Revisiting Müller 2007

Evolutionary Developmental Biology (evo-devo for short) was established as a formal discipline in the early 80s and from soon after started making its presence felt in diverse areas of biological research. In 2007, roughly 25 years after the field’s establishment, Gerd Müller contributed an opinion article to Nature Reviews Genetics in which he provided an overview of the field’s empirical and theoretical landscape, and, more importantly, explored its implications for evolutionary theory. In July of 2024, we spoke with Gerd Müller about his memories of the making of this article, his current thinking on the views he expressed therein, and, more broadly, the origins of his interest in evo-devo.

Citation: Müller, G. B. (2007). Evo–devo: extending the evolutionary synthesis. Nature Reviews Genetics, 8(12), 943-949.

Date and place of interview: Interview conducted on 24 July 2024 at the Konrad Lorenz Institute for Evolution and Cognition Research (KLI).

Credits: Hari Sridhar (background research, planning and conducting interview); Joyshree Chanam (background research, planning and editing of interview)

 

HS: Gerd, thanks so much again, for agreeing to do this interview. I’d like to start by talking about the paper’s motivation and the work that it describes. From looking at your publication list on Google Scholar, it is clear that many of the themes and ideas that are in this paper were already of interest to you for a long time before. What motivated you to bring them all together in this paper in 2007?

GM: Indeed, this was not the first contribution I had written on this subject. The paper was based on a series of previous publications, the earliest of which goes back to 1994, when I wrote an article on a new synthesis between evolutionary and developmental biology – the field wasn’t called evo-devo yet. This was published in a book by Wolfgang Wieser, in German, on the evolution of evolutionary theory, which also had articles by Peter Schuster, Rupert Riedl, and others. That, I think, was the very first time I tried to formulate the ideas I had on this subject. Subsequently, there were other papers in English. There was also a book I co-edited with Brian Hall called ‘Towards the New Synthesis’ in its subtitle, which was on early eco-evo-devo. And I had an article on evo-devo in a handbook edited by Francisco Ayala and Franz Wuketits, in 2005, who, by the way, had been an early scientific manager here at the KLI. But, you know, writings in German are rarely read by an Anglo-American audience, and the ones in English were not published in prominent places. Therefore, I thought that the Nature paper was a great opportunity to bring these ideas together. However, this was not a submitted paper. It was invited by a senior editor at Nature Reviews, Tanita Casci. Tanita was present at a talk I gave at a workshop in Venice the year before, which explored the relationship between evo-devo and evolutionary theory. After the workshop, she asked me whether I would like to write a “Perspectives Article” for Nature Reviews Genetics, based on the content of my talk. I thought this was a good occasion to make an argument for a synthesis between evo-devo and evolutionary theory. What Tanita also had in mind was to provide something like a research agenda for evo-devo, because although the field existed by that time, and was even named this way, she felt that it lacked a clear characterization of the problems it addressed and of the different areas in which research took place. Because she asked me for that, this paper got somewhat programmatic. I wouldn’t otherwise have chosen that style of presentation. But in order to condense everything in a very limited space, it had these very short sequential sections. In hindsight, I must say that she was right. We have to remember that, in 2006, evo-devo was still a very young and budding field. For the insiders, it was clear what its goals were. But for those who were not part of the evo-devo community, it was actually a bit confusing. That’s the story of how the paper came about.

 

HS: If I understand correctly, the invitation was not to write something that was focused on extending the evolutionary synthesis, but more to lay out the contours of the field itself.

GM: Yes, exactly, it should address the field of evo-devo, its empirical and conceptual content, and its relation to evolutionary theory. This was already in the title of my talk in Venice, and that is what she envisioned. The point was to not merely describe the ongoing research but, explicitly, to characterize how the theoretical outcomes of evo-devo affect the theory of evolution. The extension terminology was an almost automatic consequence.

 

HS: When did you start thinking about the implications of evo-devo for standard evolutionary theory?

GM: Very, very early on, like in the mid-80s. At that time, a lot had happened at the level of experimentation and developmental research. Things like in situ hybridization and other molecular techniques had just been invented and were being applied in a comparative manner, so that you could – for instance – compare gene expression domains in different body organizations. All that happened during this period, and the field of evo-devo blossomed very much. But very few people were interested in what it meant theoretically – what the evo-devo results meant for evolutionary theory – even though the actual founding push, if you like, of evo-devo was a theoretical consideration, namely the realization that the system that relates genetic with phenotypic evolution, i.e. development, has an important influence on what can evolve.  Hox genes and gene expression studies came later; in the late 70s and early 80s these techniques were not around. However, the question of what role development plays in organismal evolution had been formulated already at that time by the likes of John Bonner, Stephen Gould, Pere Alberch, Brian Goodwin, and others. They argued that the conventional theory of evolution treated development as a black box – this was a standard phrase then – and so “opening up the black box” became a slogan for the goals of evo-devo. Subsequently, better methods came along, and it all exploded empirically. But evolutionary biologists somehow believed that this was just about the proximate mechanisms of how development works, and that it had no decisive bearing on how organisms evolve. Because genetic evolution was thought to push phenotypic evolution along, how this was actually realized developmentally was interpreted as a different kind of question. This always puzzled me, because I saw that some of the approaches in evo-devo were able to address evolutionary problems that could not be addressed by a correlational population theoretical approach. At that time, we were starting to get beautiful results that illustrated ho­­­­­­­­­­­w development influences phenotypic and evolutionary change. Therefore, I was surprised that no one really took care of integrating these results with the standard account of evolution. Clearly, there existed overlaps between developmental and evolutionary theory where things were fine, and then there were areas where there were contradictions. In my view, these needed to be addressed.

 

HS: And at that time, were there others also who were thinking like you, about the need to address those issues?

