Full download ebook at ebookgate.

com

Adaptationism and Optimality 1st Edition

Steven Hecht Orzack

https://ebookgate.com/product/adaptationism-and-
optimality-1st-edition-steven-hecht-orzack/

Download more ebook from https://ebookgate.com

More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Polio 2nd Edition Alan Hecht

https://ebookgate.com/product/polio-2nd-edition-alan-hecht/

Optics 4th Edition Eugene Hecht

https://ebookgate.com/product/optics-4th-edition-eugene-hecht/

Partitions Optimality and Clustering Vol II Multi

Parameter 1st Edition Frank K Hwang

https://ebookgate.com/product/partitions-optimality-and-
clustering-vol-ii-multi-parameter-1st-edition-frank-k-hwang/

Optics instructor s solutions manual 4th Edition Eugene

Hecht

https://ebookgate.com/product/optics-instructor-s-solutions-
manual-4th-edition-eugene-hecht/
Optimality Conditions in Convex Optimization A Finite
dimensional View 1st Edition Joydeep Dutta

https://ebookgate.com/product/optimality-conditions-in-convex-
optimization-a-finite-dimensional-view-1st-edition-joydeep-dutta/

Minor Omissions Children in Latin American History and

Society Living in Latin America 1st Edition Tobias
Hecht

https://ebookgate.com/product/minor-omissions-children-in-latin-
american-history-and-society-living-in-latin-america-1st-edition-
tobias-hecht/

Protecting Games A Security Handbook for Game

Developers and Publishers 1st Edition Steven (Steven
Davis) Davis

https://ebookgate.com/product/protecting-games-a-security-
handbook-for-game-developers-and-publishers-1st-edition-steven-
steven-davis-davis/

Laws Mind and Free Will 1st Edition Steven Horst

https://ebookgate.com/product/laws-mind-and-free-will-1st-
edition-steven-horst/

Rethinking Superhero and Weapon Play 1st Edition Steven

Popper

https://ebookgate.com/product/rethinking-superhero-and-weapon-
play-1st-edition-steven-popper/
 Adaptationism and Optimality

The debate over the relative importance of natural selection as compared

to other forces affecting the evolution of organisms is a long-standing
and central controversy in evolutionary biology. Adaptationism is the
view that natural selection is so important, and nonselective forces so
unimportant, that accurate explanations and predictions of the pheno-
types of organisms can be obtained by simplified models in which selec-
tion is represented and nonselective forces are ignored. Adaptationists
and their critics disagree about this proposition concerning the history
of life, and they also disagree about the methodologies that are needed
to address this biological question. Many questions remain unresolved,
and the terms of the debate are still sometimes unclear.
Adaptationism and Optimality presents an up-to-date view of this con-
troversy and reflects the dramatic changes in our understanding of evo-
lution that have occurred in the past 20 years. The volume combines
contributions from biologists and philosophers and offers a systematic
treatment of foundational, conceptual, and methodological issues. The
essays examine recent developments in topics such as phylogenetic
analysis, the theory of optimality and ESS models, and the methodology
of hypothesis testing in evolutionary biology. The contributors’ dis-
agreement on fundamental aspects of this subject represents the diver-
sity of opinion that makes this controversy so highly charged. These
essays are intended to provide useful advice to “biologists in the
trenches” but also to assess the larger theoretical and conceptual issues
that form the basis of the current controversy.
This volume will serve to substantially advance the debate over adap-
tationism. It will be of interest to biologists, philosophers and historians
of biology, anthropologists, psychologists, and cognitive scientists.

Steven Hecht Orzack is President and Research Scientist at the Fresh

Pond Research Institute in Cambridge, Massachusetts.
Elliott Sober is Hans Reichenbach Professor of Philosophy at the Uni-
versity of Wisconsin, Madison.
 Cambridge Studies in Philosophy and Biology

General Editor
Michael Ruse Florida State University

Advisory Board
Michael Donoghue Yale University
Jean Gayon University of Paris
Jonathan Hodge University of Leeds
Jane Maienschein Arizona State University
Jesús Mosterín Instituto de Filosofía (Spanish Research Council)
Elliott Sober University of Wisconsin

Alfred I. Tauber The Immune Self: Theory or Metaphor?

Elliott Sober From a Biological Point of View
Robert Brandon Concepts and Methods in Evolutionary Biology
Peter Godfrey-Smith Complexity and the Function of Mind in Nature
William A. Rottschaefer The Biology and Psychology of Moral Agency
Sahotra Sarkar Genetics and Reductionism
Jean Gayon Darwinism’s Struggle for Survival
Jane Maienschein and Michael Ruse (eds.) Biology and the
Foundation of Ethics
Jack Wilson Biological Individuality
Richard Creath and Jane Maienschein (eds.) Biology and Epistemology
Alexander Rosenberg Darwinism in Philosophy,
Social Science and Policy
Peter Beurton, Raphael Falk, and Hans-Jörg Rheinberger (eds.)
The Concept of the Gene in Development and Evolution
David Hull Science and Selection
James G. Lennox Aristotle’s Philosophy of Biology
Marc Ereshefsky The Poverty of the Linnaean Hierarchy
Kim Sterelny The Evolution of Agency and Other Essays
William S. Cooper The Evolution of Reason
Peter McLaughlin What Functions Explain
 Adaptationism and Optimality

Edited by

steven hecht orzack elliott sober

The Fresh Pond Research Institute University of Wisconsin, Madison
PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE
The Pitt Building, Trumpington Street, Cambridge, United Kingdom

CAMBRIDGE UNIVERSITY PRESS

The Edinburgh Building, Cambridge CB2 2RU, UK
40 West 20th Street, New York, NY 10011-4211, USA
10 Stamford Road, Oakleigh, VIC 3166, Australia
Ruiz de Alarcón 13, 28014 Madrid, Spain
Dock House, The Waterfront, Cape Town 8001, South Africa
http://www.cambridge.org

© Cambridge University Press 2001

This book is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.

First published 2001

Printed in the United States of America

Typeface Palatino 10.25/13pt. System QuarkXPress [BTS]

A catalog record for this book is available from the British Library.

Library of Congress Cataloging in Publication Data

Adaptationism and optimality / edited by Steven Hecht Orzack, Elliott Sober.
p. cm.
ISBN 0-521-59166-X (hardcover) – ISBN 0-521-59836-2 (pbk.)
1. Adaptation (Biology) I. Orzack, Steven II. Sober, Elliott.
QH546.A285 2001
578.4 – dc21 00-040324

ISBN 0 521 59166 X hardback

ISBN 0 521 59836 2 paperback
 To Ed Ricketts,
because he loved true things
 Contents

Contributors page xi
Acknowledgments xv

Introduction Steven Hecht Orzack and Elliott Sober 1

1 A Likelihood Framework for the Phylogenetic Analysis
of Adaptation David A. Baum and Michael J. Donoghue 24
2 Adaptation, Phylogenetic Inertia, and the Method of
Controlled Comparisons Steven Hecht Orzack and
Elliott Sober 45
3 Optimality and Phylogeny: A Critique of Current
Thought Hudson Kern Reeve and Paul W. Sherman 64
4 Fit of Form and Function, Diversity of Life, and
Procession of Life as an Evolutionary Game
Joel S. Brown 114
5 Optimality and Evolutionary Stability under Short-Term
and Long-Term Selection Ilan Eshel and
Marcus W. Feldman 161
6 Selective Regime and Fig Wasp Sex Ratios: Toward
Sorting Rigor from Pseudo-Rigor in Tests of Adaptation
Edward Allen Herre, Carlos A. Machado, and Stuart A. West 191
7 Is Optimality Over the Hill? The Fitness Landscapes of
Idealized Organisms George W. Gilchrist and
Joel G. Kingsolver 219
8 Adaptation, Optimality, and the Meaning of Phenotypic
Variation in Natural Populations Kenneth J. Halama and
David N. Reznick 242

ix
 Contents

9 Adaptationism, Optimality Models, and Tests of

Adaptive Scenarios Peter Abrams 273
10 Adaptation and Development: On the Lack of
Common Ground Ron Amundson 303
11 Three Kinds of Adaptationism Peter Godfrey-Smith 335
12 Adaptation, Adaptationism, and Optimality
Egbert Giles Leigh, Jr. 358

Index 389

x
 Contributors

Peter Abrams is a Professor in the Department of Zoology at the

University of Toronto. His main interests are in evolutionary and
ecological theory, and he has recently worked on models of
the coevolution of interacting species, the evolution of frequency-
dependent characters, and interactions in food webs.
Ron Amundson is a Professor of Philosophy at the University of
Hawaii at Hilo. His research areas are the history and philosophy of
biology, and disability politics. He is currently writing a book on the
history of the relations between evolutionary and developmental
biology.
David A. Baum is an Associate Professor of Biology at Harvard Uni-
versity. His primary interests are in phylogenetic theory and in mech-
anisms of floral evolution. Recent publications have dealt with the
concept of “species,” the phylogeny and floral evolution of relatives
of cotton, and studies of the genes underlying plant morphological
evolution.
Joel S. Brown is a Professor of Biology at the University of Illinois at
Chicago. He studies the ecology of fear: How do animals balance con-
flicting demands for food and safety? How do fear responses influ-
ence the ecology of the prey and their predators? With collaborators,
his current work includes squirrels in Chicago, gerbils and owls in
Israel, mule deer and mountain lions in Idaho, and blue sheep and
snow leopards in Nepal.
Michael J. Donoghue is G. Evelyn Hutchinson Professor of Biology
at Yale University. His primary interest is in evolutionary history and
specifically in plant phylogeny. His recent publications have focused

xi
 Contributors

on patterns of convergent evolution in plants (breeding systems,

flower symmetry, plant architecture), the origin and radiation of
flowering plants, and the phylogeny of basidiomycete fungi.
Ilan Eshel is a Professor in The School of Mathematical Sciences
at Tel Aviv University. During the past 20 years, his research
has focused on closing the serious theoretical gaps between predic-
tions drawn from exact genetic models of evolutionary change and
those obtained from more intuitive and convincing game-theoretic
arguments.
Marcus W. Feldman is a Professor of Biological Sciences at Stanford
University and Director of The Morrison Institute for Population and
Resource Studies. He is also Co-Director of the Center for Computa-
tional Genetics and Biological Modelling. His research has been in
population and evolutionary genetics of complex traits, in how cul-
tural transmission affects evolutionary dynamics, in human molecu-
lar evolution, and in the demographic effects of son preference. He is
managing editor of Theoretical Population Biology and associate editor
of Genetics and Complexity. He is the author or coauthor of three books
and more than 260 research articles.
George W. Gilchrist is an Assistant Professor at Clarkson University,
a small research college in northern New York. His recent publica-
tions have focused on the adaptation of populations to a changing
climate and evolution in introduced species.
Peter Godfrey-Smith is an Associate Professor of Philosophy at Stan-
ford University. He is the author of various articles in the philoso-
phy of biology and of Complexity and the Function of Mind in Nature
(Cambridge University Press, 1996).
Kenneth J. Halama is a Senior Research Biologist for Planning Con-
sultants Research, a private company based in Irvine, California. His
research interests include the adaptive significance of phenotypic
variation within vertebrate populations and the conservation biology
of threatened and endangered reptiles and amphibians. Currently, he
is co-investigator on a project studying the ecology of the federally
endangered Arroyo toad (Bufo californicus) in the mountains of south-
ern California.
Edward Allen Herre is a Research Scientist at the Smithsonian Trop-
ical Research Institute in the Republic of Panama. Recent publications

xii
 Contributors

have examined laws governing species interactions, the evolution of

mutualisms, and the effects of pervasive density dependence recruit-
ment on seedling diversity in a tropical forest.
Joel G. Kingsolver is a Professor of Biology at the University of North
Carolina. His research focuses on the evolutionary ecology and evo-
lutionary physiology of insects.
Egbert Giles Leigh, Jr., has been with the Smithsonian Tropical
Research Institute in Panama since 1969. His primary interests are
tropical forest ecology and the interplay among different levels of
selection. He recently published Tropical Forest Ecology.
Carlos A. Machado is a Postdoctoral Fellow in the Department of
Genetics, Rutgers University. His current work includes phylogenetic
and biogeographical studies of fig pollinators and parasitic fig wasps.
He also uses these organisms to study factors that influence rates of
molecular change and levels of molecular diversity.
Steven Hecht Orzack is the President and Founder of the Fresh Pond
Research Institute in Cambridge, Massachusetts. His recent research
has focused on phylogenetic methods, sex ratio evolution in a para-
sitic wasp (using ecological, genetic, and molecular methods), life
history evolution in variable environments, evolutionary game
dynamics, conservation biology, and human population genetics.
Hudson Kern Reeve is an Associate Professor in the Department of
Neurobiology and Behavior at Cornell University. He studies the evo-
lution of reproductive skew, cooperation, and conflict in animal soci-
eties (especially wasp societies). He recently co-edited Game Theory
and Animal Behavior with Lee Dugatkin.
David N. Reznick is a Professor in the Department of Biology at
the University of California, Riverside. His research interests include
the experimental study of evolution by natural selection in natural
populations and the balance of costs and benefits associated with dif-
ferent components of the life history. He has worked primarily on
guppies from the island of Trinidad, focusing on the evolution of life
history traits in response to natural differences in mortality rates.
Paul W. Sherman is a Professor in the Department of Neurobiology
and Behavior at Cornell University. He studies the behavioral ecology
of various mammals and birds and teaches animal social behavior

xiii
 Contributors

and Darwinian medicine. He has published more than 125 papers

and books on these topics.
Elliott Sober is Hans Reichenbach Professor of Philosophy and
Henry Vilas Research Professor at the University of Wisconsin,
Madison, where he has taught since 1974. Recent books include Unto
Others: The Evolution and Psychology of Unselfish Behavior (with David
S. Wilson, 1998), From a Biological Point of View (Cambridge Univer-
sity Press, 1994), and Reconstructing the Past (1988), for which he won
the Lakatos Award for outstanding contribution to the philosophy
of science.
Stuart A. West is a BBSRC David Phillips Research Fellow at the Insti-
tute of Cell, Animal, and Population Biology, University of Edin-
burgh. He is currently working on the evolution of reproductive
strategies, especially sex and sex allocation. His research aims to (1)
address general evolutionary questions and (2) improve our under-
standing and control of pest and parasite species.

xiv
 Acknowledgments

We thank Michael Ruse and Terry Moore for encouragement, fore-

bearance, and enthusiasm. Their support made this volume possible
and greatly improved it as well. We also thank Betsy Hardinger, Joan
Omoruyi, Robyn Wainner, and Helen Wheeler for excellent assistance
during production.
It was important for us that the content of this volume reflect more
than the claims and opinions of the authors. The intent of this volume
is to promote dialog. For this reason, each chapter received three
or four reviews by other scientists. In many instances, we chose
reviewers whose views were quite distinct from those of the authors.
The efforts of these reviewers greatly improved the quality of each
chapter and of the book as a whole. We are grateful to all. The
reviewers were Carl Bergstrom, Jane Brockmann, Dale Clayton, Bill
Etges, Steve Frank, Sasha Gimelfarb, Jaco Greeff, Brian Hall, Thomas
Hansen, Gordon Hines, Kevin Johnson, Mark Johnston, Alexey
Kondrashov, Mary McKitrick, Deborah McLennan, Alan Molumby,
Paul Ode, Mark Pagel, Dave Parker, Tim Prout, David Rivers, Derek
Roff, Jay Rosenheim, Michael Ryan, Dolph Schluter, Jon Seger, Larry
Shapiro, Neil Shubin, Andrew Simons, Chris Stephens, Bill Stubble-
field, Shripad Tuljapurkar, Michael Turelli, Gunther Wagner, Stephen
Weeks, and David Sloan Wilson.
We were partially supported by National Science Foundation
(NSF) grants DEB-9407965 to S.O. and SES-9906997 to S.O. and E.S.

