Ridley, Chapter 3
Evidence for Evolution
Part 1

510 index page

Introducing a book on evolution with one chapter on "evidence for evolution" may be appealing.  But do not expect to find all the evidence summarized in this one chapter.  In a very real sense, our entire textbook, as well as all of biology, stands as evidence for evolution.

Brief Outline

-----------

Ridley, Chapter 3  Evidence for Evolution

Scientific argument.

     Ridley frames his discussion in Chapter 3 as an anti-creationism argument.  That is fairly obvious at the beginning of the chapter, and becomes explicit again toward the end.  In between he offers some interesting and relevant biology.  The biology is sound.  But, unfortunately, as an anti-creationism argument his evidence is presented rather pro forma, without real care for philosophical soundness.  He writes as if the familiarity and reliability of his conclusion rendered careful argument unnecessary.  This chapter presents the kind of argument which creationists relish tearing apart in public "debates".
     Allow me to digress to make a philosophical point here, to get a bit preachy myself.  There is a strong tendency to agree with an argument if we already agree with its conclusion.  Put a bit more strongly, if we believe that the conclusion of an argument is true, we may not be critical of the argument itself.  This is a very common human tendency, but it is contrary to the best science.  Good science depends on skeptical questioning.  Science permits us to accept as true only ideas, observations and arguments that pass very rigorous standards.  Proper controls are needed for all experiments, whether or not the results match expectations.  Similarly, arguments must address all appropriate concerns, not just some carefully selected "straw men."

Testing alternative explanations.

     The standards of science are not neatly written down somewhere on golden tablets.  They are often misunderstood, and sometimes misapplied.  The so-called "experimental method" represents an attempt to codify the scientific enterprise, but science is richer than the standard brief accounts found in the introductory chapters of so many undergraduate textbooks.  Fundamentally, science is about disciplined skepticism.  It is about subjecting observations and explanations to careful scrutiny, by considering and ruling out alternative interpretations.  That's basically the purpose for controlled experiments -- ruling out alternative interpretations for some phenomena.  But it's not always simple to know what controls are appropriate, or necessary.  You first have to realize what sources of error or misinterpretation are there, and then consider each of them.  That's what I'll be trying to do for you now, with Ridley's arguments.  He usually reaches strong and reliable conclusions, but those conclusions may not always follow from his arguments.  He omits from explicit consideration many relevant issues.  (That this is oversight rather than error becomes clear from the textbook taken as a whole.)  We need to notice, and "control for", the pieces that Ridley leaves out.
     The consideration of multiple hypotheses will in fact be our model for considering many of the problems we shall encounter in evolution.  We must consider what different interpretations or explanations should be examined, and how they can be distinguished from one another.  To use a simple example that we shall encounter repeatedly in future lessons, it won't be enough to know that changing gene frequencies can be explained by evolution.  We must consider all of the many different and independent ways that gene frequencies can change (e.g., mutation, migration, drift, and natural selection, and how each process can be affected by population size, genetic linkage, pleiotropy, and other factors).  Then we must consider the consequences and implications of each one, both separately and in combination.  We must also remember that Darwin's Origin of Species represents two distinct arguments.  First, Darwin presented a compelling argument for the occurrence of evolution, for "descent with modification" as a concept with power to explain many observations.  Second, Darwin proposed a simple causal mechanism for evolutionary change, variation and natural selection.

Supporting cases vs. critical argument.

     In Chapter 3, Ridley concentrates on arguing for descent with modification  (although in section 3.13 he stirs in natural selection as well).  In support of his chapter title, "Evidence for evolution", Ridley offers particular cases.  However, evolution is such a grand theory, it cannot be validated (or dismissed) on the basis of a few particular cases.  Each of Ridley's examples might be given an alternative (non-evolutionary) explanation (see below).  Nearly every particular fact which can be mustered in support of evolution can also be given an alternative interpretation.  The philosophical problem with those alternative explanations is that separately they don't explain very much and collectively they don't really amount to much.  Each such explanation typically doesn't extend beyond its particular case, and even then may offer no better explanation than evolution. A better title for this chapter might be, "Various examples, all of which can be elegantly explained by evolution (but which might also be explained quite inelegantly by a messy combination of several other theories)".
     The real strength of evolution is NOT that there are particular evidences which "prove" it, but that there is an overwhelmingly vast amount of information that NO OTHER explanation deals with so completely or satisfactorily.  Furthermore, evolution by natural selection could in principle be disproven in many ways - if, for example, variation were not hereditary - but that has not happened.

