 I'm Christopher Donahue, I'm the historian of the NHGRI, I also help out with the NHGRI History Genomics Program, and so Dr. Allen Love is the second speaker in the NHGRI History of Genomics and Molecular Biology lecture series, and we are extraordinarily pleased to have him. I'm going to just give a few lines of introduction, and then he's going to give his talk in the interest of time. So Allen is an associate professor of philosophy at the University of Minnesota, and director of the Minnesota Center for Philosophy of Science, and the Minnesota Center for the Philosophy of Science is the oldest and basically an extraordinarily important study center for the philosophy of science. So his research focuses on conceptual issues of biology, and much of his work has concentrated in concepts of innovation and novelty and evolutionary and developmental biology. He is also interested in issues that arise in developmental biology and functional morphology. He uses a combination of approaches to investigate a variety of philosophical questions, conceptual change, explanatory pluralism, the structure of evolutionary theory, reductionism, the nature of historical science, and interdisciplinary epistemology. Other areas of Allen's interest include the role of history in philosophical research and the nature of intuitions generated by thought experiments in the philosophical inquiry. And his talk is entitled, as you can see, Physics, Genetics, and Investigative Reasoning in Developmental Biology. And with that, I'm going to turn it over to Allen. Thank you, Christopher. Really a delight to be here. I'm going to jump right in so that I can move through the talk and have as much time for discussion as is possible. In some ways, this talk may be a little bit different in bringing up some more contemporary philosophical issues, as well as some of the historical issues. But I think that this is a great audience to hear from in terms of this material. So let me start with an outline of where I want to go. I want to talk a little bit about the opening puzzle, which is what I'm calling the Renaissance of Physics in Developmental Biology, and then contextualize that with a little bit of deeper background about the long history of interaction between physics and development, and then move farther forward into the 20th century to a time when what I term mathematical modeling fails and is, I think, really illuminating for why physics kind of backs away or is no longer as important. And part of that story is experimental tools flourishing in developmental biology. And then what I want to try and do is use that historical material to tell a philosophical story about what might be going on with the reasoning and help understand better why we got to the place that we did in the present. And then I'll close with a little bit of discussion about my own research and how some of this has been reframed by the analysis and maybe applies to some issues in philosophy more broadly that I think are interesting. So let's start with the puzzle, the Renaissance of Physics in Developmental Biology. So there's a lot of attention being paid to physics in biology right now. If anybody has been watching, you see papers coming out quite regularly. There's just one quotation from a paper a couple of years back. There's been a renewed appreciation of the fact that to understand morphogenesis in three dimensions it's necessary to combine molecular insights with knowledge of physical processes. Now interestingly that comes from this paper on the growth and form of the gut from Nature 2011. It's the Taven Lab, Harvard. The title is important because of course on growth and form refers back to a very famous book by Darcy Thompson on growth and form originally published in 1917, second edition came out in 1942, which tried to offer a mathmatized or physical style explanations of lots of biological properties including developmental ones. And so we see in this contemporary research a lot of interest in what is termed the interplay of gene expression and physical forces. And in fact I'm not going to read these quotes but I just want to sort of highlight this terminology of interplay shows up over and over again in the way it gets discussed. Here I've just tried to make that term bold in these quotes that what people are really flagging is a kind of interaction between the physics and the genetics and that's what is drawing people in. And so this terminology interplay has gotten enough cachet to then play a role in actually the repeated emphasis that's being placed on that in these different contexts by different researchers. So why this is significant has to do with the fact that for anybody who's been involved in developmental biology, genetic explanatory approaches are predominant. That's in part due to the way in which the field has developed out of some really powerful experimental techniques in the 1980s and that's some of what I'm going to talk about later and it leads to statements of this kind that you would find in a textbook, developmental biology deals with the process by which the genes in the fertilized egg control cell behavior in the embryo and so determine its pattern and its form. Now this is maybe a milder version of this statement but if we want a less mild version there's always somebody we can go to, Eric Davidson who just passed away last year. Here's the way he has phrased it, elements of the genome contain a sequence specific code for development, they determine a particular outcome of developmental processes and then most recently in the book that was published right before he died, developmental complexity is the direct output of the spatially specific expression of particular gene sets and it's at this level that we can address causality in development. So given that this is the situation and John Gerhard has summed it up for some researchers development can now be reduced to the interplay of cell-cell signaling and transcriptional regulation and some people have recognized this effect even in some sense obscuring a link between physics and biology as a consequence of the success, that's the puzzle. If you've got this genetics orientation that's so predominant and we know that physics is in the background in history where is this renaissance coming from? How is it kind of coming back onto the table in this kind of way? So let me step back in time and say a little bit about how physics and development have had some relationships in a deeper past. The most famous of these is an interaction in the late 19th century between Ernst Teckl and Wilhelm Hiss. Ernst Teckl many of you might be familiar with him appealed to evolutionary history in his famous idea on togeny recapitulates