xv
 Introduction

steven hecht orzack and elliott sober

Scientific controversies have a bad reputation these days. Instead of

being viewed as an engine of scientific progress, they are seen by
many scientists and others as clashes of unresolvable ideological dif-
ferences in the way the world is perceived and understood. One of
us long ago, as a know-it-all Harvard graduate student in conversa-
tion with other such graduate students, often referred to this or that
controversy as an example of the “dialectic of polarized fanatics,”
with the absolute but unspoken know-it-all understanding that there
was little, if anything, useful taking place. More meaningfully
perhaps, controversies are often said to generate “more heat than
light” on a given scientific issue and, to this extent, to do little to
advance understanding. To some observers, the debate over adapta-
tionism in evolutionary biology has this character. From this per-
spective, the various views about the power of natural selection to
shape phenotypes are so different that proponents of each talk past
each other and often do not even acknowledge the potential validity
of alternative views. Of course, one can still view such steadfast-
ness as essential for the progress of science instead of as indicating
an unproductive narrow-mindedness. One need not think that the
debate over adaptationism is unresolvable or agree with either view
as to the nature of ideological differences to appreciate the impor-
tance of refining our understanding of the meaning of adaptation
and of optimality models. At worst, such an endeavor can help indi-
vidual scientists in the practice of their craft, even if incommen-
surable views of the world still remain; at best, it can promote
understanding by members of the scientific community of the
meaning of the debate and an understanding of how to go fruitfully
beyond the polarized “dialectic” mentioned earlier. It is our hope that

1
 steven orzack and elliott sober

the collected essays in this volume contribute in both of these indi-

vidual and collective ways.
In organizing this volume we had two motivations. The first was
our sense that it was important to understand how the explosion of
work in evolutionary biology in the past 20 years or so has affected
the debate over adaptationism. Of most relevance are the dramatic
increases in methods and application of phylogenetic analysis, in the
application of molecular techniques, and in the formulation, analy-
sis, and testing of optimality and evolutionarily stable strategy
models. We say more about these changes later.
A second motivation was our observation that it has been common
to view this debate over adaptationism mainly in terms of recent
claims made by scientists such as Richard Dawkins, Steve Gould,
Richard Lewontin, John Maynard Smith, and George Williams. We
regard such a view as narrow and implicitly a means of portraying
the issues as of little practical interest, especially to “biologists in
the trenches.” Some people view the debate as little more than an
ephemeral battle of egos and say, “What could be less important?” In
fact, some of the issues raised in the debate are the very same issues
raised many times by several members of the evolutionary pantheon,
especially by Ronald Fisher and Sewall Wright. Yet “impractical” is
never a label for either of these greats. By comparison, one hears state-
ments to the effect that the biologists raising such issues now are
engaged in little more than intellectual game-playing (this is often
phrased much less politely); to many biologists, making a claim about
adaptationism is deemed something worse: philosophy. Such antipa-
thy may stem from a general attitude in scientific culture that is epit-
omized by one prominent physicist’s statement that “the philosophy
of science is about as useful to scientists as ornithology is to birds.”
Such views are silly at best. Still, some of these views about the debate
over adaptationism stem from the fact that much of what has been
written about it has little direct contact with and implication for the
practice of evolutionary biology. This volume is an attempt to remedy
this situation by presenting an integrated set of views about the real
biology involved in the debate. In this way, we hope to foster a mutu-
ally beneficial connection between the practice of evolutionary
biology and the debate.
We write as participants, having published several articles about
adaptationism and optimality in the past few years (Orzack and
Sober 1994a, b; 1996). To order the conceptual landscape, we first

2
 Introduction

discuss the meaning of several claims about natural selection and

then sketch our claims about adaptationism and explain how they
relate to others.
Several important issues are at stake in the debate over adapta-
tionism. The first is what this “ism” really means formally and in
practice. The second is whether and how adaptationism is testable,
how tests of specific optimality models would relate to such a test,
and how the result of a test of a specific model relates to the hypoth-
esis of optimality.
To understand more about what adaptationism means to different
scientists, it is useful to distinguish three claims about a trait. The first
is that natural selection played some role in its evolution. The second
is that natural selection was an important cause of the trait’s evolu-
tion. The third is that natural selection is the only important cause of
the trait’s evolution.
We think that almost all evolutionary biologists accept the first
claim with respect to most traits. To this extent, this claim does
not serve as a dividing line for ideology or practice. Instead, the
proverbial line in the sand is drawn between the second and third
claims we’ve outlined. Of course, talk of a line in the sand obscures
some of the variety of views about the power of natural selection
in trait evolution. When it comes to evolutionary explanation,
however, there are many biologists who look only to natural selec-
tion except under the most unusual circumstances. This goes beyond
mere invocation of natural selection as a possible evolutionary
force; nonselective explanations are excluded. On the other hand,
there are also many biologists who think of natural selection as only
one of several forces that might plausibly play an important role in a
given trait’s evolution. These views about the nature of trait evolu-
tion are substantially different. Someone with the first perspective
expects, often without specific evidence, that natural selection plays
an important role in any given trait’s evolution; those who claim that
the trait is affected by natural selection only weakly or not at all must
provide specific evidence to this effect. In short, natural selection is
guilty by default. Someone with the second perspective expects to
have to provide specific evidence that any particular evolutionary
force plays an important role in a trait’s evolution. No force is guilty
by default here.
Given that proponents of each of these views often do not even
acknowledge the plausibility of an alternative, what does it mean to

3
 steven orzack and elliott sober

talk, as we and others do, of a debate over adaptationism? Some

of this talk stems from a perspective that probably comes easier
to historians and philosophers of biology than to many biologists. Of
course, this is not a debate in the Lincoln–Douglas sense. Here, the
term debate describes the flux of ideas in the community of biologists
about the power of natural selection. This is a real debate in the broader
sense of the word and one whose resolution has significant scientific
implications. It is a debate in which score is kept partially in terms of
the triumvirate of academic goals: tenure, prizes, and grants. Facts
matter in this debate, but so do power and prestige. Although all the
contributions to this volume relate to the facts of biology, the latter
influences should be accepted as real.
Understanding what it means to refer to the debate over adapta-
tionism leads one to confront an important discrepancy between per-
ception and practice. Even well-known advocates of the optimality
approach generally do not admit to believing that natural selection
can be the only important cause of a trait’s evolution; claiming that
several evolutionary forces usually play such a role is a typical
self-description of beliefs. (The well-known producer of optimality
models who had the refreshing candor to write of himself that “his
personal discovery of natural selection constitutes the only religious
experience of his life” is perhaps an exception in this regard.) But
in fact many of these “adaptationists” practice a research program
that appears to be consistent only with the idea that natural selection
is the sole important force. On the other hand, many other scientists
often make statements about the general evolutionary importance of
several forces including natural selection but add that in a given
instance it could be the only important force. Here the discrepancy
between perception and practice is between the latter statement and
the unwillingness of these “pluralists” to agree to any claim that a
given trait is optimal. We make no judgment in regard to these dis-
crepancies between perception and practice. Our sense is that they
are motivated by a belief that appearing to be conceptually narrow is
not consistent with being a professional scientist. These discrepancies
are a welcome reminder of the human character of science and are
worthy of further study.
There is a big difference between saying that natural selection was
an important cause of the evolution of a trait and saying that it was
the only important cause. How is it possible for any evolutionary
process to be the only important cause of a trait? Surely, there must

4
 Introduction

always be some restrictions on the power of natural selection.

After all, populations are always finite, mutations always occur,
migrants always appear, and so forth. In fact, none of these facts
necessarily implies that the power of natural selection must be
reduced. For example, for some kinds of traits even small population
size can enhance the intensity of natural selection. So in a given analy-
sis the question to be answered is whether natural selection is pow-
erful enough that forces expected in a given instance to reduce the
causal role of natural selection are in fact of no significant con-
sequence. Our claim is that this question can be answered only
with a test that includes assessment of the quantitative accuracy of
an optimality model and of the nature of phenotypic variation, if any,
in the population. Of course, any such model must have some con-
straints. Does their presence make optimality impossible? No. To
think that it does confuses the potential problems associated with
trait definition with the potential nature of a given trait’s evolution.
For example, the optimum in most models of sex ratio evolution is
selected from among a set of phenotypes whose brood sizes are con-
strained to be equal. The presumption here is that brood size and sex
ratio can be “dissected at the joints.” That such a dissection could be
wrong does not imply that sex ratio cannot be locally optimal. (For
that matter, it might well be that the constraint on brood size results
from the action of natural selection.) Our point is that every opti-
mality model describes the evolution of a trait as occurring relative
to a substantive background of biological traits already in place.
Claims about background biology are fallible, of course, but this is
true of any scientific claim and so poses no special obstacle in this
context.
A common contention by pluralists is that population-genetic
theory leads one to expect that optima should not occur in nature.
The usual basis for this statement is the correct assertion that specific
conditions relating to the nature of trait determination by genes must
be satisfied in order that an optimal trait evolve. This specificity is
taken to imply rarity of occurrence. This inference is invalid because
it confounds the state of theory with the state of nature. On the other
hand, a common contention by adaptationists is that optimality
can be expected because natural selection is ubiquitous and “appro-
priate” genetic variation is always available. We find it hard to
believe that optimality should be expected just because it could
occur, as when optimality is regarded as a null hypothesis. Even the

5
 steven orzack and elliott sober

ubiquity of intense natural selection would not guarantee the evolu-

tion of optima.
What is adaptationism then? We define it as the claim that
natural selection is the only important cause of the evolution of
most nonmolecular traits and that these traits are locally optimal.
This definition builds upon our previous claim about the discrep-
ancy between perception and practice. For us, what is at stake are
the practices of biologists, and not their self-assessments. This dis-
tinction is rarely made by adaptationists and pluralists. For them,
what usually counts are the statements by biologists about their
research programs.
Several aspects of this definition place it at odds with other
common claims about adaptationism. One such aspect is the explicit
connection between adaptationism and optimality. To our knowledge,
such a connection has not been made previously; we expect that it is
viewed by many self-styled adaptationists as inappropriate, perhaps
because they think it invites having adaptationism labeled as extreme.
Yet, in our opinion, the biology practiced by most adaptationists is in
fact, rightly or wrongly, a biology in which traits are explained as
optima, although their identification as such is often not explicit.
This happens in all areas of evolutionary biology but especially
so in studies of animal behavior, behavioral ecology, evolutionary
psychology, functional anatomy, and sociobiology.
We think that adaptationism as defined here is testable by the accu-
mulation of successes and failures of specific optimality models. To
many, the truth or falsity of adaptationism has been a premise; for us,
the truth or falsity is a possible conclusion resulting from an ensem-
ble of model tests. Our point is that as such tests accumulate, the pro-
portion of those in which an optimality model appears to provide a
complete explanation for a trait tells us about the truth of adapta-
tionism. As far as we are concerned, if 90% of studies had such a con-
clusion, we would regard adaptationism as true. Is there potential
ambiguity in regard to what one should conclude if the percentage
of studies is, say, 60%? Absolutely. Of course, such an outcome might
be interpretable. But if not, it would still be an important advance in
evolutionary biology if only because the construction of any ensem-
ble requires community acceptance of standards for model testing
and for the interpretation of test results. What about the claim that
an investigator might never accept model failure as evidence for
nonoptimality because this result can always be attributed to incom-

6
 Introduction

plete understanding? Of course, an investigator might also never

accept model success as evidence for optimality. One sees then that
investigator “bias,” if any, has no special implication either way.
Nonetheless, any bias is a concern. Our hope is that its effects are
lessened as collective expectations in regard to the assessment of
test results become increasingly important over time.
The symmetry of bias in regard to test interpretation brings to
mind a larger asymmetry. Why have we framed the test of adapta-
tionism in terms of tests of optimality models? Why not frame it in
terms of tests of nonoptimality models? In a formal sense, there is
nothing to prevent this approach. However, a significant advantage
of optimality models is that testing them quantitatively generally
requires less information. After all, the presumption of many such
models is that in order to explain a trait we need to know little or
nothing about population size, about genetic constraints, and so
forth. So, in this sense, framing the test of adaptationism in terms of
testing optimality models is a test of an important scientific gambit.
Of course, we stress that the analysis of any particular trait should
not give priority to any particular kind of model; any available model
that is plausibly relevant should be considered along with the opti-
mality model. What counts is how competing models compare in
regard to explaining data.
Why do we need a test of adaptationism? Perhaps the paramount
benefit is that it could help answer a central question in evolutionary
biology, one whose resolution, even if it is contingent and provisional,
would be a major accomplishment by the field. It could also make it
easier to decide whether failure of a specific optimality model is more
likely the result of true nonoptimality or of an incorrectly formulated
model. Such an answer would also help standardize test protocols;
the present discrepancies among criteria by which optimality models
are accepted and rejected serve to lessen the biological significance of
otherwise important analyses.
We stress our belief that the optimality or nonoptimality of a given
trait is a claim to be made with reference to the result of a quantita-
tive test of a specific evolutionary model. The alternative is to have
claims that are not scientific claims in the standard sense. Yet such
claims are unfortunately common, as when a biologist says that the
optimality of a trait is a given and that it is not under test when the
predictions of an optimality model are compared with data, or when
a biologist says in a preemptive sense that a trait cannot be optimal.

7
 steven orzack and elliott sober

If testing adaptationism depends on the testing of specific opti-

mality models, what does this involve in practice? Our answer stems
from the need in this context to determine whether natural selection
is the only process that needs to be taken into account.
One must first assess the quantitative accuracy of model predic-
tions with respect to a statistical description of an individual’s trait,
because local optimality is a hypothesis about individuals and their
differences. Optimality means that the trait is the best relative to traits
possessed by other individuals with which it interacts; the result of
this functional superiority is a higher fitness. The expected conse-
quence is trait homogeneity; at most, there might be rare variants
perhaps introduced by mutation. But a snapshot could still reveal
variation among individuals if an optimal trait has several compo-
nents, any one of which might be expressed at a given time. The
necessity of such a focus on the traits of individuals and on the nature
of variation may seem self-evident; in fact, however, as we describe
later, it is very rare in practice.
The assessment of trait variation within a population is essential
for establishing the causal meaningfulness of claims for or against
optimality. Why? Consider a case in which natural selection causes
the average value of a trait in a population to be stable over time and
to match the prediction of an optimality model, although there is
between-individual trait variation because individuals with different
traits have the same fitness. A claim that each of these traits is optimal
in a different way would not be appropriate because a causally mean-
ingful claim about optimality is a claim that natural selection governs
the fate of competing traits; here, the traits’ relative frequencies could
be affected strongly by genetic drift because the traits are neutral with
respect to one another.
Our focus on quantitative analysis results from our view that
a paramount role of scientific models is to clearly delineate the
explained and the unexplained. Quantitative analysis does this with
far less ambiguity than does qualitative analysis. By the former, we
mean the statistical comparison of numerical predictions and data;
by the latter, we mean the statistical comparison of predicted and
observed trends in the data (as when a model is tested by determin-
ing whether the predicted direction of bias in a proportion matches
the observed direction). Qualitative and quantitative agreement of
data with any model means that it provides a complete local expla-

8
 Introduction

nation of the state of the trait. (We remind the reader that any math-
ematically based optimality model can generate quantitative pre-
dictions.) As described later, many tests of optimality models have
been only qualitative at best. But qualitative fit without quantitative
fit means that the model is incomplete or its structure is incorrect
in some regard. Either possibility in regard to an optimality model
makes a claim of optimality inappropriate.
The importance of assessing quantitative fit in the context of
testing adaptationism heightens the importance of standardization
of protocols and of statistical power analyses when optimality models
are tested. After all, a small sample size by itself can result in quan-
titative agreement between predictions and observations or in the
appearance of trait homogeneity. The result would be a bias toward
the acceptance of trait optimality and, thereby, of the thesis of adap-
tationism. There is no single way to deal with this problem. At least
initially, we encourage investigators to assess the extent to which
support for a claim of optimality may result from a lack of statistical
power. As analyses accumulate, sample-size standards will develop
that investigators can use when designing experiments.
Although there are hundreds of tests of optimality models in the lit-
erature and the bearing of between-individual variation and of quan-
titative testing on the question of optimality may seem obvious, we can
find only two sets of studies in which data and analyses are currently
structured so as to allow one to assess local optimality. These two sets
constitute the ensemble test of adaptationism at present.
The first set of studies focuses on the mixture of two reproductive
behaviors in the digger wasp Sphex ichneumoneus (see Brockmann and
Dawkins 1979; Brockmann et al. 1979). The authors’ optimality model
accurately predicts the observed mixture of behaviors for a popula-
tion in New Hampshire; by comparing lifetime behavioral records of
individuals, one can also demonstrate that individuals do not differ
significantly in the mix of behaviors each produces. We conclude
from the present combination of data and theory that the mixture of
reproductive behaviors in this population is locally optimal. The
authors’ model cannot be reconciled with the data from a popu-
lation in Michigan. The analysis of the New Hampshire data serves
as a reminder that optimality models can make predictions that
quantitatively match data. General claims that optimality models
cannot be accurate in this way are simply incorrect.

