell respond

The Great Oxidation Event by Adelle Schemm.

The Great Oxidation Event by Adelle Schemm.

In a recent series of posts, I reviewed Maureen O’Malley and Russell Powell’s paper in Biology and Philosophy, “Major Problems in Evolutionary Transitions: How a Metabolic Perspective Can Enrich our Understanding of Macroevolution.” Although they made several good points, I thought that some of their criticisms were off the mark and that their proposed solution to the real and perceived problems with the major transitions framework was unsatisfying.

Drs. O’Malley and Powell are both heavy hitters in the philosophy of biology, and as I usually do when I dig deeply into someone else’s paper, I invited them to respond to my criticisms. They kindly provided a thoughtful rebuttal and permitted me to post it here. I’ll have more to say later, but for now I’ll just say that they make some good points and (most importantly) fairly represent my arguments. As usual for guest posts, I have made no edits to the content of their response, only formatted and added links:

Response to Matt Herron’s posts on our paper on major transitions

We are delighted that Matt devoted some of his high-powered critical attention to our paper on major transitions in evolution (MTE), and that his thoughts featured on the beautiful and informative blog ‘Fierce Roller’. Long may his blog run!

There are a few points to which we’d like to respond, although ultimately, we suggest, we have different aims and will simply have to respect those differences. Matt’s criticisms fall neatly into Parts 1, 2 and 3 of his commentary, so that’s the order in which we’ll address them, too.

Progressivism (Part 1)

Matt says ‘Major transitions theory is not and has never been progressivist’.

We are well aware that Maynard Smith and Szathmáry (M&S) stated in the original version of MTE that they are not proposing a progressivist view, and we note as much in our paper. However, we think that any framework that starts with increasing complexity and ends up with human communication will inevitably be vulnerable to progressivist interpretations, all protests to the contrary aside. As we explained in our paper, the human communication/society ‘transition’ is the least convincing item on the list for most readers. That it is there at all, at the current pinnacle of an evolutionary complexification trend, is indicative of highly problematic assumptions about complexity and its trajectory. And of course, as we also note in the paper, there are a variety of M&S’s own statements which suggest that they really did believe in complexity as a kind of progress. We cited some of them in our original text. For example:

Maynard Smith (1988, p. 229), ‘human society is the final level of complexity so far achieved by living organisms’.

Sigmund and Szathmáry (1998, p. 439) ‘“progress”… as a series of major transitions in evolution’.

Even if one is inclined to think that M&S couldn’t have possibly meant these claims, we believe the very book cover (depicted in Part 1 of Matt’s analysis) licenses such interpretations. As we noted, and Stephen Jay Gould did too in another context, visual schemes sometimes reveal more powerfully some of the hidden assumptions that frame our ways of thinking about evolution and biological phenomena.

Many people have told us that they don’t like this criticism (on the grounds that it’s both mean and false), and that denying being progressivist is enough for our point to be refuted. But as we argued, the whole ‘type’ approach that underpins MTE 1.0, ETIs, and MTE 2.0 lends itself to picking out a progressive thread that connects the ‘obviously’ major transitions. Just to repeat that point:

‘“Ascent” iconographies are an inevitable result of such type-oriented models, because they ignore or abstract away the broader phylogenetic base of ETIs. This is even more problematic for models that culminate in human culture and communication—a questionable transition that is retained in MTE 2.0’ (O’Malley and Powell 2016, pp. 173-174).

Furthermore, even if one rejects the idea that there is a general selection bias in favour of increasing complexity—and M&S rightfully dismiss the notion that there is a ‘built-in’ tendency toward complexification—progressivist signals may still be gleaned so long as the explanatory focus remains squarely on complexity increases. We thus do not agree with Matt’s characterization in a footnote to Part 3 of his entry that M&S’s deployment of the term ‘THE major transitions in evolution’ (emphasis added) was merely ‘unfortunate’—to the contrary, that choice reflects underlying assumptions about the explanatory goals of major transitions theory that we both foreground and challenge in our paper (we’ll have more to say about this in our response to Part 3).

In any case, the worry we expressed about classic and revised major transitions theory encouraging ascent interpretations was by no means our most major point. It was an aside, but one that we thought had to be mentioned. Our main points are nicely encapsulated in Parts 2 and 3 of Herron’s critique.

Disunity (Part 2)

Disunity has been a frequent criticism of M&S’s MTE. We argue in our paper that one thing contributing to the disunity problem is that MTE theory lumps together contingent and robustly replicable events. Yet, ‘one-off events are not a problem’, says Matt, because it’s well known that contingency is a feature of evolution. Our point here is that if you want to have a theoretical (nomothetic) scheme that explains biological regularities, it cannot be about contingent events but rather about repeated and thus theorizable events. If the mitochondrion acquisition is a unique sort of individuality transition, then it can only be captured in a limited way by a general framework that ignores all the differences in different types of individuality transitions.

We think Matt agrees with this point, because he too is most concerned with how to achieve a generalizable unified account. We see this where he says, ‘focus on ETIs’, and goes on to urge readers to accept the consequence that a whole raft of minor and even totally uninteresting ETIs pervades the history of life. Dispense, he suggests, with the ‘evolutionary importance’ criterion for major transitions. Why, then, doesn’t Michod do this? Or any other ETI theorist? The answer, we believe, relates to the goals of MTE theory, which are not well served by eliminating the criterion of importance. Not many people bite the bullet in the way Matt has—that is, to maintain that any ETI event, no matter how minor, should be included. For this would mean that every endosymbiosis (perhaps even if it lasts only for a short evolutionary time), every obligatory symbiosis even if it is not endosymbiotic, and every parasitism that gets inherited vertically should also be included, insofar as these events constitute ETIs. The result would be that thousands of events would be included, perhaps millions. While all of these events are ecologically and evolutionarily interesting, many of them would not be of any general explanatory appeal. All of these events are ETIs, but very few of them reflect major shifts in the organizational structure of life on Earth. This is true, for example, with respect to most of the tertiary plastid endosymbioses: Yes, they are amazing events in their own right because of the serial nature of such endosymbioses and how the organisms involved adjust to these acquisitions. However, many of them have not had a major ecological impact, meaning that they have mostly not changed the space of evolutionary possibilities (as did, for example, the primary plastid acquisition).