phylogeny to try and understand how development works. Wilhelm Hiss rejected that outright and actually favored physical explanations where you explicitly make an analogy with physical structures to try and understand what's going on in the embryo. So if you've got physical forces operating on non-living materials and those yield certain properties then if physical forces operate on similar living entities then you should get similar kinds of behavior. And so one of the examples that was given by Hiss was this comparison of two and three day old chick brains with a rubber tube folding under the pull of a thread and the point is that you're showing analogous behavior between the living and the non-living materials to try and give an account of why the living materials would behave the way they do during embryogenesis. Now you might think oh that was just the 19th century. Let's go back to this paper on growth and form of the gut. They basically did the exact same thing. So what you have on the top is the rubber somalicrum of the gut looping which was created by putting those two different materials together and then stretching and relaxing them and then on the bottom you have the chick gut. So again you're comparing in a directly analogous way. In some respects quite similar to what Hiss was arguing more than a hundred years ago. Now when Darcy Thompson was making his arguments he was appealing to different rates of growth and geometrical relationships. He thought the similar line of reasoning applied if the physical forces generate morphologies in these viscoelastic materials then they should do the same thing when they're operating in organisms. And if you could see this visually this would be part of the reason why you should become convinced of it. That sort of the visual representation of that which we've already seen in the case both of the present paper and in the past paper that's part of what you would find persuasive. So this is the comparison in terms of jellyfish structure that you might use between the liquid splash drop and that morphology and this is the way Thompson encapsulates that. He says the living Medusa has a geometrical similarity so market and regular as to suggest a physical or mechanical element in the little creatures growth and construction. We seem able to discover various actual phases of the splash or drop and all but innumerable living types of jellyfish. These analogies indicate at the very least how certain simple organic forms might be naturally assumed by one fluid mass within another when the physical forces play their part. Now I've highlighted in blue these might be can be these statements which indicate this is the analogy that you've not made a direct demonstration you're making a claim by that. So Thompson was clear that you couldn't just have physical forces and this is an important part. This is not about somehow being in competition with genetics because one of the things Thompson was dealing with was the rise of classical genetics and the power that came out of the Morgan lab and chromosomal theory that was developing in the early part of the 20th century. And so he is clear it would be an exaggeration to see in every bone nothing more than a result of direct physical or mechanical conditions for to do so would be to deny the existence of a principle of heredity. But it would still be an exaggeration if we neglected the direct and mechanical physical and mechanical modes right that you can't neglect either one and of course that's part of what makes this interesting about how we've gotten to the point of today because it does seem like there has been that neglect at least in points of the history. Now in the middle of the 20th century just to sort of end this potted history there were further physical approaches to developmental phenomena probably the most famous as Alling Turing's model of spatially constrained reaction diffusion mechanism to get coloration patterns in animals. And again you see similar kind of reasoning in his paper from 1952. It's suggested that a system of chemical substances called morphogens reacting together and diffusing through a tissue is adequate to account for the main phenomenon of morphogenesis certain well-known physical laws are sufficient to account for many of the facts. It's not that you've actually demonstrated this in the system you've just said this kind of physical mechanism is capable of accounting for these behavioral patterns. Now I think what this highlights is a really important part of the history that we can glean that all this explanatory reasoning is analogous right so you're saying if it happens in the physical systems this way then it should happen in the biological systems this way but you actually are not demonstrating it in the biological systems in most of these cases they're how possibly explanations not how actually explanations might be can be could be you get that kind of language and then there's a sort of assumption of asymmetry between the physical and the biological explanations right so if physics is sufficient why would you need to do anything else right that's that's kind of an implicit premise in some of the the argumentation there's nothing further to do if the physics was sufficient and I think this X KCD cartoon captures that it might be a little bit hard to see I'll read it so the one says to the other you're trying to predict the behavior of complicated system fill it in just model it as simple object fill it in and then add some secondary terms to account for complications I just thought of fill it in easy right so why does your field fill it in need a whole journal anyway and then the bottom caption is liberal arts majors may be annoying sometimes but there's nothing more obnoxious than a physicist first encountering a new subject so I think this is a little bit in the background of this reasoning and this is a funny way of putting it okay so that brings us to much more recent history and this idea of mathematical modeling failing as I'm calling it's in quotation marks because the failure is not so much that the models themselves were bad but rather that they didn't get traction in the research community a lot of those were variations on Turing models which you see here some of that represented on whether you have one or two morphogens and what kinds of interactions they have between them what kinds of spatial patterns they can make and you can do the mathematical modeling for these different combinations and tune the parameters and see what possibilities are there and through the 1980s you see a lot of these papers being published and they're being published in places where developmental biologists would work so