Adaptation and The Brain Oxford Series in Ecology and Evolution Susan D Healy Full Chapter

Adaptation and the Brain (Oxford Series
in Ecology and Evolution) Susan D.

Healy
Visit to download the full and correct content document:
https://ebookmass.com/product/adaptation-and-the-brain-oxford-series-in-ecology-an
d-evolution-susan-d-healy/
OUP CORRECTED PROOF – FINAL, 11/01/21, SPi
Oxford Series in Ecology and Evolution

The Comparative Method in Evolutionary Biology
Paul H. Harvey and Mark D. Pagel
The Cause of Molecular Evolution
John H. Gillespie
Dunnock Behaviour and Social Evolution
N. B. Davies
Natural Selection: Domains, Levels, and Challenges
George C. Williams
Behaviour and Social Evolution of Wasps: The Communal Aggregation Hypothesis
Yosiaki Itȏ
Life History Invariants: Some Explorations of Symmetry in Evolutionary Ecology
Eric L. Charnov
Quantitative Ecology and the Brown Trout
J. M. Elliott
Sexual Selection and the Barn Swallow
Anders Pape Møller
Ecology and Evolution in Anoxic Worlds
Tom Fenchel and Bland J. Finlay
Anolis Lizards of the Caribbean: Ecology, Evolution, and Plate Tectonics
Jonathan Roughgarden
From Individual Behaviour to Population Ecology
William J. Sutherland
Evolution of Social Insect Colonies: Sex Allocation and Kin Selection
Ross H. Crozier and Pekka Pamilo
Biological Invasions: Theory and Practice
Nanako Shigesada and Kohkichi Kawasaki
Cooperation Among Animals: An Evolutionary Perspective
Lee Alan Dugatkin
Natural Hybridization and Evolution
Michael L. Arnold
The Evolution of Sibling Rivalry
Douglas W. Mock and Geoffrey A. Parker
Asymmetry, Developmental Stability, and Evolution
Anders Pape Møller and John P. Swaddle
Metapopulation Ecology
Ilkka Hanski
Dynamic State Variable Models in Ecology: Methods and Applications
Colin W. Clark and Marc Mangel
The Origin, Expansion, and Demise of Plant Species
Donald A. Levin
The Spatial and Temporal Dynamics of Host-Parasitoid Interactions
Michael P. Hassell
The Ecology of Adaptive Radiation

Dolph Schluter
Parasites and the Behavior of Animals
Janice Moore
Evolutionary Ecology of Birds
Peter Bennett and Ian Owens
The Role of Chromosomal Change in Plant Evolution
Donald A. Levin
Living in Groups
Jens Krause and Graeme D. Ruxton
Stochastic Population Dynamics in Ecology and Conservation
Russell Lande, Steiner Engen, and Bernt-Erik Sæther
The Structure and Dynamics of Geographic Ranges
Kevin J. Gaston
Animal Signals
John Maynard Smith and David Harper
Evolutionary Ecology: The Trinidadian Guppy
Anne E. Magurran
Infectious Diseases in Primates: Behavior, Ecology, and Evolution
Charles L. Nunn and Sonia Altizer
Computational Molecular Evolution
Ziheng Yang
The Evolution and Emergence of RNA Viruses
Edward C. Holmes
Aboveground–Belowground Linkages: Biotic Interactions, Ecosystem Processes,
and Global Change
Richard D. Bardgett and David A. Wardle
Principles of Social Evolution
Andrew F. G. Bourke
Maximum Entropy and Ecology: A Theory of Abundance, Distribution, and Energetics
John Harte
Ecological Speciation
Patrik Nosil
Energetic Food Webs: An Analysis of Real and Model Ecosystems
John C. Moore and Peter C. de Ruiter
Evolutionary Biomechanics: Selection, Phylogeny, and Constraint
Graham K. Taylor and Adrian L. R. Thomas
Quantitative Ecology and Evolutionary Biology: Integrating models with data
Otso Ovaskainen, Henrik de Knegt, and Maria del Mar Delgado
Mitonuclear Ecology
Geoffrey E. Hill
The Evolutionary Biology of Species
Timothy G. Barraclough
Game Theory in Biology: Concepts and Frontiers
John M. McNamara and Olof Leimar
Adaptation and the Brain
Susan D. Healy
Adaptation and the Brain
SUSAN D. HEALY
School of Biology, University of St Andrews, UK
1
1
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Susan D. Healy 2021
The moral rights of the author have been asserted
First Edition published in 2021
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2020949282
ISBN 978–0–19–954675–6 (hbk.)
ISBN 978–0–19–954676–3 (pbk.)
DOI: 10.1093/oso/9780199546756.001.0001
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
Acknowledgements
I want firstly to thank Paul Harvey and Bob May for generously giving me the oppor-
tunity to write this book. They had no idea quite how long it would take for me to do
realize that opportunity. I am also very grateful to the team at Oxford University
Press—Ian Sherman, Charles Bath, Helen Eaton, Bethany Kershaw, and Lucy Nash—
whose patience and goodwill I have tried and tested for over a decade. Your endless
support finally overcame my seemingly endless prevarications.
I am very grateful to a number of people who gave me invaluable advice and feed-
back on early drafts of chapters: Rob Barton, Jackie Chappell, Zach Hall, Jo Madden,
David Pritchard, Sally Street, Shoko Sugasawa, and Maria Tello Ramos. I am also very
grateful to John and Gab Chainey, Leigh and Carol Simmons, and members of the
CEB at the University of Western Australia in Perth for being wonderful hosts and
providing a fun and intellectually stimulating environment for a sabbatical in 2015,
where early versions of chapters were drafted. David Shuker read this version in its
entirety, as well as green-penning and emoji-ing several earlier drafts.
Thank you to my lab group and collaborators who have put up with my tirades on
brain size analyses for a long, long time. Some have been especially patient/encour
aging in offering a sounding board: Ida Bailey, Alexis Breen, Nora Carlson, Sophie
Edwards, Andrea Griffin, Lauren Guillette, Zach Hall, Andy Hurly, Eira Ihalainen,
Rachael Marshall, Simone Meddle, Kate Morgan, Felicity Muth, David Pritchard,
Candy Rowe, Shoko Sugasawa, and Maria Tello Ramos.
Making the miles seem few, my family have always been there supporting my life
on the other side of the planet, but they have been especially encouraging over the last
few months: Denis, Roger, Conrad, Fiona, and Tash, thank you. Helen would be
proud, again.
Thank you to Candy Rowe, without whom this book would not have begun and
who has been a fab collaborator on my thinking about brain size. But this book would
not be seeing the light of day without the day-in, day-out encouragement and positive
critique from my partner, David Shuker. He not only listened to countless harangues
on brain size and intelligence but made me explain unfathomable arguments and get
rid of the first nearly full draft. Through his constant support he made me make this
version so very much better as a result. Thank you. There is a lot of work to be done
by those two words, so the least I can do is to say them again. Thank you.
Contents
1 Introduction 1
2 The Measurable Brain 5
3 The Intelligent Brain 17
4 The Ecological Brain 35
5 The Innovative Brain 51
6 The Technical Brain 67
7 The Sexual Brain 81
8 The Social Brain 109
9 Concluding Remarks 123
References 129
Index 159
1
Introduction
Why does brain size vary among animal species? Although Aristotle may have con
sidered the heart to be the seat of the intellect, it has been a good long while since we have
known that it is our brain that enables us to think all sorts of sensible and crazy things,
that causes our dreams, good and bad, happy and sad, that helps us to plan our day, our
week, or our year. And when it begins to fail we start to lose something of ourselves.
Throughout our lives we know that it is our brain that allows us to learn to read, to
spell, to do maths, all abilities that often form the basis of what is collectively called
intelligence. Furthermore, when we look at the animals around us, our pets, the birds
in our gardens, the cows and sheep in the fields, the spider climbing the wall, the fly
buzzing at the window, we think we are smarter than they are. After all, we are the ones
doing the reading, the spelling, the maths, the inventing of machines, and the flying of
rockets to the moon. And if you did not know it already, it will be perhaps unsurpris
ing to find that we have brains that are in fact pretty large. Indeed, there are few ani
mals on the planet that have brains that are larger, especially if we allow for differences
in body size: elephants and whales do have larger brains than we do, but we are smarter
than them (witness all the things we do that they do not: maths, rockets, etc.).
For most people, the advantages of our large brains are clear and the evidence
unequivocal. But the question then is, why humans? Why don’t all animals have ‘big’
brains? Having invented fire, tools, clothes, house building, machines to transport us
over all terrains, we can now occupy almost all parts of the planet. Very few other
organisms could say the same (bar the microbiome we carry along with us). What in
our evolutionary history has led us to this preeminent position in the animal kingdom,
and has enabled us to exploit our planet almost at will (at least for now)?
Academic attempts to figure out how and when we got to be so special have largely
been the domain of anthropologists, who look at extant humans and other primates,
as well as the fossil evidence of our common ancestors. Evolutionary biologists have
perhaps been rather less interested in the specifics of human evolution, although the
genomics era has changed that to some extent, as we no longer rely just on haphazard
fossil finds. Rather, evolutionary biologists are, typically, interested in explaining pat
terns of variation, and it is difficult to see a pattern in a single species, especially with
a patchy fossil record. Given the patchiness of that record for pretty much every species,
an evolutionary biologist interested in seeing how and why some particular trait has
evolved, turns either to an experimental system (often invertebrate) where generation
times are typically short, or to the comparative method. As the name implies, this
Adaptation and the Brain. Susan D. Healy, Oxford University Press (2021). © Susan D. Healy.
DOI: 10.1093/oso/9780199546756.003.0001
2 • Adaptation and the Brain
latter method enables the comparison of trait values across a number of species,
revealing (or not) a pattern to trait values that is coincident with variation in some
possible causal factor. The comparative method in evolution goes all the way back to
Darwin, but modern comparative analyses explicitly utilize phylogenetic information
to try to separate evolutionary signals from shared phylogenetic noise. Early examples
of phylogenetically informed analyses include correlations between age at which
female mammals first reproduce and life expectancy (Harvey and Zammuto, 1985),
the brightness of male plumage in birds and parasite prevalence (Read, 1987), and the
role of diet on carnivore home-range size (Gittleman and Harvey, 1982). Since then,
phylogenetic comparative methods have been widely used to examine trait variation
across species in a vast array of traits (Harvey and Pagel, 1991). Of specific relevance
here is that the comparative method has played, and continues to play, the major role
in evolutionary discussions of why brains vary in size, and why some species have
bigger brains than others.
The first substantial comparative analyses of brain size were enabled by a number of
data sets of bird and primate brain sizes (e.g. Portmann, 1944; Stephan et al., 1981, 1988;
Jolicoeur et al., 1984; see Table 2.1). Since the early 1980s tens, if not hundreds, of com
parative analyses of variation in brain size have been published, proffering a bewilder
ing array of competing or overlapping explanations. For example, primates with bigger
brains live in larger social groups, fruit-eating primates have larger brains than do leaf-
eating primates, birds in long-term pair bonds have larger brains than those that are
less faithful, bats with small testes have larger brains than bats with large testes, bats
that fly through complex habitats have bigger brains than bats that do not need such
flight agility, whales that live in social groups have larger brains than whales that live
alone or in small groups, and many, many more.
How might one make sense of all of these correlations? Is there really no single
factor that is key to changing the size of an animal’s brain? Do birds, mammals, rep
tiles, fish, amphibia (and invertebrates) all have such different brains and lead such
different lives that we need distinct explanations for each taxonomic group?
Furthermore, does each lower taxonomic group (e.g. within the mammals: primates,
cetaceans, bats) have its own explanation?
In this book I will argue that to date all the attempts to find this one all-encompassing
explanation have failed. They have failed for one or more of a number of key reasons:
(1) it has proved statistically problematic to include all of the possible explanatory
variables in unified analyses; (2) the brain size data are themselves problematic, both
in terms of precision of measurement and in the function imputed to them (more of
this in Chapter 2); and (3) the categorization and quantification of the explanatory
variables are open to debate, as we nearly always rely on derived, ‘proxy’ variables that
we hope capture the nature of the proposed causal agent.
How to go forward, given these failures? This book represents an attempt to bring
together the complicated and diverse history of attempts to explain the evolution of
brain size across the last 50 years or so, and to examine critically what these attempts
have told us. The aims of this book are therefore four-fold. First, I will describe the
limitations of previous iterations of brain size comparative analyses. This is not so
Introduction • 3
much as to admit defeat up until this point, because some results do indeed seem more
robust than others. Rather, it is to highlight that re-analysing the same data, again and
again, with similar failings in methods and inference, is not moving us very far for
ward, however easy it is to plug data into the latest R package. Second, by laying out as
many of the possible explanations together in the same place (if not in the same table,
at least in the same book), I will provide an opportunity to stand back and see how
competing explanations do or do not fit together. For example, just how different are
explanations for brain size evolution based on foraging versus sociality, since many
social species forage together? Third, throughout I will emphasize where new data, and
new approaches, are needed to take our understanding of brain evolution truly for
ward. Part of this will involve hard thinking, especially when we use terms like ‘forag
ing’ and ‘sociality’, that in fact are short-hand for a whole variety of adaptations—sensory,
locomotor, cognitive—that we might expect will shape the size and structure of com
ponents of the brain. Finally, throughout the book, there will be the implicit, and some
times explicit, critique of the very question at the beginning of the Introduction: why
does brain size vary among animal species? If the brain is heterogeneous, controlling
countless processes (not all neural), and it outsources various of its functions to sense
organs and the periphery, is a holistic notion of brain size even a meaningful concept?
Most importantly of all, is brain size a useful proxy of intelligence?
As such, this book is both a review and a critique of the field of brain size evolution.
By the end of the book I want the reader to have gained a panoramic view of the work
we have engaged in over the last five decades or so, and what has been more or less
successful, providing pointers of where we might go next. But I am happy to be can
did about the challenge that writing this book comprises. A critique I co-authored
with Candy Rowe, well over a decade ago, whilst briefly intellectually satisfying, failed
to achieve what we hoped for at the time. Healy and Rowe (2007) provided an in-
depth critique of the study of brain size evolution, focusing in particular on meth
odological issues, with the (implicit) intention of reducing the number of analyses
undertaken that produced yet another explanation for the variation in brain size. At
that time, there were some 20, apparently independent, variables that explained a
significant proportion of the variation in brain size for both primates and birds,
although the two lists of variables were not exactly the same for the two taxonomic
groups. For at least two reasons, this slew of explanations seemed curious to us.
Firstly, about 20 years previously, there had been a similar, although substantially
smaller, range of explanations proposed to explain variation in brain size, which by
2006 appeared to have been overlooked. Secondly, most of the more recent explan
ations were being put forward in isolation with regard to the other possibilities and
unified statistical analyses were not being attempted. While it is possible that each of
these 20 or more factors may make a significant contribution to explaining the con
siderable interspecific variation in brain size across animals, it seemed at least plaus
ible to us that some of these factors might be closely correlated and, therefore, much
less relevant alone than it first appeared. It seemed that the field was due a good dose
of critical self-examination before the next big step could be taken in understanding
why brains vary so much in size.
But Healy and Rowe (2007) signally failed to have anything of this sort of effect. It
also failed to increase the quality of the data analysed and, really, has had remarkably
little impact on the content of the field, although it has been quite well cited (directors
of research asked us to write more of the same, such is modern academic life).
However, with hindsight, it seems that variation in brain size is both too interesting,
and its causal factors too little understood, for such an undertaking as ours to be any
thing other than doomed to failure.
Here, then, my aims are not nearly so ambitious. Rather I acknowledge at the outset
that there is no smoking gun to this story, no murder mystery solved in the concluding
chapter. That might seem disappointing, but I would argue that most of us acknowledge
that when we answer one problem, in doing so we often become aware of other
questions, that emerge from solving that first question. Indeed, that is the hook that
keeps us coming back and has us working evenings and weekends: a quest that might
seem endless but not fruitless. So, with that more positive outlook in mind, here
I attempt to take each explanation at face value and then see how plausible it is when
held up in the face of the proposed supportive data. Can some explanations be seen to
be less compelling than others, or less well supported? Will it become clearer what is
really needed to strengthen the case for any or all of the suggested explanations?
In the laying out of several hundred papers, I decided that the bewildering array of
explanations meant that some sort of categorization needed to be imposed. The first
was based on the decision that some papers proposed a causal explanation for vari
ation in brain size while others proposed a mechanistic explanation. These latter
papers are not considered in what follows and the reasoning is thus: development,
physiology, and other mechanistic factors will all have significant impacts on the size
of a brain and will explain how brains are able to change in size. It is therefore crucial
to understand the mechanism(s) by which brain structures can change in size across
evolutionary time. But here I focus on the potential natural selection pressures (in a
Tinbergian sense, the functional why) that may have led to some brains being larger
than others, i.e. the role adaptation plays in changing brain size. These selection pres
sures fall into five broad categories: ecology, innovation, technical, sexual, and social,
as these represent those most commonly favoured groups of explanations.
Before considering each of these in turn, however, there are two key issues that
need to be addressed: (1) the measurement of brain size, and (2) whether brain size
represents intelligence. Chapter 2, then, contains a discussion of various ways to meas
ure brain size and their pros and cons, while Chapter 3 contains a discussion of the
relationship between brain size and cognitive abilities. In the chapters (Chapters 4–8)
that follow I address each of the five selection pressures in turn. In the concluding
chapter (Chapter 9) I provide something of a summary and some suggestions for
future directions.
2
The Measurable Brain
Why do some animals have big brains? This book is all about attempts to answer this
question using comparative evolutionary analyses. The key to the success of a com-
parative evolutionary analysis is sample size, the larger the better. To detect a pattern
in the occurrence of large brains, we need lots of brains from lots of species. These
data have turned out to be pretty challenging to collect, and so the brain size data
used in the analyses described in Chapters 4–8 come from a variety of sources
(Table 2.1). Researchers have also used a variety of methods for measuring brain size,
in particular methods that allow for a high throughput. This has meant that brains
have been measured in part, or as a whole, or that they have been sliced, weighed,
scanned, and reduced to numbers of neurons in a blender. Both because of the diffi-
culties in getting large numbers of real brains, and the considerable resources offered
by museum collections in terms of historical skull collections, a substantial propor-
tion of the comparative analyses of brain size have inferred brain size from skull
capacity. In this chapter, to provide some understanding of the data that have been
used for the many comparative analyses I will discuss, I will describe the methods
that have contributed to the brain size data shown in Table 2.1. The data in Table 2.1
form the large majority of the data used in the analyses discussed in the following
chapters. I will also include descriptions of other methods that have been developed
more recently which, while they may not have contributed to the data in Table 2.1,
have been either the basis for other comparative analyses I describe later, or may be
useful sources for providing data in the future.
Brains have been measured in a variety of ways, not least because there are virtues
and drawbacks to each method. I will include a brief description of these. I will then
consider the relative merits of combining data from more than one source, or more
than one kind of measurement. Because the data available for these comparative
analyses are typically limited, combining data (e.g. brain weights combined with
endocast volumes), or transforming data from one form to another (e.g. brain vol-
ume into brain weight) has been reasonably common. Data combined using such
techniques may or may not lead to spurious associations (Healy and Rowe, 2007).
I also want to use this chapter to get inside the brain itself in order to consider what
whole brain size means, in terms of its structure and function. Here, I suggest that
we consider a brain as, at the very least, a sum of its parts, rather than necessarily
being just a single structure.
DOI: 10.1093/oso/9780199546756.003.0002
Table 2.1 The major sources of data used in the analyses in Chapters 4–8.
Data source Measurement type Taxonomic group No. of

