Saturday, July 2, 2011

The meaning of “theory” in biology

by Massimo Pigliucci
So, I’m spending a long weekend at the Konrad Lorenz Institute for Evolution and Cognition Research (KLI), where I have co-organized (with Kim Sterelny and Werner Callebaut) a workshop on the meanings of “doing theory” in biology. The basic idea is to ask questions about the (large and diverse, as it turns out) variety of activities that go under the rubric of theoretical biology, how they relate to each other, and what it is they are trying to accomplish.
My own talk got things started by highlighting some recurring trends in biological theory, and proceeding to discuss examples of different ways of engaging in theoretical biology. Let’s start with the trends. A pretty obvious and long-standing one is represented by what I think of as an obsession on the part of some biologists and philosophers of science to look for “laws” in biology. The literature is fascinating, but I am ultimately unconvinced that there are any such things as biological laws. Hell, I don’t think there are laws in physics, necessarily (only empirical generalizations). I think a good argument can be made that this search for biological laws is the result of the idea (put forth with the complicity of early 20th century philosophers of science) that physics is the “queen” of sciences, and since it always strives for the broadest possible generalization (of which laws are the epitome), then biology has to do the same in order to be taken seriously. I sincerely hope we are getting away from that kind of thinking and toward a more flexible and pluralistic way of what it means to do good science.
The second trend I noticed is more recent, though related, and it deals with attempts at producing “general” theories within the biological sciences. One of the best examples is Stephen Hubbell’s “unified neutral theory of biodiversity and biogeography,” or Sam Scheiner’s “conceptual framework for biology.” The first one is an attempt to set a general null hypothesis for community structure in ecology, while the second is an ambitious attempt at nothing less than the biological equivalent of a theory of everything. I tend to be skeptical of these grand plans as well, in Hubbell’s case because I don’t really have a high opinion of null hypotheses to begin with (and because data can too easily fit a null model even when there is quite a bit going on in the system), in Scheiner’s case because I think of “biology” as an inherently heterogeneous discipline that is ill suited to grand unifying schemes. But, of course, I could be wrong.
The central part of my talk — which was meant to be introductory to the workshop — quickly surveyed various modes of doing theory in the biological sciences, all legitimate in their own right, though of course all characterized by specific limitations and interesting problems.
To begin with, there are classical mathematical-analytical models, often explicitly inspired by theoretical physics. Fisher’s fundamental theorem of natural selection is an obvious example, and so is the Hardy-Weinberg “law” (really a mathematical truism that can be applied to describe the genotypic frequencies of a population at equilibrium if no evolutionary processes are at work disturbing that equilibrium). Typically, these models are rigorous but quickly become intractable because the number of variables affecting actual biological systems is very large.
Which brings us to the second type of modeling, statistically based (the analogy with physics here would be models in, say, statistical mechanics). This is the realm of quantitative genetics, where parameters such as means, variances and covariances are used to describe both the current state and foreseeable future of evolving populations. Quantitative genetics is extremely valuable for descriptive purposes but, I have argued in print, much less so as a predictive or explanatory approach to understand long-term evolutionary processes. Which doesn’t mean there aren’t people who (sometimes vehemently) disagree with me...
The third type of theoretical biology is the one relying on intensive computer modeling (the equivalent in physics here would be, say, climate science models). There is an increasingly fascinating literature using this approach, for instance Sergey Gavrilets’ modeling of very highly dimensional “adaptive landscapes,” or Andreas Wagner’s models of the relationship between robustness and evolvability — two fundamental properties of evolving biological lineages that are playing an increasingly significant role in what some of us refer to as the ongoing Extended evolutionary Synthesis.
Finally, there has always been a role in biology for conceptual/verbal theorizing, beginning of course with Darwin’s own “long argument” in the Origin, and continuing with the foundational books that established the Modern Synthesis during the 1940s. This non-mathematical approach, however, also includes visual models like those predominant in molecular biology — think of the kind of diagram used to summarize complex data sets concerning metabolic pathways and gene networks. Part of this heterogeneous group are also verbal/visual models of how the bio-physical properties of living cells and tissues generate organismal form, as in the work of Stuart Newman and Gerd Muller.
The very last part of my talk was then devoted to the role of philosophy of science in its particularly incarnation as — in the felicitous phrase by Hasok Chang — “the continuation of science by other means.” According to Chang, history and philosophy of science, besides being independent disciplines in their own right, can interface with science itself in the pursuit of common knowledge objectives. Chang calls this “complementary science” that “identifies questions that are excluded by specialist science. ... The primary aim of complementary science is not to tell specialist science what to do, but to do what specialist science is presently unable to do. It is a shadow discipline, whose boundaries change exactly so as to encompass whatever gets excluded in specialist science.” Examples in the philosophy of biology include discussions of species concepts and the ontological status of “species,” the role of alternative (epigenetic) systems of inheritance in evolution, and analyses of the logical structure and foundations of evolutionary theory.
Below is a brief rundown of the full list of speakers and topics featured at the workshop. The proceedings will be published either as an MIT Press volume or as a special issue of the journal Biological Theory.