journal of embryology and experimental morphology that journal doesn't exist anymore because it was renamed development okay so you know this is happening during a time when you would think oh this is what a developmental biologist is reading and you can see these are models for vertebrate limbs segmentation and the like this is Hans Meinhardt one of the key architects of some of these models in that period another group of people George Oster James Murray again lots of modeling of the limb you can see that it's being published in these standard places development developmental biology and some of it is even experimental and I've picked out one individual who I think is interesting but who hasn't received that much attention Albert Harris who was not simply drawing physical analogy but in some cases trying to do experiments with physical substrates to see if you could make this argument a little bit stronger at the same time if you go in and you look at the language in these papers you don't have to read the whole quotes just look at the blue because the blue again indicates that in most of these situations you have a similar sort of claim as we had 100 years before right how they can how these processes could be the cause right that same kind of language and here from the paper by Albert Harris this is showing some of the images that he has he has cells that are creating traction force on a thin silicone rubber and so you can see the rubber wrinkling that's what he's showing in these images our observations apply that cell locomotion and vivo could generate large tension fields and then you would oftentimes get these additional sorts of claims it would be unlike evolution not to make use of these fields to guide morphogenesis but of course you were still making a how possibly explanation by analogy. Now looking a little bit more at Harris I think helps us see the transition that was taking place during this time so this is from a letter that he wrote to his advisor John Trinkus who was a embryologist at Yale this is one of the things I'm most interested in modeling is the limb bud in the mechanical relations between the elasticity and our contractility of the ectoderm to the shaping of the bud okay so that just gives you his general physical orientation my main working assumption is that the peculiar shape of the cells of the apical ridge results from some peculiar mechanical properties and this is the key to the shaping of the whole bud. Now part of this reasoning the key to the shaping is causal reasoning right that this is the explanation why it takes the shape that it does. Harris encountered a lot of resistance and that frustration was with molecular geneticists in particular those working on the vertebrate limb. This is from a later letter he says John Saunders tells me he's on his way to the fifth international conference on limb bud development I was at the third in the series five years ago and was infuriated by the tunnel vision of the molecular types they are just so blind to mechanics they see things their way and only their way and the flood of solid data their techniques is producing reinforces and validates their narrowness so you can't argue with them. Now I don't want to get into the sociological dynamics here I want to highlight the blue text in the letter right which I think is an important feature of what was going on at the time the flood of solid data their techniques were producing and this showed up in another dimension of Harris's reasoning which was interpreting all of the Turing type pattern formation genetically especially Wolpert's French flag model which I won't go into detail but the key thing is that when Wolpert talks about positional information what he ends up doing is emphasizing genetic properties that is cells ability to interpret signals as opposed to the reaction diffusion mechanism which is a physical type explanation and this is the way Wolpert put himself cells acquire positional identities in a courted system and then interpret their positions to give the spatial patterns and this is what do they mean by the flood of solid data? So I think what Harris is acknowledging here is that from his perspective as somebody in the field at the time he is not being critical of the fact that the community is generating data that is relevant to what they are trying to do. Well in this case I think he has in mind developmental biologists who are starting to use molecular genetic methods. Is this like a speed experiment on the FGF on the limb but this is on molecular biology? And Cheryl Tickle, Cliff Taven. So this wasn't like a CDNA sequence in EST generation? No. It could have been that. So it's not that enabled these guys? Right but this is a little bit earlier so that you're right it's not the EST it's more and it's also early in situ hybridization which I'm going to come to which was coming out of the blocks at this point in time. Trinkos went to one of the first zebrafish meetings and complained that we were spending much time generating too many cheap fewns but then we just studied one. I can talk more about Trinkos, Trinkos is not the focus of this but I can talk more about him because I've looked at his stuff. So anyway just for reference point this is the French flag model where you get the different cells interpreting their identity based on where they are and differentiating now this is a nice thing about working in archives you get to see some things that you might not have seen otherwise this is a handout that Harris created and it is meant tongue in cheek differentiating cells become organized in space by first learning just where they are and then interpreting this information according to their genetic program how else could they possibly do it and of course he wants to say this is not a good explanation and he's criticizing the kind of Wolpert orientation and then it gets a little bit more extreme in another one remember if cells or groups of cells tend to round up become spherical this is caused by surface tension which is a force of nature like gravity and also it means that their components are trying to maximize their contact with one another and then of course the side parenthesis say if these turkeys knew any thermodynamics I wouldn't have to be telling them this why do I bother. So I mean this is this is just a way of sort of showing Harris's frustration in some of the ways he expressed it in cartoons that he really felt like he was having trouble communicating the significance of these physical dynamics to the developmental biologists of the time now here's one way of seeing this I think more concretely outside of Harris's own personal reaction this is from a news and view in nature in 1989 