Section-by-section commentary on Ridley's Chapter 3.

3.1.  "We distinguish three possible theories of the history of life."

     This section is, in large part, a "straw man" argument.  By making the alternatives to evolution so simplistic, Ridley easily and summarily rejects each one.  But...
     Most grand theories come with a lot of subsidiary assumptions,which give them plenty of "wiggle room".  When evidence appears inconsistent with a theory, the theory might be wrong; OR, the evidence might be incorrectly interpreted; OR, the theory's supporting assumptions might need some [slight] tweaking.  Evolutionary biologists routinely engage in such reinterpretation; the creationists do so as well.  
     Later (3.7), Ridley does admit that some creationism is more subtle than his straw men, but he does not do so until long after he has scored his initial cheap shot.  Modern "scientific creationists" would reject Ridley's caricatures.  They would even admit much of his "evidence for evolution" as fully consistent with their theory.
     The risk here is that, if Ridley's arguments are accepted as valid, then when his arguments are refuted by some moderately sophisticated creationist, the credibility of evolution will be lessened.  Indeed, various creationists have incisive rejoinders to many of the points Ridley makes in this chapter.  However, creationist arguments do have other much graver flaws, which Ridley ignores in favor of simplicity of argument.  Although creationism is NOT a legitimate scientific alternative to evolution, neither is it quite so simple-mindedly wrong as Ridley's arguments imply.

3.2.  "On a small scale, evolution can be observed in nature."

     This section does provide evidence to refute Ridley's straw man of separate creation with no subsequent change.  However, this is NOT a view held by modern creationists.
     But does the evidence Ridley presents, of phenotypic change in house sparrow populations over recent historical time, actually provide substantial support for evolution?  In chapter 1, Ridley took pains to distinguish evolutionary change from other sorts of change.  Eventually (in Chapter 4), he makes it plain that heredity is an essential element in evolutionary change.  So ask yourself, does Ridley's chosen example here of "small scale" evolutionary change include any evidence for heredity? That is, is the observed geographical variation in the mean phenotypes of house sparrow populations indeed based on evolved genetic differences in the populations?  Or might this variation result from developmental (rather than evolutionary) responses to different environments by genotypes which might be similar in all populations?
     To help see the problem, imagine a very large farm field, with deep fertile soil on one end and thin, hard, infertile soil on the other end.  Plant genetically uniform grain on one end of the field.  Then wait a few generations for the population to spread across the field.  At the end of this time, there will be tall vigorous plants on the fertile end of the field, and short scrawny plants on the infertile end.  This is the expected result, WITHOUT any evolutionary change, even if the population remains genetically uniform throughout the experiment.  From the limited evidence Ridley presents, the house sparrows could be responding in a similar way, with simple differences directly induced by the different environments in the wet east, the dry west, or the cold north.
     A simple experiment could test whether the measured changes in house sparrows are environmentally-induced or truly evolutionary (i.e., genetic).  Sample the populations in each of the various biogeographic regions.  Raise offspring from each sample in each of the other regions.  Results should establish whether or not the regional differences are inherited or environmentally induced developmental responses.  Only if the differences are indeed hereditary, will this piece of Ridley's "evidence for evolution" become complete.

3.3.  "Evolution can also be produced experimentally."

     Again, Ridley's evidence successfully refutes his straw man of absolute fixity of species.  But creationists do not argue for such absolute fixity.  They all know perfectly well that many changes can be accomplished through artificial selection.  They describe such empirical evidence as changes within a type, rather than creation of a new type (e.g., dogs and cats are "types"; selection can modify dogs into many various forms, but cannot turn dogs into cats).
     Furthermore, Ridley offers no evidence in this section that artificial selection can draw out phenotypes beyond those already potentially present in the population at the start of the experiment.  Artificial selection might work simply by paring away unwanted genotypes, without ever producing anything "new".  See Ridley's Figure 4.9 on page 84 in Chapter 4.  Unless new variation is continually supplied in appropriate amount and quality, selection will necessarily run down and further change will cease.  The examples Ridley graphs in section 3.3 do not address this problem.  Although the long-term consequences of domestic breeding programs do not seem to have encountered any obvious limits, they also have not produced the sort of "transmutation" that would impress a creationist.  (That is, we have no difficulty recognizing that Great Danes and Dachshunds, Poodles and Pekinese are all still dogs).  