Bruggeman, Frank
Netherlands Institute for Systems Biology
Systems biology and the meaning of “theory”
Callebaut, Werner
Konrad Lorenz Institute
What does it mean to do theory in biology?
Cleland, Carol
University of Colorado
Is a General Theory of Life Possible: Understanding the origins and nature of life in the context of a single example
Collins, Jim
Arizona State University
Role of theory in biology
Depew, David
University of Iowa
Rhetoric of evolutionary theory
Griesemer, Jim
University of California at Davis
Conceptual foundations of the “inexact” sciences
Gross, Lou
University of Tennessee
Selective Ignorance and Multiple Scales in Biology: Deciding on Criteria for Model Utility
Hammerstein, Peter
University of Berlin
Evolutionary game theory and the interface between evolution and economics
Kaplan, Jonathan
Oregon State University
Social impact of scientific theorizing
Leonelli, Sabina
University of Exeter, UK
Is there a difference between data-drive and theory-drive research?
Longino, Helen
Stanford University
Sociology of scientific theorizing
Love, Alan
University of Minnesota
Theory is as theory does...
Millstein, Roberta
University of California at Davis
Population genetics as the theoretical backbone of evolutionary biology?
Pigliucci, Massimo
City University of New York
Toward a broader concept of “theory”: back to Darwin?
Roughgarden, Joan
Stanford University
What might a general theory of ecology look like
Sterelny, Kim
University of Wellington, Victoria, NZ
Controversial theories in biology
Vorms, Marion
Institut d’Histoire et de Philosophie des Sciences et des Techniques
Theorizing and representational practices in Classical Genetics


  1. Thanks for this Massimo. It sounds like a really interesting meeting.

    This phrase "complementary science" -- it's felicitous alright, but maybe a bit too felicitous! I think it's a slogan broad enough to cover some really quite different approaches to HPS.

    Your vision, if I understand it right, is of HPS as complementary to science. It's a discipline that deals with foundational and conceptual questions which arise in contemporary theory, but which scientists lack either the time or the skill set to address themselves.

    On this reading, "complementary science" seems to amount to a strong form of naturalism, on which HPS is deferent to -- and more or less continuous with -- current science.

    Hasok Chang, on my understanding at least, is drawing an analogy with complementary medicine. When he talks of HPS as "complementary science", he envisions HPS as a kind of curator of neglected theories, a place where ideas dismissed by mainstream science as outdated or outré are given a second chance. We're talking about things like phlogiston theory, Aristotelian physics, maybe even alchemy.

    There is a nice manifesto for his programme in this article:

    "It is perfectly understandable that current specialist scientists would not want to be drawn into developing research programs that have been rejected long ago, because from their point of view those old research programs are, quite simply, wrong. This is where complementary science enters. Lacking the obligation to conform to the current orthodoxy, the complementary scientist is free to invest some time and energy in developing unorthodox systems."

    The thought is that, if mainstream science isn't willing to take a pluralistic attitude towards alternative theories, HPS can do the job instead. The aim is to undo the "Kuhn loss" that occurs when scientists, working within a contemporary paradigm, come to regard the questions addressed by older and alternative paradigms as irrelevant.

    On Hasok's picture, HPS is neither deferent to nor continuous with contemporary science. So, in some sense, this conception of HPS as "complementary science" represents a rejection of naturalism as it's usually understood. And I think this probably marks a significant difference between his vision of "complementary science" and yours.

    I think they are both intriguing visions for the future of HPS -- but in some ways they are so different!

  2. Jonathan, you may be' trading too much into Chang's project, or at the very least, I certainly don't see "complementary science" either as in any way similar to complementary medicine or as a rejection of naturalism. I don't think one can do any science (or, indeed, anything at all) once we do away with naturalism.

  3. But I think this just underlines the differences between your version of "complementary science" and Chang's. In the article I linked to, Chang is up-front about the tensions between his own project and naturalism. The aim is to cultivate "unorthodox systems" of knowledge that lie outside the bounds of currently accepted theories.

  4. Jonathan,

    I have now read Chang's TPM article. I don't see anywhere either a call for abandoning naturalism or a parallel / endorsement of complementary medicine. The whole thing strikes me as eminently reasonable.

  5. Sorry to clutter up your comments part of blog but I wanted to know if any of you clever people could take the time to give me some opinions on some stuff I’ve wrote at:

  6. Hi Massimo,

    First let me clarify the "complementary medicine" comment, since I can see how this might easily be misconstrued. My claim was not that Hasok endorses complementary medicine! The point was merely that, when he talks about "complementary science", he is using the word "complementary" in the same sense, viz. as implying "alternative" and "unorthodox".

    The first few paragraphs of the TPM article struck me as a pretty strong rejection of naturalism, but perhaps I'm misreading them. I will have to quiz Hasok on this.

    To me, the programme seems unabashedly Feyerabendian. The aim, as I see it, is to cultivate "unorthodox systems" of knowledge. That's not to say it's a bad thing, but it does seem to imply a tension with naturalism as I understand it. Maybe we are understanding "naturalism" differently.