very famous paper on Drosophila segmentation and if it might be hard to read but if you look within this right the first thing to highlight is this periodicity might be generated in one of two ways an elegant mechanism favored by model builders would use an intrinsically periodic patterning interaction with a gradient alright so that's the first claim and then if you go a little bit farther on you get this but the protein products of the gap genes are not so precisely localized okay so the modelers would do it this way turns out that's wrong okay and of course this is why the title is what it is it's supposed to sort of push a little bit on the if you make elegant assumptions you could be making bad assumptions Sean Carroll I think captured this explicitly in his 2005 book on evo diva with many theoreticians sought to explain how periodic patterns could be organized across entire large structures while the math and models are beautiful none of this theory has been borne out by the discoveries of the last 20 years the mathematicians never envisioned that modular genetic switches held the key to pattern formation or that the periodic patterns we see are actually the composition of numerous individual elements right so there is the could be that way just turns out it's not and since it's not what's the best way to get at it it's to do these genetic experiments and then this is a nice quote that Evelyn Fox Keller got from a proposal in the mid 90s the physics of how embryos change shape is neither an important or an interesting question and I think that that kind of statement is interesting precisely because it doesn't mesh with what we saw for a long time people have thought this was really important interesting question so it's almost like there's a moment in time when it wasn't thought to potentially be interesting and Albert Harris ended up being right in the middle of that time and experiencing that so that brings us to kind of fill out the story a little bit in this I'm going to go a little bit fast because I'm talking to an audience who knows these things in detail what we are seeing I think in this period is the application of these recombinant DNA tools in developmental biology and so the molecular biology tool kit all of that plus ways to visualize differential gene expression as we see an institute hybridization ways to determine if a gene is necessary for a process such as with knockouts and knock downs and ways to determine if a gene is sufficient for a process such as expressing it somewhere else most famously in a place that you would not expect it to be such as an eye on an antenna okay and this is all happening in the 1980s within developmental biology because that's when these tools are becoming standardized in a way that everybody can start to use them and apply them and get this solid flood of data that is being referred to by Harris now I'm going to concentrate on institute hybridization because I think it plays a special role in developmental biology as a way of visualizing it I won't go over the process because this is a cartoon version meant mostly for audiences who don't know the biology but the important part is it's a way of labeling a segment of DNA and ascertaining to what degree a particular gene is expressed at a particular time or stage in development and a particular location now if we analyze the growth of institute hybridization just by doing a pub med sort of search you can see and you can compare this with other key terms they do not show the same pattern this dramatic rise of institute hybridization basically at the end of the 80s and into the early 90s and so just to kind of map what we've been looking at right those mathematical models are flourishing in this period right there's making stripes in elegantly right it's kind of at the beginning of the growth curve and by the time you're here the physics of how embryos change shape is neither important or interesting question and I think this is a really you know helpful angle on understanding why it would be the case that people would no longer think that these kinds of questions are important in part because they're able to experimentally access genetic features of development in this way so what to make of all this right so in a sense that's a potted history because I'm you know giving you very you know quick looks at 100 years ago and then a little bit of a zooming in on about 25 30 40 years ago but what I hope I was able to convince you of is that this relationship between physics and development is old it's been around a long time that style of reasoning that's analogy is important that people have constantly appealed to it and that there was a particular moment in the 1980s when that really exploded in the face of the mathematical modelers because it turned out that in some of these experimental high profile experimental cases what actually was going on was very different and that that condition the way people thought about it well so what to make of it now I think we need to understand how important the molecular genetic approach is in developmental biology if we think about in classical genetics it was the primary investigative reasoning strategy as Ken waters has put it you discover naturally occurring or artificially produced mutants that exhibit a difference relevant to some biological process and then you carry out a genetic analysis of the mutants it's something that's still being done today but that was a real innovation at the time and it was based on the ability to track differences in a gene that caused uniform phenotypic differences in specific context change the context and the effect of the gene might change and so part of the research strategy is making sure you can control those contexts and then that was reworked molecularly obviously in the early days of classical genetics it was not a molecular approach but it was reworked as such and gave the capacity to manipulate and investigate many processes a dramatic increase in the period that we're talking about right now for developmental biology especially now what's going on in this case is what Ken waters refers to as investigative reasoning right if you think about the classical genetics examples what they could actually explain was maybe not as dramatic as what they could manipulate okay they could do these amazing triple crosses and you know get these mutants to exhibit certain phenotypes but in terms of our say understanding of how that phenotype was produced it was still relatively limited so I think that one of the things that we really need to highlight is that the transformation that we're talking about is a transformation in terms of how you do investigation not necessarily in