Full text of DOI:10.1039/jr9610004882

Volume 80, No. 1
THE QUARTERLY REVIEW OF BIOLOGY
March 2005
MALADAPTATION AND NATURAL SELECTION
Randolph M. Nesse
Departments of Psychology and Psychiatry, University of Michigan
Ann Arbor, Michigan 48104 USA
e-mail: nesse@umich.edu
abstract
The transformations George Williams initiated in evolutionary biology seem so blindingly obvious
in retrospect that they spur the question of why he saw what no one else did. While most humans are
prone to see only what theory predicts, Williams sees in bold relief whatever does not fit. Not an
adaptationist or an anti-adaptationist, Williams is better described as a maladaptionist. The challenge
of finding evolutionary explanations for apparent maladaptations has been overlooked with casualness
akin to that once typical for group selection. Suboptimal traits tend to be dismissed as illustrations of
the weakness and stochastic nature of selection compared with mutation and drift. A closer look suggests
that such constraints are only one of six possible kinds of explanations for apparently suboptimal
designs: mismatch, coevolution, tradeoffs, constraints, reproductive advantage at the expense of the
individual, and defenses that are aversive but useful. Medicine has asked proximate questions at every
possible level but has only begun to ask evolutionary questions about why bodies are vulnerable to
disease. Considering all six possible evolutionary reasons for apparently suboptimal traits will spur
progress not only in medicine but also more generally in biology. “Williams Vision” may not yield a
net benefit to the possessor, but it is invaluable for the species.
individuals who maximized their own repro-
duction. He concluded that it was theoreti-
cally possible for genes that coded for ten-
dencies to help the group at a cost to the self
to become more common if groups with
Introduction
N 1966 George Williams published a slim
volume that transformed biology. Adapta-
I
tion and Natural Selection: A Critique of Some Cur-
rent Evolutionary Thought (Williams 1966) did
not report the results of clever calculations,
experiments carried out with expensive
machines, or even decades of field studies.
Instead, it addressed a simple question with
ruthless logic. Does natural selection indeed,
more individuals with such genes succeeded
in comparison with groups with fewer. In
almost every actual instance, however, such
group selection would be so feeble as to be
inconsequential compared to the power of
selection at the level of the gene or those col-
lections of cells with identical genes that we
as most biologists then believed, shape indi- call individuals. His critique of then-current
vidual traits that benefit the species? The evolutionary thought succeeded in eliminat-
answer, in fewer than three hundred pages ing a fundamental misconception in biology.
of clear prose, was “No.” His argument was Na ¨ı ve group selection was dead.
straightforward. Williams compared the
In retrospect, his insight seems blindingly
genetic contributions to future generations of obvious. The question is how nearly everyone
individuals who sacrificed reproductive suc- else in biology went along for decades obliv-
cess (RS) for the sake of the group to those ious to such a fundamental misconception
6
2

March 2005
SYMPOSIUM IN HONOR OF GEORGE C WILLIAMS
63
about how natural selection works. While oth- der is not the eye’s perfection, but that such
ers saw a benevolent Mother Nature shaping a jury-rigged device works at all.
organisms for the good of species and even
The quirks and limits of our cognitive
whole ecosystems, Williams saw contradic- adaptations similarly compromise our ability
tions in bold relief. There is something dif- to see the world accurately. At the perceptual
ferent about how he sees the world. Is this level, Gestalt phenomena fill the gaps in par-
“
Williams Vision” an adaptation ready to tial shapes, giving the illusion of wholes. Our
spread in the population, or a flaw or a neu- minds then organize perceptual data into
tral quirk? For an initial answer, consider the categories that allow generalizations about
perfection and imperfections of the eye itself. members of the class, an efficient simplifi-
The vertebrate eye is the exemplar of cation that allows quick decisions, and that
nature’s perfection. Even a cursory examina- fosters tendencies to prejudicial assump-
tion yields admiration at the clarity of the cor- tions that individuals all share the essential
nea, the lens, and the vitreous fluid. The characteristics of their groups. The mind
mechanism that constricts the iris in bright then connects these categories in theories
light is elegant. A marvelously flexible lens is that allow us to predict the future, often cor-
exactly compressed so the eye can focus from rectly, despite inherent biases and overly
inches to infinity. A subtle mechanism regu- simple heuristics (Nisbett and Ross 1980). At
lates intraocular pressure. The rods and each stage—perception, categorization, and
cones are packed close at the fovea while the theory building—our cognitive devices dis-
specialized motion sensitive cells monitor the tort our vision. We see what we expect to see
periphery of the visual field. It is no wonder and are often blind to what does not fit our
that Paley, among others, extolled the eye as prejudices and preconceptions. Williams
a miracle of design (Paley 1802). Thanks to Vision somehow lacks these usual human
all manner of marvelous contrivances, the eye foibles. Instead of seeing only what fits the
works and works well.