GM: Oh, yes, absolutely. People like Brian Hall, for example, who wrote a book on evolutionary developmental biology, which actually coined the name for the field. There was Pere Alberch, a brilliant researcher at Harvard and a very good friend who unfortunately died very young. I just wrote a paper together with Laura Nuño de la Rosa on Pere Alberch and the conceptual ideas he had about how evolutionary theory needed to be adapted to these new findings. And there were several others. Rudi Raff, for example, was very much concerned with the significance of development for evolution. He, Brian Goodwin and several others wrote books on this subject. Also, all those who worked on heterochrony were interested in this topic as well.

 

HS: We’ll come back to the paper in a minute. I just want you to step back a little to find out about where your interests come from. Can you talk about how you became interested in developmental biology and what, in an interview you did recently, you referred to as evolutionary embryology?

GM: That’s true. My very first coining of the field that we now call evo-devo was “evolutionary embryology”, and I published a paper using that title in 1991. This had to do, on the one hand, with the experimental embryology where it derived from. On the other hand, it was a more concise term. I thought “evolutionary developmental biology” was extremely cumbersome as a term, and no one would use it. But, of course, it was the correct designation, and it took hold via the shorthand “evo-devo”. This was not embryology; it was about development and evolution. Embryology stands for a classical, descriptive approach, which evo-devo was not meant to pursue. How did I get interested? That’s a very good question. I think I have to take a step back here a little bit.

In the early 70s I was a medical student at the University of Vienna, and among all the fields in my study of medicine, I found psychiatry and psychology the most interesting. I started to go to lectures of people in that domain, because I thought maybe that was the area I was going to go into. This was precisely at the time when Konrad Lorenz returned from Germany, after retiring as the director of a Max Planck Institute. It was also the year when he won the Nobel Prize. So, I thought, I’ll go hear his talks. He gave honorary lectures at the University of Vienna. I went to the first lecture, and I was hooked. I knew this was exactly what I wanted to do, because here was the explanation for my questions, namely, evolution. He framed all his behavioural ideas in an evolutionary context. The professor who had invited him to give the honorary lectures was Rupert Riedl. Subsequently, Riedl, Lorenz, and Erhard Oeser, a philosopher of science at the University of Vienna, together organized a seminar series on biotheoretical and biophilosophical problems. The seminars were originally held at the University of Vienna, but when Lorenz grew older, and it became too cumbersome for him to travel, we moved to his villa in Altenberg, a village near Vienna. I say “we”  because, one day, Rupert saw me in the lectures and said, “You’re always here. Why don’t you come join us at the seminar we’re going to hold at Lorenz’s home?”

These seminars – later called the Altenberg Circle – were a fantastic experience for a young medical student who was not used to topics being discussed, viewed from different perspectives, analysed, and questioned. In medicine, you learned facts. There were huge tomes with chapters and subchapters on extensive numbers of medical facts, and that was what you had to learn, and you didn’t question how these facts had been established, or why they were put together in this particular way, or how the topic had been influenced by earlier concepts. So, I was totally baffled about the openness with which professors and students and visitors could discuss scientific topics. Sometimes we had external lecturers who gave a talk on something, or sometimes one of us gave a talk, which was then discussed, etc. One day, we talked about a book by Tom Frazzetta, which I think was called Complex Adaptations. In that book, Frazzetta was critical of the adaptive approach, and, as one example, he mentioned a study by a French biologist, Armand Hampé, who had performed experiments with chicken, recreating ancestral morphological patterns by interfering with the development of the limb skeleton. We discussed these experiments and we noted that, indeed, the bird limb that he had manipulated assumed a more reptile-like appearance in its skeleton. We wondered what might happen downstream of these skeletal modifications. Will other organ parts – muscles, or nerves, or blood vessels – now follow the more ancestral skeletal arrangement? We discussed how it would be very interesting to find this out, at which point I proposed that I will do it. I thought that I will simply perform this experiment, and then we were going to see.

 

HS: Were you already doing experiments then?

GM: No, no. But I was a demonstrator at the Department of Anatomy at the medical school of the university. I had access to laboratories. We bought fertilized eggs, and I had to learn how to incubate them, how to open an egg, how to get through the membranes to the embryo, and how to actually manipulate the embryo, and so on. I had, of course, no clue about what I was going to get into and how difficult all this was. Especially, I didn’t know that laboratories existed that were working with chicks and who could have taught me how to open an egg and reseal it in order to re-incubate it. But I read up on this and simply started working with eggs, and along the way improved the technique that Hampé had used, by applying different tools and materials. But in a way I reinvented the wheel. There was a whole domain of experimental embryology which, at the time I didn’t know, and, I must say, my advisors also didn’t tell me about. But it worked out, and the experience later helped me to solve other experimental questions by inventing new approaches. In this case, what I did was to perform the manipulations and then let the embryos grow older than in Hampé’s studies, so that we could dissect muscle patterns, inserting tendons, nerves, etc. I found according modifications in the muscles that were more reptile-like than in the untreated bird limbs. This is how I got into experimental embryology but also into evo-devo, because this was what, at the time, was called an ontophyletic experiment. A neurologist by the name of Michael Katz had aptly coined this term. So, through the Altenberg seminars, and the influence of Lorenz and Riedl, my interest in the relationship between development and evolution had been triggered.

 

HS: So, at this point, it would have been like something you’re doing on the side, like a hobby?

GM: Yes, because I was still studying medicine then.

 

HS: How did it become your main area of work?

GM: Well, because I found the topic fascinating – development and especially the experimental manipulation of development to investigate the mechanisms of cellular interactions, pattern formation, and so on. But by then my medical studies were already quite advanced, and towards the end of these experiments I had graduated and had accepted an assistant professor position in anatomy. This was a conscious decision, because it allowed me to continue my studies in biology, which I had started in parallel to my medical studies. At the time, the university system in Austria was very liberal; it would be totally impossible now. Eventually, this developed into a topic of dissertation in biology, and I could do this on the side while I was teaching anatomy.

 

HS: You mean your PhD Dissertation?