9
 steven orzack and elliott sober

The second set of studies focuses on sex ratio behaviors in the para-
sitoid wasp Nasonia vitripennis (see Orzack and Parker 1990; Orzack
1990; Orzack et al. 1991). For strains of wasps sampled from a
Swedish population, the predictions of optimality models for three
behaviors are not quantitatively accurate and there are significant
differences among strains in their fit to predictions. A reasonable
interpretation of these results is that natural selection is an important
force causing sex ratios to evolve in the direction predicted by the
models; however, the stronger claim of local optimality is not sup-
ported by the data.
Like all scientific statements, both of these claims about optimal-
ity are contingent in the sense that new data and consideration of
new evolutionary models may lead to a changed assessment. So,
for example, a selective model that does not invoke optimality but
that explains the Sphex data better than the present model might
lead to a revised assessment of the trait in the New Hampshire
population.
As far as we know, these tests of optimality models differ from all
others in the literature because they include both quantitative assess-
ment of model accuracy and comparisons of the phenotypes of
individuals or isofemale strains. Many other tests include only a qual-
itative prediction of the model against a population phenotype or a
phenotypic average; some others include either a quantitative analy-
sis by itself or some comparison of individuals. As a result, each of
these studies can at best support the claim that natural selection
is an important cause of a trait’s evolution. Because mustering the
evidence to support this conclusion is a significant accomplishment,
many of these “incomplete” tests, even those involving only a test of
qualitative predictions, are worthy of high praise (e.g., Milinski 1979).
But none of these tests can support a claim for or against local opti-
mality. It is here that one can best appreciate the distinct nature of our
views about how to test an optimality model because many of these
tests have been accompanied by just such claims.
It is worthwhile reflecting on the use of ensemble tests in evolu-
tionary biology. The contingent nature of all evolutionary models and
data sets is no barrier to the construction or assessment of such a test.
Otherwise, as evolutionary biologists, we could not trade in state-
ments such as “Speciation is allopatric” or “Species are genetically
polymorphic.” But we do, and appropriately so. Although such
statements are presumably not meant to be taken literally in the

10
 Introduction

sense that all species are said to have a certain property, they are
used in such a way that the implication is that they describe the vast
majority (more than 90%?) of species. Is it conceivable that more
than a very small fraction of species could actually have contributed
data to these ensemble claims? Of course not. But such statements
are taken to be based on enough independent data that we confi-
dently “fill in the gaps” and conclude that the data correctly repre-
sent a larger whole. Our point is not that these statements about
species are necessarily suspect but rather that the ensemble assess-
ment of hypotheses in evolutionary biology is nothing unusual. Of
course, this is not an entirely positive tradition. A common weakness
of ensemble statements is that there have been no clear criteria used
for construction of the ensemble; at best, one imagines that the hap-
hazard process by which this or that entity has been chosen for inten-
sive study (e.g., Drosophila melanogaster) or inclusion allows one to say
that some sort of representative sampling of species has occurred. In
the present instance, we think it best that investigators make a col-
lective effort in the form of a formally organized Adaptationism
Project to choose traits and species and to decide on methods of
analysis. Such an attempt would distinguish this ensemble test from
other such tests in evolutionary biology, which have not been for-
mally organized to our knowledge. One complex but resolvable
methodological question is the one confronting all ensemble tests
involving species: How does one properly assess a species, in as
much as it is an ensemble of populations?
For that matter, there are organizational and professional chal-
lenges to an effort to construct an ensemble test of adaptationism. It
presents a special challenge to a field with a research enterprise that
is conceptually decentralized. After all, evolutionary biology is a field
that resists the designation of model systems, at least partially in the
belief that every species is scientifically unique. This attitude clearly
has not ruled out less formal ensemble tests, but it implies that a more
formal test may be harder to organize. We also have an academic
culture in which the recognition of individual achievement is
regarded as essential. Again, this may make an ensemble test of adap-
tationism all the harder.
Two developments in evolutionary biology during the past 20
years or so likely have implications for the construction of the
ensemble test of adaptationism. The first relates to how evolutionary
models are perceived and used by the community of evolutionary

11
 steven orzack and elliott sober

biologists. The construction of population-genetic theory was the

major activity of mathematical evolutionary theorists until about two
decades ago. Although one of most influential practitioners in this
area was R. A. Fisher, the prominent advocate of the hegemony of
natural selection, this body of theory has generally been viewed by
biologists as providing a bulwark against the idea that traits can be
optimal. After all, much of the work in this area has been devoted to
understanding the genetic “details” that could get in the way of the
evolution of an optimal trait. What has happened since then? The
number of game-theoretic and more-traditional optimality models
has increased tremendously in the past 20 years; in effect, optimal-
ity “got mathematical religion.” The biologist was formerly faced on
the one hand with a body of mathematical theory that could be taken
to imply that optimality is impossible or unlikely, and on the other
hand with a body of verbal claims about the hegemony of natural
selection. Now there are two opposing claims about optimality that
have sophisticated bodies of mathematical theory backing them up.
In the past, perhaps in keeping with a general attitude in scientific
culture about the superiority of quantitative arguments, it was no
surprise that the message from population-genetic theory about
optimality held sway. This attitude about mathematics no longer
has purchase in this context.
It is clear that the acceptance of the traditional population-genetic
message about optimality has waned considerably in recent years,
and there are now more biologists than ever who practice a biology
whose organizing principle is the hegemony of natural selection.
What explains this change? One possibility is that a predisposition
in many biologists’ minds toward the idea that natural selection is
all-powerful is no longer countered by an absence of arguably sup-
portive mathematics. This predisposition may be reinforced by a
tendency of many empiricists to readily accept the message of math-
ematical theory only if it reinforces their antecedent view of the
world. What is unclear to us is what role data have played in this
transformation. An important role for data would make the con-
struction of a meaningful test of adaptationism all the easier. A lesser
role for data would make the construction of a meaningful test more
difficult but not impossible, because it would reflect a greater role for
nonepistemic influences.
Perhaps the present professional environment plays a role in
the resurgence of natural selection. Science is more competitive

12
 Introduction

than ever; it has more of a “winner-take-all” character to it than ever

before. It is the rare scientist who readily admits that his or her
career owes much if anything to “chance”; instead, one’s success is
commonly attributed to the intrinsic merit of one’s work. Science is
a meritocracy, after all. Although this is an impoverished view of
science and of what makes a scientific career, it is commonly pro-
mulgated, as is the view that natural selection is all-powerful. This
parallel between scientific culture and scientific content deserves
further study.
We note in passing that the influence of population-genetic theory
increased at a time when the professionalization of evolutionary
biology was occurring. This process may have reduced the reluctance
of many biologists to grant mathematical arguments influence.
For some, such an openness may have stemmed from a low-grade
case of physics envy. But we suspect that many others granted
influence to mathematics in the belief that it generates a more “sci-
entific” impression and therefore more respect from scientists in
other fields and to those who grant tenure and give out grants and
laboratory space.
The molecular revolution in biology also has important implica-
tions for a test of adaptationism. Our definition of adaptationism
relates explicitly to nonmolecular traits because the focus in evolu-
tionary biology has traditionally been on behavioral and morpho-
logical traits. Such traits are still the focus of much work, but the
advent of powerful methods for the study of proteins and DNA has
had dramatic consequences for evolutionary biology as a science and
as a profession. In the former case, one obvious consequence is the
growth of the study of molecular evolution, especially the study of
DNA sequence evolution. The study of DNA sequences from coding
regions is of special interest in the context of understanding the role
of natural selection and other evolutionary forces in trait evolution
because the “wobble” of the triplet code makes it easier to distinguish
between more and less selectively constrained parts of the sequence.
To this extent, there is the potential for more-precise and more-
powerful analyses of traits. In effect, the functional anatomy of each
triplet is already defined, although base pairs even hundreds or thou-
sands of positions away from one another may not be functionally
independent.
What does the growth of molecular evolution mean for a test of
adaptationism? The inclusion of molecular traits may affect the

13
 steven orzack and elliott sober

outcome of this test if evolutionary forces typically affect them

differently than they affect other traits. Of course, this remains to
be seen. These traits do differ from other traits because there is no
disjunction between the trait and the gene. Accordingly, the tempo-
ral integrity of the trait is not an issue, as it is for many traits, because
a nucleotide remains the same from one generation to the next,
barring mutation or gene conversion. Of course, as noted earlier, the
functional consequences of nucleotide substitutions can be difficult
to determine. Nonetheless, the simplicity of trait definition and inher-
itance make studies at the nucleotide level increasingly important.
There is certainly no reason to think that optimality models are
somehow inappropriate in this context. For example, population
geneticists routinely talk of purifying selection at the nucleotide level
when trying to explain levels of nucleotide diversity; this is talk of
optimality. An additional positive aspect of work at the molecular
level is that there are well-developed models of trait evolution that
do not invoke natural selection at all; such neutral models provide
optimality models with well-motivated competitors.
Molecular evolution is one of the most intellectually balkanized
areas of evolutionary biology. For some investigators, neutrality is a
null hypothesis; for others, natural selection has this status. Perhaps
not surprisingly, then, this area has also been one of the most labile
in regard to the prevailing attitude toward the hegemony of natural
selection. The rise of the neutral theory in the late 1960s resulted in a
dramatic change in the evolutionary interpretation of molecular vari-
ation in natural populations. In the space of 10 years or so, natural
selection went from being viewed as the obvious explanation for such
variation to being viewed as something to be invoked rarely and only
with extensive empirical and analytical support. Starting in the
mid-1980s, the general attitude toward evolutionary explanation has
shifted back, with selection now being regarded as generally more
important. There is still tremendous diversity of opinion. This diver-
sity is not coincident with the division of opinion about optimality
we discussed earlier. In fact, it involves mainly population geneticists,
the researchers who overall have been least receptive to claims of
optimality.
Are these changes in how natural selection is thought to affect
molecular evolution epistemically driven for the most part? We
think not. However, an ensemble test of adaptationism does not
depend on scientists putting facts above all else. In this sense,

14
 Introduction

any nonepistemic influences on studies of molecular evolution

(or of any other aspect of evolution) do not make such studies mean-
ingless in terms of constructing an ensemble test. In fact, the sim-
plicity of trait definition and the potential for a more precise
assessment of the role of natural selection in trait evolution make
the inclusion of molecular traits in the ensemble test of adaptation-
ism an important goal and a positive consequence of the molecular
revolution. Our expectation and hope is that studies of molecular
evolution will be the focus of a separate chapter in a future edition
of this volume.
The molecular revolution has had dramatic effects not only on
the science of biology but also on the profession of biology. One
consequence is the nearly universal disappearance of “Biology”
departments, especially at research universities. Now one usually
finds several departments, including one whose name includes
some permutation of “Ecology” and “Evolution” and at least one other
whose name includes some permutation of “Biochemistry,” “Cellu-
lar,” and “Molecular.” This organizational divergence reflects and
reinforces the molecularization of biology. More and more, biology is
simply research on molecules that rightly or wrongly involves little or
no investigation of the larger biological context. For that matter, a sub-
stantial fraction of molecular biologists does not view evolutionary
biology as a worthwhile science. This attitude is reinforced by the
departmental division, which lessens or eliminates direct contact with
evolutionary biologists, and by undergraduate training in biology that
now often makes molecular biology and biochemistry courses manda-
tory and courses in organismal biology optional.
This divergence, along with the substantially greater funding
available for molecular research compared with that available for
research on whole organisms, does not bode well for the continua-
tion of organismal studies, which must be an important part of an
ensemble test of adaptationism. Hence, although molecular studies
have nearly unique power and resolution, they may come to domi-
nate evolutionary biology and reduce its breadth to such an extent
that it cannot underwrite the test of adaptationism. These are mixed
consequences, indeed. It is ironic that the study of universal aspects
of biology could potentially be so narrowing.
Our claims about the intellectual, social, and organizational
aspects of evolutionary biology as a science and as a profession are
meant to provide a context for greater understanding of what is most

15
 steven orzack and elliott sober

important to us and what we expect and hope is most important

to our readers: the biology. In line with this expectation, we have
asked our diverse group of contributors to write about biology while
keeping in mind the practical-minded biologist. The focus of each
chapter concerns one or another biological issue that we mentioned
earlier.
The first three chapters of this volume concern phylogenetic analy-
sis. This focus reflects our belief that the dramatic proliferation of
studies of whether and how phylogeny affects evolutionary expla-
nation is a development with potentially major implications for a test
of adaptationism. The need for examination of these implications is
heightened by the fact that much of this work appeared subsequent
to many of the influential claims about adaptationism advanced
during the 1970s and 1980s. Indeed, some of this work was likely
motivated by some of these earlier claims. At the same time, work in
this area has proliferated to such an extent that the issues involved
go far beyond those raised earlier.
In Chapter 1, David Baum and Mike Donoghue discuss how
distinct approaches to the study of adaptation can be unified and
how such a unification can contribute to a test of adaptationism. In
the context of the phylogenetic study of adaptation, they contrast
the homology approach – the attempt to determine whether a
uniquely derived trait evolved due to natural selection for a specific
function – with the convergence approach – the attempt to determine
whether multiple independently evolved traits show a phylogeneti-
cally correct correlation with a particular selective regime. Homolo-
gists claim that correlations say nothing about causation and,
accordingly, that they say nothing about adaptation. In contrast, pro-
ponents of the convergence approach claim that only ensembles of
events provide evidence for adaptation. Studying adaptation by
looking at unique events is seen as no more than “storytelling.” Baum
and Donoghue show how the concept of likelihood provides a con-
nection between these approaches. For example, the likelihood
approach helps to defuse the common criticism of the convergence
approach that there is no objective criterion for picking the set of
taxa to be studied. However, if correlative studies are initiated for
the purposes of elucidating the origin of a specific homologous
trait, then the nature of the adaptive hypothesis for that trait serves
to define the sorts of analogous traits and selective regimes that
should be considered.