species/
individuals
(if known)
Crile and Brain mass Birds 58/407
Quiring 1940 Carnivores 30/56
Pinnipeds 4/12
Cetaceans 3/8
Bats 1/5
Edentates 5/54
Fish 38/104
Insectivores 2/69
Marsupials 1/17
Primates 19/611
Reptiles 21/126
Rodents 39/410
Ungulates 27/54
Portmann 1944, 1947 Brain mass Birds 140
Oboussier and Brain mass Ungulates 35/132
Schliemann 1966
Bauchot and Stephan Brain mass Insectivores 33
1966
Bauchot and Stephan Brain mass Primates 51
1969
Pirlot and Brain mass, some from Bats 51
Stephan 1970 fresh weight, others after
perfusion
Jerison 1973 Endocranial volume Fossil ungulates 48
and fossil 21
carnivores
Eisenberg and Endocranial volumes Bats 225/at least
Wilson 1978 converted to brain mass 900
Stephan et al. 1981 Brain mass, also volumetric Insectivores and 28/57
data for brain regions from primates 48/109
perfused brains
Gittleman 1986 Endocranial volumes Carnivores 153
converted to brain mass
Baron et al. 1996 Brain mass Bats 313
Marino 1998 Brain mass + 1 endocranial Odontocetes 21/103
volume
Dunbar and Brain volume (known) Carnivores 15
Bever 1998 Brain volume (estimated) 43
Iwaniuk and Endocranial volumes Birds (waterfowl) 55/354
Nelson 2001
The Measurable Brain • 7
Data source Measurement type Taxonomic group No. of

species/
individuals
(if known)
Garamszegi et al. 2002 Brain mass Birds 166/1,812
Iwaniuk and Brain mass and Birds 82/157
Nelson 2002 endocranial volumes
Iwaniuk 2003 Brain mass and Birds 417/2,577
endocranial volumes
Iwaniuk and Formalin fixed brain mass Birds 1,482/10,364
Nelson 2003 and endocranial volumes
Garamszegi et al. Brain mass from frozen Birds 39/584
2005a specimens
Lefebvre et al. 2006 Brain mass and Odontocetes 25
endocranial volumes
Ashwell 2008 Endocranial volumes Marsupials 182/1,257
Isler et al. 2008 Endocranial volumes Primates 232/3,813
Boddy et al. 2012 Brain mass and Primates and 930/2,000
endocranial volumes cetaceans
Møller and Erritzøe Brain mass Birds 108/2,721
2015
2.1 Measuring the brain
Although there are rapid technological changes afoot to enable live animal, in situ,
measurements of brain structure and function (e.g. Hecht et al., 2019), these are yet to
encompass sample sizes of the magnitude desirable for comparative analyses. About
half of the brain size measurements used in comparative analyses have come from brain
tissue from dead animals. But even when brain tissue is available, measurements of
actual brain size are often not made in exactly the same way. To begin with, some of the
data come from animals that have been killed in order for their brains to be measured,
while other data have come from animals that have been acquired serendipitously, e.g.
that died or were killed for some other reason rather than to acquire the brain.
If the brain is taken from an animal for the explicit purpose of measurement, it is
possible (and generally appropriate) to begin processing the brain prior to measure-
ment, and prior to removal from the skull. One common method is to kill the animal
using an overdose of anaesthetic and then perfuse it, first with an anti-coagulant, then
with a suitable perfusant (e.g. Krebs et al., 1989 used a lethal dose of sodium penta-
barbitone, perfused transcardially with physiological saline, followed by formalin/
saline and then fixed the tissue in 30 per cent sucrose/formalin; Stephan et al., 1981;
Day et al., 2005). This should mean that the perfusant will reach into even small
capillaries, resulting in a well-preserved brain that can be sliced and stained for meas-
urement. In addition to the preservation of a structurally sound brain, this method
allows for relatively straightforward and accurate, albeit laborious, volumetric meas-
urement, both of parts of the brain as well as the entire brain itself. However, that is
only possible if the interior anatomical boundaries of regions of known structure and
function can be identified (e.g. Krebs et al., 1989; Stephan et al., 1981; Day et al., 2005).
Brain map atlases can help to guide researchers even when the species of interest is
not one for which there is an atlas already (e.g. rat: Swanson, 2004; Anna’s humming-
bird Calypte anna: Stegeman, 2013; canary: Vellema et al., 2011; pigeon: Güntürkün
et al., 2013; zebra finch Taeniopygia guttata: Poirier et al., 2008). Following sectioning
into a known number of sections of known width, the area and then volume of the
brain slice or identified region(s) on the slice can be measured. In earlier studies this
measurement was achieved by projecting the section onto photographic paper, draw-
ing the borders, cutting out the paper, and weighing it (e.g. Stephan et al., 1981). The
volumes of successive sections, and thus the volume of the entire region or brain, can
then be calculated (in the Krebs et al., 1989 paper the formula of a truncated cone was
used to approximate the volumes of the hippocampus and telencephalon).
The downsides (which are not insignificant) of this method include killing the ani-
mal, and the skill and specialist equipment required (although this latter can be taken
into the field with a bit of preparation: Airey et al., 2000). Furthermore, loss of meas-
urement accuracy can arise from various parts of this process. Among them is that
slicing may damage sections, and that staining the sections will often shrink and ruf-
fle them. Moreover, there is a lot of work in preparing the tissue (sectioning the brain,
staining, and mounting the sections) and in the quantification of sections. There is
also a change in size as a result of the perfusion itself, which may vary depending on
the size of brain being prepared, and from laboratory to laboratory. But there are
significant upsides to this method. While it is an effective way to measure brain size,
volumes calculated from serial sections are especially useful for quantifying parts of
the brain (provided there is good neuroanatomical definition of functional areas).
Another advantage of this method is that prior to the collection of the brain the indi-
vidual’s sex, and sometimes age, can be identified. Furthermore, measurement of
body size can be matched to brain size on an individual by individual basis (e.g. Healy
et al., 1996; Healy and Krebs, 1993), which is important for nearly all of the analyses
that follow (see Chapters 4–8). Using species’ averages instead of brain size and body
size loses vast amounts of potentially crucial data.
Herculano-Houzel’s isotropic fractionator method of neuron counting deals with
some of these issues (e.g. Herculano-Houzel, 2010, 2011, 2016; Herculano-Houzel
and Lent, 2005), especially the time and effort spent on preparing the material (which
does require perfusion but need not always begin with live animals). This method is
unlike any of the other methods so is worth describing in a bit of detail. It works best
with tissue that has been fixed in preservative for some period of time (e.g. a couple
of weeks or longer), because the method relies on the researcher’s ability to identify
and count neurons after the brain has been reduced to its constituent cells (a soup, as
Herculano-Houzel, 2016 describes it), which is not possible if fresh, unfixed tissue is
used (the nuclei will not survive the grinding up of the tissue). A fixative like para-
formaldehyde, which cross-links protein molecules in the tissue, leaves the tissue
hard and mechanically stable. The nuclear membrane becomes especially well fixed
because it is protein rich. This fixation is useful because of the mechanical stress that
the tissue is put under during the fractionation process. Once well fixed, the whole
brain can be cut up into small pieces, or anatomically and functionally meaningful
regions can be dissected out first, which can then be cut into small pieces and fraction-
ated separately. The pieces go into Triton X-100 detergent and the solution is poured
into a homogenizer, which dissolves cell membranes but leaves the cell nuclei intact.
One then removes the nuclei and adds florescent DNA dye plus sufficient saline solu-
tion to produce a volume of liquid that can be read accurately on a measuring cylin-
der, or an immunocytochemical marker such as Neuronal Nuclear antigen. A bit of
agitation of the solution to distribute the nuclei evenly and then it just remains to
remove an aliquot for counting the nuclei using a florescent microscope. One need
not measure all of the nuclei if the agitation really has evenly distributed the nuclei
throughout the solution, and it takes only a small number of same-volume aliquots to
confirm the repeatability of counts.
The estimation of nuclei number is as accurate as is typical for stereology
(Herculano-Houzel et al., 2015; Neves et al., 2019). The upsides of this method are
several, including no need for fresh material or for histological processing, rapid pro-
cessing time (it can take just a day, depending on the amount of tissue to fractionate),
there are no concerns over issues such as tissue shrinkage (all nuclei are recovered),
the entire brain or region can be sampled, it is as accurate as other, more traditional
methods such as stereology, and, finally, counting can be automated using a flow
cytometer (Herculano-Houzel et al., 2015). The downsides are far fewer and primar-
ily concern the loss of the tissue in the fractionation process, should there be any
interest in the spatial distribution of cells. Although this method has not contributed
any of the data to the analyses that follow in Chapters 4–8, it is a method that might
enable the collection of data from large numbers of species, including from museum
collections (Burger et al., 2019). It does, however, require a better understanding than
is current of the ways in which neuron number may vary in function across species.
Furthermore, if this analysis is to be done with brain regions rather than whole brains,
it will also need some knowledge of brain anatomy.
Both the slicing of perfused brains and the counting of neurons enable some kind
of within-brain measurement of different regions, in terms of volume in the first case
and neuron number in the second. But the difficulty in obtaining material to measure
brain volume using serial brain sections or neuron counting means, however, that
relatively few brains have been put through these kinds of measurement procedures.
As can be seen from Table 2.1, nearly half of the brain data used in the analyses
described in Chapters 4–8 come from brain weights. Weighing brains is a much more
straightforward method than either perfusion or neuron counting and takes a lot less
time, even than counting neurons. It can also lead to much bigger sample sizes
through the use of already dead specimens. But there are a few important limitations.
For example, brains taken from dead animals will be in varying states of decay by the
time measurement occurs, as deterioration sets in immediately after death.