theory, it sees what does not fit. To consider
The eye is also, however, the poster child further if this is an adaptation or a maladap-
for the body’s imperfections. It works well tation, we must touch on the adaptation
when it works, but often it does not. Nearly a wars.
third of us have hereditary nearsightedness,
and almost no one over 55 can read a phone
book unassisted (except for those who have
been nearsighted for decades!). The lovely
mechanism that regulates intraocular pres-
sure often fails, causing glaucoma. Then
Adaptationists and Anti-
adaptationists
Guidance comes from the first sentence in
the contents of Adaptation and Natural Selec-
tion: “Evolutionary adaptation is a special
there is the blind spot, a manifestation of the and onerous concept that should not be
abject design failure of nerves and vessels that used unnecessarily, and an effect should not
penetrate the eyeball in a bundle and spread be called a function unless it is clearly pro-
out along the interior surface instead of pen- duced by design and not by chance” (Wil-
etrating from the outside as in the better- liams 1966:vii). This reluctance to call traits
designed cephalopod eye. Octopi not only adaptations seems to place Williams on the
have a full field of vision, but they need not anti-adaptationist side of the adaptation
worry about retinal detachment. They also wars. Does this mean that Williams is reluc-
need neither the tiny jiggle of nystagmus that tant to seek evolutionary functions for prob-
minimizes the shadows cast by vessels and lematic traits? Hardly. Most of his major con-
nerves on the vertebrate retina nor the brain tributions came from the perverse tendency
processing mechanisms that extract the visual to seek evolutionary explanations for traits
signal from the nystagmus noise. In short, the that seemed inexplicable. In the welter of
vertebrate eye is a masterpiece not of design, words spilled about adaptationism, Williams
but of jury-rigged compensations for a fun- was not in the main fray (Segerstr a˚ le 2000).
damentally defective architecture. The won- The explanation may be that his views are

64
THE QUARTERLY REVIEW OF BIOLOGY
Volume 80
not simple enough to lend themselves to or hierarchy competition, animals make
ready ridicule.
remarkably sophisticated decisions (Real and
In contrast, much of the controversy about Caraco 1986). Because physiologists and ethol-
adaptation reflects worldviews that emerge ogists see exquisite adaptations in their daily
from what scientists see in the narrow range work, their worldviews tend to presume good
of their daily work (Rose and Lauder 1996). function, leaving stochastic factors for late con-
Never was the hackneyed metaphor of four sideration.
blind men and an elephant more apt. Genet-
icists track mutations appearing, drifting, and
being selected in a profoundly stochastic pro-
cess that leads some neutral or even deleteri-
ous genes to go to fixation, while many useful
mutations either never occur or leave the gene
pool without a trace. Geneticists see interac-
tions with other genes and environments cre-
ating such phenotypic variability that it is hard
to say if a gene gives a selective advantage or
not. Likewise, the paleontologist’s daily work
involves evidence left by millions of species
come and gone. Sometimes a cataclysmic
event explains the departure of a species, but
extinction results as often from combinations
of unknown forces. Paleontologists also see
traits changing over time, sometimes for good
reasons, sometimes apparently just as the
result of the drift of traits across the eons. Con-
fronted with stochastic effects in their daily
work, it is not surprising that the worldviews of
geneticists and paleontologists tend towards
suspicion about adaptation.
Maladaptationism
Where does this leave George Williams? Is
he an adaptationist or an anti-adaptationist?