16
 Introduction

Assessment of the roles played by natural selection and other

evolutionary forces in trait evolution is the focus of our
Chapter 2. Natural selection and phylogenetic inertia have often
been portrayed as mutually exclusive explanations of traits or of
trait associations. This portrayal is often accompanied by the idea
that an inertial hypothesis takes precedence over an adaptive
hypothesis. These are ideas we reject. Claims as to the influence of
inertia and selection are causal claims, and neither of them is a null
hypothesis, any more than the hypothesis that smoking causes
cancer takes precedence over the hypothesis that asbestos causes
cancer. We go on to discuss how the test of an adaptive hypo-
thesis requires that an investigator control for the possibility of
phylogenetic inertia, and vice versa. When testing an inertial hypoth-
esis, one must use a specific adaptive hypothesis because neither
hypothesis has precedence. Our ideas concerning how comparative
data can be used to test adaptive hypotheses connect with larger
questions about the meaning and testability of adaptationism. As
described earlier, we claim that adaptationism can be tested by an
ensemble of test results and that it is not something one should
assume true or false.
The previous chapters have the assumption that phylogenetic
analysis has an important role in the analysis of adaptation and, by
extension, in the debate over adaptationism. In Chapter 3, Kern Reeve
and Paul Sherman question the validity of this assumption and argue
that historical methods are often inappropriately used to assess the
roles that forces such as natural selection and genetic drift have
played in the evolution of a trait. They argue that claims that phylo-
genetic inertia accounts for a particular trait’s presence in an
organism are not meaningful. They also discuss three reasons why
researchers should shy away from a sole emphasis on Hennig’s par-
simony method when reconstructing evolutionary history: Trait evo-
lution may be positively correlated among lineages, different states
of the trait may have different extinction rates, and a trait value may
be “unstable,” that is, the probability that a population will make a
transition to a new trait value between lineage-splitting events is
greater than the probability that the population will retain the same
trait value.
The next two chapters concern our theoretical understanding of
optimality and evolutionarily stable strategy models. This focus
reflects our belief that the dramatic proliferation of studies of such

17
 steven orzack and elliott sober

models is also a development with potentially major implications for

a test of adaptationism. As in the case of work on phylogeny, much of
this work appeared subsequent to many of the influential claims about
adaptationism advanced during the 1970s and 1980s. As noted earlier,
at that time mathematical theory in evolutionary biology was gener-
ally interpreted as having only one implication for the debate over
adaptationism. Work in this area has proliferated to such an extent that
this is no longer true. Both chapters exemplify this change in the past
20 years in the nature of mathematical evolutionary theory.
In Chapter 4, Joel Brown defends the claim that a game-theoretic
approach can go very far toward providing a complete picture of evo-
lution. As such, his analysis exemplifies an adaptationist view of evo-
lution. Using this approach, he attempts to explain three important
aspects of evolution: the Fit of Form and Function (the fact that
animals and plants seem well adapted), the Diversity of Life (the fact
that many species are polymorphic and that millions of species vary
spectacularly in size and shape), and the Procession of Life (the
waxing and waning of clades over millions of years).
Chapter 5 concerns our theoretical understanding of evolutionary
dynamics and how they relate to the debate over adaptationism. Ilan
Eshel and Marc Feldman contrast short-term evolution (the dynam-
ics of the relative frequencies of a finite, fixed set of genotypes) with
long-term evolution (the “trial and error” process by which muta-
tion introduces new genotypes into the population). Analysis of the
dynamics of genetic systems has traditionally concentrated on the
short-term process because it can be directly observed. Eshel and
Feldman contend that the two processes obey radically different
quantitative rules and that it is incorrect to derive the qualitative
laws governing long-term evolution from analyses of the short-term
process. The authors focus on an important difference between the
two processes – the tendency toward an optimum – and introduce a
general adaptivity property that can be thought of as the long-term
counterpart of Fisher’s theorem. Unlike the latter, it holds for any
genetic system subject to random mating and frequency-independent
viability selection. This property guarantees that during the process
of long-term evolution, any such system converges to an optimum,
if it converges.
The next three chapters concern the rationale and meaning of tests
of optimality models. This focus reflects our belief that the dramatic
proliferation of tests of such models is also a development with

18
 Introduction

potentially major implications for a test of adaptationism. As before,

much of this work appeared subsequent to many of the influential
claims about adaptationism advanced during the 1970s and 1980s. All
these chapters discuss the important issues arising from the interplay
of model predictions and data.
In Chapter 6, Allen Herre, Carlos Machado, and Stuart West
provide a case study of how the analysis of specific optimality
models and data sets relates to the debate over adaptationism. The
authors discuss models predicting optimal sex ratios when a finite
number of individuals contribute offspring to a given local popula-
tion and data on sex ratios produced by fig-pollinating wasps. Their
results shed light on the question as to whether optimality theory is
more appropriately used as a measure of the precision of adaptation
or whether the organisms’ responses are more appropriately used as
a test of the validity of the theory. Herre, Machado, and West go on
to discuss the meaning of qualitative and quantitative tests of models
and explain how they can contribute to the refinement of data and of
theory.
In Chapter 7, George Gilchrist and Joel Kingsolver discuss a differ-
ent aspect of the interplay between data and theory. They first describe
how the adaptive landscape has been a powerful metaphor in the
study of selection and adaptation, especially as it has provided a ratio-
nale for using optimality models. Yet little attention has been paid to
the topography of real or theoretical fitness landscapes. The authors
describe several recent studies of the curvature of an adaptive land-
scape and its relationship to the degree of phenotypic or genetic vari-
ation in a population. Their examples include models of the classic life
history interaction between body size and age at first reproduction and
Gilchrist’s recent models of thermally sensitive mating and oviposi-
tion rates. The results illustrate the complementarity of fitness land-
scapes and genetical models. They go on to discuss the limits that the
topography of adaptive landscapes may place on our ability to derive
quantitative predictions from optimality models and to suggest that
the debate over adaptationism will be best served by a broader look at
the topography of fitness landscapes.
In Chapter 8, Ken Halama and Dave Reznick discuss the relevance
of variation among individuals to the debate over adaptationism
and optimality. Evolutionary ecologists have traditionally viewed
phenotypic variation among individuals in a population as the
necessary material for evolutionary change, but ironically they have

19
 steven orzack and elliott sober

paid little attention to evolution within populations. However, recent

investigators have demonstrated that multiple adaptive phenotypes
can exist within populations. A population may be polymorphic for
one or more traits, with each morph achieving maximum fitness in
a specific part of the species’ realized niche. Halama and Reznick
describe the theoretical framework surrounding the adaptive nature of
differences among individuals and then review laboratory and field
studies that demonstrate functional differences among individuals
from the same population. They then discuss the roles of phenotypic
plasticity and of genetic correlations in creating variation among
individuals.
The nature of the debate over adaptationism is addressed in
several ways in the final four chapters. In Chapter 9, Peter Abrams
argues that the role of adaptationism generally and the role of opti-
mality models in particular have been overstated and misunderstood
by proponents as well as opponents of these approaches and that
there has been an unfortunate tendency to employ weak evidence to
support optimality models. He proceeds to examine whether opti-
mality models make different claims than do other adaptive expla-
nations. Many proponents and opponents of such models regard
agreement with quantitative predictions as being necessary to judge
the validity of an adaptive scenario based on an optimality model.
But Abrams argues that quantitative predictions are seldom possible
and have not been employed in many widely accepted adaptive
explanations. He also discusses the meaning of “genetic constraints”
and explores what it means when constraints cause optimality
models to make different predictions than typical population-genetic
models. Abrams goes on to assess the social dynamic of the debate
over adaptationism, which he describes as a positive feedback
process in which proponents and opponents have lost sight of the
reason for using optimality models.
Critics of adaptationism often invoke development as the “piece”
of biology whose absence from evolutionary explanations renders
them incomplete. In Chapter 10, Ron Amundson focuses on two
approaches to the problem of integrating the study of adaptation and
the study of developmental biology. The functional approach analyzes
developmental constraint as a pattern of trait covariance. The other,
structural, approach posits development as an explanatory mode
independent of adaptation. The explanatory mode approach is appeal-
ing because the notion of explanatory presupposition has received

20
 Introduction

considerable attention by philosophers. Theoretical clashes in the

history of science are marked by divergences of explanatory presup-
position, and the function versus structure conflict has occurred
repeatedly in the history of biology. Several features of the present
debate over adaptationism make sense in light of this model. One
is the isolation of embryology from modern evolutionary theory;
another is the claim by adaptationists that structuralist explanations
don’t explain anything. Amundson discusses the divergent explana-
tory presuppositions that seem to underlie the two explanatory
modes, and he attempts to determine whether developmental
biology can be assimilated into studies of adaptation via a concept
such as genetic constraint defined in quantitative-genetic terms,
whether it will remain peripheral to those studies, or whether some
other outcome is possible.
In Chapter 11, Peter Godfrey-Smith argues that there are three forms
of adaptationism found within the community of evolutionary biolo-
gists. One form is the view that natural selection is the key to under-
standing the “fit” between organisms and their environments. Natural
selection is the most important evolutionary force because it is the
force that is most important in answering this special question. This
does not necessarily imply a belief in the claim that selection is
omnipresent and omnipotent. A second form of adaptationism is the
view that natural selection is the only evolutionary force usually
needed to explain data from natural populations. The third form con-
cerns the best procedure for the evolutionist to follow when it comes
to evolutionary explanation: One should start with a selective expla-
nation as a null hypothesis. This is not a view about what evolutionary
theory is for or a view about the state of nature; instead, it is a view
about how to organize one’s scientific work. Godfrey-Smith argues
that evidence for or against one form of adaptationism often is not
evidence for or against the other forms and that this is a source of
confusion and misunderstanding that has significantly hindered our
understanding of the role of natural selection in trait evolution.
In Chapter 12, Egbert Leigh discusses several general issues relat-
ing to the validity of adaptationism and to the development and
testing of optimality models. He argues that adaptation is self-
evident, largely on the analogy between organisms and human arti-
facts and from the fact that building a self-replicating entity is still
beyond our powers: rigid standards must be met to accomplish
such a task. He then reviews three claims about adaptation: D’Arcy

21
 steven orzack and elliott sober

Thompson’s claim that adaptation is reflected in the clearly demon-

strable form of mechanical fitness for the exercise of a function,
Fisher’s claim that an organism is adapted to a particular environ-
ment only insofar as we can imagine different environments to
which it would be less well adapted, and Levins’s claim in terms of
fitness sets. Leigh then describes how one determines the appro-
priate set of possible phenotypes for an optimality model, how one
establishes the adaptive function, and how one determines the
appropriate level of selection. The establishment of adaptive func-
tion is especially problematic because it must be defined relative to a
particular niche. In this context, Leigh discusses two central issues:
whether each species requires a different niche and whether we
can justify coherence in adaptive function within- and between-
species in a world where evolution and speciation occur. Finally, he
discusses why it is important to test any given adaptive explanation,
inasmuch as forgoing a test and invoking the omnipotence of natural
selection do not add to understanding. A test contributes to our
understanding of the definition of traits and of morphological inte-
gration, and it adds to our understanding of the levels of selection
involved in adaptation.
Of course, there is more to be said about adaptationism and opti-
mality than all of what is said here. Nonetheless, we and the other
contributors to this volume have the ambitious hope that it leads
toward a greater understanding of evolutionary biology, of science,
and of nature. If so, we will have been amply rewarded.

LITERATURE CITED

Brockmann, H. J., and R. Dawkins. 1979. Joint nesting in a digger wasp

as an evolutionarily stable preadaptation to social life. Behaviour 71:
203–245.
Brockmann, H. J., A. Grafen, and R. Dawkins. 1979. Evolutionarily
stable nesting strategy in a digger wasp. Journal of Theoretical Biology 77:
473–496.
Milinski, M. 1979. An evolutionarily stable feeding strategy in sticklebacks.
Zeitschrift für Tierpsychologie 51: 36–40.
Orzack, S. H. 1990. The comparative biology of second sex ratio evolution
within a natural population of a parasitic wasp, Nasonia vitripennis. Genetics
124: 385–396.
Orzack, S. H., and E. D. Parker, Jr. 1990. Genetic variation for sex ratio traits
within a natural population of a parasitic wasp, Nasonia vitripennis. Genetics
124: 373–384.

22
 Introduction

Orzack, S. H., E. D. Parker, Jr., and J. Gladstone. 1991. The comparative biology
of genetic variation for conditional sex ratio adjustment in a parasitic wasp,
Nasonia vitripennis. Genetics 127: 583–599.
Orzack, S. H., and E. Sober. 1994a. Optimality models and the test of adapta-
tionism. American Naturalist 143: 361–380.
Orzack, S. H., and E. Sober. 1994b. How (not) to test an optimality model. Trends
in Ecology and Evolution 9: 265–267.
Orzack, S. H., and E. Sober. 1996. How to formulate and test adaptationism.
American Naturalist 148: 202–210.

23
 Chapter 1

A Likelihood Framework for the Phylogenetic

Analysis of Adaptation
david a. baum and michael j. donoghue

The common core of all studies of adaptation is the assessment of the

role of natural selection in character origin and/or maintenance. A
range of information seems pertinent to such studies, including phy-
logeny, ecology, developmental biology, physiology, biomechanics,
ethology, and genetics. At present, however, we lack a conceptual
framework for integrating these diverse types of data. Our primary
mission in this chapter is to develop one such framework based on
likelihood ratios. As we hope to show, this likelihood approach to
the study of adaptation not only clarifies the hidden assumptions
of adaptationist studies but also provides a common language for
communication among disciplines. In particular, we believe that
this approach will clarify the interrelationship between studies of
uniquely evolved characters and the phylogenetic distribution of
analogous variation.
This chapter is not about the definition of adaptation. We happen to
prefer a definition of adaptation as characters that evolved via natural
selection for some specified biological role (Gould and Vrba 1982;
Sober 1984). This historical view challenges us to decipher the causes
of the fixation of a particular state in a particular ancestral lineage. As
difficult as this might be, we think it is valuable to have a definition
of adaptation that encourages such investigations rather than one
that leads us to restrict our attention to character function in extant
organisms. However, even biologists who prefer an ahistorical defini-
tion (e.g., Fisher 1985; Reeve and Sherman 1993) are generally still
interested in knowing how and why particular traits evolved. There-
fore, regardless of one’s favorite definition of adaptation, improved
methods for inferring the historical action of natural selection may be
welcome. Such methods are the subject of this chapter.

24
 Phylogenetic Analysis of Adaptation

PHYLOGENETIC APPROACHES TO THE STUDY

OF ADAPTATION

Among those who maintain that phylogeny bears on the study of

adaptation, two distinct schools have emerged. Coddington (1994)
dubbed them the “homology” and “convergence” approaches. The
homology approach is characterized by a focus on a particular
character in a particular lineage and the attempt to test hypotheses
of adaptation using phylogenetic and functional information. This
approach is primarily practiced by systematists and tends to use the
parsimony framework described more fully in the next section. The
convergence approach, in contrast, is primarily involved in looking
at broad phylogenetic patterns of correlation among characters
(discrete or continuous) or between characters and environmental
variables. The central objective of the convergence approach is to
establish whether such correlations exist when one takes into account
the inferred phylogeny (e.g., Felsenstein 1985; Harvey and Pagel
1991). The convergence approach has been popular among ecologists
and behavioral biologists.
Whereas the homology approach is directly concerned with evalu-
ating the historical action of natural selection on a particular character
and lineage, this is not so obviously the case for the convergence
approach. After all, it can be argued that detecting phylogenetic
correlations is an end in itself, analogous to discovering biological
laws (Pagel 1994a). And however strong a correlation may be, it can-
not determine whether any individual evolutionary change was or
was not caused by natural selection (Wenzel and Carpenter 1994).
Nonetheless, there are many researchers involved in the study of
broad phylogenetic patterns who seem to be specifically interested in
assessing the role that selection played in the origin of characters (e.g.,
Ridley 1983; Sillén-Tullberg 1988, 1993; Donoghue 1989; Maddison
1990; Maddison and Maddison 1992). Indeed, we suspect that part
of the attractiveness of the convergence approach for evolutionary
biologists at large is an intuition that observing a general pattern of
correlation in n - 1 cases sheds some light on the adaptive status of the
nth case. However, although we may intuit the existence of such a link,
a formal treatment has not been achieved and methods of analysis
have not been developed. One objective of our chapter is, therefore, to
lay the foundation for such a rigorous treatment of information flow
between classes of analogous characters and individual cases.

25
 david a. baum and michael j. donoghue

THE PARSIMONY FRAMEWORK

The use of phylogenies in evaluating hypotheses of adaptation was

formalized by Greene (1986), Coddington (1988), and Baum and
Larson (1991) based on the conceptual framework of Gould and Vrba
(1982). The approach that emerged sought predictions of adaptive
hypotheses that could be evaluated with phylogenetic information
combined with appropriate functional studies. Two predictions were
highlighted: (1) that the character enhances performance of the spec-
ified biological role relative to the antecedent state and (2) that the
character evolved in a lineage whose selective regime was such that
enhanced performance of that biological role was favored. The former
test of current utility primarily involves functional studies of the char-
acter compared with its inferred antecedent state. Here, phylogenetic
information helps identify the character’s precursor condition and
the lineage in which the transition occurred. The second test evalu-
ates historical genesis. It initially involves identifying aspects of the
selective regime that determine whether improved performance of
the biological role is selectively advantageous. These aspects of the
selective regime are then mapped onto the phylogeny using parsi-
mony (in combination, perhaps, with biogeographic and paleoeco-
logical information) so as to evaluate whether selection would have
favored the derived over the antecedent state on the lineage in which
the character is thought to have evolved (Figure 1.1). If both tests are
passed, the character is inferred to have evolved via natural selection
for the biological role; that is, the adaptive hypothesis is corroborated.
If either test is failed, the adaptive hypothesis is refuted.
This phylogenetic approach has been criticized because phyloge-
nies can be difficult to reconstruct accurately (Reeve and Sherman
1993; Frumhoff and Reeve 1994). Even if we have a correct phylogeny,
the use of parsimony to reconstruct ancestral states is fallible, espe-
cially if we make incorrect assumptions about the rate of character
evolution (Frumhoff and Reeve 1994; for related discussion, see
Maddison 1995; Schultz et al. 1996; Schluter et al. 1997; Ree and
Donoghue 1998). Furthermore, pleiotropy and epistasis may make it
difficult to interpret phylogenetic patterns without careful functional
and genetic studies (Lauder et al. 1993; Leroi et al. 1994). However,
as formulated, these criticisms of the parsimony approach are weak
because any scientific inference will fail if the underlying assump-
tions are not met. In this case the application of the parsimony

26
Figure 1.1. A hypothetical example of the parsimony approach to the study of adaptation
(see Baum and Larson 1991). A phylogeny is shown for a plant group, some of which have
the ancestral condition, white petals, whereas others have the derived condition, red petals
(shown in gray). Parsimony is used to locate the branch on which petal color changed. To
test the hypothesis that red petals evolved due to natural selection for the biological role
of attracting pollinating birds, we first must show that red petals perform better than white
petals in bird-pollinated plants. If this is shown, and if parsimony reconstruction of the pol-
lination mechanism (the selective regime as defined by Baum and Larson 1991) implies
that red petals evolved on a bird-pollinated lineage (top panel), then the adaptive hypoth-
esis is supported. In contrast, if red petals evolved in a bee-pollinated lineage (middle
panel), then the character cannot have evolved via natural selection for bird attraction and,
therefore, it is not an adaptation (it is an exaptation sensu Gould and Vrba 1982). If bird
pollination and red petals evolve on the same branch (bottom panel), the result is equivo-
cal; the adaptive hypothesis is neither supported nor rejected.
 david a. baum and michael j. donoghue

method assumes that the phylogeny is correct, that character and

selective regimes are correctly reconstructed on the phylogeny, and
that the evaluation of character performance was accurate. Therefore,
the possibility of obtaining an incorrect result using the parsimony
approach is not grounds for rejecting a role for phylogenetic analysis
in the study of adaptation. Although mistakes can be made, in the
long run one expects errors to be corrected through further phyloge-
netic and functional studies (Larson and Losos 1996).
It seems clear to us that the protocol described by Baum and
Larson (1991) will be useful in many situations. However, we think
that the parsimony framework has limitations. Specifically, it can
give only a yes-or-no answer, and there is no way to quantify the
degree of support for an adaptive hypothesis (Pagel 1994a; Garland
and Adolph 1994; Doughty 1996). As a consequence of this lack of
a statistical outlook, we see four issues that are not adequately
addressed within the parsimony framework.