how you offer an explanation and here this is something that's important for how philosophers think about science because philosophers typically assume that scientific knowledge is structured by explanatory reasoning and that research programs are organized around filling out a theory okay that's kind of like well what would what else would scientists be doing what turns out I think a lot more okay and it's oftentimes these investigative strategies which guide day to day reasoning what is that practical know-how about doing experiments being able to maintain stocks produce mutants having basic descriptive knowledge of some causal regularities and evaluating how that's going to help you leverage further research okay not so much having an explanation but having a tool that lets me do more now I think one of the things that is going on is that the explanatory potential of the physical approaches as we track them did not change a lot through the 20th century but what has changed is the precision of the physical manipulations that you can do in developing systems and that has increased the practical knowledge that you can look at those physical variables during embryogenesis okay it's not because we somehow now have a very rich physical explanation of development it's that we can manipulate development physically in a way that was simply not possible and was not what the mathematical modelers were doing so I think a central reason for the renaissance of these physical approaches is their amenability to experiment them in a vision on analogy with mutational analysis in the genetic approach you're applying the same kind of causal reasoning to the physics that you did to the genetics that's what changed okay this kind of reasoning I won't go over it in detail it's been described in some philosophical literature as causal reasoning in terms of difference making where variables represent causes in which you manipulate the value under certain contexts and then figure out different ways in which you can establish certain things will happen under certain conditions it's a allows you to presume context as long as you can hold it stable and that's the key part of it is being able to control variables and you can recognize many causes because you can manipulate one cause and hold other things fixed and then you can hold that cause fixed and manipulate something else so that's what researchers are deliberately doing they're constructing experimental situations where they can do those kinds of interventions and establish whether or not a particular factor makes a difference in a particular context so one of the things that has been highlighted in this discussion is the difference between an actual difference maker that is something that is making a difference in a population as opposed to something that could so something that could right well it might in a population if you were able to see the value of the variable change if it actually does it's because it changes and that makes the difference okay the important thing is that until recently all of that discussion about physics was about potential right it was about this could possibly make a difference right but we haven't experimentally established that it actually does all the mathematical modeling is in terms of potential difference making and that was the primary strategy so I think this helps explain the rise of physical approaches because in a sense what's going on is you're actually able to manipulate physical variables and show that they make a difference and that just wasn't possible in most experimental systems before that now very briefly let me illustrate this with a case that's now a few years old on fluid forces in cardiogenesis here what was important was the ability to do quantitative and vivo imaging that allowed you to track the flow and then modify that flow via some kind of occlusion that was stable and reliable and you could then measure the effects and importantly it's directly analogous to the measurement of gene expression differences through imaging you have to be able to see it so similar to an in situ hybridization you have to see what the intervention does and you change the valuable variable by over expression or knockdown so the occlusion is doing something similar in that way so this is just some of the images from the paper showing the schematic of the occlusion here's some of the way the authors themselves described it embryos with impaired cardiac flow demonstrated three dramatic phenotypes their hearts did not form they liked heart looping this is what's fascinating this latter phenotype is reminiscent of the zebrafish jackal mutant which demonstrates abnormal blood flow I think this is a really important part of the reasoning here is showing we're doing something just like the geneticists are doing in creating mutants but we didn't modify a gene we modified a physical force that's what wasn't being done before so I think developmental biology uses the genetic approach because it lets you get at actual difference makers but that doesn't require it to be a genetic one if you can manipulate physical variables that way they too are susceptible to that kind of analysis and if you can do with physical forces then you could show it and so that directly parallels gene expression visualization and mutation analysis that standard in the genetic approach I think this is a lot of what accounts for why physics gets a kind of renaissance in developmental biology it's not because it somehow got recognized again as an important explanation it's because we could for the first time manipulate it in a way that we could the way genetically we did with the systems already all right so that brings me to my last section and this is going to explain the pictures of the zebrafish that were skeletons that were on the first slide so for those zebrafish researchers who were curious what what I was doing I'm going to explain so one of the projects I've been engaged in is trying to look at how biologists put together physical or generic explanations with genetic ones and something I've been interested in it shows up in this case and in others but you'll notice that the way my research is framed is in terms of explanation right and to some degree I was initially misled in trying to understand this current situation because I was looking for those explanatory models rather than looking at investigative reasoning okay now you might think well why why were you misled well if you look biologists seem to be saying we needed integration right field needs increased integration between cell biology biomechanic analysis integrating biomechanics with genetic analysis right but what was happening was I was