Making Sense of Evolution explores contemporary evolutionary biology, focusing on the elements of theories—selection, adaptation, and species—that are complex and open to multiple possible interpretations, many of which are incompatible with one another and with other accepted practices in the discipline. Particular experimental methods, for example, may demand one understanding of “selection,” while the application of the same concept to another area of evolutionary biology could necessitate a very different definition.
Spotlighting these conceptual difficulties and presenting alternate theoretical interpretations that alleviate this incompatibility, Massimo Pigliucci and Jonathan Kaplan intertwine scientific and philosophical analysis to produce a coherent picture of evolutionary biology. Innovative and controversial, Making Sense of Evolution encourages further development of the Modern Synthesis and outlines what might be necessary for the continued refinement of this evolving field.
|Publisher:||University of Chicago Press|
|Product dimensions:||6.00(w) x 9.00(h) x 0.90(d)|
About the Author
Massimo Pigliucci is associate professor of ecology and evolution at Stony Brook University. Jonathan Kaplan is assistant professor of philosophy at Oregon State University.
Read an Excerpt
Making Sense of Evolution The Conceptual Foundations of Evolutionary Biology
By MASSIMO PIGLIUCCI JONATHAN KAPLAN
THE UNIVERSITY OF CHICAGO PRESS Copyright © 2006 The University of Chicago
All right reserved.
Chapter One Natural Selection and Fitness
After the "Force" Metaphor
I HAVE CALLED THIS PRINCIPLE, BY WHICH EACH SLIGHT VARIATION, IF USEFUL, IS PRESERVED, BY THE TERM OF NATURAL SELECTION. -CHARLES DARWIN, ON THE ORIGIN OF SPECIES
Natural selection is often conceived of as a "force" that acts on populations through the differences in relative fitness among individuals within each population. Recent work by such authors as Matthen and Ariew points out difficulties with the force metaphor and the attendant concept of fitness. This is far from being a trivial semantic quibble; the way in which one conceives of natural selection and fitness strongly influences what one will take to be evidence for natural selection having occurred, and how one thinks of the difference between drift and selection. Following other authors, we suggest that one must distinguish between informal and formal fitness, and between informal and formal selection. Doing so not only makes clear why the force metaphor is misleading, but also has implications for our ability to distinguish selection from drift, for measurements of selection in natural populations (chapter 2), for the levels of selection debate (chapter 3), for our conception of constraints on evolutionary change (chapter 4), and for the way that the metaphor of adaptive landscapes is used (chapter 8).
WHAT IS NATURAL SELECTION?
When Darwin (1859, 80ff) coined the term "natural selection," he did so in hopes that an appeal to a process that was well known-artificial selection-would make the process to which he was referring clearer. Just as animal or plant breeders can change the distribution of traits in a population by selectively breeding those animals or plants with traits deemed desirable, Darwin argued, the distribution of traits in natural populations of organisms changes over time as those organisms with traits better fit to the environment succeed in reproducing at a higher rate than those with traits less well fit, at least where the differential fit between organism and environment is shared between parents and offspring (box 1.1). It is the increasing "fit" between the organisms and the environment in populations that explains "the extraordinarily complex and intricate organization of living things" (Bell 1997). Darwin reasoned that just as, over time, artificial selection can produce radically different breeds from the same ancestral population, natural selection, given the enormously greater amount of time available, can produce different species, adapted to different local conditions or to different ways of living, from the same ancestral populations. Natural selection, then, is the explanatory principle accounting for the fit between organism and environment, as well as being an integral part of the explanation of the variety of organisms.
Still, there are important dis-analogies between artificial and natural selection. Artificial selection is undeniably a process, the result of which may be change in a population (given certain assumptions, such as the heritability of the trait in question); there is someone doing the selecting, and it is this that permits us to say unequivocally that selection is taking place. Contrast this with Futuyma's (1998) definition of natural selection as "any consistent difference in fitness (i.e., survival and reproduction) among phenotypically different biological entities" (349, emphasis in original). Here it seems as if natural selection is not a process, but rather a fact about the population in question-that is, the existence of an average difference in fitness between certain entities. Of course, Futuyma is quick to point out that natural selection can be decoupled from evolution, both because it is possible for a population to evolve through nonselective mechanisms, and because natural selection can occur without evolution (if, for example, the variation in the traits under selection is not heritable) (1998, 365). But the idea that natural selection is a fact about populations still sits awkwardly with Futuyma's claim that natural selection "is one agent of change in the pattern of variation" in populations, and hence one of the causes of evolution (365); facts about populations and agents acting on populations are generally conceived of as two different things.