GM: Yes, the PhD dissertation. I did it on the side, Rupert Riedl was my advisor, and chick limb development became the topic, in particular the reestablishment of ancestral patterns, as we called it. I collected three papers for that dissertation and wrote a theoretical text around it. I was teaching human anatomy then – a lot, by the way, a huge amount of teaching – but there remained enough time during which I could do my research. I was free to do whatever I wanted, so I could pursue experimental embryology for my dissertation.

 

HS: Coming back to the paper, what are your memories of writing this paper – what was that like, and when and where you did the writing for this paper?

GM: I don’t have very clear memories of this, but since I had written on the subject before, I think the writing process was not so difficult. The only difficulty was compressing it all into a very condensed space. That is why it took this programmatic form. I wrote the paper at the Department of Theoretical Biology. In the meantime, I had moved from Medicine to the Natural Sciences faculty. I had received a full professorship there – after strong competition with other candidates – and I was hired as a developmental biologist, initially. However, during the time I arrived, a restructuring of the Austrian university system took place. The idea was to abolish the large institutes that existed below the faculties level. The full professors were asked to propose new departmental structures. I proposed a Department of Theoretical Biology. I found some mathematical biologists who were interested, some developmental biologists as well as evolutionary biologists from the former Institute of Zoology. I got them all together, and we founded a department that was focussed on the theoretical biology of evo-devo. This is where I was at the time when I was writing this paper. It was an extremely busy period, because I was trying to get a few projects off the ground, writing grant applications, solve issues with the new departmental system, and whatnot, you know…the usual thing. At the same time, I was heading the KLI, and I had two very young children!

 

HS: This would have been 2006?

GM: Yes, exactly. The workshop I mentioned took place in Venice, in 2006, and the paper appeared in 2007.

 

HS: Do you remember anything about the peer-review of this paper?

GM: In principle, the peer-review went smoothly. There was one reviewer who, if I remember correctly, lamented a lot about the fact that – in his eyes – I downplayed the role of genetics in evolution. And this in a journal devoted to genetics! He was quite outraged by this. But the other reviewer was much more favourable to my approach and actually predicted what the first reviewer was going to say, namely that it will be found that there is too little on genetics. But I was able to explain the reasons for why I took this approach, which, in fact, never intended to invalidate genetics. As an aside: together with other colleagues, once in a workshop on evo-devo we proposed to draw up signs that we would hold up during the discussions and that would say, “I believe in genes” while discussing non-genetic issues! We did believe in genes. However, we did not believe in a reductionist explanation of phenotypic evolution that was solely based on genes. That’s the only thing. I think that because I explained what evo-devo was really about, this was eventually accepted. Everybody could see that all the theoretical arguments were rooted in empirical research, showing that evo-devo was considering factors beyond genetics, and this was what I was talking about. Luckily, the editors were open minded and accepted this kind of reasoning, even for a journal devoted to genetics.

 

HS: I was curious about the two examples that you chose to illustrate the paper – the two figures. I was wondering, why you chose those examples.

GM: I can tell you exactly about this. The first figure I was using comes from one of the very first morphodynamic models in evo-devo that related shape changes of a developmental region to gene activation domains. This particular model was about the evolution of teeth and on how tooth shape influenced cusp formation. The model integrated signalling areas and shape of the tooth anlage, and it showed that modification of the shape can be used to predict where certain gene activation centres would arise. This was then compared to actual tooth development. So, there was the model that was used to create predictions, and these were tested on actual development in mice and voles. The model was able to predict where these signalling centre changes – the enamel knots – would arise. I used this model to demonstrate that natural selection on the shape and proportions of organisms elicits specific morphodynamic responses in the developmental system that is undergoing change. Natural selection acts as a general factor here, but the specific response and the phenotypic result that is the consequence of natural selection would be dictated by the developmental system and not by selection itself. This model was a beautiful example, which at the time was also one of the very first computational models that allowed parameter changes in a virtual developmental system.

 

HS: Was it easy to get permissions to reuse the figure?

GM: This was very easy because I knew Jukka Jernvall, who did this work. He, by the way, is now my successor as the president of the Euro-Evo-Devo Society.

The second figure is by my close collaborator Stuart Newman who had developed this schematic for a paper we had written together with Gabor Forgacs on the role of developmental physics in evolution. The purpose of this figure was to show that when you have given cell properties, such as differential adhesion, or polarity, or physical qualities, or whatever properties cells might have, when they come together in a multicellular context and combine with more overarching patterning mechanisms, such as chemical diffusion, or oscillation, or gradients, then, through this combination of the individual cell behaviours with the general patterning mechanism, certain arrangements of cells – certain structural motifs – result automatically, without the need of having them specified precisely by a genome that would say: you create a hollow sphere, you create a segmented sequence of cells, you create a star like arrangement, or whatever. Rather, such patterns arise automatically from the multicellular context. We called this – or Stuart actually, in an earlier paper  – called these forms generic. They are generic upon the properties with which those cells are endowed and would therefore represent primary configurations. Natural selection would then be able to seize on such generic forms, modify them, make things longer or wider or change the segment numbers or whatever. But in this mode, causally, the developmental properties are primary, and natural selection comes in as a stabilizer of developmental pathways, since such early generic forms would be very susceptible to external influences. A slight change of pH or calcium content, for instance, might change adhesivity and you could get a different form. But as a way of stabilizing against environmental perturbations, one could say that natural selection leads to a streamlining of the genetic-developmental arrangement that brings these forms about that, initially, were a generic consequence of the material properties of those organismal tissues.

So, I was using this figure to illustrate how the propensities of developmental systems contribute to the emergence of evolutionary innovations in the phenotype. Evo-devo really is about the phenotype. This is also something that is very often misunderstood, because population genetics and quantitative genetics work with population phenomena. Here, the question and the explanandum are different. It’s about how does the phenotype originate, and how is it transformed in evolution. This emergence of innovations is one of the topics in evolutionary biology to which evo-devo makes a genuine contribution, and which has not been addressed by the standard theory.

 

HS: I want to go on with the names of people who you have acknowledged to get an idea of your connections to them and what they brought to this paper.

GM: Yes, absolutely.

 

HS: The first name there is Werner Callebaut.