Matt points out that ‘plastid acquisition is discussed as a major transition’ (because we say it hasn’t been), but in many cases the minor plastid acquisitions are ignored. When the whole gamut of plastid acquisitions is noted as a series of major transitions, it is often only done so casually by people not focused on analysing and revising MTE (we include in this category some of the references Matt notes). One exception is West et al. (2015), who are setting up a new table of events (‘major’ transitions in individuality based on degrees of dependence and cooperation). They have a little to say about all the secondary and tertiary plastid acquisitions, but bundle them up with the primary plastid acquisition. All plastid acquisitions pass the criteria West et al. use to rule out a lot of symbioses, despite many secondary plastid symbioses not persisting very long evolutionarily. West et al. (2015) do this in the name of unity, which we question as the main aim of MTE (see the next section). Overall, however, despite these quibbles, we agree wholeheartedly with the general point Matt is reinforcing: namely, that anyone with a singular focus on ETIs cannot ignore plastid endosymbioses. How they are included is our next topic.

What are ‘major transitions’ about? (Part 3)

The final point concerns how narrow the focus of major transitions theory should be and, relatedly, whether the key thing that needs explaining is ‘major transitions’ or ‘transitions in individuality’. Matt argues that we lay an unfair burden on MTE and ETI theorists by arguing that the point of MTE theory is to explain why biology is like it is today (the ‘Big Picture’ story we note in our paper). That is too grand and too unfeasible a project, he says. Shifting the focus to metabolism – something we argue has to be included in the Big Picture story – is not the answer, because the reasons for including it appeal to these unreasonably grand explanatory ambitions. Minor ambitions for major transitions are enough: stick with ETIs, Matt maintains, even if they are minor and multiple. As long as they produce ‘new individuals’, such events fit the same ‘natural kind’ schema that all other ETIs do. And since having a unified and complete scheme is what theorizing MTE is all about, this is the only plausible way forward.

Most respectfully, we are arguing that this is in fact the problem. At least half the appeal of the ETI-based version of MTE is that it is focused solely on major ETIs. Stuff it full of minor ones, and much of that appeal goes down the drain (but keeps the ‘baby’, as Matt advises). Moreover, the Big Picture story that we take to be the goal of MTE does not, contra Matt’s reading, require that we explain all of the biological world—that would indeed be unreasonable. Rather, what we call the ‘Big Question’—the question of how it is that we got here from there—aims to identify and explain the large-scale structure of life and its organization, of which transitions in hierarchy constitute just one important aspect. This organizing research question may be vague, but it is not unhelpfully so—in fact, it is the broadness of the Big Question that opens the door to alternative, mutually informative approaches to understanding life at its grandest scales of organization. We believe that this grand ambition rightfully lies at the heart of MTE and its many revisions, and we believe that metabolic organization and explanation should play a more integral role in theorizing answers to the Big Question, including answers that focus on major ETIs.

Finally, there’s the ‘natural kind’ problem, which we are very pleased Matt draws so much attention to. We agree that if major events are not natural kinds, this bodes poorly for their theoretical unification. However, if one accepts that the goal of MTE is provide an answer to the Big Question, then the focus should be on explanatory events regardless of whether they are tokens (instances) of natural kinds. In any case, as we argue in our paper, integrating explanatory strategies that focus on metabolism need not result in a fatal disunity, so long as the resulting theoretical pluralism is of a manageable sort. Integrating replication-based and metabolism-based explanations has proved difficult, but we see no reason to prioritize one over the other in MTE theory, as each gets at important features of the large-scale structure of life on Earth and, presumably, of life elsewhere. Explanatory power, rather than theoretical unity, should be the gold standard for MTE theory. A comprehensive pluralistic explanation is preferable to a more limited explanation that can lay claim to theoretical unity. We can therefore acknowledge that the biological world has an unruly causal ontology—one that resists the human psychological penchant for elegance and order—without conceding theoretical defeat.

In conclusion

This is a fabulous discussion of our paper and we are deeply honoured, both to be taken to task on various unclear bits of our argument and to be allowed to respond to this tasking. As is probably inevitable, we stand by what we said, but we think Matt has done an excellent job of showing why what we say really matters. Where people align themselves in any such debate comes down to what they want to get out of MTE theory. At the very least, we have shown that the current literature is not particularly clear on these fundamental aims, and that MTE theory still has inconsistencies to resolve regardless of the aims on which it settles.


References & stable links:

Maynard Smith J (1988) Evolutionary progress and the levels of selection. In: Nitecki MH (ed), Evolutionary Progress. University of Chicago Press, Chicago, pp 219–230

O ’Malley MA, Powell R (2016) Major problems in evolutionary transitions: how a metabolic perspective can enrich our understanding of macroevolution. Biology & Philosophy, 31, 159–189.

Sigmund K, Szathmáry E (1998) Merging lines and emerging levels. Nature 392:439–441

West SA, Fisher RM, Gardner A, Kiers ET (2015) Major evolutionary transitions in individuality. Proceedings of the National Academy of Sciences, 112, 10112–10119.

Leave a Reply