tending to read those as explanation not investigation right and so I think that an important part of the analysis here is really reframing some of these questions in terms of investigation that is the call for bringing together physics and genetics is not necessarily a call for bringing them together an explanation but rather bringing them together in investigation in the tools that we use to get at systems that doesn't mean there isn't integration or some kind of thing to think about there but it's very different in its structure I think so that leaves an open question of why the explanatory power of physics has not been more actively countenanced by developmental biologists and I'm still exploring that I think that's an open question but I have a hypothesis and that's that the standards of ex investigation explanation are aligned in genetic approaches to development that is there's a sense in which they run on the same rails what's not clear is that the standards of investigation explanation run on the same rails for physics and if that's the case that would be one reason for this kind of discrepancy now you might say well what would align them and here I think the explanation might be very pragmatic and if we go back to that fluid flow paper right and they talk about the experiments that lead to those effects they emphasize this approximately 40% of congenital heart defects involve a valve abnormality our defects in hearts with no genetic lesion suggests that a critical role was played by blood flow induced forces during normal heart development and suggests that altered hemodynamics may contribute to the cardiac phenotype in some cardiac mutants and perhaps also birth defects so here it's not that you're offering some kind of comprehensive physical explanation which are actually saying this might be medically relevant this might have medical payoff a pragmatic relevance and in fact that's why those fish are there because they are an example of this being done recently in zebrafish where they were able to produce mutants that exhibited this very dramatic spinal curvature by changing the fluid flow of the cerebral spinal spinal fluid and seeing a kind of model of idiopathic scoliosis in the zebrafish and so if you look at the way it's framed here in this most recent paper irregularities in cerebral spinal fluid flow represent the underlying by cell biological cause of idiopathic scoliosis right and so this idea that you need to reexamine the anatomy the physiology the genetics in terms of flow so maybe it's medical impact that could really cause some of the realignment in investigative and explanatory standards and that brings out something that may be more of interest to philosophers but I think it's worth talking about in in this context which is many philosophers distinguish between epistemic reasons or criteria like accurate representation and pragmatic reasons like successful medical treatment and they think that the epistemic considerations are more important that you know it's more important that you somehow represent things correctly whether or not you can change the world or treat people would be secondary but if we remember that investigative reasoning yields knowledge that's typically categorized as pragmatic that's going to suggest that there are standards that are dependent on pragmatic criteria in a way that philosophers have not paid as much attention to and so I think they may need to be blended together because what we're seeing is that investigative reasoning is central in lots of sciences I've talked about developmental biology and so I think we might want to resist pulling these two things apart and actually start talking about them together that in many cases we are finding that a joint justification in terms of pragmatic and epistemic might be what is going on and if so that might be a way to understand how you could align the investigative explanatory standards for physical approaches to development instead of the how possible sufficiency that we saw through a lot of the history we can focus on how actually justified by a combined pragmatic epistemic rationale where that rationale includes the explicit ability to treat medical pathology so I won't go back over but this is just a reminder of us moving through the sort of renaissance of current physics through the history to an interpretation of why that might have been so and I'm happy to take your questions in that regard thank you so questions comments I don't think you touched on this but I'd be curious if there's any correlation between people who formally trained in physics and then crossed into yeah well no biology really or any of the more biological sciences and sort of how they described the challenges or how they described how I mean yeah and what sort of influences the cross training might have had on their views of some of the so I think this is a really important point because it is quite clear that in many of these laboratory contexts the ability to manipulate the physics is coming out of a kind of hybrid of having a physicist there who has some kind of training but that training is about the manipulation of the physical variables not manipulation of the animal or the plant and so you need a fusion of those two things and so in most of these cases that I've studied you have that clearly represented somebody has come into the lab with a physics background and has some of that but they themselves are also only able to do what they're doing because of the experiment experimental expertise that those working on the model organisms already have in terms of how to manipulate some of the genetics as well but none of the none of the prominent developmental biologists that you talked about none of them have formal training in physics almost all of the prominent people in these situations are primarily genetic oriented in their training and then they have picked up the physics through a postdoc or through bringing something for somebody else in so there are very few researchers Lance Davidson would be one but in his case he ends up working in engineering more so than in developmental biology now this is a really important point for another reason and that is that in talking with some developmental biologists about this off the record there are complaints that some of the physicists who have come in don't really know how to do the experiments right and and so you get this issue about whether or not the standards are really being met and there you see the genetic standards the standards for what counts as a genetic manipulation and experiment are in some sense governing what will count as a legitimate experiment with respect to physical factors I think some of that's