GM: Werner was a philosopher of science from Belgium and was also the very first scientific director of the KLI. After I became president, I asked him to not only visit occasionally from the University of Ghent, but to actually stay permanently at the KLI, and luckily he agreed. He was a fantastic analytical mind and was very interested in evolutionary theory and the related philosophical debates. He wrote a book on the naturalization of philosophy called “Taking the Naturalistic Turn.” He was also interested in evo-devo and the growing role of evo-devo in evolutionary biology. He wrote articles on these topics from a philosophical perspective. And he argued very early on that evo-devo represented a kind of epistemological shift from a correlational population theoretical approach to a causal mechanistic approach in evolution. This is what he concentrated on, and because he was my colleague at the KLI, I talked a lot with him about evo-devo. He likely read one of the first drafts of my paper.

 

HS: How did you get to know him?

GM: He had come to the KLI already in the early 1990s to work on evolutionary epistemology. I got to know him there, and, as I said, he later stayed permanently. We became very good friends, coauthored articles, founded the journal Biological Theory, organized workshops, and went to conferences together. His untimely death in 2015 was a major intellectual loss for the KLI and a sad loss of a personal friend.

 

HS: Manfred Laubichler.

GM: He’s a biologist and a historian of science. He’s at ASU. Manfred was a student of Günter Wagner, and I knew him already since his student days. He also had a very keen and very early interest in in evo-devo and published about it. We edited a book together – one of the Vienna Series in Theoretical Biology volumes – on modelling in biology. We also once taught a course together at the University of Vienna, and he is a member of the KLI’s Scientific Advisory Board. I had many discussions with Manfred on evo-devo and related topics, which is why I included him in the acknowledgements.

 

HS: At the time of this paper, was he still here?

GM: No, Manfred was in the US. The course we taught together was later, maybe around 2010 or so. He stayed for a longer visit and we compressed the course into a two weeks module.

 

HS: Stuart Newman.

GM: Stuart is a cell biologist at New York Medical College, with an outstanding expertise in the physics of development. Throughout the many years since I met him, he was, if you like, my most inspiring collaborator. We published a series of papers together, on innovation, on physics in development, and on what we call inherency, i.e. the causal importance of the internal properties of organisms for evolutionary change, and on other theoretical themes in evo-devo. So, in many ways, this paper was also inspired by him. I regard Stuart as one of the leading intellectuals in theoretical biology, and, after Werner Callebaut’s death, he has been the editor-in-chief of the journal Biological Theory published by the KLI together with Springer.

 

HS: Do you remember when you first met him?

GM: I met Stuart through co-organizing a symposium with him at a conference on evolutionary biology in Budapest in 1996.

 

HS: Massimo Pigliucci. 

GM: Massimo is a philosopher of science at the City College of New York, but when we interacted he was at Stony Brook. This is interesting. I met him shortly before this paper was published, and we found that we shared a number of views, especially on the shortcomings of the Modern Synthesis theory of evolution. I think I entered his name in the proofs, more as a reference to a kindred spirit who was equally interested in this topic area and with whom I had discussed these themes after the paper was written. Later in the interview we might talk about the EES [Extended Evolutionary Synthesis], and then I can mention more about him.

 

HS: Where did you meet him?

GM: I met him at the ‘Ishkabibble’ Conference (ISHPSSB) in Exeter in 2007.

 

HS: Jeff Schwar[t]z.

GM: He is a physical anthropologist at the University of Pittsburgh who had come to Vienna’s anthropology department for many years, which is where I met him for the first time. Later he also joined us at the KLI several times for workshops and visiting periods. Jeff had published a lot on human evolution. He was very critical of the adaptationist and gene centric view of the Modern Synthesis. He also was the convener of the symposium in Exeter, where Massimo and I met. The acknowledgement is about the general views we exchanged on non-adaptationist thinking, but we did not discuss this specific paper.

HS: Günter Wagner.

GW: Günter is an evolutionary biologist who, at the time, was at Yale. He is one of my closest and earliest friends in science. At the time when I started my dissertation with Rupert Riedl, he was already a doctoral student with him. He was also, I would say, my very first partner in advocating evo-devo as a distinct field in biology. And he was, of course, an outstanding theoretician of evo-devo with whom I had countless discussions about the conceptual issues that shaped the field. At the time, he had already been in the US for many years, and we had no direct interaction on this particular paper. I should say, that he probably wouldn’t have agreed with everything I was saying in it [laughs].

 

HS: Your department.

GM: At that time, I headed the Department of Theoretical Biology at the University of Vienna, and I included it in the acknowledgements in a general sense, because there were a number of people in the department with whom I continuously discussed evo-devo. After all, the department was about developing tools and concepts for evo-devo research. We developed 3-D gene expression techniques in developing embryos and things like that. One of the department members whom I thanked implicitly, then, was Philipp Mitteröcker, who is now my successor at the KLI.

 

HS: My next question is about the title. The part “extending the evolutionary synthesis”  is something that has attracted a lot of debate and controversy subsequently. Is this the first time, you used this phrase in a publication?

GM: That’s a good question. I believe so. But the “extending” terminology has a very interesting genesis. As I said, in 2007, Jeff Schwar[t]z organized a symposium at ISHPSSB meeting in Exeter, titled “What happened to evolution after the synthesis?” And during this meeting, Massimo Pigliucci and I discovered that we both had a paper in press that had the word “extending” or “extension” (of evolutionary theory) in the title. This was really interesting, and we thought it was probably more than a coincidence that in several fields of evolutionary biology the notion that the standard theory needed some kind of expansion or revision had become a recurring topic. Therefore, we decided there in Exeter to organize a KLI workshop together on the evolution of evolutionary theory in 2008, titled “Toward an Extended Evolutionary Synthesis.” It was meant to produce a volume that should be ready for the Darwin year in 2009 – the 150th anniversary, which we didn’t make, the book appeared in 2010 – but this is how the phrase extended evolutionary synthesis came into being.