Standardizing the time between death and measurement will be difficult, if not
impossible (and often not reported). Furthermore, brain weights will depend on
whether they are weighed fresh or after drying, freezing, or having been kept in pre-
servative. The material, for example, used by Garamszegi et al. (2005a, 2005b) came
from birds weighed after freezing, whilst Crile and Quiring (1940) and Garamszegi
et al. (2002) used birds post mortem without freezing. It is unclear how comparable
such data are.
Finally, it is possible in some cases to measure brain dimensions (length, height,
and width) externally. For example, Kotrschal et al. (2013) used dorsal photographs
taken through the semi-transparent skull plate of newborn fish, while Gonda et al.
(2009) measured not only whole brains of nine-spined sticklebacks Pungitius pungitius
but also the bulbus olfactorius, the telencephalon, the optic tectum, the cerebellum,
and the hypothalamus, from digital photographs.
Largely because of the difficulty of acquiring live animals, the other half of the data
used in most of the analyses (Table 2.1) do not come from any of these methods.
Sacrificing a large number of animals for comparative analyses on brain size and
structure, always uncommon, is now even rarer. Although one may hope that scan-
ning, e.g. via Magnetic Resonance Imaging (Hecht et al., 2019), which is an even
more accurate and less invasive solution than those described above, will eventually
solve this problem, scanning methods are currently expensive, access to such facilities
is limited, and animals need to be transported to the scanner. Given that measuring
actual brain size is a significant barrier to collecting large amounts of data, one solu-
tion has been to measure the box in which it is carried, i.e. the skull. Indeed, this looks
like by far the most straightforward way to collect lots of data. However, it does
assume that skull volume is a close proxy of brain size.
2.2 Measuring skull volumes
The skull has significant advantages over the brain with regard to its measurement,
both biologically and logistically. For example, once the animal reaches adulthood,
the size of its skull does not change (at least in determinate growers), even as the ani-
mal ages or after it dies. The significant logistical advantages to measuring skull size
or capacity include the relatively straightforward preparation of the skull for meas-
urement (irrespective of time of death of the animal): one only needs to remove the
skin/hair/feathers and then the remains of the brain without concern for keeping it
for measurement. Alternatively, using skulls allows measurements of animals that
have been both recently and long dead, including now extinct species and fossils,
which may increase the sample size, both intra- and interspecifically, to a significant
extent (e.g. Aiello and Dunbar, 1993; Burish et al., 2004; Watanabe et al., 2019; Knoll
and Kawabe, 2020).
Both the internal capacity of the skull and its external surface have been used to
provide brain size measurements. External measurements are typically made using
calipers, whereby multiple lengths are measured (e.g. head height, length, breadth:
Møller, 2009), or by a tape measure wound around the head at a consistent position
(if this seems very old-fashioned, it is). This latter, ‘low-tech’ method has been used
primarily for measuring skull sizes within groups of humans (e.g. Mortimer et al.,
2003; Wang et al., 2019), and is especially quick and easy. However, using calipers or
a tape measure may make comparisons across species a little challenging, for instance
when those species have skulls that differ subtly or dramatically in shape, or include
features such as heavy brow ridges and other changes to the skull used for defence
(e.g. in those mammals whereby the males literally engage in head-to-head combat).
The crudity of using a tape measure may seem replaceable by using calipers, which
might allow greater precision of measurement of sub-components of the skull.
However, due to variation in the shape of skulls, especially through modifications
caused by major features such as eye sockets and jaw attachments, these external
measures may contain significant measurement error, both within as well as across
species. Perhaps more importantly, changes in external features may not map straight-
forwardly to differences—if any—of the brains inside (Figure 2.1).
By far the most common method for measuring skulls has been to assess their
internal capacity by filling the skull case with lead shot, glass beads, seed, and so on.
Although use of this method began in the nineteenth century, such measures are still
being taken today (e.g. Iwaniuk and Nelson, 2001, 2002, 2003; Marino, 1998, 1999;
Iwaniuk, 2001; Isler et al., 2008; Bininda-Emonds, 2000). And it is certainly simple.
The skull is filled with the lead shot (for example), which is then decanted into a cylin
der for measurement (or weighing: e.g. Isler et al., 2008). This is a hugely appealing
method due to its ease: pour the material into the skull, tamp it down a bit, top up as
necessary, and then tip out into the measuring cylinder. It can be a rather crude
method though, as described by Stephen Jay Gould in his Mismeasure of Man (1996),
and researchers reusing skulls by filling with one kind of material may find that they
tip out both their material and that from a previous measurer. But it does have con-
siderable advantages over all of the alternatives: it is very cheap, relatively quick to do,
the measuring apparatus can be transported to the skulls (which allows access to
specimens in geographically widespread collections), and it is relatively straightforward
Fig. 2.1 Left plate: The left-hand side of a puffin Fratercula sp. skull (with bill). Ruler for scale.
Middle plate: The left-hand side of the skull of a domestic cat. Right plate: Top-down photographs
of skulls of a puffin and a domestic cat (same skulls as in the photographs to the left and middle).
Source: Photographs by Steve Smart.
to assess repeatability in the measure. For example, Iwaniuk and Nelson (2001) esti-
mated a sampling error of 1.2 per cent, after measuring a single specimen 20 times,
although repeatabilities are best tested across multiple specimens. Similarly, Eisenberg
and Wilson (1978) measured four individuals twice, although they did not provide
error measures, while Ashwell reached accuracies of within 0.02 ml for volumes less
than 2.5 ml, and around 1.0 ml for volumes up to 100 ml, although it is not clear how
much retesting is required for this level of accuracy.
Other, more twenty-first-century, measures, such as x-ray photography (Madden,
2001) or computerized tomography (CT) scanning have also been used to measure
skull volume. For example, Madden’s x-ray photography involved firstly aligning the
x-ray source so that photographs of skulls were taken from a constant angle, and then
taking two orthogonal images, one lateral and one dorso-ventral, with a metal bar of
known length placed along the centre of the skull. The length, width, and height of the
skull cavity (rather than measures taken from the exterior) were measured and these
measures were used to calculate brain-cavity volume using the formula for an ellips
oid (π × length × width × height). A variety of authors have used CT scanning to do
something similar (e.g. Sakai, 2011; Logan and Palmstrom, 2015; Carril et al., 2016;
Macrì et al., 2019).
Although these modes of measurement are still relatively expensive and involve
equipment that is not readily moved from skull collection to skull collection, they are
beginning to transform paleoneuroecology (Knoll and Kawabe, 2020). They will also
become cheaper and more available, which will be important for the amount of data
required for a comparative analysis.
Whether measured inside or out, the skull is the most accessible and non-invasive
source of data for estimating the brain size of living animals. Be it by calipers or tape
measure, external measures of the skull can be taken from a live animal restrained
briefly (although one would want to calibrate such head size measures with skull
capacity and the size of the brain itself) and an animal need not be alive at all for
measuring its skull volume.
2.3 Variation in brain size measurement
Whichever method is used, it is important that measurements are accurate, and a

number of authors have been able to use the same individuals to compare the meas-
urements made by different methods. For example, Iwaniuk and Nelson (2002)
weighed brains from 24 budgerigars Melopsittacus undulatus, then filled the skulls of
those birds (the pieces were put back together after the brain had been removed) and
transformed these endocranial volumes with the value for density of fresh brain tis-
sue (1.036 g/ml). These authors found a strong correlation (Z = 7.77, n = 24, P < 0.01,
r = 0.94) between the two sets of measurements. Caution should perhaps still prevail,
however, as Taylor et al. (2006) found a considerable difference between data collected
from MRI scans of koala Phascolarctos cinereus brains and data collected from
cadaver specimens. Furthermore, Logan and Palmstrom (2015) encountered a similar
issue when they measured 40 grackle skulls by using CT scans, external linear meas-
ures, and then by filling them with glass beads (they also repeated these measure-
ments twice). These authors found high consistency between the two measurements
obtained by CT scan and bead volume, but not for the volumes calculated by external
linear measurements (skull length × height × width). The volumes calculated from
the linear measurements also did not correlate well enough with volumes calculated
from CT scans to resolve individual differences in endocranial volume. There was a
similar problem with bead filling. Logan and Palmstrom also found that males had
larger volumes than did females, irrespective of measurement method, and that there
was an effect of the bird’s age, which depended on sex (smaller volumes for immature
than for adult males, but larger volumes for immature than for adult females). This
variation among studies in the consistency of volumes measured by different tech-
niques suggests that more effort of this kind should be made, especially in terms of
making multiple measurements on the same animals.
But in many instances, the researchers are not in a position to make these kinds of
comparison because their data have come from the literature, rather than from brains
or skulls they have measured themselves. However, it is still possible to check that
data that come from different sources are comparable, and there have been more
checks of this kind because, due to the dependence of a comparative analysis on big
sample sizes (the more the better), many researchers have combined data from more
than one source. Sometimes this has meant a combination of the same kind of data
(e.g. brain weight, but just measured by different researchers). Alternatively, this has
meant combining different measures (e.g. brain weight together with endocranial
volume). How much the differences in measurement method contribute to variation
in the data is often not clear though (Sol and Lefebvre, 2000).
Although most authors who have combined data across sources have not had
access to different measurements on the same animals, sometimes they have been
able to combine data for individuals of the same species, which allows for some kind
of assessment of the differences in the two data sets. For example, Burish et al. (2004)
combined 138 brain and body weights from Portmann’s (1947) data with 28 species
from Boire and Baron (1994), which included 12 species also measured by Portmann.
But because Portmann reported his species brain components as ‘indices normalized
to the brain core’ of a member of the order Galliformes of comparable brain size, this
merging of data sets was not straightforward. The difference between 11 of the species
in common was resolved by taking an average for each, while for a twelfth species, the
value for the wild chicken was preferred over that for a domesticated chicken, given
data on telencephalic reduction in domesticated chicken (Rehkämper et al., 2003).
Others have looked for correlations between the data sets to be combined. For
example, Møller and Erritzøe (2015; see also Garamszegi et al., 2005b) found that the
data they measured were strongly correlated with data from the literature (e.g. F(34,
206) = 119.37, P < 0.001; r = 0.980 with Crile and Quiring, 1940; Portmann, 1947;
Armstrong and Bergeron, 1985). Similarly, Iwaniuk and Nelson (2001) found no
differences between the volumes of fresh brains (calculated by dividing brain weight
by the density of fresh brain tissue) and endocranial volume data taken from different
individuals of five species of waterfowl (paired t-test, t4 = 0.25, p = 0.81; Wilcoxon

signed ranks, t5 = 0.41, p = 0.69). Iwaniuk and Nelson (2002) and Boddy et al. (2012)
combined brain mass data together with endocranial volumes by calculating estimated
brain mass for a number of bird species for which they had endocranial volumes
(brain mass = endocranial volume × the density of fresh brain issue; there was no
difference between the values from the two measures in Iwaniuk and Nelson’s com-
parison of 81 species: t81 = −0.42, p = 0.68). Finally, others have used allometry to
impute missing data. For example, given that the scaling of the size of the neocortex
with whole brain size is known, then species without neocortex values can be included
in analyses if there are whole brain data (whether mass or endocranial volumes)
available for those species (e.g. DeCasien et al., 2017).
It appears, then, that when the data sets contain different individuals as well as
brain sizes measured by different methods, there is more evidence of compatibility
among the data sets than when different kinds of measures are taken from the same
individual. This is a little curious but may be because Logan and Palmstrom (2015),
in particular, applied a rather stringent criterion for accepting that the measures were
the same, which was the ability to identify individuals in the comparison. When they
simply looked for correlations between species averages, they did find quite a strong
relationship between CT scan data and bead volume data. But because there have
been so few cross-method comparisons, it is very difficult to determine what impact
they might have on the outcome of comparative analyses. However, if one or the other
is to be used as a proxy of an individual’s cognitive ability (for example), then it seems
important to know which is the suitable proxy and which is not.
2.4 Body size measurement
Whether the measurements of brain volume or weight or endocranial volume are

accurate, it is crucial to understand that body size is the best explanation for brain size:
the bigger the animal the bigger its brain is likely to be. And yet, with the exception of
the Møller and Erritzøe avian data (some 3,000 individuals), most of the comparative
analyses are based on data that come from a range of sources, comprise few individ
uals per species, and, frequently, the data on body size come from yet other sources
in the literature. I am not going to discuss the pros and cons of using absolute or rela-
tive brain size, or of some extrapolation of brain size such as the encephalization
quotients of Jerison (1973: EQ = brain mass ⁄ 0.12 × body mass2⁄3) or of Martin (1984:
EQ = brain mass ⁄ 0.059 × body mass0.76). Not only has this issue been discussed else-
where, but here the pertinent point is that body size data typically do not come from
the same individuals as those from which the brain size data were collected. Rather,
the body size data come from diverse literature sources. This requires adult body size
to be an invariant feature of a species, even though sexual dimorphism is an obvious
source of variation, at least for many mammals (see Chapter 7). Body size also
depends on the animal’s geographic location, the season, its age, and more. How this
affects our understanding of the variation in brain size is not clear. Iwaniuk and
Nelson (2001) were unusual in considering this issue, as for some of the museum
specimens (waterfowl) for which they collected endocranial volumes they also had
body weight data. They did not find a difference between the museum body weights
and body weight data for the same species cited in the literature (t40 = 1.55, p = 0.13),
but a single check on this potential source of variation is not enormously comforting.
2.5 Measuring parts of the brain