1. If both the character and the selective regime are inferred to have
changed on the same branch of the phylogeny, then the method
returns an equivocal answer because it cannot be stated with con-
fidence that the character evolved after the selective regime as
demanded by the adaptive hypothesis (Baum and Larson 1991;
Figure 1.1). This makes the method inapplicable to characters that
are under such strong selection in the derived selective regime that
they evolve too quickly for cladogenesis to capture the historical
ordering of events. Such strongly selected adaptations are of great
interest, so it is reasonable to wish that we had the methodologi-
cal tools to study them.
2. The approach requires that we commit to a particular tree and to
a particular reconstruction of the characters and selective regimes
on that tree. We know that any inference of a phylogeny or of
ancestral states, given a phylogeny, is subject to error. However,
the parsimony approach does not take into account such uncer-
tainty, and it does not distinguish between cases in which we are
very confident of our historical inferences versus those in which
we are unsure.
3. The parsimony approach focuses exclusively on character origin.
As discussed earlier, a character that we observe in a living taxon
has not only evolved but has also been maintained in at least one
lineage through to the present day. However, although under-

28
 Phylogenetic Analysis of Adaptation

standing the role of selection and other evolutionary forces in this

maintenance would seem to be part of a complete adaptationist
program, the parsimony methods developed to date are not
equipped to shed light on this issue.
4. The approach provides no avenue by which information from else-
where on the tree can influence our interpretation of a particular
case. For example, despite the intuitive importance of observing
the repeated coincidence of red petals in bird-pollinated plants,
this information does not factor into the test of the adaptive status
of red petals in Lobelia cardinalis. Thus, the current approach
implies that the study of the evolution of a particular character is
conceptually disconnected from the study of repeated events
(Coddington 1994). We suppose, in contrast, that there is some
conduit whereby information from repeated events would influ-
ence our strength of belief in a specific adaptive hypothesis and,
likewise, that adaptationist studies on particular unique characters
would influence explanations of broad phylogenetic patterns.

A LIKELIHOOD FRAMEWORK

Likelihood ratios have recently been used to study directionality

in character evolution (Sanderson 1993), to evaluate ancestral states
(Schluter et al. 1997), to detect shifts in diversification rate (Sander-
son and Donoghue 1994), to detect the correlated evolution of two
discrete characters (Milligan 1994; Pagel 1994b, 1998), to compare
alternative models of molecular evolution (Felsenstein 1981; Yang
et al. 1995; Huelsenbeck and Rannala 1997), and to see whether there
has been cospeciation of hosts and parasites (Huelsenbeck et al. 1997).
The principle underlying these methods is that the better supported
of two hypotheses is that which, if true, would have been more likely
to generate the observed data (Edwards 1992). The measure of how
much better one hypothesis is than another, the support, is given as
the natural logarithm of the likelihood ratio (Edwards 1992).
Unlike traditional frequentist statistics (Garland and Adolph 1994;
Doughty 1996), likelihoods can be applied to singular observations
(Sanderson 1995). As a result, even unique historical events can
serve to statistically discriminate between two rival hypotheses. This
aspect of likelihoods is possible because the rival hypotheses are
evaluated within the context of an explicit probabilistic model. The
need to provide such detailed models might be seen as limiting the

29
 david a. baum and michael j. donoghue

applicability of likelihood approaches. However, it is commonplace

to explore a range of models to determine the area of parameter space
within which one hypothesis is favored over the other. This amounts
to specifying the background assumptions that need to be met to
prefer that hypothesis and is, we think, a great improvement over
approaches such as parsimony whose background assumptions are
hard to identify (Lewis 1998).
To see how a likelihood framework can be applied to the testing
of an adaptive hypothesis, consider the logic of the parsimony
approach of Baum and Larson (1991). In this framework, an adaptive
hypothesis is supported if a character enhances performance of the
specified biological role relative to its antecedent state, and if this
enhanced performance was selectively favored at the time of charac-
ter origination. If both of these things are true, then the parsimony
approach assumes that selection for the specified biological role did,
indeed, cause character fixation. However, although such an obser-
vation strengthens an adaptive hypothesis it is by no means defini-
tive. Even if a derived character has a selective advantage relative to
its antecedent state, genetic drift or selection for a different biologi-
cal role might have been partly or wholly responsible for its fixation.
Thus, to evaluate an adaptive hypothesis we must take into account
the possibility that the character would have evolved without selec-
tion for the hypothesized role. That is, instead of just looking at
evidence for adaptation, we must also consider evidence in favor of
alternative hypotheses. Support for the adaptive hypothesis is, there-
fore, best captured by the ratio of the likelihood that the character
would have evolved via natural selection for the specified biological
role versus the likelihood that it would have evolved without such
selection.
Consider a situation in which we are studying a particular char-
acter (e.g., red petals in Lobelia cardinalis), we have a phylogeny in
hand, and we believe we know the branch along which red petals
evolved. In the likelihood framework we would first seek to evalu-
ate the likelihood of the adaptive hypothesis (Ladapt): that red petals
would have evolved in the lineage given that there was selection for
enhanced attraction of birds. Second, we would seek to evaluate the
likelihood of an alternate hypothesis (Lalt): for example, that red petals
would have evolved in the lineage via genetic drift, without selection
for bird attraction. The measure of support for the adaptive hypoth-
esis is the likelihood ratio Ladapt/Lalt.

30
 Phylogenetic Analysis of Adaptation

A likelihood ratio greater than 1.0 would indicate that the data
favored the adaptive hypothesis, whereas a likelihood ratio less than
1.0 would mean that the data argue for the alternative hypothesis.
The magnitude of the ratio is a measure of the objective support
for the favored hypothesis (Edwards 1992). Various approaches are
available to interpret these ratios. Edwards (1992) suggested that
a likelihood ratio convincingly favors a hypothesis when the natural
logarithm of that ratio exceeds a score of 2.0 (see also Schluter et al.
1997). Alternatively, one can compare the ratio to a null distribution
to evaluate its “significance.” In some situations one expects log like-
lihood ratios to follow a c2 distribution, with the number of degrees
of freedom determined by the difference in the number of free vari-
ables in the two models (Felsenstein 1981). In some cases, such as
when the two hypotheses are not nested, a c2 distribution is not
the appropriate null expectation. In those cases the likelihood ratio
can be evaluated by Monte Carlo simulation (e.g., Sanderson and
Donoghue 1994; Pagel 1994b). It should be noted that there is some
question as to how “P-values” should be interpreted in likeli-
hood statistics (Sanderson 1995).
Interestingly, the parsimony approach described earlier can be
rephrased as a special case of the likelihood method with certain sim-
plifying assumptions. In effect, it is assumed that Lalt is close to zero,
meaning that the strength of support for the adaptive hypothesis is
basically determined entirely by Ladapt. If the character confers a per-
formance advantage for the biological role relative to the antecedent
state (i.e., it passes the test of current utility) and if the character
evolved at a time when enhanced performance of the biological role
was selectively favored (i.e., it passes the test of historical genesis),
the character is seen as having a high likelihood of evolving (i.e., Ladapt
takes on an unspecified high value). Given that Lalt is assumed to be
close to zero, the likelihood ratio in this case would be large and the
hypothesis would be supported by the data. If, on the other hand,
the character either lacks current utility or evolved at a time when
improved performance of the biological role was not favored, then
Ladapt takes on an unspecified low value, similar in magnitude to Lalt.
With Ladapt/Lalt close to 1.0, the null hypothesis cannot be rejected and
the adaptive hypothesis is not justified by the data.
The simplifying assumptions built into the parsimony formulation
may be reasonable in many circumstances, and, thus, we see no
grounds for rejecting that methodology outright. However, the

31
 david a. baum and michael j. donoghue

assumption that Lalt is low will not apply, for example, if there are rival
biological roles that could equally account for the origin of a character.
Therefore, the parsimony method might give unwarranted support for
the adaptive hypothesis (as noted by Leroi et al. 1994). Genetic drift
also must be considered as a potential reason for the evolution of a
character, especially if populations at the time of fixation were small.
Similarly, Ladapt need not always be large – for example, if the perfor-
mance advantage is minimal. Thus, even if the tests of current utility
and historical genesis are passed, an adaptive hypothesis might be
only weakly supported by the data. Clearly, it would be desirable to
have a framework for evaluating Ladapt and Lalt directly.

ESTIMATING THE LIKELIHOODS FROM

MICROEVOLUTIONARY DATA

Let us imagine that we have identified the branch of a phylogeny along

which a character of interest evolved, that is, the derived character is
believed to have arisen through mutation and become fixed in the focal
lineage. Under the likelihood approach we wish to know the likeli-
hood that the character would have evolved on that branch given
either the adaptive or the alternate hypothesis. If we know a lot about
the genetics and performance characteristics of the character and
about the population biology of the focal lineage, we could, in prin-
ciple, use a population-genetic model to estimate Ladapt and Lalt.
One of us (Baum, unpublished) has developed a simple transi-
tion matrix approach for cases of randomly mating hermaphroditic
diploids with constant selection strength, unigenic characters, and
effective population size less than 25. This model assumes that the
mutation rate is unaffected by selection and that one knows lineage
duration (in number of generations), population size, and the selec-
tion coefficient for the character in the relevant selective regime (that
of the focal lineage). The probability of the original mutation occur-
ring in any generation is assumed to be equal (i.e., equiprobable prior
probabilities of mutation). Given a mutation in generation i, the gene
frequency in that generation is taken as 1/2N (assuming mutation
rate is much lower than 1/2N). The probability of going to fixation
by the end of the lineage (Pfix) can be obtained using a transition
probability matrix taken to the power of the number of remaining
generations (Baum, unpublished). One can then obtain the overall
likelihood that the character would have evolved by summing Pfix

32
 Phylogenetic Analysis of Adaptation

across all generations in the focal lineage. Since Pfix is influenced by

the strength of selection invoked, the likelihood will differ under
the adaptive and alternate hypotheses. It should be noted that one
does not need to know the actual mutation rate because, if it is
assumed to be the same under both hypotheses, it cancels out of the
likelihood ratio.
This microevolutionary model was explored for focal lineages of
different length (5 to • generations) with different population sizes
(5 to 20 diploid individuals) and different selection coefficients (0.025
to 0.4). It was found that the likelihood ratio increases (meaning an
increase in the support for the adaptive hypothesis) as the strength
of selection increases, as population size increases, and as the dura-
tion of the lineage decreases (Baum, unpublished). These results are
intuitive. The stronger the estimated selection on the character, the
higher is its likelihood of going to fixation in the lineage. The smaller
the population size, the greater is the effectiveness of genetic drift
relative to selection. The shorter the lineage, the greater the advan-
tage of having selection to drive the character to fixation. The likeli-
hood ratio also increases if one assumes that the character is encoded
by more than one gene (Baum, unpublished).
This microevolutionary model is useful for highlighting the factors
that affect the magnitude of the likelihood ratio and, hence, the
strength of an adaptive hypothesis. However, it would be difficult
to obtain good estimates of the input parameters (population size,
selection coefficient, etc.). Difficult, but not impossible. Phylogenetic
and paleontological data can help one learn about the ecology and
population biology of the organisms on the focal lineage. Additionally,
one might be able to look at genetic variation in the descendants of the
ancestral population to learn something about the effective population
size. To determine the genetic architecture and selective consequences
of the character in the ancestral population, one would probably study
the development, genetics, and performance of the character in extant,
descendant populations as compared with suitable surrogates for the
antecedent state. In addition, it may be possible to obtain information
about character performance by reference to analogous characters
occurring in different taxa, or from taxa that are currently experienc-
ing a selective regime similar to that inferred for the ancestral popula-
tion. Even when we can only narrow the range of possible parameter
values, it might still be possible to obtain enough precision to justify
support or rejection of the adaptive hypothesis.

33
 david a. baum and michael j. donoghue

Even when we cannot obtain estimates of some of the parameters

of the microevolutionary model, we can always make conservative
assumptions. For example, assuming that the character is encoded
by a single gene and that the lineage in question was infinitely long
will reduce the magnitude of support for the adaptive hypothesis.
Thus, if one still obtained a convincingly large log likelihood ratio
one might conclude that the data at hand strongly supported the
adaptive hypothesis.

ESTIMATING LIKELIHOODS FROM

ANALOGOUS CHANGES

A category of simplification that warrants special attention is the

treatment of analogous characters as replicate experiments. Suppose
we are studying the evolution of a particular character, such
as the red petals of Lobelia cardinalis, and testing the hypothesis
that it is an adaptation to a particular biological role, such as
attracting pollinating birds. In the absence of microevolutionary
information we might be able to extract valuable information from
the phylogenetic distribution of analogous characteristics, such as
red petals in other angiosperms. For such information to be relevant,
one would have to assume “transitivity”: that independent
occurrences of red petals in other lineages are similar to the case of
Lobelia in regard to genetic basis and functional consequences, and
that the ecology and population structure of the various lineages
do not differ greatly from those of the focal case. Under these
assumptions each lineage of the phylogeny can be considered an
independent opportunity for red petals to evolve. Suppose one
were also prepared to assume that the tree is made up of two
types of lineages: those in which bird attraction is selectively
favored (i.e., the plants are at least partly bird-pollinated) and
those in which attracting birds is not beneficial (e.g., those in
which birds cannot effectively transfer pollen). If these assump-
tions apply, then the probability of gaining red petals (Pg) in bird-
pollinated (BP) lineages should have a direct relation to the likeli-
hood of red petals evolving given that they are favored by selection
for bird attraction (PgΩBP). This likelihood should be a reason-
able prior expectation of Ladapt for Lobelia cardinalis. Similarly, Pg in
lineages that are not bird-pollinated (NBP) should be a reasonable
predictor of Lalt.