Before I go on, it is really important to understand that there is a difference between the extension of something that exists already, and an extended theory version, in which “extended” is used in the sense of comprehensive, broad, inclusive, which is not meant to be the additive extension of a particular version of the general theory that already exists, such as the Modern Synthesis. This has led to a lot of confusion. In my paper on evo-devo, I still use “extending” in the sense of, you know, an addition to the standard theory through evo-devo. But through the subsequent reactions, and due to the workshop we held, and the people who spoke there, and the new results in evolutionary biology, we developed the feeling that, in fact, theoretical expansion happened not just in evo-devo. There were several other fields that are either new or are critical of certain aspects of the standard theory, such as inclusive inheritance, niche construction theory, epigenetics, and the whole environment-development interaction, etc. Eco-evo-devo was not known at that time as a term, but direct impacts of the environment on development were already considered. And evo-devo itself had further evolved. So, there were four or five domains in which calls for a theory revision were voiced. And now you get this phenomenon in theory development – like in biological evolution – that you can add only so many new elements and still keep the original theory intact. When too many alterations are made to a theory structure, the system kind of flips into a new steady state, as we would say in developmental biology, because the causal interactions between its individual elements assume a different logical arrangement. This is why, at that point, we called the theory “extended,” and not a linear “extension” of a previous framework. This is, in my interpretation, the relationship between the gradual extension of a specific theory version and an extended new theory structure, which then becomes an alternative. Which, by the way, does not deny the factors that composed the standard theory, but puts them in a different relationship with the other processes that are now included.

In summary, I am not sure that this paper was the first use of the “extension” terminology, since Massimo used it at the very same time. But the idea was in the air, others were talking of an expanded synthesis, or a new synthesis, etc.

HS: That’s really interesting. There’s a subtle but really important difference between the way it’s used here, and subsequently, say in the 2015 paper led by Kevin Lala.

GM: Yes, indeed, but it already showed in the title of the 2010 volume I edited together with Massimo called “Evolution- The Extended Synthesis”.

 

HS: At this point, when you decided to invite these people from different fields for the workshop, were you already aware that these people were sort of challenging the standard theory of evolution?

GM: Yes, but we didn’t know all of them personally. Of course, we invited many of those who were known to hold critical views on the standard theory, but at least two or three contributors did not share this position. We kept the workshop very open, there were exciting contributions, and, altogether, it worked very well. But we cannot claim that the book represents a uniform view. Also, by the way, this book, although it established the term Extended Evolutionary Synthesis (EES), does not contain a full account of the EES. Rather, it was a collection of concepts from different fields in which significant theoretical change was underway. Massimo and I wrote an introductory piece about how these elements come together, but it did not spell out a concrete new theory structure, as Kevin [Lala] correctly points out in his interview on your website.

 

HS: How did you know of these challenges in these other fields? Was it something that you were actively looking for, e.g. that in niche construction there is a challenge to standard evolutionary theory? Similarly, in inheritance?

GM: I was following the evolutionary literature and knew in which areas the major debates took place and who the advocates for revision were. In genetics, I was following the rise of epigenetics, because of my interest in developmental biology, and I noted that a push for revision came not only from developmental epigenetics, but from inheritance and ecological epigenetics as well, as I spell out in an article that has just appeared in a new book On Epigenetics and Evolution, edited by Carlos Bosagna. In my chapter I talk about the different versions and usages of epigenetics and their relations to the EES. About niche construction I learned from John Odling-Smee, who was present in the extended evolution workshop at Altenberg, and through him I got to know the important work by Kevin Lala whom I met later at the KLI, and who became a strong ally in the fight for theory revision. In our first meeting we discovered that we had very similar experiences with criticisms of the received position in our respective fields.

 

HS: Did Massimo Pigliucci know Odling-Smee?

GM: Maybe he knew him, but I am not sure. There were many other participants whom we knew and who had been at the KLI previously, such as Eörs Szathmary, Eva Jablonka, Stuart Newman, and Günter Wagner. There was also Greg Wray, who later was the leader of the opposing team in our debate paper in Nature. He’s a very good friend, we just disagreed on this a little bit, I met him a few weeks ago in Helsinki. From the evo-devo community there were also Alan Love and Marc Kirschner as well as several others from the fields of genetics, genomics, paleontology, and philosophy of biology, some of which we knew personally and others through their publications.

 

HS: And so, once the workshop happened, that’s when you and Massimo realized that this was bigger than just an extension, and that it therefore needs to be called extended synthesis.

GM: Yes. Of course, this is my version of portraying the thing. Maybe others would say there is no difference between “extension” and “extended”. I sometimes get angry remarks on this. See, I don’t care. I mean, this is the way we had conceived it. If others have other ideas, then that’s fine.

By the way, a similar confusion happens with the term “synthesis” in the EES title. We don’t claim this to be a synthesis in the formal sense. This is a label we use. We used it because Massimo and I, at the same time of preparing our book, were the editors of a new reprinting of Huxley’s volume on the Modern Synthesis. Since these two books were going to appear back to back  – same format, same trim size – we also meant to emphasize their similarity in bringing different fields in evolutionary biology together by using the Extended Synthesis title. We used “synthesis” not in a different sense from how Huxley used it, which was also no synthesis in the formal sense, but a joining together of theoretical concepts from formerly independent fields. This is what is happening here as well, an integration of formerly separate theory domains. Therefore, I don’t understand why people get so hung up on the term “synthesis”, like those who write, “the last thing we need is another synthesis.” The point is that you have to call a new theoretical framework something. Thus, we chose this title, but not because a more formal synthesis was the goal.

 

HS: It makes sense. You were just using the term that had already been used in this context.

GM: Yes, precisely.

 

HS: Do you have memories of reactions or responses to this paper when it was published? Did it attract attention of any kind?

GM: I must say that the reactions I got were generally very positive. The paper was very well received as an overview and a programmatic depiction. The criticism was, as this one reviewer foresaw, that genetics figures not prominently enough. I can live with that. But the point I was trying to make, namely that the field of evo-devo produces theoretical insights that affect our understanding of organismal evolution and the structure of evolutionary thought, was well received.