All of the discussion above, and almost all of the comparative analyses themselves, are
predicated on the assumption that brain size is a coherent entity and that selection
acts on it as a whole. But, as will become even more obvious in the examples described
in the following chapters, brains can, and do, vary a lot between individuals in the
same species and among species to little outward effect with respect to whole brain
size. For example, the hippocampus, a part of the brain that processes spatial infor-
mation, can be >10 per cent larger in some songbirds than it is in other species while
there is no detectable difference in whole brain size due to the difference in hip-
pocampal volume (e.g. Krebs et al., 1989; Healy et al., 1996). In such instances, adap-
tations will only be detected if brain regions can be delineated and quantified
separately. This can be done either by slicing or, increasingly, by functional scanning
(e.g. Hecht et al., 2019).
2.6 Summary
Although there are a variety of ways that have been used to measure brain size, each
with their advantages and disadvantages, most of the data used in the comparative
analyses to be described in Chapters 4–8 come from brain masses or from endocra-
nial volumes. Both of these are relatively easy-to-use methods, cheap, quick, and
entailing very little expert skill. They can be applied to lots of material and, for endocra-
nial volumes in particular, enable the use of museum specimens. CT scanning seems
likely to replace filling skulls with beads and seeds, but more data on the relationship
between the volume of the skull and the volume of the brain for the same individual
would be useful to confirm that the two measures really do provide the same result. The
real crunch comes in Chapter 3, however, when I consider the evidence for whether
brain size, however it is measured, is a good or useful proxy of cognitive ability.
3
The Intelligent Brain
Why is brain size and its variation interesting? For most of us the brain is a remark
able structure that somehow manages to enable us to run our lives, something that we
notice especially when things start to go wrong. But we are also especially interested
in its ability to enable us to be smart. From knowing that our brain is the seat of our
intelligence, it is very easy to move to thinking that brain size represents intelligence,
and that the bigger the brain, the more intelligent the animal (or person). Conversely,
we also flip this around to have a starting assumption that the more intelligent the
animal, the bigger its brain should be. This seemingly intuitive relationship is con
firmed by data from humans: we have big brains for our body size and we know that
we are very intelligent, some would say the most intelligent species on the planet. Ergo
big brains equal big smarts. We have of course also been interested in explaining just
what caused humans to be especially intelligent, with history littered with a multitude
of possible explanations for how and why we came to be so intellectually special.
Comparative analyses of brain size have almost entirely been built on just these
assumptions, i.e. that brain size is a good measure of intelligence, and that animals
with bigger brains are more intelligent. Many of these analyses begin with proposing
that a particular selective pressure sets a context in which the smarter the animal the
better it will do in getting food, mates, territories, and so on, ultimately gaining
greater fitness. In Chapters 4–8 I discuss how well those varied contexts explain vari
ation in brain size. However, in this chapter I want to take a step back and consider
the evidence that the correlations we might see between brain size and intelligence
indicate some sort of causal relationship. I begin then with an assessment of the
assumptions and evidence surrounding brains and intelligence. I follow that up by
outlining a framework for the steps needed to establish evidence for a causal associ
ation between intelligence and brain size, illustrated by using the relationship between
food storing and hippocampal enlargement in birds as a model system. Finally, hav
ing discussed whether brain size represents intelligence, I also consider how we might
go about determining whether variation in an ecological or social context might (or
might not) demand greater intelligence.
3.1 Brain size and intelligence
Whether brain size represents intelligence is a question that has been debated for
decades (e.g. Hodos and Campbell, 1969; Macphail, 1982) and is a debate that has
DOI: 10.1093/oso/9780199546756.003.0003
concerned differences both between species and differences within a species. But
although we can see that some species have bigger brains than others (leaving aside
the issues of measurement we met in Chapter 2), the really big challenge has been to
show that species differ in their intellectual abilities (henceforth cognitive abilities).
In practice, this is where the debate has been focused: are we able to determine that
one or other species (or one or other individual) is cognitively more gifted (more
‘intelligent’) than another? But even before we begin to make these kinds of com
parisons, we encounter another debate, in this case in terms of deciding what we
mean by ‘cognition’ itself.
3.2 What is cognition?
Pinning down a definition of cognition turns out to be not that easy. This was amply
demonstrated in 2019 when the editor of Current Biology (Geoffrey North) asked
11 researchers who work on and write about cognition to define cognition (Bayne
et al., 2019). The editor suggested it was time for some clarity and agreement on the
definition of cognition, and how to demonstrate its existence. The answers from
the 11 researchers had only some overlap in their definitions, some of which require
reasoning, or planning, while others specify the acquisition, processing, storing, and
retrieval of information (Shettleworth, 2010). Given the diversity of definitions from
the handful of experts, it is perhaps not surprising that many authors of comparative
analyses of brain size have tended to skip reference to actual cognitive abilities and go
straight to behaviours that they consider indicative of cognitive abilities, such as
increased range size (Chapter 4), innovation (Chapter 5), tool use (Chapter 6), mate
quality (Chapter 7), or sociality (Chapter 8). While this leap is understandable, to get
at the causal relationships between brain size and cognitive abilities we do need to be
able to measure, and then compare, cognitive abilities.
And here is where I need to come off the fence and provide some definition of
cognition that I find appropriate. Many authors use Shettleworth’s definition, which
covers all the ways that animals take in information through the senses, which they
then process, retain, and decide to act on (Shettleworth, 2010). This holistic approach
has learning and memory at its core, with perception at the information-acquisition
end and decision making at the output end. Historically, it was the inclusion of mem
ory to the already considerable body of work on animal learning that led to animal
cognition as a field.
That cognition involves a continuum from perception through to decision making,
in my view leads to the diversity in the definitions of cognition, and to the range of
vague adjectives that many append to ‘cognition’, such as high, low, simple, advanced,
or complex. What is often intended (almost always implicitly) is a comparison with
human abilities, such as counting or theory of mind, which are always considered to
be evidence of ‘high’, ‘advanced’, or ‘complex’ cognition. Such a view, though, disre
gards the stages of information flow that must be in place before such abilities can be
demonstrated. Given that this book is all about explanations for brain size, it is also
crucial to look at the role that the brain, and its component parts, plays in providing
The Intelligent Brain • 19
the functional structures for cognition. What can then be seen is that the information
flow that makes up the whole of cognition is also evidenced in connectivity within the
brain: information comes in from the sensory structures, goes to primary sensory
areas, and then gets mixed up in various regions, interacts with stored information
(memories), and the results go via motor systems to cause the animal to behave. Some
parts of the brain are described as primary sensory or motor regions, but there are
multiple parts of the brain that receive information from multiple other parts. The
hippocampus has been a prime example, dealing as it does with information called
spatial information, information that does not emanate from any one single sense
or modality.
Given that I consider cognition to be a process (and one rather like a flow diagram),
cognitive abilities can be specific to sub-sets of information that vary in kind, such as
spatial information or numerical information. Specifying the kind of information
acquired enables, at least in principle, the quantification of the amount of information,
and the quality of the information, and if enough is known of the function of relevant
brain regions, the processing of that information can also be identified. This means
that a brain could evolve to be better at processing some kinds of information, but not
necessarily all kinds of information. And similarly, some parts of the brain could be
used primarily, or solely, to process specific kinds of information, while others might
have a hand in processing information from lots of different sources. As noted above,
learning and memory are essential mechanisms that enable the processing of informa
tion and can also be specific, such as in spatial learning and memory. Of all the variety
of cognitive abilities, identifying when an animal has learned and remembered some
information has been made possible and now often quite straightforward by decades
of ground-truthed, robust methodological techniques.
Characterizing perceptual capabilities has also been enabled by rigorous method
ologies. Near to the output end of the flow diagram, the picture is much more mixed,
however. For abilities such as timing and counting, perhaps because work on rodents
has been to the fore in both, identifying and quantifying such abilities in a range of
animals is, again, almost straightforward. However, for abilities such as decision mak
ing, planning, and theory of mind, there is much yet to be developed, not least because
the focus has been so strongly on humans and other primates. As researchers are
increasingly attempting to address questions concerning these abilities in more and
more species, the data being produced from this growing taxonomic range suggest
that this narrow focus has indeed been far too narrow. However, with much of the
methodology still appropriate for humans only, or other primates at a pinch, and
therefore very far from the comparative wealth required, for the purposes of this
book, I will not consider comparative data on cognitive abilities from this far end of
the flow diagram.
3.3 Comparing cognitive abilities
One of the first places to look perhaps for differences in cognitive abilities is in
the presence or absence of a particular ability. Although early work tended to
underestimate animal capabilities, and the absence of many cognitive traits was often
presumed, there is a lot of comparative psychology that provides evidence for the
presence of specific cognitive abilities in non-human animals (e.g. whether animals
can count or time intervals: Henderson et al., 2006; Davis and Pérusse, 1988; Dacke
and Srinivasan, 2008; see Healy, 2019 for a review). Such demonstrations have
increased as researchers have become more attuned to the relevant biology of their
chosen species. For example, testing may need to take place at warmer temperatures
than is usual in a laboratory (e.g. 29°C for testing reptiles: Wilkinson et al., 2007) or
take place outside the laboratory altogether (e.g. testing spatial cognition in wild,
free-living hummingbirds: Healy and Hurly, 1995). In the course of this work over
recent decades, we have repeatedly seen that most animals are far more capable than
once thought, not only with classic examples of animal stupidity—such as the three-
second memory capacity of goldfish—increasingly going by the wayside (although
this one has actually long gone for goldfish, e.g. Rodriguez et al., 1994) but also that
wild, free-living hummingbirds can count (Vámos et al., 2020), male bowerbirds use
visual illusions to entice females to their bowers (Kelley and Endler, 2012), while pos
sums Trichosurus vulpecula (Sargisson et al., 2016) and long-tailed bats, Glossophaga
soricina (Toelch and Winter, 2013) can time intervals.
Demonstrating the presence of a cognitive ability can be challenging, and even
more so if one is really looking for a human-like ability in a non-human species. But
with an increasing ecological and behavioural knowledge of a wide range of taxa,
from observations and experiments in the wild and the laboratory, we now have a
much more detailed and nuanced understanding of cognitive abilities in animals, at
least in terms of presence/absence.
But if demonstrating the presence of a cognitive ability has proved challenging,
then demonstrating variation in said cognitive ability leaves the presence/absence
question looking like child’s play. This is not least because of all the factors that can
impact on the behaviour of an animal, even when one is explicitly attempting to test
its cognitive abilities (see also Figure 5.5). In the endeavour to bring as much stand
ardization as possible to cognitive tests (and thus remove or greatly limit the impact
of extraneous variables), almost all tests of cognitive abilities have taken place in a
laboratory setting. Here the experimenter can hold constant all sorts of environmen
tal factors that might impinge on an animal’s performance (e.g. temperature, lighting,
humidity, time of day), and the experimenter can try to standardize a variety of
aspects of the experimental context (e.g. testing in boxes of the same size, shape, or
colour, providing food rewards of the same size and energetic value, depriving ani
mals of food prior to a food-rewarded test for the same length of time, and so on).
Further standardization is sought by training animals to perform a task (any task is
likely to have at least some unfamiliar components) by providing all subjects with
either the same number of trials to acquisition, or requiring all subjects to reach the
same minimum level of performance (termed the criterion), regardless of the number
of training trials. All of these are put in place to provide a level playing field such that
any variability in performance that results can be explained by variation in the ability
of the individual (Griffin et al., 2015).
But training and testing animals in this way can be hugely restrictive. The number
of species (and individuals) that can be kept (usually in captivity), trained, and tested
is very small in comparison to the number that are required for a comparative ana
lysis of the kind used to look at brain size. Furthermore, the kinds of tests used in
these sorts of studies (often originating in experimental psychology) are also typically
standardized to the point that they may or may not be very species-appropriate for
most animals. For example, it is common to use an eight-arm radial maze to test spa
tial learning and memory in animals. This apparatus has a central arena into which
the animal is placed, with eight (traditionally, but it can vary) arms down which the
animal moves to find food (or not) at the end of each arm. This is a relatively straight
forward apparatus in which to test rodents (although one must be careful to ensure
they do not climb out as the arms are typically open-topped) but can cause problems
when testing other species (e.g. the size of the arena for pigs: Laughlin and Mendl, 2000,
or the issue of escape when testing birds and other flying animals: Bond et al., 1981).
However, at least the radial maze has the animal performing in a way that might seem
relatively natural. But tests such as delayed-matching-to-sample tests are commonly
used to test visual or spatial memory in birds, such as pigeons (for an example see
Figure 3.2). In this method, the animal is presented with a sample, to which it
responds, and after a delay it is presented with a choice of options. The correct choice
typically results in a food reward. This method, whereby the options are generally
presented as images on touchscreens or as options such as coloured lights associated
with pressing/pecking levers, requires a considerable degree of training, not least
because it is far from being ‘natural’.
This combination of restrictions on testing has meant that, until recently, the range
of species for which there was any information on their cognitive abilities was very
narrow. And the problems do not end with the narrowness of the taxonomic range of
species tested, because even when the playing field is apparently level in these labora
tory experiments, variation in motivation to do the task, the previous experience of
the animal, and more may all affect performance anyway (Rowe and Healy, 2014).
These problems have been especially challenging for comparisons among species, as
it is unlikely that different species will view the same task in the same way, in terms of
saliency or motivation (Macphail, 1982).
I will use the example of the search for cognitive differences between food-storing
and non-storing songbirds to illustrate not just these difficulties but also that mean
ingful variation in cognitive abilities can indeed be identified. It might just take a bit
(actually, quite a lot) of work. The food-storing example also serves to illustrate how
to show a causal relationship between the brain and cognition, and a cognitive
demand imposed by a specific context (in this case, the ecology of the animal).
3.4 Food storing, spatial cognition, and the hippocampus
Food storing is a rather unusual behaviour because most animals, when they encoun
ter a rich patch of food, eat as much as they can, and store the food on their body as
fat. But a handful of species eat a little of the food, and then remove the remaining
food to one or more locations. Some of these animals store their food in larders (e.g.
a number of rodents), piling up all the excess food in one location, such as their burrow,
which they then have to defend. On the other hand, when small (around 9 g) British
songbirds, such as marsh tits and coal tits (Poecile palustris and Periparus ater), find
excess food, they take each piece of food away and hide it in separate locations
(Perrins, 1979). Because these birds can store many food items (100s–1000s, typically
seeds), for a time there was a debate as to how they relocated them. That they might use
memory came from an experimental study in which marsh tits in Wytham Woods,
just outside Oxford, United Kingdom, were provided with radioactively labelled seeds.
These labelled seeds enabled Cowie et al. (1981) to locate the birds’ stores and then
add other seeds hidden at 10 cm and 100 cm from the bird’s own stored seed. Of 93 of
the 121 seeds where one seed only had gone between the experimenters’ three-hourly
visits to check seed survival, it was the stored seed that had disappeared. Furthermore,
the birds collected their own stored seeds within eight hours but took more than
14 hours to remove the seeds hidden at 10 cm, while seeds hidden at 100 cm survived
for just over 20 hours of daylight. These data suggested that, however unlikely it might
seem, such small (and therefore small-brained) birds could remember the locations of
their stores. Confirmation was provided by tests in the laboratory (e.g. Sherry
et al., 1981; Sherry, 1984; Healy and Suhonen, 1996). In these experiments, birds
stored food in specified locations around an experimental room and then some time
later (minutes, hours, or days) had the opportunity to retrieve them. The number of
locations in which the birds searched were many fewer than the number they would
have searched if they had looked at random.
These data showed that retrieval of stored food was done, at least in part, by the
memory of the locations of the stored food. The next question was whether this
demand for remembering multiple locations of stored food had led to an adaptation
in spatial memory. If yes, then because the closely related species blue tits Cyanistes
caeruleus and great tits Parus major do not store food, they should face less demand
for the kind of spatial memory capacity that food-storing species have, all else being
equal (and their ecologies are, otherwise, rather similar). These tit species provided
an opportunity to test whether contextual variables were the sole explanation for dif
ferences in cognitive performance between species (Macphail, 1982) or whether, as
was becoming increasingly discussed (Domjan and Galef, 1983; Kamil, 1987), that
memory might vary due to the animal’s ecology.
The prediction from this hypothesis was that, in tests of spatial memory, food
storers should do better than the non-storers. In one of these tests, food-storing
black-capped chickadees Parus atricapillus and non-storing dark-eyed juncos Junco
hyemalis were presented with four blocks of wood on a wall of an experimental room
(Brodbeck, 1994). Each block was painted in a different colour/pattern and had a hole
drilled in it, which was covered by a piece of black Velcro (see schematic in Figure 3.1).
A bird sitting on the perch immediately below could access the hole by lifting the
Velcro. One of the blocks contained half a piece of peanut. In the search phase of a
trial, a bird searched the blocks until they found the food, which they removed and
(a) (b)
Fig. 3.1 A schematic of an experimental design to investigate cue use. In this example, food-
storing and non-storing birds were presented with four differently coloured blocks, each of
which had a flap covering a hole in which food could be hidden. In the first search phase (a) of
this example, the bird would have found a piece of peanut in the yellow block. It was allowed to
eat that peanut for 30 seconds and then it left the room. Five minutes later, in the second phase
when the bird returned to the room (b), the yellow feeder had been switched with the feeder
with the blue stripes. None of the feeders contained food. The researchers looked to see which
feeder the bird visited first. The yellow feeder bears the appropriate visual cues while the feeder
with the blue stripes is in the appropriate location.
ate for 30 seconds before being returned to their home cage. Five minutes later, the
bird was allowed back into the room to search the blocks for the remainder of the
piece of peanut. In the five-minute retention interval the experimenter had emptied
and switched the rewarded feeder with one of the unrewarded feeders. If the bird
remembered which feeder contained the peanut, it was faced with looking in the
feeder that looked the same as before, or in the feeder that was in the same spatial
location as before. The food storers preferred to look in the feeder in the correct loca
tion, while the non-storers were as likely to look in the feeder with the correct colour/
pattern as they were to look in the feeder in the correct location. Food storers, then,
seemed to prefer to use spatial rather than visual cues while non-storers did not (see
also Clayton and Krebs, 1994b). But this kind of preference for one cue type or
another can have multiple explanations and is also likely to be dependent on experi
ence and context. For example, non-storing great tits prefer to use spatial cues if they
visit a reward location on multiple occasions (Hodgson and Healy, 2005). On the
other hand, other non-storers sometimes behave like food storers. For example, after
a single visit to a rewarding location, rufous hummingbirds Selasphorus rufus (which
do not store food) prefer to use spatial cues when returning to a choice of flowers (e.g.
Healy and Hurly, 1995; Hurly and Healy, 2002). And even food storers will use colour
rather than spatial cues, when colour cues provide more useful information (LaDage
et al., 2009b). As there is now an abundance of evidence that animals pay attention
to those cues that are most reliable or stable (e.g. Biegler and Morris, 1996; Hurly
et al., 2010), and that they may develop a hierarchy of preferences for the cues they
use and when they use them, the initial difference between food storers and non-
storers in cue preference is not necessarily due to whether or not they store food.
In order to reduce the training required for wild-caught birds to be tested in the
laboratory, the first experiments to test memory rather than cue use were designed to
be analogous to food storing rather than of the more traditional kind of test used for
model organisms such as rodents or pigeons. Because non-storing birds do not store,
great tits (non-storers) and coal tits (storers) were tested with an experimental design
called ‘window shopping’. In this experiment, very much like that of Brodbeck above,
food was hidden in a hole in a wooden block under a cloth flap (Shettleworth and
Krebs, 1986). The additional feature, however, was to cover the hole under the cloth
with a transparent plastic cover. Multiple blocks were distributed around artificial
‘trees’ in an experimental room. In the first phase of a trial, the birds searched around
the blocks by lifting the cloth covers. When they lifted a cloth cover, they could see
whether the block contained food, but if it did then the plastic cover prevented them
from removing the food. When the bird returned to the room in the second phase of
a trial, this time when it lifted the cloth cover, it could retrieve the food because all the
plastic covers had been moved to make the food available. By observing which blocks
the birds visited, it was possible to see that the storing and non-storing species dif
fered in how they discriminated between remembered locations (Krebs et al., 1990).
But the differences were not as compelling as first expected. Indeed, the differences
between the two species in their memory was really rather subtle: having visited sites
with and without food in the first, searching phase of a trial, on their return for the
second, recovery phase, both food storers and non-storers were actually equally good
at relocating the sites they had seen to contain the food. Where they differed was in
their ‘mistakes’: the food storers’ mistakes were to visit new sites that they had not
visited in the search phase (see also rufous hummingbirds: Hurly, 1996), while the
mistakes the non-storers made were to return to sites they had earlier seen to be
empty. And there were also several alternative explanations for the differences in
‘retrieval’ performance, including that the great tits found it harder to discriminate
rewarded from unrewarded sites than did the coal tits. This was plausible if, for
example, the coal tits had spent more time inspecting the rewarded locations than the
great tits (although these data were not collected).
A variety of experiments followed, including an attempt to repeat the window-
shopping experiment but without the plastic windows (Healy and Krebs, 1992a). In
this experiment, the birds were provided with one of the seven blocks put out in the
room rewarded, and the retention interval between the first (search) and second
(retrieval) phases varied from one minute, to 30 minutes, to 24 hours. However, in
this experiment, if anything the non-storing blue tits were better at remembering
rewarded locations as the interval increased. More traditional tests of spatial memory
also failed to produce compelling evidence for a difference between food storers
and non-storers. For example, although food-storing black-capped chickadees
(a)
(b) delayed spatial match-to-sample (c) delayed spatial non-match-to-sample
sample phase sample phase
retention interval retention interval
choice phase choice phase
correct choice correct choice
Fig. 3.2 (a) Photographs showing a great tit in its home cage undergoing a spatial delayed-
non-matching-to-sample trial. In the left-hand photograph, the bird is viewing a single square
in the sample phase on a touchscreen. In this phase, when the bird pecks at the square, the
square disappears, the bird does not receive reward and the retention interval ensues. In the
right-hand photograph, the bird is viewing two squares in the second, choice phase. If the bird
pecks at the square that is not in the same location as that occupied in the sample phase, the
squares disappear and the reward (a small piece of peanut) is delivered down the tube onto a
small tray to the left of the screen. An intertrial interval with no images on the screen ensued
until the next trial began. (b) A schematic of the experimental design for a spatial delayed-
match-to-sample experiment. In this design, the subject is rewarded for responding to the
image in the same location as that occupied in the sample phase. (c) A schematic of the experi
mental design for a spatial delayed-non-match-to-sample experiment. In this design, the sub
ject is rewarded for responding to the image that was not in the same location as that occupied
in the sample phase.
outperformed dark-eyed juncos Junco hyemalis on a spatial-non-matching-to-sample