3.4.  "Interbreeding and phenotypic similarity provide two concepts of species."

     Understanding the origin of species requires some concept of what a species is.  It turns out to be astonishingly difficult to provide a good, all-round workable definition for species (see Chapter 15 for an extended discussion).  The reasons why are easy to understand in evolutionary terms, but here Ridley adopts a simplistic sense of the reality of species that invites confusion of argument.
     Ridley gives two species concepts, a reproductive-interbreeding concept (commonly called the "biological species" concept, attributed to Ernst Mayr) and a phenotypic-similarity concept (allowing distinctions among species to be made from dead museum species).  Both concepts have well known exceptions and either one can cause great difficulties if pushed too hard.
     Ridley admits in this section that although artificial selection can yield extensive morphological change (e.g., dog breeds), artificial selection "has not produced much evidence for new reproductive species" (p. 47).

3.5.  "Ring 'species' show that variation within a species can be extensive enough to produce a new species."

     Ring species constitute some the nicest evidence for how allopatric speciation (speciation by geographic separation) may occur.  But, by definition, a ring species is still a single species.  (Definitions can sometimes, all by themselves, create great difficulties for clear thinking.)  Furthermore, as an argument against a creationist viewpoint, the mere existence of ring species provides no evidence of evolutionary change over time.  Evidence given in this section does not show change over time, and so does not address the possibility that populations of the ring species described here (a California salamander) might have always been just the way they are now.

3.6.  "New reproductively distinct species can be produced experimentally."

     The phenomenon of creating new species by artificial production of polyploid hybrids is certainly interesting. A process of hybridization has also been important in the origin of many natural species, particularly plants. But this is NOT a conventional Darwinian evolutionary process, such as Ridley has been attempting to illustrate with house sparrows and ring species and domestic breeding. It is much more closely akin to mutationism (i.e., sudden formation of a new species by a single genetic macromutation).

3.7.  "Small-scale observations can be extrapolated over the long term."

     This concept is closely related to the philosophical principle of uniformitarianism.  What we can observe now is fundamentally similar to what has happened in the past.  Of course, this principle is not itself a direct source of evidence for evolution.  It is only a method for interpreting such evolution.
     In Ridley's words, "It is possible to imagine, by extrapolation, that if the small-scale processes we have seen were continued over a long enough period they could have produced the modern variety of life."  Unfortunately, it also possible to imagine the opposite, that extrapolation of processes we have seen might NOT be sufficient to produce the modern variety of life.

Occam's Razor

     We need an additional principle before we can use uniformitarianism to construct a compelling argument for evolution.  One of these is Occam's Razor (after English philosopher William of Occam, ca. 1350), "Essentia non sunt multiplicanda praeter necessitatem" or "Essentials shall not be multiplied beyond necessity".  This principle asserts that one should not complicate an explanation unnecessarily.  Applied to evolution, it is generally taken that we should not deny the appropriateness of extrapolation (which would then require some additional essential process, like "intelligent design" or "orthogenesis") without good reason.  No sufficiently good reason has yet been found.  Nevertheless, like any other sharp tool, Occam's Razor can be dangerously misused; its application must always be treated with some degree of skepticism.

3.8.  "There are homologous similarities between groups of living things."

     Homologies (similarities of structural form which do not depend on any functional necessity) may indeed be "evidence for evolution".  But homologies can also be considered as evidence for non-evolutionary theories (such as the pre-evolutionary concept that homologies are evidence of essential forms that reflect a divine plan).  As evidence for evolution, homology works best only as one piece in a "consilience" argument (see Chapter 3 Notes, Part 2), which Ridley does not make explicit (although he provides a weak version in section 3.12).