 

HS: I want to now ask you a few questions about what you’ve said in this paper and the way you’ve organized the field. Given that it’s now been 17 years since the paper was written,  I was interested in knowing how you feel about it today, considering everything that’s happened since then. One of the things you do in the paper is to identify four major research programmes of evo-devo: 1. comparative embryology and morphology; 2. evolutionary developmental genetics; 3. experimental epigenetics and, 4. theoretical and conceptual work. I would ask you two things in relation to this. One, would you still characterize the field in this way? And the other is what have been the relative fates of these areas? Do you think that there’s been more work in certain areas over others, and do you feel that there’s a need for some of these areas to get more attention?

GM: The first thing I would say is that a characterisation of a field like I provided in this paper is a very subjective thing. This is my way of partitioning things, but one could do it in entirely different ways. My goal here, in separating out these programs was to demonstrate that evo-evo is not a monolithic block with one set of questions or merely an empirical program of deciphering how the black box works. On the one hand, its questions are specifically related to evolution. This is very important, because sometimes people think that when they do research on gene regulation in one single organism, it is already evo-devo. But a concrete evolutionary question must be part of the approach, regardless of which methodologies it uses. I think my characterizations were useful to that end, to demonstrate that evo-devo has several avenues of attack, so to speak, and therefore this partitioning was sort of a didactical approach. I’m not wedded to it at all.

At that time, evo-devo really wasn’t that well established yet, although things were changing. Whereas many thought it was primarily about gene regulatory evolution, I meant to highlight that there are other domains of research and other questions that constitute significant parts of evo-devo, especially regarding the distinction between questions that address the evolution of development itself from questions that address the role of development in evolution. The evolution of developmental mechanisms represents a different set of questions from how developmental systems and the properties of these systems influence what can arise phenotypically in evolution. Thus, evo-devo is based on a dialectical set of questions. Whereas the former part is much better known – the part that looks at how development evolves, how, for instance, developmental gene expression changes, making certain organisms different from others, because of, say, Hox gene modifications. That’s wonderful, but it’s not the same question as how the collective behaviours of the cells that are affected by the Hox genes – in the context of other cells and tissues – actually determine what will happen in a modification of that system, say an eye, or a limb, or some other organ.

I think that a partitioning as I’ve undertaken there, is still useful today. For instance, we’re using it to a certain extent when we plan the programs of the Euro-Evo-Devo meetings. There, we’ve always tried to incorporate all these different areas of evo-devo – have symposia on comparative organismal development, on gene regulatory evolution, on epigenetics, on theory development, etc. This doesn’t happen by chance; it only happens through conscious design. If  the meetings would be organized simply by the number of submissions in certain areas, within a few years you would have pure genetics meetings! So, in that sense, there is even a practical use for that kind of partitioning of the field, because then you can structure certain projects such as meetings, etc. Today, I would make one change to those four research programs. I wouldn’t call the third program, “experimental epigenetics” any longer, because in the meantime, epigenetics has become so much wedded to the molecular genetics terminology. At the time when I used this term, I meant epigenetics in the sense of epigenesis, which is the term for a certain concept of development. Now, epigenetics is almost exclusively linked to DNA methylation, histone acetylation, micro-RNAs, etc. When you hear the term epigenetics, you immediately think about those things and not about what I meant to define with this category of the evo-devo program.

 

HS: I’m guessing that also reflects how the field has progressed over time, that there’s been a lot more work on genetics compared to the other research programmes? 

GM: Yeah. But that, again, is my perspective. Others in the history of evo-devo would maintain that evo-devo only started at the moment when Hox genes were discovered, and therefore it had always been a genetical field. Then you can draw a totally different history.

 

HS: But that’s not how you see it?

GM: No, that’s not how I see it. Not at all. Because historiographically it is clear that the core issues of evo-devo started to be discussed long before Hox genes were discovered, as Laura Nuño and I recently documented in the Evo-Devo Reference Guide.

How much research these four programmes have attracted? Well, there are, of course, different dynamics in these different areas. The comparative embryology and comparative morphology programme might not be called this way today, but there are numerous organismal projects in evo-devo that I think would fall under this category. They do not require the use of sophisticated molecular tools but rely on various forms of the comparative approach. In a way, I had been thinking that these kinds of projects were waning out, but at the evo-devo conference I attended in Helsinki a few weeks ago I noticed that this is not true. There were quite a number of very interesting projects that would fall under this label.

The evolutionary developmental genetics programme is expanding, of course, at an ever more rapid pace, and it would now include GWAS [Genome-Wide Association Studies] and genomics. It is a very successful program but also creates a difficult situation for the evo-devo field. At the evo-devo conferences you increasingly get talks where these beautiful GWAS clouds are shown, of huge numbers of dots in different colours. One cloud after the other follow at an incredible pace. It’s really important to determine, while you are watching, whether this sorts of approach addresses any evo-devo problem at all. If it does, fantastic. But often times the clouds remain in the clouds [laughs].

Experimental epigenetics, like I was saying, has taken on a new meaning. But there is a lot of epigenetics now in evo-devo, with different approaches addressing the different aspects of epigenetics I have mentioned earlier.

In the theoretical program, besides conceptual and mathematical approaches, I include computation and modelling. These techniques have gained by an enormous amount. There are many in-silico studies and wonderful models and simulations of development that were unthinkable at the time I wrote this paper, and they can be placed into evolutionary contexts. Therefore, I would say that this field is also very much alive and kicking.

 

HS: Similarly, I wanted to ask you about, the major theoretical themes that you identified: modularity, plasticity and innovation. Today, would you still characterize it in the same way? Would you add anything new?