task (see example Figure 3.2), juncos outperformed the chickadees on a continuous
spatial alternation task (in this task the animal needs to respond alternately to the left
image or object and then to the image or object on the right to gain reward: Hampton
and Shettleworth, 1996b).
All in all, after a handful of experiments it was already looking more difficult to
demonstrate the spatial cognitive advantage of food storers than had been antici
pated. After all, food storers store food a lot, hiding and retrieving hundreds or
thousands of food items across a year. Surely the memory ability required to make
this behaviour successful was considerably more than the memory required by
non-storers? And surely it was appropriate to test memory abilities in these animals
using the naturalistic kinds of experiments described above? Furthermore, surely
using species that were very closely related reduced to a very great degree other fac
tors (e.g. the kinds of food appropriate as rewards, hiding food under cloth, and so
on) that might have obscured memory differences?
One elegant experiment, however, suggested that the difference might lie in being
able to remember spatial locations for longer (Hilton and Krebs, 1990). In that experi
ment, tits were presented with a version of an eight-arm radial maze that used light to
form the ‘arms’. A bird flew into a dark room illuminated by a single light set on a
platform in the middle of the room. Once it had settled on the perch on the platform,
the room lights were switched on and the bird could see that it was surrounded with
an array of eight platforms, each bearing a container of food. The bird could then fly
to any of these platforms and eat the food it found there. Once the bird was sitting on
one of these radial platforms the lights were turned off except for that on the central
perch. The bird then flew back to the centre and the process was repeated (Hilton and
Krebs, 1990). The birds’ memory was tested in this task through imposing a delay
between the birds visiting some of the platforms but not all of them. The delay varied
from 30 seconds to 24 hours and after each interval the food storers tended to do bet
ter, an effect that was, however, pronounced only after 24 hours. It appeared that the
non-storers were just as good as the storers at remembering locations of food, at least
when the time over which they had to remember the information was very short.
However, the food storers gained an advantage when the time to remember was
stretched out to 24 hours (in the wild coal tits and similar food storers store and
retrieve food across days and weeks). Although the authors interpreted these data at
the time as showing no spatial memory benefit to a larger hippocampus, in light of
data that were to come (see below), maybe there was a bit more to these data than was
first realized.
Nonetheless, the lack of compelling evidence for spatial memory differences
between storers and non-storers was very much as Macphail had suggested: when all
the extraneous variables that could affect spatial memory were much reduced, and
the animals were tested in species-appropriate ways, there were no differences in cog
nition between storers and non-storers, whatever their ecological differences might
have suggested. It is a bit ironic, then, that it was the employment of a test that was
very much more traditional than almost all those that had gone before, that suggested
that the cognitive advantage of storers over non-storers was more subtle than
expected. In particular, it was not so much that storers had better spatial cognition
overall, but rather that food-storing success required very specific spatial abilities.
The first such ability was that food storers remember spatial locations more accurately
than do non-storers. This reprises the data from Cowie et al. (1981) where wild, free-
living marsh tits were accurate to within less than 10 cm in retrieving their stores. In
a laboratory test, great tits, blue tits (both non-storers), and coal tits were presented
with a spatial delayed-matching-to-sample task on a touch screen (Figure 3.2;
McGregor and Healy, 1999). In this task, the bird received a reward for touching the
sample image. After a 30 second delay with a blank screen, a bird was presented with
a choice between an image in the same location on the screen (for which they were
rewarded a small piece of peanut) and either one, two, or three distracter images,
which varied in whether they were ‘near’ or ‘far’ from the rewarded location. In the
choice phase, if the bird chose the image for which they had been rewarded in the
sample phase, they received another reward. In spite of the small scale of a touch
screen, the food storers were more accurate in their choices than were the non-storers,
irrespective of the presence of one to three distractors. A second feature of the birds’
performance that emerged from this experiment, which was consistent with the
Hilton and Krebs’ data described above, was that food storers could wait longer to
make a choice (sometimes the birds took a while to peck at the images in the choice
phase) and still make more accurate choices than did the non-storers, irrespective of
the number of distractors present during the choice phase. This suggested that food
storers could remember locations for longer than could non-storers.
In a subsequent experiment, three components of spatial memory (duration,
capacity, and accuracy) were explicitly tested in coal tits and great tits, this time using
a delayed non-matching-to-sample task, again presented on a touch screen (Biegler
et al., 2001). Birds were presented with images in the sample phase that differed in
number (one, two, three, or four) or in distance (near or far) from each other. Once
the bird had touched all the images, there was a delay that began at 1 s but was increased
by 0.7 s in the following trial if the bird made a correct choice. If the bird made an
incorrect choice, the delay (the retention interval) was reduced by 0.3 s. In this way,
the retention interval was titrated until birds maintained a level of performance at
around 70 per cent correct. This allowed some estimate of the duration over which
the birds could remember the locations of the sample images. Although the experi
mental conditions could not have been much further from reality for the wild-caught
birds tested (the durations were in the order of seconds, the number of items were a
small handful, and the accuracy was tested over a few centimetres), the outcome of
the experiment was clear: food storers outperformed non-storers only with regard to
the duration over which they could remember the locations of images. Indeed, even
when they had to remember only one item, food storers could remember that single
item for longer than could the non-storers (Figure 3.3a). The memory advantage was
also specific to the memory for spatial location. In an unpublished second experi
ment, in which the birds were tested for their ability to remember which picture(s)
they had just seen, food storers and non-storers performed equally well, irrespective
of the number of images in the sample phase (two–four), their proximity, or the dur
ation over which the birds could maintain 70 per cent correct (Biegler and Healy
unpub. data; Figure 3.3b).
Finally, these were the data that strongly supported the original hypothesis: that the
spatial demand imposed by the need to remember the locations of stored food is
associated with an ability to remember locations for longer. But they were reached
only after considerable effort and some ingenious experimental designs. Why did the
search continue in the face of multiple instances where the data rarely offered com
pelling support? The search continued because all along, the researchers suspected
that genuine differences between storers and non-storers lay in their brains. That
Storer Non-storer
(a) (b)
25 25
Final retention interval (seconds)
Final retention interval (seconds)

20 20
15 15
10 10
5 5
0 0
1 2 3 4 1 2 3 4
Number of images
Fig. 3.3 (a) Data from a spatial delayed-non-match-to-sample experiment with coal tits
(storers, closed circles) and great tits (non-storers, open circles), showing the average final reten
tion interval (± s.e.) reached by each group as the number of images increased. (b) Data from a
delayed-non-match-to-sample experiment with coal tits (storers, closed circles) and great tits
(non-storers, open circles), showing the average final retention interval (± s.e.) reached by each
group as the number of images increased. In this experiment the images were coloured pictures
and the birds had to choose the new picture that appeared in the choice phase.
Source: (a) Biegler et al., 2001; (b) Biegler and Healy, unpublished data.
suspicion, and thereby the belief that there really should be a cognitive difference
between food storers and non-storers, was confirmed by the evidence that food
storers have a much bigger hippocampus than non-storers.
3.5 The hippocampus and spatial memory in food storers
By the 1980s it was pretty clear that the hippocampus in mammals was a very import
ant part of the brain for spatial memory (e.g. O’Keefe and Nadel, 1978; Morris et al.,
1982). It also seemed like a good place to look for spatial memory in the food-storer’s
brain. If food storing is spatially demanding, and spatial demand causes changes in the
brain, then perhaps the hippocampus would be bigger in food storers than in non-
storers. Support for this hypothesis was shown via lesions to the hippocampus of black-
capped chickadees, which showed that this part of the brain was involved in spatial
memory for stored food, as lesioned birds were unable to relocate food they had
recently stored. Lesioned birds were still able, however, to find food in locations that
they had learned to associate with a colour label (Sherry and Vaccarino, 1989).
These lesion data led to two comparative analyses of hippocampus size, both of
which found that food storers had bigger hippocampal volumes than did non-storers
(Krebs et al., 1989; Sherry and Vaccarino, 1989). Importantly, differences in hippo
campus volume between storers and non-storers did not, however, play out in terms
of differences in whole brain size between these species. This observation is extremely
relevant for much of what follows in the coming chapters, namely that ecological
selection can shape the size of different parts of a brain, without necessarily the whole
brain size changing in response (i.e. revealing) to that selection.
There is a final, necessary link to complete the argument that food storing creates a
demand for greater spatial memory and so causes a change in the size of the part of
the brain that provides that capability (i.e. the hippocampus, albeit not exclusively).
That link is to demonstrate not just that damage to that part of the brain means that
the storer cannot remember where to look for its stores, but also that these hippocampal-
lesioned animals cannot solve other spatial memory tasks. Hampton and Shettleworth
(1996a, 1996b) did just this by showing that damage to the hippocampus in black-
capped chickadees did, indeed, impair the birds’ performance on spatial memory
tasks (a spatial non-matching-to-sample task very like that used by Biegler et al.,
2001).
It is important to recognize that even these data do not show that a bigger brain
means better cognition. They show that storing needs spatial memory, that storers
can remember spatial locations for longer than non-storers, and that storers have a
bigger hippocampus, which when damaged causes storers to perform badly on spatial
tasks and at retrieving their stores. In fact, the clincher would be the demonstration
that individual storers with a bigger hippocampus outperform storers that have a
smaller hippocampus on spatial memory tasks. But the story is now pretty strong
without this last piece of the puzzle. It is also important to realize that to get to such a
compelling conclusion required more than the original correlation between food
storing and hippocampus size. Yes, it was assumed that food storing demanded spa
tial memory, and yes, the hippocampus was considered to be the prime brain region
dealing with spatial information. But it was identification of a relevant selective pres
sure (i.e. one that clearly requires cognition), together with appropriate experimental
design and the relevant neuroscience, that allowed us to show that natural selection
can act on both the brain and its ability to remember information.
3.6 The relationship between whole brain

size and cognition
Food storing, duration of spatial memories, and hippocampus size provide a robust
model for approaching the question of the relationship between cognitive ability and
a given part of the brain. But while food storing appears to have driven an increase in
hippocampus size, the need for increased spatial memory has not led to a larger brain
in food storers. Perhaps this is because the cognitive demand is met just by the
increase in hippocampus size, without a need for other brain regions also to increase.
Alternatively, perhaps the advantage of the larger hippocampus has allowed small
decreases in other brain regions (as yet unidentified).
What cognitive demand or cognitive ability might have led to much more substan
tial changes, such that they are evident in an increase in the whole brain? Executive
function, a rather nebulous term for a suite of cognitive and behavioural characteris
tics that include working memory and inhibitory control, is thought to enable a range
of abilities such as attention, organization, planning, and self-monitoring. Such a cog
nitive attribute should therefore rely on a number of brain regions. And indeed,
domestic dogs from breeds with bigger brains (estimated from a scaling function on
data from breed averages) did perform better than did dogs from breeds with smaller
brains on tasks that tested executive function (exemplified in one experiment by self-
control: Horschler et al., 2019). But although these data show a correlation between
executive function and brain size, they cannot show that dogs get better at executive
function when their brains get bigger. This is because correlation does not mean caus
ation. Both the changes in executive function and the changes in brain size could, in
theory, be caused by a third unknown factor. This is an issue with all correlations
between cognitive tests and brain size of course (and anything else for that matter),
regardless of whether the correlations come from within- or between-species com
parisons, or from comparing performance on a single task or a range of tasks. To truly
make a causal link between cognitive ability and brain size, one has to do experi
ments, and—more importantly perhaps—look at the brain itself.
Fortunately, it is becoming increasingly possible to do just this, such as collecting
MRI data (from which the size of the whole brain and of different regions can be
measured) from dogs. Hecht et al. (2019) examined dogs from a range of breeds that
have been selected for a variety of behaviours (e.g. boxers for fighting and guarding,
cocker spaniels and retrievers for bird retrieval, old English sheepdogs for herding,
and Cavalier King Charles spaniels for companionship and silly ears). These
researchers found considerable volumetric variation in various regions of the brain
among these breeds (Figure 3.4). Particularly promising was that variation in size of
particular regions correlated rather well with the roles those regions, grouped together
in networks, are known to play in terms of vertebrate cognition.
For example, border collies have relatively larger regions in the reward system
(e.g. the nucleus accumbens and cingulate gyrus; no surprise for collie owners per
haps). Cavalier King Charles spaniels, on the other hand, have relatively larger social
action and interaction regions (e.g. the premotor and the prefrontal cortex), while old
English sheepdogs have relatively larger olfactory and visual regions (e.g. the olfac
tory peduncle and primary visual cortex), in spite of their excess facial hair (or per
haps because of it?). Perhaps unsurprisingly, Hecht et al. also found a lot of variation
in the relationship between body size and brain size, as might be expected given the
variety of morphological attributes, as well as behavioural traits, that humans have
been selecting in dog breeds. Indeed, these data yielded all sorts of interesting correl
ations, including variation in the relationship between external head shape and brain
shape (smaller dogs have more spherical brains). But it is the promise of these data for
gaining access to measuring brain size and functional regions within a brain that are
especially exciting. These data also underline the importance of regional variation
Movement, eye movement,