He is nothing so simple. If anything, he is a
maladaptationist. He has spent his entire
career preoccupied with one apparent mal-
adaptation after another. Very early, he won-
dered why senescence should exist at all.
Organisms can, he pointed out, regenerate
damaged tissues and even whole limbs.
Maintaining a body that has already devel-
oped should surely be a simpler matter than
developing a completely new soma from two
strands of DNA. Why has selection not cre-
ated bodies that repair themselves so they
endure, if not forever, at least much longer?
In a seminal paper, Williams suggested that
selection would maintain the frequency of
genes that gave benefits early in life even if
those same genes had pleiotropic effects that
caused steady and fatal damage to older
organisms (Williams 1957). To illustrate this
principle, he imagined a gene that changed
calcium metabolism in ways that speeded
bone healing but that also caused steady cal-
cification of the coronary arteries. Even in
the absence of senescence, other forces will
By contrast, the daily work of physiologists
brings them face to face with adaptations
(
Alexander 1996). A careful examination of
osmotic balance or blood pressure regula-
tion reveals mechanisms as precise as they
are robust (Vander et al. 1985). As the focus reduce the population with age, thus making
becomes more fine grained, wonderful selection stronger at younger ages and mag-
details emerge, such as bacterial navigation nifying the effect (Hamilton 1966; Austad
strategies (Dusenbery 1996), the exquisite 1994; Hawkes et al. 1998). Medawar had seen
contrivances plants use to lure insects and the same problem (Medawar 1952), but Wil-
anoint them with pollen (Proctor et al. liams made it explicit and initiated a new
1
996), and the pony fish’s glow that camou- perspective on senescence that remains on
flages it from deeper predators (Williams center stage (Williams 1999; Masoro and
997). Scientists who study comparative Austad 2001).
physiology cannot help but be impressed at
In the same article, Williams (1957)
1
the diversity of adaptations even in closely addressed the mystery of menopause. What
related species (Schmidt-Nielsen 1990). could be more senseless than premature fail-
Likewise, ethologists see adaptations at every ure of the female reproduction system, and
turn (Alcock 1998). Whether their attention what could explain survival beyond the age
is on bee communication, prepared learn- of reproduction? Where others saw a system
ing, bumblebee foraging, mating strategies, simply wearing out, Williams saw a major

March 2005
SYMPOSIUM IN HONOR OF GEORGE C WILLIAMS
65
question. His answer was that stopping indi- under the actual curve after maturity to the
vidual reproduction might maximize RS if area under a curve for which mortality
continued reproduction at advanced ages remained constant after maturity gave a very
might jeopardize the success of existing chil- rough estimate of the impact of senescence
dren. This provocative idea, containing the on reproductive success. The results for all
seeds of kin selection, spurs much continu- the life tables I could find for species in the
ing research (Austad 1994; Hawkes 2003).
wild showed vast differences in the pre-
My own path to Darwinian medicine sumed force of selection acting on senes-
began as a sophomore undergraduate when cence in mammals compared to birds, and
senescence suddenly struck me as a great interestingly, for rotifers (Nesse 1987, 1988).
mystery. The rate of aging is heritable and The method is not of great practical utility,
closely related species sometimes have very but it did inspire the idea that recent
different life spans, so it seemed that selec- changes in the average life span for a species
tion should either eliminate senescence, or might result in very different numbers of
at least maximize life span for each species. genes contributing to senescence. A species
Why does senescence persist? After several that has experienced a recent decrease in
months wandering in computerless libraries life span, perhaps because of new predation
with index cards in hand, I never found Wil- risks, might well show a long span of life in
liams’s article from ten years previously, but the laboratory in which the effects of senes-
I did arrive at an evolutionary answer: senes- cence were mediated by only a few new muta-
cence was essential to ensure the steady turn- tions. By contrast, a species like Homo sapiens,
over of individuals so that a species could whose life span has increased rapidly, would
evolve and not go extinct in a changing envi- be expected to have many modifier genes
ronment. This at least provided an evolu- pushing the expression of many deleterious
tionary explanation, but over the years it effects later and later, producing a formida-
kept seeming to me that this might not be ble obstacle to any attempt to slow the pro-
quite right. In medical school, I was taught gression of human senescence.