GM: Yes. First, I should say that this section collected those topics of phenotypic evolution on which evo-devo has a major bearing. Maybe it was not really ideal to name that section “theoretical themes.” The real theoretical topics come in the subsequent section, which I called “Theoretical implications.” With “theoretical themes” I meant to characterize the conceptual areas to which evo-devo can make a significant contribution, and I think that modularity, plasticity, and innovation are good examples. Modularity has continued to be an important topic. There are books on modularity and a number of high level publications. The same is true with plasticity. Developmental plasticity and phenotypic plasticity figure in the titles of an increasing number of meetings and publications. Research on innovation, my personal favourite topic, is also present, but is not growing to the same extent as the other areas. You could say that modularity and plasticity have also been addressed by the standard approaches to evolution, especially in genetics, whereas innovation is a theme that truly arose in conjunction with evo-devo, and here evo-devo makes a very genuine contribution. Therefore, I would uphold these example areas, but I wouldn’t say that these are the only theoretical themes that are important.

 

HS: You were saying that maybe theory was not the right word.  In the next section, you talk about major departures of evo-devo from the standard theory and discuss evolvability, emergence and organization…

GM: Yes, these I prefer to see as the true theoretical challenges resulting from the evo-devo approach. With some caveats: I still believe that evolvability is a central topic in the sense of how development influences the generation of variation and the plastic phenotypes that arise as a consequence. But even in the field of evolvability, there is a certain division between the more developmentally oriented evolvability and the genetics based evolvability. And sometimes the term evolvability is used to pull evo-devo into the population genetic domain, only to argue that, implicitly, it had always been part of the evolutionary account. I’m a little wary of this usage of evolvability. On the other hand, it is, of course, correct that evo-devo contributes to our understanding of evolvability, if we define it as the capacity of organisms to generate selectable variation. Here development plays a central role, a role that is different from mere genetic variation.

Emergence is also very important, since all complex systems, when subjected to change, have emergent properties. They will react in some way to – in this case – the selectional challenge, but natural selection cannot foresee the developmental consequences, that is how the inherent properties of those systems that are affected by selection are actually going to determine the phenotypic outcome. Therefore, I like to say that the specificity of the phenotypic response is dictated by the developmental system under change. Emergence, of course, is a difficult term. It has a number of tricky philosophical connotations. But in evo-devo we use it in a very practical sense. For example, if natural selection affects, in development, the proportions of an area in which two inductive domains independently exist, then selection may bring those domains closer and closer to each other, and at a certain point there will be an interaction that is not “foreseen,” if you like, by that proportional change. Therefore, these emergent properties will have a decisive influence on what will happen in a developmental system that is undergoing evolutionary modification.

Organization is also an incredibly important topic. The truth is that the standard evolutionary account has no theory of organization, that is it includes no explanation for why and how certain phenotypic motifs and anatomical solutions arise whereas others don’t, nor why these structural entities are arranged in specific – also lineage specific – complexes. Obviously, an increasing number of independent or linked structural entities in a system become functionally and genetically interrelated and integrated over time, and evolution modifies and hones these interactions, but also fixates certain ones, which then cannot be undone by future rounds of selection. Evo-devo is the only field in evolutionary biology that addresses the origins of phenotypic complexity. In an approach I called “organizational homology” I have argued that this increase of phenotypic organization even reflects back on genomic organization, so that the organizational properties of the organism and of the phenotype actually have a causal role in organizing the genotype. For all these reasons, I maintain that organization is a crucial theoretical topic of evo-devo.

 

HS:  In the conclusions section of the paper, you point to two major future challenges of evo-devo. One is with regard to empirical testing of emerging new concepts, and the other is the integration of evo-devo with formal evolutionary theory. Could you reflect on how well you think these challenges have been tackled in the time since this paper was published?

GM: Let me start by saying that, in biology, a majority of research is caught up in the themes of a certain lab, of a research focus that exists in certain institutions, etc. And although scientists like to believe that research should be theory-driven, it rarely is. Only very few people actually start out with testing theories – while unconsciously they always do – which is cumbersome and often fails, and then you may not have any results to show, and so on and so forth. So, at the outset of posing research questions, people are often not bent on testing specific hypotheses that would be derived from an evo-devo concept. Nevertheless, in their work they are embedded in these conceptual domains and their approaches reflect back on conceptual development. Empirical work has exploded in evo-devo in the past one and a half decades. Thus, many of the challenges have been tackled very well. Evo-devo has made enormous advances, especially in the molecular domain, in the epigenetics domain, and in the computational domain.

With regard to theory integration, of course there are factions in evolutionary biology that take these approaches seriously. Here, by necessity, I will mention the Extended Evolutionary Synthesis [EES], because it currently is the only theoretical evolutionary framework that takes evo-devo and other theoretical advances into account and seeks integration with the population model. There exists not a singular formally unified framework of how the EES works, but there are individual formal models in each of its subdomains, like developmental patterning, or epigenetic inheritance, or niche construction, and so the EES represents more a family of models rather than a single formal entity. This, I believe, is a good state of affairs, and I am very sceptical of the vision of a single grand formal theory. Evo-devo has an important role in the EES as one of the cornerstones of this new approach. It contributes to our understanding of the generation of variation, of evolvability if you like, and defines rules for what can arise phenotypically under certain selectional conditions. The second major domain in which the EES introduces a new approach is the environment-organism relationship. Organisms – from bacteria to humans – influence the environment through their activities, and that environment will be the selective environment for future generations, which is captured by the model of niche construction. But the environment also feeds back directly on plastic development. This aspect does not exist in the standard theory, namely a direct environment-development interaction, something Mary Jane West-Eberhard called environmental induction. And the third area in which the EES differs from the received theory is inheritance, where, instead of solely genetic inheritance, the EES acknowledges three other forms of inheritance: epigenetic inheritance, behavioural inheritance, and cultural inheritance, as Eva Jablonka points out so beautifully in her writings. While acknowledging the conventional evolutionary processes that are constantly at work, the EES presents a different logic of the causal relations among these evolutionary factors, and, therefore, its predictions are different from those of the standard theory. With the classical theory, one could make predictions about what is going to be maintained under certain selectional  regimes, whereas with the EES you can additionally make predictions about what features are possible to arise under such conditions. From these predictions, new hypotheses can be derived that can be tested empirically. So, heuristically, the EES model is fruitful and pluralistic. Since the hypotheses derived from the standard theory have been tested 1000s of times, no significantly new insights can be expected from that. But the EES can take us forward to solving evolutionary problems that had not been addressed in the past.