Mesolimbic reward system
Social action & interaction
Fear, stress, & anxiety:

Olfaction & gustation
& spatial navigation
Olfaction & vision

Drive & reward:
Component 1
Component 2
Component 3
Component 4
Component 5
Component 6
HPA axis
–3 –2 1 2
Dachshund
German short-haired pointer
Maltese
Keeshond
Silky terrier
Yorkshire terrier
Golden retriever
Labrador retriever
Beagle
English pointer
Basset hound
Border collie
Welsh corgi
Old English sheepdog
West highland white terrier
Springer spaniel
Cavalier King Charles spaniel
Weimaraner
Greyhound
Whippet
Doberman pinscher
Bichon frise
Miniature schnauzer
Boston terrier
Bulldog
Wheaten terrier
Standard poodle
Lhasa apso
Siberian husky
Boxer
Fig. 3.4 Relationship between morphologically co-varying regional brain networks and
phylogenetic tree for domestic dogs. The circles indicate factor loading.
Source: Hecht et al., 2019, figure 4.
within brains, noted in Chapter 2 and which will be a recurrent theme throughout
this book.
Promising as this methodology is, however, these data remain correlational even if
the selective breeding that generated the breeds is a kind of experiment. Ideally, there
would be some kind of manipulation (e.g. lesions as done with the black-capped
chickadees, but one that is less likely to be conducted in dogs) in which the target
brain region is removed or reduced, and where the target behaviour changes con
comitantly. But being able to see into brains of living animals in this way could make
it much easier to see what is happening in the brain when the animal is doing a cogni
tive task, or behaving in a way that we might think requires extra cognitive abilities.
We might also get a handle on how much of the brain is involved, and whether we
should be looking for changes in whole brain size.
3.7 Cognition and fitness
The final piece in the jigsaw that is the demonstration that cognition/brain size is an
adaptation is to show that increased cognitive abilities (and increased brain size) lead
to increased reproductive success (Figure 3.5). For example, males with better spatial
cognition may be able to remember more locations of reproductive females (or are
better at remembering where their female partners are to be found) and, thereby, are
more likely to father more offspring (cognition as a sexually selected trait: see also
Chapters 4 and 8). Although it has long been assumed that being smarter must be
beneficial, for selection to favour increased cognitive abilities, those increased abil
ities (and their associated brain size) must increase reproductive success. Put another
way, being smarter than you need to be will be selected against: neurons and cognitive
processing do not come for free.
However, because most researchers interested in cognitive performance test cogni
tive abilities in the laboratory, and because demonstrating fitness benefits (ideally, the
number of offspring that survive to reproduce) is never easy, possible fitness benefits
of better cognitive performance have only begun to be examined relatively recently.
But there is some suggestion that better cognitive abilities can translate into fitness
benefits. For example, male New Zealand robins Petroica longipes that were better at
remembering which one of eight wells contained a mealworm were more likely to
raise two or more independent offspring per nest than were more forgetful males
(Figure 3.6; Shaw et al., 2019). Additionally, female Australian magpies Cracticus tibi-
cen dorsalis that generally did better on several cognitive tasks (inhibitory control,
associative learning, reversal learning, and spatial memory) produced more offspring
Cognitive ability
(e.g. spatial memory)
Exclude
attention, Neural damage
perception, (e.g. hippocampal
motivation lesions)
Fitness
(more offspring surviving to Performance on relevant tasks Neural activity
reproduction (e.g. tests of spatial memory) (e.g. in
hippocampus)
Size of brain
region
(e.g. hippocampus)
Selective pressure
(e.g. food storing)
Fig. 3.5 Diagram of the components needed to demonstrate a causal relationship between a
selection pressure (in this example, food storing), a cognitive ability (spatial memory), and size
of the relevant brain region (hippocampus).
(a) (b)
18 12
>2 >2
Fledglings per nest
10 6
1 1
26 19
0 0
20 18 16 14 12 10 20 18 16 14 12 10
(c) (d)
15 9
Independent offspring per nest
>2 >2
13 9
1 1
26 19
0 0
20 18 16 14 12 10 20 18 16 14 12 10
Male memory performance Female memory performance
Fig. 3.6 The average (± s.e.) memory performance (measured as the number of lids opened
during test trials) associated with the number of fledglings per nest (A and B) and the number
of independent offspring per nest (C and D) for males (A and C) and females (B and D).
A smaller value (on the right of each x axis) represents a more accurate performance in the
memory test (i.e. fewer lids opened). The number of nests associated with each mean is shown
above each point.
Source: Shaw et al., 2019. Reprinted with permission from Elsevier.
(Ashton et al., 2018). For neither case yet are there data on whether the production of
a greater number of offspring resulted in a greater number of offspring surviving to
reproduction. Nor how these smarter adults fared in their own subsequent lifetime
reproductive success. This is a concern because Cole et al. (2012) found that great tits
that were better problem solvers laid more eggs but then fed their offspring less often
and were more likely to desert their nests (see Chapter 5).
Furthermore, grey mouse lemurs Microcebus murinus that performed well on
a spatial memory task were not more likely to survive (let alone reproduce) than
were more forgetful grey mouse lemurs (Huebner et al., 2018), and male spotted
bowerbirds Ptilonorhynchus maculatus that did well on a reversal learning task and
on a spatial memory task did not attract more mates (Isden et al., 2013).
There are still far too few data at this point to make much comment about these
mixed results. But it might be useful to recognize that there is more than one way to
achieve reproductive success, even in a single species: an individual might be smarter
than conspecifics and thus better at finding food (and then go on to have more
babies). But bigger conspecifics might be able to threaten their smart colleague into
giving up their extra food, or humans, at least, richer conspecifics could just buy more
food. And we know in humans at least that being rich is not necessarily correlated
with being smart!
3.8 Conclusions
It would be very useful and convenient if brain size was a good proxy for cognitive
abilities, because it might allow for comparisons across individuals and species that
could confidently be interpreted as being due to cognitive variation. But appealing as
this notion is, even demonstrating the cognitive ability that ties food storing to hippo
campus size, when it seemed self-evident that there had to be some cognitive advan
tage conferred by the enlarged hippocampus, involved significant challenges. And that
association was between a brain region with a clear function and a behaviour that
demonstrably required cognition. The cognitive ability turned out to be relatively (!)
easily quantified and accessible for comparison within and between species. But after
song learning, this relationship was the first compelling instance of selection having
shaped cognitive abilities and the relevant brain region (an adaptive specialization:
Shettleworth, 1998a). It is a long way from being a demonstration that a larger brain
confers greater cognitive ability. It is likely to be as challenging, if not more challenging,
to associate brains and brain regions convincingly with selection arising from more
complicated situations, such as multi-individual social interactions or searching for
and finding mates. And yet, as we are about to see, researchers have certainly tried.
I will end with a comment on ‘intelligence’. Brains have regions of both separate
and distributed neural function, and so there is no overall thing that a brain does that
could be called intelligence. Moreover, most of what a brain does is to run the body
with well-known functions largely associated with well-known regions. Put simply
brains are not all about intelligence and body size explains brain size better than any
thing else. Intelligence, and the tasks crafted to test it, is an anthropocentric and often
anthropomorphic concept, difficult or impossible to define except in reference to the
measures designed to test it. As such, intelligence is something that emerges from a
brain and its body in terms of performance on cognitive tasks, hopefully ecologically
relevant ones. In this book, I will use terms such as ‘intelligent’ and ‘smart’ as con
venient shorthand for ‘very good performance on a cognitive task’. Use of such short
hand does not suggest a belief in such notions as ‘general intelligence’ or the idea that
‘intelligence’ is anything other than a sum of cognitive parts shaped by natural selec
tion. As such, and to the extent that selection has shaped the cognitive abilities of all
animals, all animals are as smart as each other.
4
The Ecological Brain
Natural selection has shaped the organisms we see around us. If we take even a curs
ory glance at a couple of different species of birds or mammals, it is easy to see how
their different lifestyles shape their bodies. For instance, look at the size and location
of their eyes, or the size and location of their ears, noses, and forelimbs. The eyes of
some animals face forward while others have eyes that appear to enable them to see
behind themselves without moving their head. Some animals have a long nose and
little ears, and others a big nose and big ears. Some species have long, slim legs and
others have short and stubby ones. In these examples the animals with eyes facing
forward may be predators rather than prey, while the animals with long noses and
little ears might rely heavily on their sense of smell, perhaps because they are noctur
nal, and the animals with long, slim legs might run faster than those with short,
stubby legs, and be better able to escape predators.
Thanks to more than a century’s worth of experimental biology, we now know that
an animal’s form is the result of the selective pressures placed on it by that animal’s
ecology. This is an insight that goes back to Darwin (of course). For example, preda
tory species such as eagles, wolves, and tigers do have forward-facing eyes. The con
siderable binocular overlap allows for the visual perception of depth, helping such
animals move effectively toward and capture their prey. Their prey, on the other hand,
are more likely to have laterally placed eyes, which provide a much greater, all-round
field of vision, increasing the chance of detecting the predator. The size as well as the
placement is often a useful indicator of the importance or effectiveness of a morpho
logical structure: for example, kiwis are nocturnal, rely heavily on their sense of smell,
and have only very small eyes (Martin et al., 2007), while most owls, also nocturnal,
have relatively large eyes and are instead heavily dependent on vision for hunting
(Martin, 1982).
Given the success of the adaptationist programme since Darwin in explaining so
many features of organisms, it is no wonder, then, that examining whether ecology
had likewise shaped the brain was one of the first explanations for variation in brain
size to be tested by comparative analyses (Mace et al., 1980; Clutton-Brock and
Harvey, 1980). Indeed, because the role of ecology has been at the forefront of all the
various hypotheses of brain size evolution, it is the first one for me to address. Its
prominence is not simply because of Darwin (1859), but also because it seems intui
tive that—given the impact of ecology on the rest of an animal’s form—ecology would
also have shaped the brain, at least to some extent. This importance of ecology in
explaining brain size variation has also been accepted (to varying degrees) for almost
DOI: 10.1093/oso/9780199546756.003.0004
all of the comparative analyses I will discuss in this book, as they have included ‘ecol
ogy’ as a variable in some form or other, even while examining the importance of
other factors on brain size evolution. I should also say that every single comparative
analysis has included body size: the only point of consensus across all analyses is that
body size will explain most of the variation in brain size. I would also argue that body
size can, itself, be considered an ecological variable. In this chapter, therefore, I exam
ine what the evidence actually looks like for the role of ecology and ecological selec
tion in shaping animal brains, and in causing changes in brain size within and across
species.
4.1 What is ‘ecology’?
Before diving into that evidence, to make some sense of what ‘ecology’ has meant in
the following comparative analyses, it might help to know how it has been defined
and what variables have been used in those analyses to represent ecology. Whilst ecol
ogy is usually defined as something like the interaction between organisms and their
biotic and abiotic environments, in practice only certain kinds of ecological inter
actions have fed into ‘ecological’ explanations for brain size. And some of them are
arguably not especially ecological.
The foremost of the ecology variables has been foraging, represented most often as
the kind of diet in some form or another, and entered into the analysis as a categorical
variable. Sometimes diet represents the food itself, e.g. fruit, foliage, invertebrates
(Healy and Guilford, 1990), or sometimes diet breadth, which in Navarette et al.
(2016), for example, was a count of the number of food types eaten, in that instance
up to 13 categories. Sometimes a classification of dietary mode, e.g. folivorous,
frugivorous, insectivorous (Mace et al., 1981; Eisenberg and Wilson, 1978), or grazer,
browser, grazer/browser (Shultz and Dunbar, 2006) is used, and at yet other times
some sub-set such as percentage of fruit (Barton, 1996) or percentage of prey in the
diet (Barton and Dean, 1993).
Other factors have also been used to capture some aspect of an animal’s ecology to
a greater or lesser extent. These include the timing of activity, i.e. whether a species is
diurnal, crepuscular, or nocturnal (Healy and Guilford, 1990; Barton et al., 1995), the
habitat type or complexity of habitat the animal lives in, i.e. whether a species typic
ally occupies forest, woodland, grassland, or aquatic environments (Gittleman, 1986),
and habitat zonation, i.e. whether a species is typically arboreal, terrestrial, or aquatic
(Budeau and Verts, 1986).
Taking a different perspective, sometimes one or more life history variables may
feature as ‘ecology’. These include breeding system, i.e. whether a species is monog
amous, polygynous, promiscuous (Tsuboi et al., 2017; Sawaguchi, 1989), breeding
group type or size, e.g. whether a species is single-male or multi-male (Gittleman, 1986;
Pawlowski et al., 1998), and parental care, such as the number and sex of the indi
viduals caring for the young (Gittleman, 1994; Iwaniuk and Nelson, 2003).
Developmental mode (for example, whether a species produces altricial or precocial
Another random document with
no related content on Scribd:
opportunity to slay their game at noon, but are very apt to hide the
carcass and come back to devour it in the cool of the evening.
The products of fermentation are so repulsive to the higher animals

that only the distress of actual starvation would tempt a monkey to
touch a rotten apple or quench his thirst with acidulated grape-juice.
Poppy fields need no fence; tobacco leaves are in no danger of
being nibbled by browsing cattle. Nature seems to have had no
occasion for providing instinctive safeguards against such out-of-the-
way things as certain mineral poisons; yet the taste of arsenic,
though not violently repulsive (like that of the more common, and
therefore more dangerous, vegetable poisons), is certainly not
attractive, but rather insipid, and a short experience seems to
supplement the defects of instinct in that respect. Trappers know that
poisoned baits after a while lose their seductiveness, and old rats
have been seen [59]driving their young from a dish of arsenic-
poisoned gruel.
Certainly no animal would feel any natural inclination to seek arsenic