that natural selection was irrelevant; senes-
Also in 1957, Williams saw another mys-
cence is inevitable, we learned, because wear tery, cooperation, especially in insects. He
and tear damage the body. Shortly thereaf- collaborated with his wife on an article that
ter, in my quest to find a solid foundation for foreshadowed the revolutions of the 1960’s
understanding mental disorders, I made (Williams and Williams 1957), and the solu-
contact with a wonderful group of evolution- tions they proposed were close to those that
ary biologists at the University of Michigan. reshaped our understanding of social biol-
Soon, I told them about my continuing pre- ogy. Cooperation in humans, and especially
occupation with senescence. “What!?” they tendencies towards what appeared to be
said, “You have never read Williams, 1957?!” goodness, posed yet another mystery, one
I soon did, and the world has never looked that Williams addressed unflinchingly (Wil-
the same since. If evolution shaped senes- liams 1989).
cence, could it also help explain our vul-
Sex was next. Before most biologists rec-
nerability to cancer, schizophrenia, and ognized that there was a problem, Williams
impacted wisdom teeth? We needed to bring documented the huge fitness costs of sexual
evolutionary biology into the medical curric- versus asexual reproduction and proposed
ulum.
First, however, I wanted to test Williams’s
possible solutions (Williams 1975).
The pattern is clear; Williams saw the world
pleiotropy theory. It should be possible to through evolutionary lenses, but instead of
estimate the force of selection acting on fitting perceptions into the theory, the lenses
senescence by comparing actual survival highlighted every exception. Others have had
curves for organisms in the wild to hypo- this same tendency, notably Charles Darwin,
thetical curves based on the assumption of whose attention was constantly drawn to facts
no senescence. Taking the ratio of the area that did not fit, whether honeybee coopera-

66
THE QUARTERLY REVIEW OF BIOLOGY
Volume 80
tion or flowers with spurs longer than the species seems remarkable. Geckos have toe
tongue of any known pollinator.
pads covered with setae that make use of van
der Waal forces of molecular attraction to
Six Reasons for Disease Vulnerability allow them to walk on ceilings. Other lizards
reproduce asexually in expanding clones
By 1986, I was preoccupied with an evolu-
that are fine in the short run but liable to
tionary perspective on disease; it made no
extinction from epidemic disease. Some
sense that the body so well designed in some
trees monitor the wavelength of light and
respects could be so poorly designed in oth-
allocate more resources to upward growth if
ers. I went looking for an evolutionary biol-
the wavelength indicates the presence of
ogist to help me understand why selection
nearby competitors. Others have limbs so
had left us so vulnerable to disease. Williams,
fragile that they lose many in every wind-
it turned out, was working on the same prob-
storm. The human body concerns us most.
lem and was looking for someone who knew
Our eyes are wonderful but fundamentally
about disease. Our shared interest in senes-
fragile bundles of jury-rigged design. We
cence brought us together, and we soon
have heart valves that stay in an exacting
started work on “The Dawn of Darwinian
leak-proof profile for a hundred years, but
Medicine” (Williams and Nesse 1991). The
far earlier most of us will get Alzheimer’s dis-
magnitude of the knowledge gap in medicine
ease, coronary artery disease, breast cancer
soon became clear. Medical research has
or prostate cancer. Near the early peak of
asked about every conceivable proximate fac-
health, we can run for miles, but at the same
tor that might explain why some people get
age we get impacted wisdom teeth, acne, and
sick and others do not, but evolutionary ques-
the onset of anxiety and mood disorders.
tions about disease were hardly even recog-
These jarring contrasts between remarkable
nized. This now seems less surprising. The
adaptations and examples of hopelessly poor
medical goals of prevention and treatment
design demand attention. They suggest that
are advanced most directly by finding out
our thinking about maladaptation has been
what causes disease in individuals. Seeking
too casual.
evolutionary explanations for diseases seems,
Biologists have recognized the opportuni-
and in most cases is, of less immediate prac-
ties as well as the problems in the study of
tical benefit. Furthermore, diseases represent
adaptation and maladaptation and proposed
failure, and selection shapes things that work,
a variety of strategies to exploit them (Mayr
not things that fail, so the relevance of selec-
tion is not readily obvious. We began making
progress when we changed the object of
explanation from disease to vulnerability to
disease. This apparently small change in per-
spective made all the difference. For disease
after disease we asked ourselves, “Why isn’t
the body better designed?”