Sober's (1984) detailed development of a force metaphor for evolutionary change is one way of trying to make the view of natural selection as a process cohere with the view that takes it to be a fact about populations. If fitness is defined as average differences in reproductive success among members of a population, then, Sober suggests, natural selection can be viewed as the force resulting from those fitness differences (1984, 46). This "force" is then conceived of as one of the causes of evolution; it is the only such force that can result in the evolution of adaptations. However, as evolution in this view is conceived of as heritable changes in the composition of a population, other forces need to be evoked to explain evolution more generally. For example, since genetic drift can change the frequencies of alleles within a population, it must be a force as well, as must, say, migration events, genetic linkage, and mutations (Sober 1984, 38). The actual changes in a population are the result of the combination of these forces, just as the actual changes in the velocity of physical objects are the result of the combination of forces in Newtonian physics (1984, 50). This talk of combining forces in a Newtonian way is not supposed to be just metaphorical; Sober and Lewontin (1982) state that "[s]election for X, and against Y, and so on, are component forces that combine vectorially to determine the dynamics of the population" (160) (fig. 1.1).
This interpretation of population changes being the result of various forces also fits in very well with one of the standard visual metaphors used to understand the relationship between fitness and evolution; namely, the fitness landscape. It is natural to think that in order to explain the movement of a population on the landscape, one must appeal to forces pushing it in various directions; selection pushes the population up "hills" of higher fitness, drift pushes it randomly, and so forth (fig. 1.2). In the absence of any force, the population is assumed to remain in place on the landscape.
The force metaphor, however, is misleading in important ways. It is not just that the metaphor is not perfect (something everyone agrees with), nor is it just that there are ways in which the comparison between "forces" in evolutionary biology and Newtonian forces breaks down. If the only places where the comparison failed were not significant to understanding selection, nor to empirical work in evolutionary biology, and if the comparison were intellectually fruitful where it did not break down, then the failures of the metaphor would not matter. We argue here that this is not the case; rather, the places where the metaphor fails and where the comparisons break down are conceptually important, and the study of selection and evolution is hampered by the (often implicit) use of the force metaphor. Further, even where the metaphor is not obviously inadequate, other ways of looking at the relationship between fitness, natural selection, and evolution are able to do all the work done by the force metaphor without engendering the conceptual problems it does. The force metaphor simply is not more "fruitful" than other ways of understanding selection and fitness. Moreover, carefully distinguishing different uses of "selection" and "fitness" permits one to avoid the force metaphor altogether, while simultaneously pointing toward new (and more fecund) solutions to philosophical problems as well as more productive avenues of empirical research in biology.
TWO WAYS OF THINKING ABOUT FITNESS AND NATURAL SELECTION, REVISITED
Natural selection is supposed to result whenever there are heritable differences in fitness between biological entities in a population (see box 1.1). Matthen and Ariew (2002) make an important distinction between informal fitness (box 1.2) and predictive (or formal) fitness. In their characterization, vernacular fitness refers to a biological entity's overall relative ability to survive and reproduce; it is thus linked to the idea of there being a "fit" between the organism and the environment. In this view, those organisms in a population that are, overall, better fit to the conditions of their existence may be expected to do better (be more reproductively successful), on average, than those organisms that are less well fit to those conditions (see also Ariew and Lewontin 2004). In this interpretation, the "fit" between the organism and the environment emerges from the continuous interactions between the two; that is, particular organisms interact with particular aspects of the environment in particular ways, and those organisms better suited to succeed in their interactions with those aspects of the environment are better fit to that environmental regime.
This sense of vernacular fitness can play little or no role in actual research programs because it is difficult or impossible to determine the (overall) vernacular fitness of an organism. There are too many environmental variables and too many traits, and they interact in ways that are too complex. This problem is not just epistemic; the vernacular fitness of an organism is supposed to be about how well it is fit to its environment, not about how it actually fares. That is, vernacular fitness is about the propensity of the organism to succeed in environments like the ones it finds itself in. The actual reproductive success of the organism, which emerges from its interaction with the environment in all its details, does not tell us how fit the organism is in the vernacular sense, because random factors that do not reflect the organism-environment fit also contribute to reproductive success (e.g., an individual may be potentially very fit given its typical environment, but extremely unlucky in being hit by lightning). But neither is it obvious how to interpret the idea of the organism's average reproductive success in relevantly similar environments: what is to count as an environment that is "relevantly similar" for the purpose of evaluating vernacular fitness? If we demand that many of the details be maintained, the idea of fitness is in danger of collapsing back into actual reproductive success; if we permit too coarse an analysis, on the other hand, we will miss too much of what makes one organism in a population more or less fit than another. Nor is the overall fitness of individuals particularly valuable for thinking about changes in populations over time-the features of the population that change over time are not obtainable in any obvious way by summing over the reproductive success of individual organisms; instead, they must be obtained by comparing the success of individuals with and without particular variants of particular traits (or clusters of traits).