or alcohol for its own sake, and there is no reason to suppose that
man, in that respect, differs from every known species of his fellow-
creatures. Our clerical temperance lecturers rant about “the lusts of
the unregenerate heart,” the “weakness of the flesh,” the “danger of
yielding to the promptings of appetite,” as if Nature herself would
tempt us to our ruin, and the path of safety could be learned only
from preternatural revelation. But the truth is that to the palate of a
child, even the child of a habitual drunkard, the taste of alcohol is as
repulsive as that of turpentine or bitterwood. Tobacco fumes and the
stench of burning opium still nauseate the children of the habitual
smoker as they would have nauseated the children of the patriarchs.
The first cigar demonstrates the virulence of nicotine by vertigo and
sick headaches; the first glass of beer is rejected by the revolt of the
stomach; the fauces contract and writhe against the first dram of
brandy. Nature records her protest in the most unmistakable
language of instinct, and only the repeated and continued disregard
of that protest at last begets the abnormal craving of that poison-
thirst which clerical blasphemers ascribe to the promptings of our
natural appetites. They might as well make us believe in a natural
passion for dungeon air, because the prisoners of the Holy
Inquisition at last lost their love of liberty and came to prefer the
stench of their subterranean [60]black-holes to the breezes of the free
mountains.
The craving for hot spices, for strong meats, and such abominations
as fetid cheese and fermented cabbage have all to be artificially
acquired; and in regard to the selection of our proper food the
instincts of our young children could teach us more than a whole
library of ascetic twaddle. Not for the sake of “mortifying the flesh,”
but on the plain recommendation of the natural senses that prefer
palatable to disgusting food, the progeny of Adam could be guided in
the path of reform and learn to avoid forbidden fruit by the symptoms
of its forbidding taste.
[Contents]
B.—REWARDS OF CONFORMITY.
There is a tradition that the ancient Thessalians made it a rule that

the guests of their banquets must get drunk on pain of expulsion. To
let anyone remain sober, they argued, would not be just to the
befuddled majority, of whose condition he might be tempted to take
all sorts of advantage. If the evils of drunkenness were undeserved
afflictions, it would certainly be true that sobriety would give an
individual an almost unfair advantage over the rest of his fellow-men.
He would be an archer trying his skill against hoodwinked rivals, a
runner challenging the speed of shackled competitors. There is not a
mechanical or industrial avocation in which sobriety does not give a
man the advantage which health and freedom confer over crippling
disease. For the baneful effects of intemperance are by no means
limited [61]to the moments of actual intoxication, but react on the half-
lucid intervals, and even on the after years of the reformed toper.
Temperance, in the widest sense, of abstinence from unfit food and
drink, would be the best gift which the fairies could bestow on a
favorite child, for the blessing of frugal habits includes almost all
other blessings whatever. Spontaneous gayety, the sunshine of the
unclouded soul, is dimmed by the influence of the first poison-habit,
and the regretful retrospects to the “lost paradise of childhood” are
founded chiefly on the contrast of poison-engendered distempers
with the moral and physical health of earlier years. Temperance
prolongs that sunshine to the evening of life. By temperance alone
the demon of life-weariness can be kept at bay in times of fiercest
tribulation: Undimmed eyes can more easily recognize the gleam of
sunshine behind the cloudy. The prisoners of the outlawed
Circassian insurgents admitted that, in spite of hunger, hardships,
and constant danger, their captors contrived to enjoy life better than
their enemies in the brandy-reeking abundance of their
headquarters. The myth of the Lotos-eaters described a nation of
vegetarians who passed life so pleasantly that visitors refused to
leave them, and renounced their native lands. The religion of
Mohammed makes abstinence from intoxicating drinks a chief duty
of a true believer, and that law alone has prevented the physical
degeneration of his followers. With all their mental sloth and the
enervating influence of their harem life, the Turks are still the finest
representatives of physical manhood. At the horse [62]fairs of
Bucharest I saw specimens of their broad-shouldered, proud-eyed
rustics, whose appearance contrasted strangely with that of the
sluggish boors and furtive traffickers of the neighboring natives. After
twelve hundred years of exhaustive wars, alternating with periods of
luxury and tempting wealth, the descendants of the Arabian
conquerors are still a hardy, long-lived race, physically far superior to
the rum-drinking foreigners of their coast towns. For more than six
hundred years the temperate Moriscos held their own in war and
peace against all nations of Christendom. Their Semitic descent
gave them no natural advantage over their Caucasian rivals; but they
entered the arena of life with clear eyes and unpalsied hearts, and in
an age of universal superstition made their country a garden of
science and industry. Their cities offered a refuge to the scholars and
philosophers of three continents, and in hundreds of pitched battles
their indomitable valor prevailed against the wine-inspired heroism of
their adversaries.
Frugality has cured diseases which defied all other remedies. For
thousands of reformed gluttons it has made life worth living, after the
shadows of misery already threatened to darken into the gloom of
approaching night. Luigi Cornaro, a Venetian nobleman of the
sixteenth century, had impaired his health by gastronomic excesses
till his physicians despaired of his life, when, as a last resort, he
resolved to try a complete change of diet. His father, his uncles, and
two of his brothers had all died before the attainment of their fiftieth
year; but [63]Luigi determined to try conclusions with the demon of
unnaturalism, and at once reduced his daily allowance of meat to
one-tenth of the usual quantity, and his wine to a stint barely
sufficient to flavor a cup of Venetian cistern-water. After a month of
his new regimen he regained his appetite. After ten weeks he found
himself able to take long walks without fatigue, and could sleep
without being awakened by nightmare horrors. At the end of a year
all the symptoms of chronic indigestion had left him, and he resolved
to make the plan of his cure the rule of his life. That life was
prolonged to a century—forty years of racking disease followed by
sixty years of unbroken health, undimmed clearness of mind,
unclouded content. Habitual abstinence from unnatural food and
drink saves the trials of constant self-control and the alternative
pangs of repentance. “Blessed are the pure, for they can follow their
inclinations with impunity.”
[Contents]
C.—PERVERSION.
The poison-habit, as we might call the craving for the stimulus of

unnatural diet, is the oldest vice, and in some of its forms has been
practiced by almost every nation known to history or tradition.
Thousands of years before Lot got drunk on home-made wine, the
ancestors of the Brahmans fuddled with soma-juice; Zoroaster
enacts laws against habitual intoxication; the art of turning grape-
juice from a blessing into a curse seems to have been known to the
nations of Iran, to the Parsees, and to the [64]first agricultural
colonists of the lower Nile. Nunus, the Arabian Noah, is said to have
planted vineyards on the banks of the Orontes; the worship of
Bacchus was introduced into Asia Minor several centuries before the
birth of Homer. The origin of the opium habit antedates the earliest
records of Chinese history; for immemorial ages the Tartars have
been addicted to the use of Koumis (fermented mare’s milk), the
Germanic nations to beer, the natives of Siam to tea and sago-wine.
Intoxication and the excessive use of animal food were prevalent
vices, especially in the larger cities, of pagan Greece and Rome.
Yet the ancients sinned with their eyes half open. Their recognition of
dietetic abuses was expressed in the word frugality, which literally
meant subsistence on tree fruits—or, at least, vegetable products—
in distinction from the habitual use of flesh-food. The advantages of
temperate habits were never directly denied; the law of Pythagoras
enjoins total abstinence from wine and flesh, and the name of a
“Pythagorean” became almost a synonym of “philosopher.” In all but
the most depraved centuries of Imperial Rome, wine was forbidden
to children and women. The festival of the Bona Dea commemorated
the fate of a Roman matron who had yielded to the temptation of
intoxicating drink, and was slain by the hand of her stern husband.
Lycurgus recommends the plan of letting the pupils of the military
training-schools witness the bestial conduct of a drunken Helot, in
order to inspire them with an abhorrence of intoxication. The bias of
public opinion [65]always respected the emulation of patriarchal
frugality and frowned upon the excesses of licentious patricians.
But the triumph of an anti-physical religion removed those

safeguards. Mistrust in the competence of our natural instincts
formed the keystone of the Galilean dogma. The importance of
physical welfare was systematically depreciated. The health-laws of
the Mosaic code were abrogated. The messiah of Antinaturalism
sanctioned the use of alcoholic drinks by his personal example—nay,
by the association of that practice with the rites of a religious
sacrament. The habit of purchasing mental exaltation—even of a
fever-dream—at the expense of the body, agreed perfectly with the
tendencies of a Nature-despising fanaticism, and during the long
night of the Middle Ages monks and priests vied in an
unprecedented excess of alcoholic riots. Nearly every one of the
thick-sown convents from Greece to Portugal had a vineyard and a
wine cellar of its own. The monastery of Weltenburg on the upper
Danube operated the largest brewery of the German empire. For
centuries spiritual tyranny and spirituous license went hand in hand,
and as the church increased in wealth, gluttony was added to the
unnatural habits of the priesthood, and only the abject poverty of the
lower classes prevented intemperance from becoming a universal
vice. As it was, the followers of the Nature-despising messiah lost no
opportunity to drown their better instincts in alcohol. They could
plead the precedence of their moral exemplars, and vied in sowing
the seeds of bodily diseases [66]which their system of ethics
welcomed as conducive to the welfare of a world-renouncing soul.
Among the slaves of the Scotch kirk-tyrants the long-continued

suppression of all healthier pastimes contributed its share to the
increase of intemperance. On the day when the laboring classes
found their only chance of leisure, outdoor sports were strictly
prohibited. Dancing was considered a heinous, and on the Sabbath
almost an unpardonable, sin. The tennis-halls were closed from
Saturday night to Monday morning. Bathing was sinful. Mountain
excursions, strolls along the beach, or in the open fields, were not
permitted on the day of the Lord. Dietetic excesses, however,
escaped control, and thus became the general outlet for the cruelly
suppressed craving for a diversion from the deadly monotony of
drudgery and church-penance. For “Nature will have her revenge,
and when the most ordinary and harmless recreations are forbidden
as sinful, is apt to seek compensation in indulgences which no
moralist would be willing to condone, … and the strictest observance
of all those minute and oppressive Sabbatarian regulations was
found compatible with consecrating the day of rest to a quiet but
unlimited assimilation of the liquid which inebriates but does not
cheer” (Saturday Review, July 19, 1879). “Everyone,” says Lecky,
“who considers the world as it really exists, must have convinced
himself that in great towns public amusements of an exciting order
are absolutely necessary, and that to suppress them is simply to
plunge an immense portion of the population into the lowest depths
of vice.” [67]
Clerical despotism is still a potent ally of intemperance. In hundreds
of British and North American cities the dearth of better pastimes
drives our workingmen to the pot-house. They drink to get drunk, as
the only available means of escaping tedium and the consciousness
of their misery. Nature craves recreation, and the suppression of that
instinct has avenged itself by its perversion.
[Contents]
D.—PENALTIES OF NEGLECT.
Dietetic abuses have contributed more to the progress of human

degeneration than all other causes taken together. Our infants are
sickened with drastic drugs. The growth of young children is stunted
with narcotic beverages; the suppression of healthier pastimes
drives our young men to the rum-shop; intemperance has become
the Lethe in which the victims of social abuses seek to drown their
misery. The curse of the poison-habit haunts us from the cradle to
the grave, and for millions of our fellow-men has made the burdens
of life to outweigh its blessings. There is a doubt if the “years” of
Genesis should be understood in the present meaning of the word;
but historians and biologists agree that the average longevity of our
race has been enormously reduced within the last twenty centuries,
and intemperance is the chief cause of that decrease. Our average
stature has been reduced even below that of the ancient natives of
an enervating climate, like that of the lower Nile, as proved by
D’Arnaud’s measurements of the Egyptian mummy-skeletons. On
our own continent, outdoor life in the struggle with the [68]perils of the
wilderness has somewhat redeemed our loss of physical manhood;
but what are the men of modern Europe compared with their iron-
fisted ancestors, the athletic Greeks, the world-conquering Romans,
the Scandinavian giants, the heroic Visigoths? Like a building
collapsing under the progress of a devouring fire, the structure of the
human body has shrunk under the influence of the poison-habit; and
there is no doubt that the moral vigor of our race has undergone a
corresponding impairment—appreciable in spite of the recent revival
of intellectual activity and the constant increase of general
information.
The tide is turning; the victims of anti-physical dogmas are

awakening to the significance of their delusion; the power of public
opinion has forced the dupes of the alcohol-brewing Galilean to join
the crusade of the temperance movement; diet-reform has become a
chief problem of civilization; but the upas-tree of the poison-habit is
too deeply rooted to be eradicated in a single generation, and the
task of redemption will be the work of centuries. As yet the probing of
the wound has only revealed the appalling extent of the canker-sore.
The statistics of the liquor traffic have established the fact that the
value of the resources wasted on the gratification of the poison-vice
far exceeds the aggregate amount of the yearly expenditure for
educational, charitable, and sanitary purposes—nay, that the
abolition of that traffic would save a sum sufficient for all reforms
needed to turn earth into a physical and social paradise. And yet that
waste expresses only the indirect [69]and smaller part of the damage
caused by the curse of the poison-habit. The loss in health and
happiness cannot be estimated in coin; but if the sum thus expended
in the purchase of disease were devoted to the promotion of arson
and robbery, the utmost possible extent of the consequent mischief
would probably fall short of the present result. The stimulant habit in
all its forms clouds the sunshine of life like an all-pervading poison-
vapor. Alcohol undermines the stamina of manhood; narcotic drinks
foster a complication of nervous diseases; opium and tobacco impair
the vigor of the cerebral functions. The excessive use of animal food,
too, avenges itself in all sorts of moral and physical disorders. It
inflames passions which no prayer can quench. “Alas! what avails all
theology against a diet of bull-beef?” Father Smeth wrote from the
Sioux missions; and the almost exclusive use of flesh food has,
indeed, afflicted our Indians with the truculence of carnivorous
beasts. The same cause has produced the same effects in western
Europe. The carnivorous saints of medieval Spain delighted in
matanzas and heretic-hunts, as their carnivorous ancestors in the
butcher sports of the circus, and their British contemporaries in bear-
baits and Tyburn spectacles.
[Contents]
E.—REFORM.
The consequences of intemperance have at all times provoked

protests against the more ruinous forms of the poison-habit, but the
advance from special to general principles is often amazingly slow;
and even now the cause of temperance is hampered by the
[70]shortsightedness of reformers who hope to eradicate the Upas-
tree by clipping and hacking its more prominent branches. They
would limit prohibition to the more deadly stimulants, not dreaming
that the fatal habit is sure to reproduce its fruit from the smallest
germs; that the poison-vice, in fact, is infallibly progressive, ever
tending to goad the morbid craving of the toper to stronger and
stronger poisons or to a constant increase in the quantity of the
wonted stimulant: from cider to brandy, from laudanum to morphine,
from tonic bitters to rum, from a glass of wine to a dozen bottles,
from beer and tobacco to the vilest tipples of the dram-shop.
“Principiis obsta” (Resist the beginnings) was a Latin maxim of deep
significance. The cumulative tendency of the stimulant vice may be
resisted, but only by constant vigilance, constant self-denial,
constant struggles with the revivals of a morbid appetency, all of
which might be saved by the total renunciation of all abnormal
stimulants whatever, for only in that sense is it true that “abstinence
is easier than temperance.”
We must accustom our boys to avoid the poison-vice as a loathsome