1
983; Reeve and Sherman 1993; Rose and
Lauder 1996). In particular, Crespi proposes
bringing phylogeny, teleonomy, develop-
ment, and genetics together in an integrated
quantitative approach to the study of the
selective landscapes that result in maladapta-
tion: “Selection defines the landscape on
which populations move, and to the extent
The usual answer in medicine, as well as in that environments and selective pressures
much of biology, has been that natural selec- change our landscape becomes a sea where
tion is an undirected process whose influence waves rise, sink and move like water in a bath”
is weak because of the stochastic nature of (Crespi 2000:625).
mutations and drift, so we should not expect
Our more medically-derived evolutionary
bodies to be anywhere near optimal. This per- approach to maladaptation asks why selection
spective predicts that most traits in most spe- has not made the body better. The vagaries of
cies will be suboptimal if not severely flawed. a stochastic process and the constant pressure
However, examination of actual traits for spe- from mutation and drift are one possibility,
cific species reveals some that are extraordi- however, there are five other reasons why
narily close to optimal, while others seem so selection leaves bodies with designs that
hopelessly ill wrought that the survival of the appear substandard.

March 2005
SYMPOSIUM IN HONOR OF GEORGE C WILLIAMS
TABLE 1
67
host as well as pathogen. For instance, auto-
immune diseases arise from malfunctions of
the inflammation system, and even the nor-
mal immune system exacts constant costs and
accumulating damage that result in disease.
Natural selection cannot accomplish other
potentially protective changes even given
plenty of time. Any change that improves the
performance of one trait is sure to harm
another. Such tradeoffs are intrinsic to all
designs, natural or human made. A car that
gets 40 miles per gallon will not reach 60
miles per hour in 6 seconds. The thin distal
ends of the human radius and ulna are fragile
and prone to fracture. If they were thicker,
however, we would not be able to rotate our
wrists with such adaptive felicity.
Six Evolutionary Explanations for
Vulnerability to Disease
•
•
•
Selection is slow
•
Mismatch between design and envi-
ronment
•
Competition with a pathogen or
other organism
Selection cannot solve some problems
irrespective of time
•
•
Tradeoffs
Constraints peculiar to living
systems, e.g. path-dependence
We misunderstand what selection
shapes
•
Traits that increase RS at the cost of
disease vulnerability
Other limits are peculiar to living systems.
Selection can only tinker with designs that
already exist. Because a species cannot go
into the shop for a few generations of retool-
•
Aversive defenses are readily mis-
taken for diseases
One obvious reason for disease vulnerabil- ing, many structures are fundamentally sub-
ity is the time lag between environmental optimal with no hope that selection can fix
change and adaptation by natural selection. them. The inside-out vertebrate eye is the
Periods of cataclysmic change, such as the exemplar; the human spine is another pain-
past 300 years, pose challenges for most ful reminder of what selection cannot do. In
organisms (Committee to Review the U.S. Cli- addition, the process of selection is stochastic,
mate Change Science Program Strategic Plan so useful alleles are lost and disease alleles
2
004) and lead to mass extinctions (Thomas persist, causing thousands of genetic diseases.
et al. 2004). We humans die young from heart The variation in many traits, such as leg
disease thanks to diet and exercise prefer- length, arises from multiple genes whose
ences and metabolic systems shaped in an interactions with each other and environ-
environment where constant ready access to ments ensures that some individuals will be at
salty, hot fatty meat was beyond imagination the maladaptive tails of the distribution for
(
Eaton et al. 2002). Certain South American that trait (Houle 1992). Then there is the
plants are in even more dire straights since the genetic code itself. Any information code will
extinction of the giant mammals that ate their accumulate errors, but diplodity makes
thorny fruits and spread the seeds ( Janzen organisms intrinsically vulnerable to diseases
and Martin 1982). Even in the absence of eco- that result from heterozygote advantage, and
logical catastrophes of the sort now unfolding, linkage disequilibrium guarantees that some
environments vary from year to year, making traits will be suboptimal (Crespi 2000).
adaptation somewhat suboptimal for most
organisms in most environments.