3.9.  Different homologies are correlated, and can be hierarchically classified."

     The Linnean system provides a consistent framework for hierarchical classification.  Like homology itself (section 3.8 above), this consistent correlation of homologous structures offers evidence for evolution.  But, like homology, this same evidence can be (and, before Darwin, was indeed) interpreted in support of non-evolutionary views.  And like homology itself, this evidence works best as part of an explicit "consilience" argument (see Chapter 3 Notes, Part 2).

3.10.  "There is some fossil evidence for the transformation of species."

     The evidence presented here is, finally, strongly suggestive of evolutionary transmutation.  Unfortunately, as Ridley notes, "such examples are rare".  They also tend to fall into the same "within-kind" transformation that fails to impress creationists.  Once again, this evidence really works well only within a "consilience" context (see Chapter 3 Notes, Part 2), wherein one simple explanation works for a huge variety of data.

3.11.  "The order of the main groups in the fossil record suggests that they have evolutionary relationships."

     For many casual observers, this appears to be the strongest evidence for evolution.  How else could one explain the fossil record?  Even better is to recognize (as does Ridley's quote from Haldane on p. 64) that the fossil record offers an endless supply of "experimental tests" for evolution.  Every new fossil offers an opportunity to refute evolution, by revealing some supposedly descendent form in strata predating the ancestors ("a rabbit in the Precambrian").  Yet no such contradiction has ever been found.
     However, non-evolutionary explanations can also explain the same observations, albeit in a strained and uncompelling manner.  Thus, once again, this evidence works best in a "consilience" framework (see Chapter 3 Notes, Part 2).

3.12.  "Summary of the evidence for evolution."

     Here, at last, Ridley presents the "consilience" argument -- that all sorts of data can be explained by one and the same theory, evolution by natural selection.  It's not any one special source of evidence, but the fact that so many sources of data (including those appearing in later chapters) all jump together, concurring in consistency with this one explanation, that makes evolution such a strongly supported theory.  See Chapter 3 Notes, Part 2.
     In contrast, creationism offers NO consistent, testable explanation for observable patterns, only rejection of evolution.  That was a credible position when considerable evidence seemed to stand against evolution (e.g., when the age of the Earth was thought to be at most a few million years, and when inheritance was generally understood as blending of traits).  But for many decades, creationism has become more and more conspicuously a case of special pleading on behalf of a belief that is grounded in religion rather than science.

3.13.  "Creationism offers no explanation for adaptation."

     Most of this chapter has been devoted to evidence that evolution can occur.  The argument in this section, that only natural selection can explain adaptation, addresses a different concern, how adaptation can be explained.  Note that transmutation of species does not automatically explain adaptation.  Nor is natural selection the only mechanism that could explain adaptation.  Both the "inheritance of acquired characters" hypothesis and the "intelligent design" hypothesis do so.  However, in both of these, the causal mechanism (striving for improvement, or design by an intelligent agent) is itself an unexplained assumption.  Indeed, inheritance of acquired characteristics has been abundantly DISconfirmed. Intelligent design is much harder to disprove; however, unlike natural selection whose elements and operation can be readily demonstrated, there is no positive empirical evidence for "intelligent design" apart from the phenomenon, adaptation, that is to be explained).

3.14.  "Modern 'scientific creationism' is scientifically untenable."

     Here Ridley returns to the "straw man" argument with which he began the chapter.  Modern "scientific creationism" is indeed scientifically untenable.  But Ridley oversimplifies and misrepresents scientific creationism, so we are left without knowing what the latest arguments are and without appreciating how those arguments might be met.
     Finally, Ridley's closing assertion that "No important religious beliefs are contradicted by the theory of evolution" is extravagantly arrogant and inappropriate for a textbook such as this.  Taken at face value, this assertion is blatantly false, unless we grant to Ridley the prerogative of defining which religious beliefs are "important" and which ones are not.  At best, some religious people do indeed accept that none of their own personal beliefs are contradicted by the theory of evolution.

Part 2.  Additional comments on the evidence for evolution.

510 index page

-----------

Comments and questions: dgking@siu.edu
Department of Zoology e-mail: zoology@zoology.siu.edu
Comments and questions related to web server: webmaster@science.siu.edu


SIUC / College of Science / Zoology / Faculty / David King / ZOOL 510
URL: http://www.science.siu.edu/zoology/king/510/mr03a.htm
Last updated: 4-Jan-2000 / dgk