In a biologically more realistic, but less satisfying, fashion, one might think instead about an entity's relative ability to survive and reproduce given some particular trait (or some combination of a relatively small number of traits) and some set of physical processes that interact with the trait in question. One organism can be more fit than another in this, which we dub the informal sense of fitness, if the physical processes that interact with the traits of interest are discriminate with respect to the variations in those traits; that is, if organisms with one variant of a trait are more likely to survive and reproduce than are organisms with another variant of that trait, given the interaction of a particular physical process with the trait. If, on the other hand, the processes that interact with the trait in question are indiscriminate with respect to variation in that trait, there will be no informal fitness differences between the organisms when those processes and traits are considered. Notice that this way of thinking about fitness does not take into account how frequently organisms interact with the physical processes in question; it fails to take into account any differences in the mean frequencies with which organisms with different variants of the trait might interact with the relevant physical processes. This version of informal fitness is only about the expected success of an organism given a particular relationship between a trait and some particular physical process. It is focused on the "individual" nature of fitness, but does not address the likely reproductive success of the organism overall. In this way, this version of informal (or individual) fitness differs from the usual vernacular view of fitness, and is much more tractable than is the vernacular version.
In order to make clear the sense in which this informal version of fitness is tractable, we need to be clear about what "discriminate" and "indiscriminate" mean in this context. If, given a particular kind of physical process the organisms will face, the variations in a trait among those organisms make a difference to their abilities to survive and reproduce, the process is discriminate with respect to that trait, and there is an informal fitness difference between the organisms (with respect to that process and that trait). On the other hand, if the variations in the trait do not make a difference given the process, there will be no difference in the informal fitness of the variants. Note that whether a trait is relevant to informal fitness differences depends on the physical processes that trait is interacting with. For example, with respect to death by predation, running speed may be a trait in which differences result in differential success; however, with respect to death by lightning strikes, running speed is very likely not a trait in which variations matter to survival and reproduction. So, in this example, with respect to running speed, predation is a discriminate physical process, whereas lightning is an indiscriminate process. Similarly, mitosis is a process that is usually indiscriminate with respect to particular alleles; however, in the case of segregation distortion (discussed in chapter 3), it can be discriminate with respect to different alleles.
This much more limited conception of individual fitness-the fit between a given trait and a physical process-suggests that the idea of an organism's overall fitness (its vernacular fitness in the above sense; its overall propensity for reproductive success) could be given by something like the net result of the interactions of all the discriminate and indiscriminate processes with one another and with all of the organism's traits, given something akin to the average expected environment for that organism. But again, there are good reasons to reject the notion of "overall fitness" in this context. Consider, for example, how difficult it is to make sense of the idea of an average expected environment for an organism, especially when we consider that the environments faced by organisms with different variants of a trait may well be different because of that very same variation (as in the case of "niche-constructing" traits). Even if this idea of overall fitness could be made conceptually coherent, it is certainly intractable in practice, and can play little or no part in our understanding of either the actual or the mean population dynamics in any real instance. That is, the concept of an overall (vernacular) fitness can play no role in either our ability to understand what has actually transpired in some particular population or our ability to understand what is likely to transpire in populations of that kind.
Excerpted from Making Sense of Evolution by MASSIMO PIGLIUCCI JONATHAN KAPLAN Copyright © 2006 by The University of Chicago. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of Contents
Prelude - Evolutionary Biology and Conceptual Analysis
1. Natural Selection and Fitness: After the "Force" Metaphor
2. How (Not) to Measure Natural Selection
3. The Targets and Units of Selection: Individual Events and Population Dynamics
4. Studying Constraints through G-Matrices: Is Evolutionary Quantitative Genetics Looking in Vain for the Holy Grail?
5. A Quarter Century of Spandrels: Adaptations, Adaptationisms, and Constraints in Biology
6. Functions and For-ness in Biology
7. Testing Adaptive Hypotheses: Historical Evidence and Human Adaptations
8. Slippery Landscapes: The Promises and Limits of the Adaptive Landscape Metaphor in Evolutionary Biology
9. Species as Family Resemblance Concepts: The (Dis-)Solution of the Species Problem?
10. Testing Biological Hypotheses: The Detective versus the Statistician
Coda - A Philosophical Dialogue