backing off but it's there and it's there for the reasons that you talked about that you have this differential expertise so how far back do you push the line right so you make a physical block in the heart and parts of the heart don't form right and parts of the heart don't form because there's a genetic cascade that's a result right so yep so when do you stop calling it physics and you push it back to saying this is a genetic response right so this is part of the reason why I see this project as combining the two that is it's not about sort of fighting is it really physics is it really genetics but it's somehow about putting the two together and some of that has to do with time that is how time is organized in the way you study if I start my study here then I'm going to look at factors as they move forward but I could have started back further and then the causal variables that I want to focus on could be different they could be more they could be genetic they could be physical and like what's interesting is that at least in the in the different cases that I've looked at no matter where you slide that sort of temporal window and you can you can find those interactive dynamics possible in fact some of the fascinating stuff coming out of work by Michael Levin at Tufts on the role of electrical gradients I think is a good example of this in symmetry breaking in particular shows that the physical dynamics can be of a certain kind they can be very early in in a process I think at the end of the day the important point is not sort of a fight over whether it's best characterized one way or the other but how we construct models that allow us to put them together and that that actually turns out to be somewhat tricky because you in this is a problem with the on growth and form of the gut paper which is that so this paper begins by talking about how we really need to combine them and then the whole paper is about the physics of the gut and it's not about combining them and so this is I think part of the reason why this is intriguing as a philosopher is I think there's a desire to combine them but actually doing it is quite tricky and hard and that's still being done and some of the strategies I think involve choice of windows of time and how judicious that is like you know if you narrow it in a certain way you can actually get a good account of the back-and-forth dynamics but if you widen it too far or if you try to take into account to to larger region spatial region of an embryo or something like that but those are I mean it's part of the difficulty for sure it's good question so your discussion has been about almost entirely embryological development and I'm wondering if do you or do the people involved in this debate or these controversies regard embryological development as a special case of all development including later development of a fetus and development of a young child because the reason I asked is because you know physicians have known for centuries if not millennia that physical forces relate to development you know the child who doesn't walk will not end up at age 21 with normal leg bones they don't grow because of the physical you know something related to weight-bearing and so on there's many other examples so is this something special why should it why should it not be the case why why should this be special why should it not be obvious that physical forces are related to some forms of development well so I think that the so the recognition that the forms are related or that physical forces are related in some sense to development is always been there right so this is part of why it's a puzzle right I mean in a sense all the way back in the 19th century people recognizing physical forces play some kind of role and that's manifested in different models over time Darcy Thompson Alan Turing and these others and I think you raised the point that it's also when you think of development on a longer life cycle type scale it's clearly relevant in these other contexts I think it has to do with the experimental manipulation context of developmental biology I think that's one of the key things going on here is that the degree to which the community built up a set of standards for what counts as an experiment that's legitimate that you can say this experiment gives me a result I can trust and once that standard was set up then you had a benchmark that's allows you to say okay the degree to which I can acknowledge the physics playing a role is a degree to which I can show similar to the way I've already shown in genetic manipulations that it makes a difference in this way and that's a relatively new invention now in the case of the physical forces for later stages of say human development or any animal for that matter there you have an issue about a different kind of investigative context that that's not happening in a laboratory in the same way you're oftentimes doing longitudinal prospective studies following cohorts again I think that there's going to be a question about whether or not the standards that establish one type of cause are applicable across the board but I think that the case I've been describing there's this key element in the recent history that makes the kind of putting physics at arms length it helps explain why that happens so something that I haven't understood maybe I just wasn't getting it but it sounds like you're saying the you know the precision tools for looking at the impact of physical determinants of development weren't there until recently for the modern era for now you said now they now we have more precise tools to do these experiments we didn't have them before not sure why so much precise tools so it's the the issue has a lot to do with whether or not the way you can manipulate the value of the variables meets this particular strategy about creating mutants and then being able to recreate them in same fashion over and over again it's that's the parallel so so it's not causal reasoning generally it's causal reasoning in genetics and it has these particular dimensions that I think show up in this experimental reasoning and that's what was not being done with the physical forces so little confused maybe I'm dense but could you just give an example or two of things that people are doing now that measure the effects of physical variables that couldn't have been done 30 years ago that couldn't have been done 30 years ago okay so some of it so if you go back to the case that I described about the cardiac fluid flow in the zebrafish heart okay that could have been done 30 years ago in part because it required the ability to have the high-speed camera imaging to be able to track the flow and the changes in the flow