Other traits shaped by selection seem mal-
adaptive only because our vision is self-cen-
When the relevant environment is a path- tered. Most physicians imagine that selection
ogen or another competing organism, there shapes organisms for health and happiness. It
is no catching up as coevolutionary arms does, when that advances the interests of the
races create cycles of escalating responses and genes. However, a gene that increases RS will
counterresponses. Antibiotic resistance is the spread even if it reduces health, happiness, or
dramatic exemplar. Worse yet, the armaments longevity. Thus, we all age faster than we would
and defenses shaped in these arms races take otherwise, and young men have tendencies to
resources away from other essential tasks and do wild and dangerous things (Wilson and
yield mechanisms likely to be dangerous to Daly 1985). Even more troubling, a gene that

68
THE QUARTERLY REVIEW OF BIOLOGY
Volume 80
advances its own interests at the expense of
Medicine is adopting these ideas cautiously.
individuals will sometimes spread, despite They have aroused little criticism, and many
opposing mechanisms, possibly including expressions of surprise that all of this is not
recombination (Burt and Trivers 1998; Haig already well known. However, the fast growth
2
002).
of courses on evolution and disease has not
Finally, there are defenses. Cough, fever, been in medical schools but in biology and
pain, nausea, vomiting, pain, anxiety, and anthropology departments. The tight pack-
fatigue are not abnormalities, even if they do ing of the medical curriculum leaves it rigid,
motivate people to seek medical help. They and medical school funding systems generally
are aversive, almost certainly for good reason, demand federally funded research that prom-
but they are not themselves problems, they ises to help specific diseases, so it will likely
are solutions (Nesse and Williams 1994). The be some years yet before evolution is recog-
tendency to misinterpret them as problems is nized as an essential basic science for medi-
so pervasive it has been called “The Clini- cine.
cian’s Illusion.” Clinicians are prone to think
The biggest impediment may be the tacit
defenses are abnormal because they are metaphor physicians and medical researchers
expressed in the presence of disease, and gen- use to understand the body. Descartes initi-
eral medicine uses drugs to block them with ated a revolution when he described the body
remarkable safety. This apparent safety is as a machine. Ever since, physicians have
partly because our environment is safer than viewed disease as arising from flaws in the
it once was, but it is also because selection bodily machine. The early years of medical
shapes regulation mechanisms according to training make the machine seem spectacu-
the costs and benefits of responding versus larly well-designed. In the clinic, however, it
not responding to ambiguous cues. A signal seems like a slip shod piece of work. Either
detection analysis shows that normal systems way, the tendency is to view the body as a
express many false alarms. This “smoke detec- machine designed by an engineer.
tor principle” explains why so many defenses
The metaphor of the body as machine is
can be blocked safely, and why we should now restraining progress. Bodies are not
urgently look for rare dangers from drugs machines designed by engineers; they are
that block defenses (Nesse 2005).
somas shaped by natural selection. The vul-
Together, these six evolutionary explana- nerabilities that lead to disease do not arise
tions for vulnerability provide a powerful from miscopying blueprints manufacturing
framework for thinking about not only dis- errors. Instead, they arise from mutations and
ease but any apparently substandard design developmental sequences shaped by natural
feature in any organism. The tendency to selection. Some vulnerabilities manifest only
explain apparently suboptimal features by in a modern environment, others arise from
mentioning the vagaries of selection and its host-pathogen coevolution. Some reflect the
limits is casual thinking little different from limits of what selection can do. Some reflect
the complacent acceptance of group selec- our difficulty accepting that bodies are not
tion. A closer look at what selection shapes for us, but for our genes. Those genes are not
requires consideration of all the possibilities part of a single “normal genome,” and no sin-
for each trait in question. This challenge is gle body design defines normal for a species.