34
Another random document with
no related content on Scribd:
meno sonora e che l’eco ne morisse; poi gli nacque un dubbio che
ritornò a tormentarlo con insistenza assidua:
— Ma.... o io sogno.... o questa è musica che conosco!
E non sognava e ben presto se ne convinse per sua personale
esperienza.
La Giovine Turchia aveva cambiato i suonatori, ma la musica era la
stessa.
Giunse a Derna il nuovo kaimakan con le sue donne.
Fu ricevuto a gran festa; si installò nel palazzo del comando come i
reucci da fiaba.
E cominciò la serie delle mirabili cortesie.
Inchini a destra, sorrisi a sinistra e strette di mano e parole soavi,
mentre in cuor suo cantava l’antica antifona turchesca:
— Accidenti ai giaurri!
Sorrisi e parole oh! tante quanti i datteri su la palma.
E il nostro connazionale cominciò a bene sperare. Le necessarie
formalità per la legittimazione de’ suoi terreni sarebbero state
compiute ed egli avrebbe potuto mettersi all’opera; senonchè....
Ecco; la prima volta ch’egli ne parlò al kaimakan (erano diventati
intimi amici), questi fece le più alte meraviglie. Come mai a
Costantinopoli si aspettava tanto? Come spiegare una così dannosa
incuria? Non era l’Italia una nazione amica, anzi una fra le migliori
amiche della Turchia?
— Non dubiti, scriverò, scriverò; la cosa sarà sbrigata entro un
termine relativamente breve. È bene si cominci la cultura di queste
terre. Lo Stato deve concedere, facilitare, stimolare, difendere.... —
e giù una fiumana di parole entusiastiche.
Il nostro connazionale se ne andò sfregandosi le mani. Era giunta la
volta buona.
Che cara persona quel kaimakan, e che esatta visione delle cose e
quale modernità d’intendimenti!... Con una persona tanto compita
non conveniva insistere, non bisognava mostrarsi seccatori; egli
avrebbe agito per conto proprio ottenendo sicuramente un risultato
favorevole.
E il nostro amico comincia ad aspettare, non trascurando pertanto di
usare le dovute cortesie a quel caro giovine turco.
E passan due mesi, passan cinque mesi, ne passano otto....
L’italiano guardava timidamente il riverito rappresentante della
potenza islamitica; ma questi non capiva, sorrideva, aumentava le
cortesie, parlava dei canali di Marte; poi un giorno in cui al
connazionale nostro rinacque quel tale dubbio di cui abbiamo detto
sopra, un giorno in cui il suo buonumore era relativo, si decise a
parlare:
— Dica un po’, e le mie terre?
— Quali terre?...
— Quelle di cui abbiamo parlato, via!... Quelle che debbono essere
segnate a catasto!
— Ah! sì, mi ricordo, perdoni. Ha ragione, ha ragione; ma ho scritto,
vede, ho scritto e ho fatto sollecitudine!... Non capisco come non mi
abbiano ancora risposto. Ma abbia la cortesia di aspettare.... guardi,
lei è qui, leggerà la mia nuova lettera, poi la spediremo. Va bene?
— Benissimo.
E la nuova lettera è scritta nei termini più soddisfacenti ed è affidata
al primo piroscafo che la porti a Costantinopoli. La risposta non si fa
aspettare, senonchè....
Una sera, dopo l’arrivo del postale da Costantinopoli, il nostro
connazionale ti vede il kaimakan, ma con tale una faccia nera da
sembrare uno spaventapassere; fa per tirar di lungo; ad un tratto si
pente, fa un cenno al nostro amico, il quale, quando gli è vicino, si
sente dire in tono cavernoso:
— Venga; dobbiamo essere soli!
E vanno. Quando sono soli, il kaimakan estrae di tasca una lettera,
ma prima di consegnarla all’amico nostro gli dice:
— Lei mi deve promettere di non compromettermi.
— Comprometterla?
— Sì. Si tratta di un affare gravissimo. La mia posizione ne va di
mezzo.
— Ma dica, dica e si fidi della mia discrezione.
Una pausa. La lettera è consegnata.
— Legga!...
E il nostro connazionale legge strabiliando, poi si stringe fra le spalle
e se ne va raumiliato.
Nella lettera in questione il Governo centrale rimproverava
aspramente il kaimakan per il suo poco patriottismo (parole testuali)
e lo consigliava di non insistere troppo su certi punti, minacciandolo
altresì di punizioni severissime.
Che dire dopo tutto ciò? Con quale coraggio insistere?...
E qui torna a proposito il terzo senonchè....
La commedia era troppo evidente; i machiavellucci da strapazzo non
conoscevano la buon’arte toscana dell’inganno politico, chè non
erano soccorsi nè dall’alto ingegno, nè dalla scaltrezza, nè dalla
furberia dei nostri uomini maggiori; non sapevano essere che astuti
come le volpi e come i contadini, e cioè di una grossolana astuzia la
quale non saprebbe ingannare non già un cane da fiuto, ma più
inesperto bamboccio in fatto di politica.
Di questo si avvide il nostro connazionale, ma a quali ripari poteva
ricorrere?
La commediola fra il kaimakan e il Governo di Costantinopoli era più
che evidente: il kaimakan non aveva fatto che il giuoco del Governo
e, per non avere ulteriori noie dal suo amico personale, era ricorso
allo strattagemma perfidiosetto della lettera minacciosa.
Necessariamente l’amico avrebbe abboccato all’amo, e pace e patta!
Questa storiella si è svolta nell’anno di grazia dell’Egira 1287 e più
precisamente nel nostro 1909.
Potrei documentarla e far nomi. In essa non è una parola aggiunta,
nè un particolare esagerato.
La persona di cui ho parlato attende ancora, se pure, come ne
espresse desiderio, non ha lasciato in asso tutto e non ha
abbandonato la Cirenaica.
Il sentirsi le mani legate, il vedersi ostacolare sordamente ogni
iniziativa è cosa che finisce per stancare un uomo d’azione, il quale
vede nel mondo altri campi aperti alla propria energia.
Se ha resistito con la perseveranza della nostra razza che non si
lascia infiacchire, vorrei augurarmi ch’egli potesse da solo (e l’Italia
ha sempre fatto da sola oltre il suo Governo decorativo) ottenere il
risultato che merita, ma che mi sembra tuttavia molto dubbio; se poi
ha abbandonato il campo, i Giovani Turchi si fregheranno le mani
gioiosamente assegnando la diserzione a un nuovo trionfo del loro
sistema.
Essi vogliono ostacolare in qualunque modo l’opera e l’impiego del
capitale italiano in Cirenaica, e siccome non sono forti e non possono
opporsi con la violenza, si adornano di sorrisi e di salamelecchi, e
con mille scuse e con inchini profondi ci mettono soavemente alla
porta.
Essi non hanno per ora nè capitali, nè energie da impiegare nello
sfruttamento di queste terre; ma che importa? il popolo è bestia: si
nutre di Allah e di una manciata di grano, e muore convinto che così
era scritto. Poi: favorire un russo, un inglese, un patagone, sì; ma un
italiano, no.
In Tunisia i francesi ci trattano come bestie da soma, ci negano le
scuole, ci fanno una colpa di essere italiani, e noi zitti; in Cirenaica
siamo perseguitati e messi alla porta, e noi zitti; in Tripolitania
succede altrettanto, e noi raccomandiamo il silenzio e ci umiliamo.
La linea di navigazione, sussidiata dal Governo, la quale partendo da
Catania tocca la Tripolitania, la Cirenaica, l’isola di Creta, Smirne e
Costantinopoli (linea istituita a solo beneficio dei turchi), ha dato fino
ad ora risultati magnifici. La nostra penetrazione, come abbiamo
veduto a Derna e come accade a Bengasi e a Tripoli, si svolge
indisturbata, e gli ingenui che si rivolgono ai nostri consoli per aver
schiarimenti, li trovano fermi in un gesto ieratico, come il dio indiano
accosciato sul fiore di loto; le mani sul ventre dorato.

Bengasi.

Abbiamo costeggiato l’immenso altipiano deserto, animato solo, a

grandi distanze, da qualche gruppo di palmizi; ora, fra le nebulosità
del cielo e del mare, in una pianura rossiccia si distingue il minareto
di una moschea; è l’unica cosa che soverchi all’intorno.
Bengasi appare fra una desolata natura, e, quando si discende,
l’impressione non varia. Camminiamo fra la sabbia, nella quale il
piede si affonda fino alla caviglia; il caldo è soffocante; tutto è arido,
sitibondo.
Sorgono qua e là alcune case meschinissime, una montagna di sale,
una piazza deserta.
Passano torme lacere di beduini incappucciati e grondanti sudore. Le
vie sono disselciate, sabbiose, cosparse di immondizie e di
pozzanghere di un’acqua nerastra che tramanda un fetore
insopportabile.
Non un aspetto gaio, non un’ombra piacevole vi invita a riposare;
tutto è riarso, stanco, inebetito nella gran calura.
Alle ombre brevi dei muri riposano lunghe fila di arabi taciturni;
qualche lento cammello dalla bocca bavosa e dagli occhi malinconici
passa nel sole ondulando.
Cammino a fatica.
Nelle strade solitarie gli arabi che riposano su le soglie mi guardano
col disprezzo indefinibile che è in fondo agli occhi di tutta questa
gente quando squadra un infedele.
Attraverso il Suk, che non ha altro carattere se non quello di una
grande miseria. Tutto è misero, lacero, esausto. È il paese dello
squallore.
Oltrepasso la cinta della città, e fra la sabbia ardente mi interno in
un villaggio di neri.
Tutta una tribù si è stabilita alle soglie di Bengasi e vi ha elevato le
sue misere capanne.
Le donne, per vezzo, hanno infilato nel naso un corallo; sul nero
della loro faccia la piccola macchia rossa fiammeggia come una
ferita.
In questo caldo meridiano quasi tutti dormono. Vedo donne e
fanciulli sdraiati all’ombra delle loro capanne su la sabbia.
Intorno intorno si levano i ciuffi dei palmizi.
Questo villaggio africano si chiama Zraib.

La storia di un pellegrino.

Qualche giorno fa capitò a Bengasi un bell’uomo alto, tarchiato,

adusto dal sole; vestiva alla foggia algerina.
Parlava correttamente il francese e l’arabo e, quantunque si
dichiarasse mussulmano, si vuole fosse di nazionalità francese. Oltre
il tipo fisico, molti altri indizi davano a credere che la sua terra
d’origine non fosse l’Africa.
Aveva adottato il nome di Rafaat Safi.
Scese all’Albergo Cirenaica, condotto da un maltese e, nei discorsi
che ebbe a fare col proprietario dell’albergo stesso, si mostrò
fanatico dell’islamismo. Tanto fanatico da destare qualche dubbio
circa la sua sincerità.
Il giorno che seguì al suo arrivo si diresse al muftì (prefetto
mussulmano), gli espose la sua intenzione di studiare gli ordini
religiosi maomettani e gli chiese una scorta che lo accompagnasse e
lo guidasse verso l’interno.
Il muftì gli fece osservare tranquillamente che non era affatto
necessario dirigersi verso regioni inospitali, dato il còmpito che si
proponeva; ch’egli poteva rimanersene a Bengasi, dove lo studio gli
sarebbe riuscito più facile, soccorso dall’aiuto e dal consiglio diretto
del muftì stesso.
Rafaat Safi non si lasciò convincere e partì solo.
Che voleva in realtà? dove andava? Nessuno seppe delle sue
intenzioni.
Ieri, al di là di Deriana, nelle vicinanze di un antico castello diroccato,
fu trovato il cadavere del pellegrino.
Gli sciacalli ne avevano fatto festa, poi ch’egli era caduto per via e
non si sa come.
Certo, qualcuno sorride nell’ombra.
 Dalarna. — Vecchie case di contadini.

Sempre cortesie.

A Bengasi, come a Derna, si rinnovano le cortesie dei Giovani Turchi

verso i nostri connazionali. Ormai ho detto di che si tratta, nè vale
citare esempi sopra esempi per avvalorare le mie affermazioni.
Il terreno di un italiano, per citare un caso tipico, è occupato da gran
tempo da un enorme mucchio di sale di proprietà governativa, e chi
sa mai quando sarà liberato, nonostante le preghiere e le insistenze
del proprietario stesso.
Un altro caso tipico, benchè di indole diversa, è quello
dell’ambulatorio mantenuto a spese del Governo italiano.
Il nostro Governo mantiene in questo inferno un giovine medico
valentissimo e valorosissimo, il quale, per voler adempiere
scrupolosamente l’incarico avuto, si buscò già una malattia dalla
quale non è ancora completamente guarito.
Una volta i nativi pareva gli serbassero gratitudine per il suo
sacrificio continuo e per lo zelo e per la valentìa; ora le cose sono
cambiate a tal punto ch’egli deve riguardarsi dal percorrere certi
quartieri in ore poco opportune.
Risate, frizzi e sassate lo accolgono molte volte, tanto che si direbbe
ch’egli fosse qui ad esercitare un ufficio ben diverso da quello che
esercita in realtà.
Anche tale mutamento nei sentimenti della popolazione indigena,
deve assegnarsi al frutto della nuova politica dei Giovani Turchi.

In mare.

È il tramonto. Ho lasciato or ora la vasta piazza nella quale, in

quest’ora meno aspra, erano usciti a passeggiare a piedi o a cavallo i
deportati politici.
Un grande negro dalla faccia gioviale mi accompagna verso il
piroscafo. Tutto arrossa: cielo e mare. I ciuffi dei palmizi su la
spiaggia sembrano pennacchi neri. Uno stagno si infoca; le
lontananze si addolciscono.
Dalla bianca fortezza sparano undici colpi di cannone ad indicare che
incomincia il digiuno del Ramadan [8]; rulli lontani di tamburi e suoni
di piccole cennamelle dànno lo stesso avviso ai fedeli.
Il minareto che si leva nel pulviscolo d’oro dell’aria, dorato egli
stesso e tenue tanto da sembrare una dolce forma fantastica, pronta
a dileguare col dileguar del sole, si incorona di piccole luci scialbe a
festeggiare l’inizio del mese sacro.
Ricordo i versetti del Corano che fecero del Ramadan un precetto
fondamentale dell’islamismo.
“O credenti! È scritto che osserverete il digiuno come lo osservarono
i vostri padri, e ciò perchè temiate il Signore.„
E ancora:
“Il mese di Ramadan, durante il quale il Corano è disceso dal cielo [9]
per essere la guida e la luce degli uomini, e la regola dei loro doveri
è il tempo destinato all’astinenza....„
“Voi potete, durante la notte del digiuno, accostarvi alle vostre
spose. Esse sono la veste vostra e voi siete la loro. Dio sapeva che, a
questo proposito, avevate trasgredito al suo comandamento. Egli vi
ha guardato benevolmente; vi ha perdonato....„
“Potrete bere e mangiare fino all’ora in cui vi riuscirà distinguere, alla
chiarità del giorno, un filo bianco da un filo nero. In seguito
digiunate fino alla notte....„
Il gran negro che rema vigorosamente, seduto verso la poppa del
battello, mi guarda sorridendo.
— Tu non digiunare.... — mi dice; e la sua bocca, quando parla, si
apre come quella di un coccodrillo.
— No — gli rispondo.
— Tu essere porco! — risponde garbamente. — Tu brugiare
inferni!... Tu malate pesta critare cane!... Infitèli saltare inferni; non
morire, saltare sempri giorni, notta! Me, ciartini piratisi!...
E ride e ripete beato:
— Me, ciartini piratisi!
E sarebbe davvero un gran torto, poveretto, se non dovesse trovare,
valicando la gran soglia, il giardino paradisiaco che lo fa tanto gaio.

Il “Ramadan„ in viaggio.

Il piroscafo è animatissimo, viaggiano in prima classe: il fratello del

Governatore di Tripoli con la sua signora, un bimbo e una serva
nera; un impiegato turco e qualche greco. L’impiegato turco, che ha
il passaggio gratuito fino a Costantinopoli e una commendatizia per il
nostro ambasciatore, non viaggia per diporto, fugge.
Pare lo attenda una grave punizione per aver egli favorito o tentato
favorire, in cosa di non troppo grave importanza, gli interessi italiani.
Il tramonto muore.
Dice il Corano, a proposito del Ramadan:
“.... Chiunque vedrà codesto mese deve osservare il precetto.
“Colui che sarà ammalato o in viaggio digiunerà, in seguito, un
numero uguale di giorni.„
Questa la ragione per la quale la maggior parte dei viaggiatori
mussulmani non digiuna.
La maggior parte, ma non tutti. I più vecchi mostrano uno zelo
eccessivo.

Un tramonto sul mare.