disease, rather than as a forbidden luxury which could ever be
indulged without paying the penalty of Nature in a distressing
reaction, far outweighing the pleasures of the morbid and momentary
exaltation. We must teach them that the artifice by which the toper
hopes to cheat Nature out of an access of abnormal enjoyment is
under all circumstances a losing game, which at last fails to produce,
even for the moment of the fever-stimulus, a [71]glimpse of happiness
at all comparable to the unclouded sunshine of temperance.
But before we can hope to redeem the victims of the poison-vender,

we must learn to make virtue more attractive than vice. We must
counteract the attractions of the rum-shop by inviting reforming
topers, not to the whining conventicles of a Sabbath-school, but to
temperance gardens, resounding with music (dance music, if “sacred
concerts” should pall) and the jubilee of romping children, and
shortening summer days with free museums, picture galleries,
swings, ball grounds, and foot-race tracks. The gods of the future will
contrive to outbid the devil.
It would be unfair, though, to depreciate the services of the Christian

ministers who in a choice between dogma and reform have bravely
sided with Nature, and, defying the wrath both of spiritual and
spirituous poison-mongers, of rum-sellers and heretic-hunters, are
trying their utmost to undo the mischief of their antinatural creed, by
frankly admitting that a man can be defiled by “things that enter his
mouth,” and that the sacrament of eucharistic alcohol should be
abandoned to the rites of devil-worshipers.
But the religion which pretends to inculcate a peace-making spirit of

meekness has been strangely remiss in opposing the excessive use
of a diet which is clearly incompatible with the promotion of that
virtue. In Christians, as in Turks, Tartars, and North American
Redskins, a chiefly carnivorous diet engenders the instincts of
carnivorous beasts, and a Peace Congress celebrating its banquets
with sixteen courses of flesh food might as well treat a vigilance
[72]committee to sixteen courses of opium. “Frugality” should again
be promoted in the ancient sense of the word; in a community of
reformants temperance and vegetarianism should go hand in hand.
Or rather, the word “temperance” should be used in the extended
sense that would make it a synonym of Abstinence from all kinds of
unnatural food and drink; and Dr. Schrodt’s rule should become the
canon of every dietetic reform league. “Avoid,” he says, “all drinks
and stimulants repulsive to the palate of an unseduced child, but
also all comestibles that need artificial preparation to make them
palatable.” The first part of that rule would exclude opium, tobacco,
alcoholic beverages, tea, coffee, absinthe, fetid cheese, and caustic
spices. The second would abolish many kinds of animal food, but
sanction milk, butter, eggs, honey, and other “semi-animal”
substances, condemned by the extreme school of vegetarians.
“From the egg to the apple,” is an old Latin phrase which proves that
the frugality of the ancient Romans never went to such extremes.
Milk, eggs, and vegetable fats, in their combination with farinaceous
dishes, might amply replace the flesh food of the northern nations,
and, considering the infinite variety of fruits and vegetables known to
modern horticulture, there seems no reason why a vegetarian diet
should necessarily be a monotonous one. The Religion of Nature will
require the renunciation of several deep-rooted prejudices, but its
path of salvation will in no sense be a path of thorns. [73]
[Contents]
CHAPTER V.
SKILL.
[Contents]
A.—LESSONS OF INSTINCT.
The organic faculties of each species of animals are marvelously

adapted to its peculiar mode of life, but only in the lower creatures
the skilful exercise of those faculties appears to be an inborn gift.
The young bee builds its first hexagon with mathematical precision.
The young ant needs no instructor to aid her choice of proper
building-material, of proper food to be stored for winter use or
distributed in the nurseries of the larvæ. The young butterfly, an hour
after issuing from the shell of the chrysalis, can use its wings as well
as at the end of the summer, and displays the same skill in steering
its way through the maze of a tangled forest.
Young birds, on the other hand, have to acquire such

accomplishments by long practice. Instead of driving them back to
their nests, their parents encourage their attempts at longer and
longer flights, and seem to know that occasional mishaps will prove
a useful lesson for future emergencies. The mother fox carries half-
crippled game to her burrow and sets her cubs a-scampering in
pursuit, allowing the best runner to monopolize the tidbits. Young
kittens practice mouse-catching by playing with balls; puppies run
after grasshoppers, young squirrels play at nest-building by
gathering handfuls of leaves and moss. A British naturalist, who had
domesticated a young beaver, one day caught his pet building a dam
across the floor of his study. The little engineer [74]had dragged up a
cartload of books, papers, sticks of wood, etc., and piled them up to
best advantage, placing the heavier volumes in the bottom stratum
and the lighter ones higher up, and filling out the interspaces with
letters and journals. Every now and then he would “stand off” to
scrutinize the solidity of the structure and return to mend a
misarrangement here and there.
Children manifest early symptoms of a similar instinct. Infants of two

or three years can be seen squatting in the sand, excavating tunnels,
or building prairie-dog towns. Young Indians insist on the privilege of
breaking colts; the youngsters of the Bermuda Islanders straddle a
plank and paddle around with a piece of driftwood, if their parents
are too poor to afford them a canoe of their own. To a normal
American boy a tool-box is a more welcome present than a velvet
copy of Doré’s Illustrated Bible. Swiss peasant lads practice sharp-
shooting with self-constructed cross-bows. The old English law
which required the son of a yeoman to practice archery for three
hours a day was probably the most popular statute of the British
code. On new railroads, bridges, etc., artisans, plying their trade in
the open air, are generally surrounded by crowds of young rustics,
who forego the pleasures of nutting and nest-hunting for the sake of
watching the manipulations of a new handicraft. Even in after years
the instinct of constructiveness frequently breaks the shackles of
etiquette, and princes and prelates have defied the gossip of their
flunkeys by getting a set of tools and passing whole days in the
retirement of an [75]amateur workshop. The emperor Henry I.
invented a number of ingenious hunting-nets and bird-traps.
Mohammed II., the conqueror of Constantinople, forged his own
chain-armor. Charles V., the arbiter of Europe, preferred
watchmaking to every other pastime. Cardinal de Retz delighted in
the construction of automatons. Peter the Great was the best ship-
carpenter of his empire.
[Contents]
B.—REWARDS OF CONFORMITY.
The English word king, like Danish kong and German König, are
derived from können (practical knowledge), and the first ruler was
the most skilful, as likely as the strongest, man of his tribe. Skill,
whether in the sense of bodily agility or of mechanical cleverness,
established the superiority of man over his fellow-creatures, and is
still in many respects a test of precedence between man and man.
Supreme physical dexterity is always at a premium, in peace as in
war, in the sports of princes, in the pastimes of pleasure-seekers, in
the adventures of travelers, in moments of danger, in camps, in the
wilderness and on the sea, as well as in smithies and workshops.
Conscious skill and agility form the basis of a kind of self-reliance
which wealth can only counterfeit. In a cosmopolitan sea-port town of
western Europe I once overheard a controversy on the comparative
value of protective weapons. Revolvers, stilettos, air guns, slung-
shots, and bowie knives found clever advocates, but all arguments
yielded to the remark of an old sea-captain, who had faced danger in
four different continents. “There’s a use for all that, no [76]doubt,” said
he, “but, I tell you, mynheers, in a close row the best thing to rely
upon is a pair of quick fists.” For the efficacy, even of the best
weapons, depends to a large degree on the expertness of the
handler, the panoply of a weakling being as unprofitable as the
library of an idiot. “Presence of mind” is often only the outcome of
such expertness, and in sudden emergencies theories are shamed
by the prompt expedients of a practical man. In war the issue of a
doubtful campaign has more than once been decided by the superior
constructiveness of an army that could bridge a river while their
opponents waited for the subsiding of a flood. The conquest of
Canada was achieved by the skill of a British soldier who devised a
plan for hauling cannon to the top of a steep plateau. The fate of the
Byzantine empire was decided by the mechanical expedient of a
Turkish engineer who contrived a tramway of rollers and greased
planks, as an overland road for a fleet of war ships. By the invention
of the chain grappling-hook Duilius transferred the empire of the
Mediterranean from Carthage to Rome.
Even for the sake of its hygienic influence the development of

mechanical skill deserves more general encouragement. Crank-work
gymnastics are apt to pall, but in pursuit of a favorite handicraft even
an invalid can beguile himself into a good deal of health-giving
exercise, and, besides, the versatile development of the muscular
system reacts on the functions of the vital organs, and thus explains
the robust health of active mechanics often laboring under the
disadvantage of indoor confinement. The poet [77]Goethe, whose
intuitions of practical philosophy rival those of Bacon and Franklin,
records the opinion that every brain-worker should have some
mechanical by-trade in order to obviate one-sidedness, and mental
as well as physical debility. Every handicraft reveals by-laws of
Nature which no cyclopedia can teach an inquirer; manual labor is a
school of practical wisdom, and sound “common sense,” as the
English language happily expresses the sum of that wisdom, is a
prerogative of farmers and mechanics far, far oftener than of
speculative philosophers.
Nor are such benefits limited to emergencies from which wealth

could dispense its possessor. An amateur handicraft is the best
safeguard against the chief bane of wealth: ennui, with its
temptations to folly and vice. Nabobs can do worse than imitate the
example of Carlo Boromeo, who spent every leisure hour of his
philanthropic life in practical landscape gardening, and turned a large
and once barren lake-island into the loveliest paradise of southern
Europe. “Heroum filii noxae,” “the sons of the great are apt to be
nuisances,” would be less true if Goethe’s advice were heeded by
our fashionable educators, and the benefits of his plan would extend
to emergencies for which fashionable accomplishments afford only a
dubious safeguard. “A mechanical trade,” says Jean Jacques
Rousseau, “is the best basis of safety against the caprices of
fortune. Classical scholarship may go begging, where technical skill
finds its immediate reward. A distressed savant may recover his loss
in the course of years; a skilful mechanic need only enter the next
workshop and show [78]a sample of his handiwork. ‘Well, let’s see
you try,’ the reply will be; ‘step this way and pitch in.’ ”
Thus, too, gymnastic agility is the best safeguard against

numberless perils. A mother who hopes to protect her boy by
keeping him at home and guarding him from the rough sports of his
playmates, forgets that her apron-strings cannot guide him through
the perils of after years; and a better plan was that of Cato, the
statesman, warrior, and philosopher, who, in the midst of his
manifold duties, found time to instruct his young sons in leaping
ditches, and swimming rapid rivers, in order to “teach them to
overcome danger that could not be permanently avoided.”
[Contents]
C.—PERVERSION.
The absurd contempt of mechanical accomplishments is due partly

to the direct influence of anti-physical dogmas, partly to the indirect
tendency of that caste spirit which has for ages fostered the
antagonism of wealth and labor. The opulent Brahmans of ancient
Hindostan thought themselves so immeasurably superior to the
children of toil that a Sudra was not permitted to approach a priest
without ample precautions against the defilement of the worshipful
entity. The temples of high-caste devotees were closed against low-
caste believers. The very breath of a Sudra was supposed to pollute
articles of food to such an extent that a Brahman had always to take
his meals alone.
The secret of such prejudices was probably the supposed

antagonism of body and soul and the imagined necessity of
emphasizing that contrast by [79]constant insults to the
representatives of physical interests and occupations. For in Europe,
too, the propagation of an anti-physical creed went hand in hand with
the systematic depreciation of secular work, excepting, perhaps, the
trade of professional manslaughter, the military caste, which here, as
in India, found always means to enforce respect by methods of their
own. During the most orthodox centuries of the Middle Ages
industrial burghers were valued only as tax-payers; peasants were
treated little better than beasts of burden—in many respects
decidedly worse, for after drudging all day for an inexorable master,
the serf had often to work by moonlight, in order to get a little bread
for himself and his family. The proposition to join in any manual
occupation (the handling of a whip, perhaps, excepted) would have
been resented as a gross insult by every little baron or priest of
Christian Europe. Paul Courier describes the indignation of a French
nobleman who caught a tutor instructing his boys in botany and the
secret of improving trees by grafting: “Going to make a clown of him?
You had better get an assistant-teacher with a manure cart.” The
manual-labor dread of several medieval princes went to the length of
employing special chamberlains for every detail of their toilet: a chief
and assistant shirt-warmer, a wig-adjuster, a hand-washer, a foot-
bather, a foot-dryer. German barons thought mechanical labor an
incomparable disgrace—more shameful, in fact, than crime—for the
same Ritter who would have starved rather than put his hand to a
plow, had no hesitation in eking out an income by [80]highway

Adaptation and The Brain Oxford Series in Ecology and Evolution Susan D Healy Full Chapter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Adaptation and The Brain Oxford Series in Ecology and Evolution Susan D Healy Full Chapter

Uploaded by

Copyright:

Available Formats

Adaptation and the Brain (Oxford Series

in Ecology and Evolution) Susan D.

Oxford Series in Ecology and Evolution

The Ecology of Adaptive Radiation

Adaptation and the Brain

2 The Measurable Brain  5

3 The Intelligent Brain 17

4 The Ecological Brain  35

5 The Innovative Brain  51

6 The Technical Brain  67

7 The Sexual Brain  81

8 The Social Brain  109

9 Concluding Remarks  123

2 • Adaptation and the Brain

4 • Adaptation and the Brain

6 • Adaptation and the Brain

Data source Measurement type Taxonomic group No. of

The Measurable Brain • 7

Data source Measurement type Taxonomic group No. of

2.1 Measuring the brain

8 • Adaptation and the Brain

The Measurable Brain • 9

10 • Adaptation and the Brain

time measurement occurs, as deterioration sets in immediately after death.

2.2 Measuring skull volumes

The Measurable Brain • 11

12 • Adaptation and the Brain

2.3 Variation in brain size measurement

Whichever method is used, it is important that measurements are accurate, and a

The Measurable Brain • 13

14 • Adaptation and the Brain

individuals of five species of waterfowl (paired t-test, t4 = 0.25, p = 0.81; Wilcoxon

2.4 Body size measurement

Whether the measurements of brain volume or weight or endocranial volume are

The Measurable Brain • 15

2.5 Measuring parts of the brain

3.1 Brain size and intelligence

18 • Adaptation and the Brain

3.2 What is cognition?

The Intelligent Brain • 19

3.3 Comparing cognitive abilities

20 • Adaptation and the Brain

The Intelligent Brain • 21

3.4 Food storing, spatial cognition, and the hippocampus

22 • Adaptation and the Brain

The Intelligent Brain • 23

24 • Adaptation and the Brain

The Intelligent Brain • 25

(b) delayed spatial match-to-sample (c) delayed spatial non-match-to-sample

sample phase sample phase

retention interval retention interval

choice phase choice phase

correct choice correct choice

outperformed dark-eyed juncos Junco hyemalis on a spatial-non-matching-to-sample

26 • Adaptation and the Brain

The Intelligent Brain • 27

28 • Adaptation and the Brain

Final retention interval (seconds)

3.5 The hippocampus and spatial memory in food storers

The Intelligent Brain • 29

3.6 The relationship between whole brain

30 • Adaptation and the Brain

The Intelligent Brain • 31

2 The Measurable Brain 5

3 The Intelligent Brain 17

4 The Ecological Brain 35

5 The Innovative Brain 51

6 The Technical Brain 67

7 The Sexual Brain 81

8 The Social Brain 109

9 Concluding Remarks 123