 

HS: In my next question, I wanted to actually ask you about, you know, certain specific lines from the paper. My sense is that you’ve sort of answered it already in your other responses, but maybe we could just go with these lines to see if you have anything more to add. You say, “Given the correlations between differences in phenotype with differences in gene activation, a major line of evo–devo concentrates on developing a theory of evolving gene regulatory networks¹⁰. Further experimental proof will be necessary to determine the extent to which gene regulatory change has a causal role in evolution.”

GM: Ah, the further experimental proof. My point here was that the ever more detailed elucidation of molecular and gene regulatory interactions in extant embryos – you can do this research only on extant organisms – does not necessarily tell you much about the causalities at work when a certain trait arose in the first place, and why it is constructed the way it is. This is what I meant to say here. You can show experimentally how gene regulation works today, but since we know about developmental systems drift, it is clear that the gene regulatory apparatus can have been modified over vast periods of time and looks different from the time when the structures arose that it regulates today. Therefore, it can be too quick a jump to unwarranted conclusions to take an extant process and argue that it indicates the causal basis of how a structure came into being. Most likely it only reflects the extant version of how the development of this structure is regulated. This is why I said the causal role of certain gene regulatory processes for the origin of a character needs to be demonstrated.

 

HS: What kind of experimental proof did you have in mind?

GM: I think this experimental proof can only come from laboratory systems where you test the behaviours of cells that are endowed with certain properties that can be modified through genetic, or physical, or whatever manipulation. You can then show that a modification of, let’s say, the genome actually does the trick of giving rise to a form that we now call, I don’t know, a or whatever. You cannot test this directly, because you cannot do molecular genetics on fossilized embryos. But it can be done indirectly by testing the behaviours of extant cells, showing which genetic or nongenetic modification of their behaviours can give rise to a certain structural arrangement.

 

HS: Have such studies happened?

GM: Stuart Newman, whom I mentioned earlier, has been working very successfully along those lines.

 

HS: In the Conclusions you also say, “The evolution of modularity as an adaptive principle, if confirmed, should enhance a population’s ability to generate heritable phenotypic variation.” What do you mean by, “if confirmed”?

GM: The principle of modularity is thought to be of major importance because it makes evolution much easier. If modules can be reused again and again, it becomes easier and faster to modify an organism. And modularity has been shown to exist in numerous publications and books. I didn’t mean to say “if confirmed” as if these mechanisms were only hypothetical  but rather in the sense that better empirical examples could be developed. Maybe I should have said “if further confirmed.”

 

HS: And has this happened?

GM: Again, research has moved very much into the genetic domain, and modularity in gene regulation is absolutely confirmed. Therefore, it will also be true for development. But I must say, I don’t know what exactly we mean by “confirmed”. It’s a tricky term. The important thing is that modularity has been found on all levels of biological organizations, from the genome to the structure of macromolecules, protein-protein interactions, gene regulation, multicell composites, morphology, etc. I think it’s undeniably one of the basic organizing principles in organismal evolution. You could also say that there are many different avenues in how to generate modularity, such as – in the genetics domain – via gene duplications, or co-option of regulatory pathways, etc. Selection, of course, has a major role in creating modularity, favouring robustness or the response to environmental conditions. I think that’s obvious, and evo-devo is clearing up the developmental aspect of this. By “confirmed” I meant that it will be necessary to determine how much impact modularity has on the rates and the patterns of evolution. We need not confirm that it exists but rather its importance in evolutionary change. Maybe that’s the best way to reformulate my expression.

 

HS: Would you consider this one of your favourite papers, among those you have written?

GM: Well, it’s one of my most successful papers in terms of citations. But I wouldn’t regard it as one of my personal favourites, mostly because I find its programmatic attitude or style a bit clumsy. It doesn’t read as well as I like a paper to read. But, as I explained, due to the length limitations and the request to clarify the research agenda of evo-devo, it happened this way, and I think it achieved what it meant to achieve. But it wouldn’t be a favourite. I have papers on limb development, or on polydactyly, or papers on novelty, or even other papers on the EES, which I prefer. One paper I like a lot, which was cited very little, is on homeotic pelvic segments in regenerating amphibian tails, something that had not been demonstrated before, namely that you could create homeotic features in vertebrates that are on the same scale as a bithorax mutant in Drosophila. But this paper is completely neglected. I would count it among my favourites, but I don’t mind that the paper we are discussing has become sort of a classic.

 

HS: How do you relate to the themes and the topics that are presented in this paper today? Is this something that you still think about? Is this something that you still work on? What is the place of the topic of this paper in your life today?

GM: I rarely return the paper itself, because it was written for a different purpose. But it has helped me a lot to structure my own thoughts about evo-devo and why it is important. So, it is there in the background. Since it’s neither an empirical paper nor a major theoretical insight, but rather an agenda or a programmatic collection, it doesn’t play much of a role in my work today. But it sometimes works as a reminder of how certain topics become more important over time. With that number of citations, I will always be reminded of what it was like at the time when I wrote this paper, when evo-devo was not yet such an established field. If it has contributed a little to raising the awareness of the theoretical impact of evo-devo, then its purpose has been achieved.

 

HS: I don’t have any more questions, but before we end the interview, I wanted to ask you whether there is anything else you would like to add about this paper that wasn’t covered by the questions I asked?

GM: Maybe I should add that this paper helped me to also see the changes in other domains. It was through this paper that I got to know Kevin [Lala] and others who had read it. I’ve had a wonderful experience with these colleagues from other fields, such as in epigenetics with Eva Jablonka, or with Kevin Lala and John Odling-Smee in niche construction. That we discovered the parallels in our thinking also has to do with this paper to some extent. One could say that it served as one of the triggers for the EES endeavour.