Due giovinette, abbassato il velo che le nasconde, guardano la

corolla solare discesa sulle acque per un attimo di splendore.
Ed è tanto bella che, per sua dolcezza, mare e cielo si congiungono
smorendo nel color d’amore.
Non sorride, ha l’ansioso languore degli estremi abbandoni; è
penetrata tutta quanta di paurosa letizia come la vergine che più non
si oppone e piega temendo e sospirando forte sotto la carezza
dell’amato; e così si invermiglia e negli occhi vaghi, larghi e fissi,
traluce l’anima turbata nel desiderio del dolce dolore e nel timore
della violazione.
E la corolla cede; ecco, il miracolo avviene. Il cielo l’ha perduta e la
segue con un accenno di stelle, mentre il mare tutta la possiede e ne
gode e se ne feconda nel suo ardore, che non muta segno.
È la sera, la languida sera per i begli occhi innamorati; fioriscono dal
vuoto i misteri adamantini della tenebra, le luci che chiamiamo stelle
e che immaginiamo come un diadema sopra la fronte di un gioioso
Iddio.
Le stelle, il brivido della tenebra.
Tu le fissi e se l’anima tua si disfrena e schianta ciò che la vincola, e
più non riconosce nè parola, nè senso, nè limitazione, e per le
bianche vie delle meteore si scaglia fuori dal tepido e lucente circolo
dell’atmosfera negli abissi tenebrosi dell’immensità; se l’anima tua
inutilmente ardita vuol raggiungerle, ecco non sa che il terrore, chè il
piccolo cuore la chiama dalla sua zolla, la chiama prigione entro le
vene per la sua vita rossa.
Vita rossa, mistero pari al mistero delle costellazioni!
Oh! affaticarci proni su l’infinito inconoscibile e tessere nostre
ghirlande e vederle sfiorire; e elevare bianche torri fra le nuvole e
vederle rovinare! Oh nostra pena diuturna, nostra angoscia
terribilmente vana, nostro ansimare e ascoltare e scrutare!
Altri compone sue città portentose e ad un ghigno dileguano; altri
sorridendo tesse mirabili fedi e ad una parola si oscurano.
E l’ansia del cammino ne sospinge.
La sera è morta. Ecco il Carro di Boote ch’io guardavo già sopra i
pioppi della mia casa; ecco l’accenno.
Cammina, poeta, cammina; si accendon le stelle; qualcuna che forse
non raggiungerai ti attende lontana.
SUL MARE DEGLI ULISSIDI.
Accampamento di Lapponi.
 Un fiord nell’estremo nord dell’Europa.

Il viaggio del sogno.

È, questo, il viaggio del sogno. Conviene essere soli, dimenticare il

presente, darsi tutti all’onda delle memorie, vivere della magica
eredità del passato e niente più. Di isola in isola si ritrova il silenzio,
la solitudine, il deserto. Attraversate miseri villaggi fra aride terre,
gruppi di squallide case fra giri di montagne rocciose; procedete a
dorso di mulo per sentieri impossibili, chè strade non ve ne sono; vi
internate sempre più fra il silenzio e la desolazione; chiedete
ospitalità a qualche montanaro per trascorrere la notte, chiuso in un
sacco che vi preservi da assalti importuni; riducete il vostro vitto a
pane, uova e formaggio; continuate la strada solo, sempre più solo
per giungere ai luoghi sacri, ai ruderi informi, dai quali vi proverrà il
brivido tragico delle grandi evocazioni e nessun’altra vita che non sia
quella lontana che rifulge nei millennii, nessun’altra voce che non
provenga dalla storia più luminosa del genio umano vi sarà dintorno
a distrarvi, a richiamare l’attenzione vostra su l’ora che fugge.
Due parole sole vi rimarranno nella mente per sentirle ripetere di
continuo, anche al più timido presentarsi di un vostro desiderio: den
exei: non c’è, non l’abbiamo, non esiste; una negazione, insomma,
l’indice dell’assoluta mancanza di tutte le cose di prima necessità.
Conviene ridurre ad una misura più che minima i propri bisogni.
Chiedete un alberguccio qualsiasi? den exei; chiedete del sapone?
den exei; un po’ d’acqua che non sia torbida? den exei; nulla c’è che
non sia la tranquilla ospitalità che vi viene offerta di tutto cuore.
Dividono con voi ciò che hanno, e se sono poveri non è colpa loro.
Vivono nell’abbandono, nello squallore, e la dolcezza del cielo e del
clima non basta alla loro prosperità.
D’altra parte da quando vi imbarcate su questi vaporetti che
approdano, unici, alle isole dell’arcipelago greco, cominciate ad
abituarvi alle privazioni, alla pulizia relativa. È una specie di
iniziazione. Compite un vero pellegrinaggio con tutti i suoi disagi, ma
non si approda a Cerigo, a Milo, a Santorino per puro diporto; non si
viene quaggiù alla ricerca degli alberghi inglesi o tedeschi.

L’isola di Venere.

Abbiamo toccato Cerigo, l’antica Cythera, nella quale fiorì il culto

della Venere fenicia. La terra della voluttà è apparsa in un mare
chiaro quando l’alba rompeva all’oriente; è fiorita con l’alba dai fondi
marini. Ero solo, i pochi viaggiatori dormivano sul ponte l’uno presso
l’altro, avvolti nelle loro coperte. Il canto acuto di un gallo, chiuso in
una stia, a poppa, si levava a interrompere il monotono pulsare delle
macchine, il cigolìo delle catene, lo scricchiolìo degli assiti. C’era
ancora una piccola lanterna accesa su la prua. Il timoniere ed il
pilota, fermi sul ponte di comando, scrutavano l’orizzonte senza
parlare. Una pesantezza di sonno era tuttavia su la piccola nave
grigiastra. Il mare non aveva movimento, e su quell’immota serenità
l’isola apparve quasi all’improvviso, quando il sole si affacciò sul
mare. Una montagna azzurrina su le acque corse da sùbiti bagliori.
Avrei voluto vederla non più di così; non avvicinarmi; non saperne
l’arida povertà. Toglierle il colore, studiarne troppo da presso lo
scheletro è un uccidere in noi il fantasma nato per il fascino di una
fresca giovinezza appassionata che cantò remotamente la sua gioia
bambina. Il nido di Cythera aveva allora il color delle viole e del
rosso bronzo; i culmini erano accesi; le coste pallide smorivano nel
mare magnifico. Le ombre e le penombre potevano sembrare
tuttavia foreste e giardini; immensi roseti per la bionda Iddia che le
ore trassero al trionfo dei cieli. Nessuna cosa velava il sorriso del
mare, il sorriso dei grandi occhi turchini, dai quali traluceva il
desiderio e la limpida anima giovanile aperta come un fior d’oro
sotto il sole; nulla s’interponeva. Nella quiete immensa sbocciava
l’isola dell’amore ricinta dalla luce dell’aurora, dal grande abbraccio
del mare, tutta serena, senza asperità, senza particolarità, come
qualcosa che era e non era nella lontananza, che poteva da un
attimo all’altro dileguare, essere assorbita come un giuoco di nebbie
dal concavo orizzonte. Incerta e affascinante a simiglianza della voce
lontana che il tempo non seppe vincere e che ascoltiamo tuttavia col
brivido di un bacio. Meglio era trascorrere, passar oltre con nella
mente gli echi di un antico convivio, i ritmi delle elegie e degli iambi
che servirono a misurare, a dar vita impareggiabile agli impeti della
poesia erotica e simposica; meglio era dileguare ricordando il
fantasma di Cythera e non più.
“Aveva un ramo di mirto e i bei fiori del rosaio, e si trastullava: la
chioma le ombrava le spalle....„
Così appare ad Archiloco la donna inutilmente amata, appare in un
segno di bellezza semplice e divina e dilegua nell’amarezza della
realtà; così a noi l’isola dolce dell’aurea dea dell’Amore.
Quando vi abbordammo, il mattino era nei cieli. Un piccolo molo,
qualche veliero, una tranquillità stanca, forse non turbata mai
dall’affanno del lavoro; e intorno intorno l’ininterrotta aridità.
Camminai per sentieri dirupati fra montagne riarse. Ricordo il canto
di un giovine pastore seduto al sole sul culmine di una roccia
rossastra; un canto di una malinconia inesprimibile, che mi seguì per
lungo tempo fra le solitudini montane; ricordo, in un cortiluccio di
una casa solitaria, tre belle giovinette che, per un attimo,
rievocarono innanzi a me, inconsciamente, l’antica vita ellenica.
Ero uscito dal mio sacco di difesa dopo una notte quasi insonne;
avevo aperto le finestre; guardavo sorgere il sole fra le montagne. Di
repente apparvero in un frullo, nel cortiluccio sottostante, i capelli
quasi disciolti, le vesti scomposte, nude le braccia e il collo, tre
giovinette belle, fresche come il frutto di luglio su la rama. Avevano
gli occhi umidi e grandi sul volto bruno; umidi di gioia, umidi di
sorriso e una languida soavità accompagnava i loro gesti. Si fecero al
pozzo, riempirono d’acqua le loro secchie di rame rosso, se ne
irrorarono fra strilli e risate, poi, presesi per mano cantarono:
“Io mi lavo con fede davanti al Figlio di Dio e davanti alla stella del
mattino che è la pupilla del mondo....„
Scomparvero, non mi fu possibile rivederle, furono non più di un
bagliore, eppure, per loro virtù, rivisse innanzi a me fuggevolmente
l’anima omerica:
“Quivi giovinetti e fanciulle, che portano in dote le mandre di buoi,
danzavano tenendo le mani l’una nel carpo dell’altra.... Ora
correvano con abili piedi assai facilmente.... ora ricorrevano in fila gli
uni verso gli altri....„
Echi remotissimi che riempiono l’anima di nostalgica dolcezza, e solo
per simili echi si è animato fulgoreggiando tutto il mio viaggio
solitario.
Di Cythera, o meglio di Cerigo, non rammento se non scarsi ruderi
informi, grandi cespi di mirto e di gelsomino e otri di miele, di un
ottimo miele che sa di timo; e rammento altresì una ricetta per un
filtro di amore donatami da una vecchia, nei dintorni di Hapsali. Sì
come le feci l’elemosina, la scarna creatura volle farmi compartecipe
di un suo segreto che potrebbe risalire, forse, al tempo del molle
Adonis. A me non serve e lo comunico ai lettori:
“Si riempie di acqua di rose, per metà, una bottiglietta di cristallo e
vi si pongono a macerare alcuni grani di incenso e qualche foglia,
disseccata, di mirto e di santoreggia. Si chiude ermeticamente la
bottiglietta e la si espone al sole per quarantott’ore, trascorso il
quale periodo si filtra il liquido e vi si aggiungono trenta gocce del
succo di un cedro di trecento anni e la rugiada di nove rose rosse e
di sette rose bianche raccolte all’alba che segue l’apparire del primo
quarto di luna; poi un chiodo di garofano e l’ultimo anello della coda
di un serpente disseccato.„
Nient’altro.
La vecchia dagli occhi pallidi mi assicurò che tale ricetta era
infallibile.
— Provala — mi disse —; non avrai più desiderio....
E si allontanò zoppicando sotto il crepuscolo violaceo.

Da Cythera a Mitilene.

Da Cerigo a Milo, da Milo a Santorino, da Santorino a Paro a Scio a

Metelino abbiamo percorso in tutti i sensi il classico mare degli
Ulissidi.
Di isola in isola, sostando a lungo in piccoli scali quasi abbandonati,
in compagnia di povera gente che si industria come può, ma che ha
precisa coscienza della propria condizione e non si abbrutisce
nell’ignavia che nulla cura; parlando rado e sempre di politica chè
anche l’ultimo fra i greci si interessa della vita pubblica del proprio
paese e ci si appassiona; passando da piroscafo in piroscafo senza
mutare per questo, sistema di vita: dalla Morea, sono giunto a
Metelino lungo le coste dell’Asia Minore.
Si navigava su mari di cobalto, di azzurro tanto intenso che a volte,
nell’ombra, diventava nero e avevamo sempre in vista qualche isola,
qualche scoglio, qualche aggruppamento di montagne coniche.
L’orizzonte era sempre interrotto da una forma singolare, che ne
frastagliava la linea piatta. Nelle lunghe ore silenziose si seguiva
l’apparire e lo scomparire degli scogli granitici e gialli e rossi; nudi,
rivestiti di sole e battuti dalle onde. Avevano ai piedi una bianca
orlatura di spume; al vertice, nell’incendio solare, un volo di
procellarie. Passavano oltre, più remotamente, sotto i confini del
cielo, altri vapori. A volte non si vedeva che un tenuissimo
pennacchio di fumo su l’ultimo mare. Poi piccole barche da pesca
con la vela d’artimone tutta bianca e la piccola vela del trinchetto di
un rosso violento e si andava, si andava senza tregua lentamente,
monotonamente. A bordo era padrone il sonno. La gente mangiava e
dormiva.
Tutto era propizio al sonnecchiare, e il silenzio e l’ampio
ondulamento e il caldo. Così di tappa in tappa, dalle coste di Milo
tutta fiorita di gelsomini a l’immenso cratere invaso dal mare che
forma la baia di Santorini; dalla montagna circolare di Paro, l’isola
dolce che dette i natali ad Archiloco, alla scoscesa Scio, sacra a
Minerva, si approdò a Metelino. Anche il bel nome antico le fu tolto,
e l’“insula nobilis et amœna„ di Tacito, la patria di Alceo e di Saffo,
Lesbo dalle belle figlie, fra terremoti e guerre e invasioni nulla serba
dell’antico splendore.
Quando entrai a Metelino (Midillii la chiamano i Turchi), era quasi
notte. Su la porta di una casa equivoca un turco grattava una guzla,
un istrumento a corde dal suono aspro, chioccio, stridente, e alcune
voci cantavano. Era un canto turco, ciò è a dire un indefinibile
gargarismo, una specie di singhiozzo modulato sgarbatamente su
alcune note discordi che si rincorrevano a gran furia accavallandosi,
sopraffacendosi, urtandosi come tre fiere nemiche rinchiuse in uno
spazio breve. Da prima ascoltate stupefatto, cercate un filo
conduttore, una ragione di tanto frastuono, un qualsiasi disegno
melodico, poi restate disorientato finchè non fuggite stordito. Questa
è la musica turca.
Tale frastuono mi accolse quando entrai nella meschina città senza
garbo e senza carattere, percorsa da marinai turchi e dalle eterne
ammantate che scivolavano rasentando i muri, simili a goffi spettri
claudicanti o ad otri rigonfi dall’equilibrio instabile.
Ricordai i silenzi di Paro, il calmo raccoglimento delle isole sperdute
fra cielo e mare, e le voci e i suoni e le cose che stavano in sì grande
contrasto con l’alba remota che venivo pensando, mi riuscirono
odiose. Non mai come a Metelino sentii il profondo abisso che separa
due anime, due popoli, due razze.
La lira di Orfeo, che la leggenda fece approdare a quelle rive, doveva
essere sepolta bene a fondo sotto la terra sacra, nell’isola più
musicale fra quante ne ebbe la Grecia. Tutto è morto, tutto è
scomparso nel gran buio dei millennii.
Il fanatismo vinse la calma serenità antica, la sopraffece, la turbò, la
sconvolse. I greci che rimangono non sono meno fanatici dei turchi;
a suprema difesa, nella loro debolezza, si avvincono al loro Dio, e il
fanatismo è una sciocca follìa che annebbia ogni chiaro entusiasmo.
Più che altro, pare sia rimasto a questo popolo smembrato,
perseguitato, combattuto senza tregua nella pace e nella guerra, ciò
che era un tempo la rarissima eccezione: la malinconia di Esiodo. È
un popolo malinconico, in fondo, cupamente malinconico. Lo dicono
le sue canzoni e tutta quanta la sua vita. Quante cose non ha visto
morire? Quale soccorrevole Minerva ha sopravvissuto per la sua
fortuna? Egli favoleggiò bensì, un tempo, quando il giovine sorriso
dell’Iliade era nell’anima sua; favoleggiò di un bel Dio dall’arco di
argento che vinse, lottando, lo spirito delle tenebre; ma con quanta
tenebra non ha ripagato di poi quella luce che non fu mai più
sopravanzata?
Sulla terra nella quale l’assurdo e l’incomprensibile furono sempre
sdegnati, doveva stabilirsi appunto il regno dell’assurdo e
dell’incomprensibile, e il popolo più fecondo, perchè credette alla
libertà, doveva piegarsi e immiserire sotto il fanatismo mussulmano
perennemente sterile. Sterilità che fece il deserto. La gran vita solare
fu oscurata nei secoli. L’anima gioiosamente chiara tramutò, si
intristì, decadde e, ai più sperduti, rimase unicamente una inconscia
eredità di costumi.
Solo l’irrequietezza, la turbolenza, la mutevolezza, il cinismo,
caratteri che furono già degli antichi, si mantengono nei greci
moderni.
Dolorosa eredità.
Allorchè, nei tempi più belli di Atene, alla domanda:
— Chi governa il Destino?
Eschilo fece gridare a Prometeo:
— La Parca e le Furie!... — vide nel tempo, profetizzò. Eschilo fu
l’Elia del popolo greco.