you and the the implantation of the bead might have been possible if you had that but you wouldn't have been able to reliably track what it was doing other than messing things up I think and so that's the kind of thing I'm highlighting that's a well you know I'm thinking that there are there are things that maybe they were done I'm just curious about this but one could have looked at decades or centuries ago in small or development of small organisms the effect of centrifugal force develop you know development in centrifugal force yes in space we've been going into space taking organisms there since 1960 surely somebody must have been looking at the development of organisms actually they did and so so here you have an issue about whether an experiment has been done and whether a community has coalesced around a set of methods so in the 18th century I'm gonna forget his name he he designed basically a water wheel for plant growth and was able to show by changing the the gravity dynamic that it changed to the way the plant grew so he established there's a physical force effect on the development of the plant so yes you're right you can you know show that that doesn't spawn a sort of community of research working on it it's kind of a one-off and I think one of the things that's important here is about about how communities develop around certain standards and that's true especially of a couple of figures if you think about the people in the early 20th century who we highlighted people like Darcy Thompson or Alan Turing these are people who kind of they're independent geniuses they're oftentimes working across different fields they're not well ensconced in any particular community and and so I think that this development of a community standard is an important part of the story not it was somehow physically impossible to ever do an experiment of that kind though clearly some of the technological innovations are you know dramatic I think it's it's also a story about the social organization of the research that's important I don't know if that no I think that it raises a lot of thoughts because what I see in science now is a lot of fad science you know ideas get ingrained they may be the wrong ideas they may not be the best ideas but they just take hold it's like an epidemic of ideas and then a whole bunch of scientists waste time doing minimalist research and not addressing big questions I'm an infectious disease guy I see that all the time yeah it's a you know sociological phenomenon I guess something about group behavior I do think that you I mean one of the things that is true of post-World War II science that starts to be funded centrally by the federal government you you clearly have a different dynamic than you had prior to that when you have different funding structures and so you're going to have susceptibility to group think and things like that that emerge as the community gets larger and is reliant upon say a central source of funding or something like that so it's not going to be the case that simply because a community has coalesced or is chasing a particular item that that somehow automatically justified I think in the present case what's of interest to me is the fact that you have something that has a long history of people thinking this is important it's interesting people offering these mathematical models and yet there's a this really distinctive phase transition that happens where all of a sudden it kind of fades but it fades in part right at a time when this set of experimental tools really solidifies the community of developmental biology which of course includes model organism entrenchment you know in terms of which models you're committed to working with it and the same thing so that's that's it that's not going to be the same if I move into infectious disease or other places are going to be different historical dynamics that will create that and that does create fads I'm sure the developmental biologist the room will say there's plenty of fads to highlight as well can I just ask one last question yeah with respect to this perhaps group think or people going off and trends and fashionable areas of thinking reasons of thinking what do you think is involved in that is it professional societies or disciplinary training or disciplinary identity why does this happen because most of us at least nowadays most of us are individual scientists we get individual are one type grants we don't necessarily have to be sheep and a big herd what's making this happen whereas it didn't happen very much in the 1800s or too much to a lesser extent to a lesser extent I mean I think that's a that's a really good sociological question that I can't give a good answer to and in short order in part because I'm not a sociologist and so don't know well enough the kind of social dynamics that would be relevant but what I do think is that the kinds of things that we're highlighting for how groups of scientists might behave are clearly going to not be only manifested in science they're going to be patterns of social behavior that humans adopt and that once we have the you know relevant analytical tools you know we can make the comparisons about why under certain dynamics like the last say 30 or 40 years you see this trend occurring in a way that you didn't see it so visibly before that it's also linked to specific technological breakthroughs right I mean you know a PCR CRISPR cheap sequencing these things drive specific you know it's looking for your keys under the lamppost even though you lost them over there because the lights that lights here that story I think these some of these fans are driven by there's all of a sudden a new way to look at the problem you've been tackling for generation and so everyone moves to this new technology because they haven't they didn't have access to that way of answering a question before and so I think some of its it's not even cheap mentality it's just like all of a sudden there's a release of tension you know you've been blocked it to addressing this question for so long that all of a sudden you can do it everyone moves into that area because they can for the first time which is what the solid data yeah absolutely that's yeah that's right that's absolutely right and that said that this idea that that technology is driving this group thing is also applies for 18th century optin microscopy most of the physicists that got into developmental biology they were all microscopists they were developing new ways to look they weren't manipulated they were manipulating right so any and we're slightly over time so any any further quick questions all right thank you very much for for coming and it was a wonderful talk one thank you