difficult, especially since definitive experi- Instead, collections of genes interact with
mental methods are rarely available. Also, environments to create variable phenotypes
evolutionary explanations for vulnerability that are more or less successful in surviving
are often multicausal and this offends human and reproducing copies of their genetic mate-
cognitive preferences. For example, our vul- rial. An evolutionary view disappoints our
nerability to atherosclerosis likely arises from human preference for essential types, and it
environmental novelties, inflammation sys- frustrates our wish to define normal. A body
tems shaped to protect us from pathogens, is a body, not a machine.
tradeoffs, constraints, and perhaps even ten-
The emphasis in the field of Darwinian
dencies that increase RS at the cost of health. medicine and this article has been on human

March 2005
SYMPOSIUM IN HONOR OF GEORGE C WILLIAMS
69
disease and implications for medicine. As oth- nation but with our theories. Whatever the
ers have noted, however, the serious investi- outcome, maladaptationists may make faster
gation of apparent maladaptations offers progress than adaptationists or anti-adapta-
opportunities across biology (Crespi 2000). tionists.
Why is DNA repair not better? Why does the
Whether Williams Vision benefits individ-
kiwi lay a single huge egg? Moreover, why, fol- uals is a different question. It may be rare for
lowing the persisting questions asked by Wil- good reason. The route lacks the pleasures of
liams, do sex and senescence persist despite seeing the world as a wondrous whole and
their costs? And exactly why does the Michi- calls attention to the dark side of nature. It
gan lily hold its blossom erect and open to rarely runs adjacent to vast pools of funding
damaging rain. Every species has dozens of and does not lead to ready recognition. Wil-
apparent maladaptations ripe for study.
liams Vision may be accurate, but there are
What about Williams Vision itself? Is it an few indications that it is an adaptation that is
adaptation spreading quickly, or a maladap- likely to spread soon. It seems more like a
tation ripe for study? Although the trait seems quirk, one that happens to benefit the whole
still to be rare, it does seem to offer a pow- species.
erful route to scientific progress. Careful
examination of phenomena that do not fit
our theories usually reveals nature’s subtle
acknowledgments
Thanks to George Williams for being the best collab-
ways that we failed to comprehend. Occasion- orator and teacher one could have, and to Barbara
ally, however, the fault is not with our imagi- Smuts for very helpful suggestions on this manuscript.
REFERENCES
Alcock J. 1998. Animal Behavior: An Evolutionary
Approach. Sixth Edition. Sunderland (MA): Sin-
auer Associates.
Alexander R M. 1996. Optima for Animals. Revised Edi-
tion. Princeton (NJ): Princeton University Press.
Austad S N. 1994. Menopause: an evolutionary per-
spective. Experimental Gerontology 29(3–4):255–63.
human longevity. American Journal of Human Biology
15(3):380–400.
Hawkes K, O’Connell J F, Blurton Jones N G, Alvarez
H, Charnov E L. 1998. Grandmothering, meno-
pause, and the evolution of human life histories.
Proceedings of the National Academy of Sciences USA
95(3):1336–1339.
Burt A, Trivers R. 1998. Genomic conflicts in genomic Houle D. 1992. Comparing evolvability and variability
imprinting. Proceedings of the Royal Society of London
B 265:2393–2397.
Committee to Review the U.S. Climate Change Sci-
ence Program Strategic Plan NRC. 2004. Imple-
of quantitative traits. Genetics 130:195–204.
Janzen D H, Martin P S. 1982. Neotropical anachro-
nisms: the fruits the gomphotheres ate. Science
215:19–27.
menting Climate and Global Change Research: A Review Masoro E J, Austad S N. 2001. Handbook of the Biology
of the Final U.S. Climate Change Science Program Stra- of Aging. San Diego (CA): Academic Press.
tegic Plan. Washington DC: National Academies Mayr E. 1983. How to carry out the adaptationist pro-
Press. gram? American Naturalist 121(3):324–334.
Crespi B J. 2000. The evolution of maladaptation. Medawar P B. 1952. An Unsolved Problem of Biology. Lon-
Heredity 84:623–629. don: H K Lewis.
Dusenbery D B. 1996. Life at Small Scale: The Behavior Nesse R M. 1987. Pages 45–64 in An Evolutionary Per-
of Microbes. New York: Scientific American Library.
spective on Senescence, edited by R M Nesse. Boston
Eaton S B, Strassman B I, Nesse R M, Neel J V, Ewald
(MA): D Reidel Publishing.