Ciò ch’io vidi a Metelino mi fece triste. Fu come chi giunga

festosamente dove era già un giardino e non ritrovi che aride sabbie.
Solo quando l’isola si allontanò sul mare e fu tutta azzurra come un
nido di sogno nell’immensità, potei pensarla e vederla quale era;
solo allora mi rifiorirono nella mente gli echi dei lontani hymenei e gli
sparsi frammenti di una poesia immortale.
Cadeva un crepuscolo soave; ero solo. Passò una piccola barca; mi
giunse una voce. Un giovine e una giovinetta erano ritti su la prora.
Non udii ciò che dicevano; mi ritornarono nella mente le parole di
Saffo:
“Stammi in faccia, caro, e spandimi la grazia che hai negli occhi....„
Sui confini del mare la luce moriva come in una ghirlanda.
Tutte le rose della Pieria eran cadute dal cielo a coronare il fantasma
che dileguava.

Una croce in uno scoglio.

Atene è dileguata fra il fulgore violaceo de’ suoi monti dispogli.

Visione indimenticabile, ebbrezza senza pari!
L’Acropoli, a somiglianza di una enorme torre mozza elevata a
prodigio in queste terre d’oro, è scomparsa. Passano le ombre delle
nubi su le montagne.
Navighiamo verso il canale di Corinto.
In quale diafana chiarità, sotto il fumo del vapore, si perde Atene!...
Naviga, sopra di noi, nel cielo, un velario di nubi bianche. E tanto
sono belli il cielo ed il mare e l’oro del sole è sì vivo, che gli scogli
lontani sembrano gemme.
Di sopra le basse colline spuntano strani profili di nubi; piccole
barche con due vele rosse trascorrono silenziose.
Ecco Helena dal colorito di ambre; Helena, la piccola ateniese mi
ritorna alla memoria.
Ricordi, piccina, le belle meduse nel golfo di Corinto? Navigavano su
le acque chiare, simili a sugheri, le meduse rotonde, e tu le indicavi
col braccio ignudo, con piccole grida.
C’erano due inglesi che andavano in viaggio d’amore e sorridevano
alla tua fanciullezza pensosa come io ti sorridevo; io che pensai in te
il destino della tua patria.
Poi Atene spuntò e tu stavi raccolta su la prora, immersa nella
magnifica chiarità orientale e avevi negli occhi il cuore, il cuore della
tua terra che si affissa verso l’avvenire.
Passammo vicino a uno scoglio. C’era una croce bianca. Non so
perchè una nube di tristezza ti passò sul chiaro volto di bambina e
non ti interrogai: vidi, nel segno, il dolore della tua terra.
Un raccoglimento ed una fede.
E ti benedissi in cuor mio, piccola ignara, che impersonavi un destino
tragico e stupendo. Adesso, come allora, appare lo scoglio dalla
croce bianca, ma tu sei lontana.
Prima che la gran terra dilegui, a te viene il mio cuore benedicendo,
poi che tu mi sei la giovine Grecia che attende.
VERSO I MARI DI GHIACCIO.

Buöreb lä jode, go oro.

(Meglio è andare anzichè
stare).
Proverbio lappone.
 VERSO I MARI DI GHIACCIO.
Una ferrovia in Lapponia.

Un canto dei Lapponi, e cioè degli uomini che abitano le ultime terre
verso il silenzio polare, dice ad un dipresso così:

Solo le navi degli Dei raggiungono la stella del nord,

ma tu non ti affidare, fratello,
non ti affidare alla tua forza.
Nessuno ritorna dalle montagne della bianca stella,
dalle montagne di ghiaccio
che vegliano i palazzi infernali!

E in queste parole è espresso il sentimento che coglie ciascuno di noi

allorchè il pensiero rievochi la titanica sfinge eretta alle soglie del
polo a conservarne intatto il tragico mistero.
Ma noi, gente dei paesi felici che il sole non abbandona mai,
abbiamo del lontano orrore iperboreo una visione incerta, vaga come
le immagini del sogno, indefinita come i paesaggi crepuscolari,
diffusi fra un fluttuare di nebbie; non ne sentiamo l’immediatezza,
non ne avvertiamo la minaccia se non a grandi intervalli, allorquando
l’ardimento di qualche eroe nostro si spinga ad avventurarsi verso
l’ultimo mistero del mondo. All’opposto i Lapponi, come gli
Esquimesi, i Ciukci, gli indigeni delle coste americane polari, vivono
la loro povera vita randagia ai piedi della grande sfinge e ne sentono
il contatto e ne sopportano l’assidua violenza, sì nelle tenebre senza
aurora, come nel pallido giorno privo di tramonti. Essi stanno alle
soglie delle terre ignote, nei paesi tristi e sterminati, in una
solitudine che non ha confine.
E i loro rarissimi canti significano, la maggior parte delle volte, o una
timida speranza o la paura che sovrasta in un impero incontrastato.
Tale la sensazione ch’io ne ebbi accostandoli, vivendo con loro
qualche settimana nella più interna parte delle loro terre verso il mar
Glaciale.
Era mio proposito, partendo da Stoccolma, di inoltrarmi quanto più
mi fosse possibile nel cuore della Lapponia, e ciò unicamente per la
gioia della mia conoscenza e per quel sentimento nostalgico che
sospinge eternamente colui che ha l’anima del viandante ad un
punto più remoto sempre più remoto su le vie della terra e dei mari.
Un tempo non si poteva intraprendere tale viaggio se non seguendo
le coste della Norvegia; ora il Lappland Express, che giunge fino a
Narvik, nel fjord di Ofoten, e cioè verso il 70º parallelo, facilita di
gran lunga il cammino al viaggiatore.
Tale linea ferroviaria, costrutta dallo Stato svedese, è l’unica al
mondo che sorpassi il circolo polare. Essa si lancia attraverso le
steppe deserte del nord ed è protetta per tratti lunghissimi da
gallerie artificiali affinchè le terribili tempeste dell’inverno non
abbiano a seppellirla sotto uno strato di quattro e a volte di cinque
metri di neve. Con tutto ciò dall’ottobre al maggio, lungo il tratto che
va da Boden a Narvik, cioè dal circolo polare fin quasi al 70º
parallelo, nonostante una squadra di oltre un migliaio di operai,
impiegata a mantenere libera la linea, non è raro che i treni vengano
arrestati dalle tempeste, che in quelle latitudini piombano improvvise
e violentissime, e siano quasi sepolti sotto la neve.
Le grandi miniere di Kiruna, il paese delle montagne di ferro,
rendono possibile il mantenimento di tale ferrovia, chè altrimenti non
potrebbe sussistere perchè l’umanità, per quanto la si voglia sedurre
con la visione di una bellezza tragica ed inattesa, preferirà sempre il
Cairo e magari l’Equatore alle singolari dolcezze di un paese nel
quale la notte si prolunga per tre e quattro mesi, confortata da una
temperatura di 40 e 50 gradi sotto zero.
Comunque sia, l’industria che non ha troppe sottili preferenze, visto
il proprio tornaconto, ha sfidato le difficoltà maggiori e con la sua
tenace gagliardia, con l’ardimento che non cede, le ha superate
trionfandone.
Per opera sua ai piedi del Kirunavaara, dove fino a pochi anni or
sono non era che il silenzio dell’immensa steppa, è sorta una città di
9000 abitanti (cifra enorme per quelle latitudini) e per opera sua la
ferrovia che attraversa la Lapponia meridionale può prosperare.

Tristezze nordiche.

Fra i miei appunti di viaggio trovo la nota seguente: — 5 agosto —

Vännäs. — Il cielo è sempre grigio. La vegetazione impoverisce. Fra
poco entreremo in Lapponia. —
Forse fu la mala ventura che si fece mia assidua compagna, ma il
fatto si è che dal giorno in cui sbarcai a Trelleborg, nel punto più
meridionale della Svezia, fino al giorno in cui giunsi ad Enontekis e a
Kautokeino, nel cuore della Lapponia, la pioggia, il vento e il freddo
furono miei assidui compagni. Conobbi l’estate più nella volontà degli
uomini che non in quella della stagione. Vidi le signore nelle più
ardite vesti estive, nude le braccia e il principio del seno, ma il
termometro segnava modestamente 12, 15 gradi e qualcosa più ed
anche molto meno. A Stoccolma, ad esempio, alla sera, tempo
permettendolo, si andava a frescheggiare su le terrazze aperte dei
grandi caffè. Frescheggiare è la giusta parola, perchè non appena
seduti, il cameriere si affrettava ad offrirci, oltre alla bibita, una
coperta rossa, nella quale era necessario avvolgersi per difenderci un
poco dalle intemperanze estive.
Tale non è il clima consueto della Svezia, che gode fama di
dolcissimo durante l’estate. Per mia disgrazia io assaporai un frutto
di eccezione e lo assaporai tanto più compiutamente quanto meno
avevo preveduto un simile cataclisma.
Man mano che procedevo verso il nord i miei poveri vestiti, troppo
nostalgicamente italiani, mi dimostravano la loro completa
insufficienza finchè, prima un ampio scialle da viaggio, poi un
pastrano imbottito, non li corressero in modo definitivo. Data la
quale correzione mi fu concesso osservare i luoghi ed i costumi con
occhio più tranquillo. Così il 5 agosto 1907, essendosi fermato il
Lappland Express alla stazione di Vännäs, scrivevo la malinconica
nota.
Avevamo percorso fino allora il Norrland, fra le dense nebbie delle
valli ed il pallido sole delle cime. La natura aveva assunto un aspetto
di grandiosità severa, ma troppo continua, troppo eguale per l’anima
nostra mutevole, che mai non s’appaga perchè educata ad una
varietà incomparabile.
Gli occhi nostri facilmente si saziano di un aspetto sì per la rapida
facilità di coglierlo sotto tutte le sue linee, come per l’indubbia
certezza di trovarne innumerevoli altri dissimili; ora tale facilità e tale
certezza nuocciono alla compiuta penetrazione di un paesaggio
nordico. Là dove la vita si attarda, dove le cose si fanno oscure e
solenni, dove gli uomini hanno un’anima diversa, più tranquilla, più
dolce, più profonda, è necessaria a noi una continua moderazione,
una vigilanza assidua su ogni giudizio, su ogni moto d’impulso, su
ogni repentina impressione; conviene temperare il nostro ardore per
bene intendere le cose che pur passandoci da presso non ci sono
vicine e l’intimo senso delle quali ci sfugge. A primo aspetto gran
parte della Svezia farebbe a noi un’impressione desolatamente
monotona, sicchè saremmo tratti, con tutto il garbo possibile, a
parlarne come di un paese che si ripete e si ripete senza varietà,
quasicchè la natura, esausta dal troppo creare, fosse giunta laggiù
con poche idee e le avesse svolte in una continuità che opprime. E
questo sarebbe un errore causato da una impressionabilità mal
contenuta e dalla luce e dai colori e dai rapporti interamente diversi
e disposti in armonie insuete.
Noi proveniamo dalla terra dei contrasti, dalla terra in cui ogni
arditezza è possibile, in cui i giuochi delle mezze tinte, delle più tenui
gamme di colore, son pur sempre vivi, distinti, apertamente chiari e
rilevati. La nostra natura, che ha breve il sonno invernale, si desta
nell’ebbrezza primaverile prorompendo come in un magnifico
tumulto, impaziente di vivere, di donarsi alla carezza del sole, di
espandersi in una festosità piena di ardore. Il dolce miracolo si
compie in poche notti quando l’aria raddolca, così come fioriscon le
stelle nel sereno, quasi inavvertitamente. V’è giorno in cui gli alberi
aprono i loro bocci bianchi e vermigli, in cui le selve, i campi, si
coprono della soave veste di giade della primavera, in cui un infinito
tremito di vita dilaga in un impeto improvviso quasi per forza di
prodigio. A troppa gioia son usi gli occhi nostri perchè possano a
tutta prima intendere il lento languore, la profonda malinconia di un
paese nordico, colto nella pienezza di una triste estate che non ha
autunno e non ha primavera. E tale languore e tale profonda
malinconia provengono appunto da quelle indefinibili sfumature, da
quei rapporti degli alberi coi pallidi cieli, con le acque dei laghi
deserti, con la luce diffusa che li fonde e li tramuta lievemente e li
vela e li inargenta, ma appena, in tocchi di una leggerezza a noi
ignota. E per questo giuoco sottilissimo, ecco: ciò che pareva
uniforme svaria, si moltiplica in mille aspetti, in un seguito di
apparenze multiformi estremamente instabili, ma di una triste
soavità che rivela tutta l’anima di quella terra dolente.
Ricordo quella corsa ininterrotta attraverso il Norrland con una
particolare sensazione di tristezza. Il giorno non finiva mai. Il riposo
della notte non c’era concesso.
Verso le dieci di sera il cielo si rasserenò, apparve tutto azzurro. Il
sole volgeva allora al suo breve tramonto, tanto breve da essere
vinto da una subita aurora che gli si accendeva a lato. Quell’eterna
luce mi eccitava talmente da togliermi il sonno. Gli altri viaggiatori si
erano già ritirati nelle loro cabine, ero rimasto solo. Ci fermammo ad
Asträsk: una piccola stazione in legno in una landa. Nessuna casa io
vidi, solo la luce di un lago lontanissimamente. Una linea fulgida; un
bagliore in un deserto. Le abetaie eran finite coi monti. Guardai
tutt’intorno per quanto era vasto l’orizzonte, ma non scorsi traccia
del paese. Asträsk non esisteva, era un nome ed una casa in una
infinita solitudine. Le grandi, le forti selve erano scomparse.
Incontrammo ancora gruppi d’alberi, ma sempre più miseri e
contorti: il circolo polare si avvicinava. A Bassuträsk una bimba
scalza, e il vento freddo le faceva tutti rossi quei suoi piccoli piedi, si
avvicinò al treno per offrirmi alcuni fiori raccolti lungo gli stagni in
quel paese sinistro. Non disse parola, tese la mano verso di me e mi
guardò senza sorridere offrendomi quel suo dono con una luce
talmente dolorosa, in fondo a que’ suoi grand’occhi di bimba, ch’io
non potrò mai più dimenticare.
Discesi, l’interrogai. Aveva un fratello e la mamma; il padre le era
morto a Kiruna dove lavorava su le montagne del ferro. Un giorno
l’abisso l’attrasse, gli occhi gli si annebbiarono, precipitò nel vuoto,
colto dalla vertigine. Da quel tempo esse vivevano nelle terre di
Bassuträsk in una casa sola, al limite di una selva, lontana miglia e
miglia da ogni altro luogo abitato. Ogni mattina ella, ch’era la
maggiore, vestiva il suo piccolo fratello e partivano per la scuola.
Percorrevano dieci chilometri, su la neve, durante l’inverno, fra gli
stagni nell’estate. Nonostante una notte di tre mesi, l’inverno era
preferibile. Allora scivolavano via su gli skidor e dieci chilometri
erano un niente.
Scompariva appena il lume della loro capanna che già si
intravvedeva fra gli alberi e la neve il lume della scuola. Avevano
avuto tre maestre. Una era morta, l’altra era impazzita dalla paura,
la terza era sempre triste e parlava della Scania, di un paese lontano
dove il sole ritornava ogni giorno.
La scuola era una piccola casa rossa sperduta tra i boschi, le
abitazioni più vicine essendo ad otto o dieci miglia di distanza. Del
resto in tutta la regione di Bassuträsk sorgevano forse una ventina di
case, sparse sopra un raggio di venti miglia.
Le maestre erano giovani e quasi sempre sole. Giungevano dalla
Svezia meridionale, dai paesi felici, e poche potevano resistere a
tanta solitudine, a quell’isolamento pauroso. Dal novembre alla fine
di febbraio scendeva la notte, la terribile notte polare, squarciata
solo a quando a quando dalla tremula luce rossiccia dell’aurora