P W, Williams G C, Weder A B, Eaton III S B, Lin- Nesse R M. 1988. Life table tests of evolutionary the-
deberg S, Konner M J, Mysterud I, Cordain L.
002. Evolutionary health promotion. Preventive
ories of senescence. Experimental Gerontology 23(6):
445–453.
2
Medicine 34(2):109–118.
Nesse R M. 2005. Natural selection and the regulation
of defenses: a signal detection analysis of the
smoke detector principle. Evolution and Human
Behavior 26:88–105.
Haig D. 2002. Genomic Imprinting and Kinship. New
Brunswick (NJ): Rutgers University Press.
Hamilton W D. 1966. The moulding of senescence by
natural selection. Journal of Theoretical Biology Nesse R M, Williams G C. 1994. Why We Get Sick: The
1
2:12–45.
New Science of Darwinian Medicine. New York: Times
Hawkes K. 2003. Grandmothers and the evolution of
Books.

70
THE QUARTERLY REVIEW OF BIOLOGY
Volume 80
Nisbett R, Ross L. 1980. Human Inference: Strategies and Williams G C. 1957. Pleiotropy, natural selection, and
Shortcomings of Social Judgment. Englewood Cliffs
NJ): Prentice-Hall.
the evolution of senescence. Evolution 11(4):398–
411.
(
Paley W. 1802. Natural Theology. London: Wilks and Williams G C. 1966. Adaptation and Natural Selection: A
Taylor. [Reprint. 1970. Westmead (England):
Gregg International Publications.]
Critique of Some Current Evolutionary Thought. Prince-
ton (NJ): Princeton University Press.
Proctor M C F, Yeo P, Lack A. 1996. The Natural History Williams G C. 1975. Sex and Evolution. Princeton (NJ):
of Pollination. Portland (OR): Timber Press. Princeton University Press.
Real L, Caraco T. 1986. Risk and foraging in stochastic Williams G C. 1989. A sociobiological expansion of
environments. Annual Review of Ecology and System-
atics 17:371–390.
Reeve H K, Sherman P W. 1993. Adaptation and the
goals of evolutionary research. Quarterly Review of
Biology 68(1):1–32.
Evolution and Ethics. Pages 179–214 in Evolution and
Ethics: T H Huxley’s Evolution and Ethics with New
Essays on Its Victorian and Sociobiological Context,
edited by J Paradis and G C Williams. Princeton
(NJ): Princeton University Press.
Rose M R, Lauder G V. 1996. Adaptation. San Diego Williams G C. 1997. The Pony Fish’s Glow: And Other
(CA): Academic Press.
Clues to Plan and Purpose in Nature. New York:
Schmidt-Nielsen K. 1990. Animal Physiology: Adaptation
BasicBooks.
and Environment. Fourth Edition. Cambridge: Cam- Williams G C. 1999. The Tithonus error in modern
bridge University Press.
gerontology. Quarterly Review of Biology 74(4):
Segerstr a˚ le U C O. 2000. Defenders of the Truth: The Bat-
405–15.
tle for Science in the Sociology Debate and Beyond. New Williams G C, Nesse R M. 1991. The dawn of Darwin-
York: Oxford University Press.
ian medicine. Quarterly Review of Biology 66(1):
Thomas C D, Cameron A, Green R E, Bakkenes M,
1–22.
Beaumont L J, Collingham Y C, Erasmus B F N, de Williams G C, Williams D C. 1957. Natural selection of
Siqueira M F, Grainger A, Hannah L et al. 2004.
Extinction risk from climate change. Nature (427):
individually harmful social adaptations among sibs
with special reference to social insects. Evolution
11:249–253.
145–148.
Vander A J, Sherman J H, Luciano D S. 1985. Human Wilson M, Daly M. 1985. Competitiveness, risk taking,
Physiology: The Mechanisms of Body Function. Fourth
Edition. New York: McGraw Hill.
and violence: the young male syndrome. Ethology
and Sociobiology 6:59–73.

March 2005
SYMPOSIUM IN HONOR OF GEORGE C WILLIAMS
71