What were your very earliest ancestors like? I do not mean your
great-great-great-grandparents. I mean the earliest life on the planet. In
principle we all have a unique lineage of ancestors that runs all the way back
to the origin of life. What was life like then—and is the supposed life on
Mars our cousin? These are the problems palaeontologist Bill Schopf faces.
Unlike my more recent ancestors these earliest remains had no skeletons so did
not fossilise well. They have also been crushed by rocks for so long that little
remains.

It is, as Schopf points out, an embarrassing fact that, until quite recently,
all the fossils we knew were relatively large and complex. We now know these
more complicated forms date from only the last sixth or so of life’s history on
Earth: the earliest life we presently have evidence for dates from around 3500
million years ago. To study these early microfossils is to study the majority of
evolution.

Schopf doesn’t really believe that even if we did not know about these life
forms their absence would be a serious problem for the idea of evolution: I
don’t know what my great-great-grandparents looked like, but I don’t doubt they
existed. So Schopf’s motivation is curiosity: he wants to know more about life’s
earliest representatives.

And he’s in a good position to enlighten us. Schopf has been central to the
science of unearthing (literally) these microfossils, and Cradle of
Life is the story, told with authority, of what we know and how we know it.
The how is especially important. Small grey blobs under a microscope could be
almost anything—they may never have been alive, they may be modern
single-celled contaminants or they may truly be fossils. Even if they are
fossils, they may not be contemporary with the rocks. Schopf doesn’t want to
sweep these difficulties under the carpet. Quite the opposite. He lays before
the reader the methods by which the various alternatives are eliminated, but
beyond that, far from wishing to assert anything, he wants to encourage the
reader to develop the studied scepticism that he feels is the stuff of good
science.

Indeed, for the most part, the book is not about the science of the early
history of life but about science and scientists. We learn, for example, of how
his one-time boss (the improbably named Elso S. Barghoorn) came to submit a now
classic paper to the journal Science, only after a comparable paper
from a competing colleague arrived on his desk, also from Science, for
him to review (tut-tut). For years in the late 19th century many experts,
notably John William Dawson, asserted that a particular formation in a rock was
a large foraminiferan (a type of single-celled organism), it was not until the
same formation was found to have been ejected from Mount Vesuvius that
geologists agreed that this was nothing more than deformed limestone.

He shows how the characters of the scientists guided its course and how
entrenched views are hard to shift. (Dawson never would admit he was wrong.)
What if his accounts are anecdotal, they are charming, fascinating and well
researched. See, for example, Schopf’s weighing up of the political pressures
surrounding the scientists claiming to have found life on Mars. (Schopf is not a
believer.)

In fact, the history is relayed not for curiosity’s sake but as a warning:
don’t believe what the men in white say, be sceptical. But, in the end, the
great thing about the studied scepticism of science is that it all comes out in
the wash.

It has been a while since I read a book with so much good sense, put over in
so amicable a style. If I ever were to discover my
great-great-great-grandparents I hope they turn out to be as wise as Schopf.

ONCE upon a time altruism was not a problem for evolutionary biologists.
Doing good unto others was, as Vero Wynne-Edwards argued in the early 1960s,
“good for the species” and this, it was supposed, is why we do it. After the
work in the 1960s and 1970s of Bill Hamilton and John Maynard Smith, such a view
has come to be considered fundamentally wrong. Philosopher Elliot Sober and
biologist David Sloan Wilson would like us to think again.

The notion that selection cannot act principally for the good of the species
stems from the realisation that cheats can prosper. Imagine two species, one in
which all the individuals are nice to each another, the other in which they are
all selfish. The first species will probably have higher “species fitness”,
although it is unclear what this means. Now imagine a cheat that is happy to
have its back scratched to get rid of fleas, but which doesn’t bother to return
the compliment. Such individuals will do better than their fellows, as they gain
by the altruism of others while not suffering the costs of being altruistic.
Selfishness wins, so the species fitness cannot have decided the evolutionary
trajectory that leads to altruism.

Wilson and Sober point to an alternative scenario. Imagine a species with a
population in which groups form and dissolve. Different groups have different
fitnesses dependent on the proportion of altruists. If 80 per cent of a group is
altruistic, it does better than a group with 20 per cent. So, they argue, while
selection within groups favours selfishness, those groups with many altruists do
better. The component of fitness that can be considered as that deriving from
differences between groups can be enough to keep altruism going.

From this premise, they argue that human evolution is dominated by “group
selection” and that we act for group benefit against our selfish interests. To
arrive at this conclusion, however, the authors both redefine what the term
“group selection” means and paper over a critical gap in their argument.

From their viewpoint, that the selection acting on a trait can be divided
into a within-group component and a between-group component, they build a
mathematical framework for describing what happens to genes over time. They then
surround this framework with a language. Whenever the between-group component of
fitness comes into play they call the process “group selection”. This is a much
broader definition than the one normally used. While their mathematical
framework leads to the same conclusions as those reached by Hamilton and Maynard
Smith, their language relabels the results. Hamilton considers relatedness
between individuals as key to the evolution of some forms of altruism. Most
people label this process “kin selection”, but Sober and Wilson include it under
the heading of group selection. Similarly, selection on organisms with two
genomes—what most mortals call “individual-level selection”—they
insist is also group selection.

What, you might ask, is there to object to if both analyses come to the same
conclusions, differing only in the language used to describe the result? For one
thing, Sober and Wilson’s language is a recipe for confusion. Why use their
language when a decent one associated with a different but equivalent
mathematical protocol already exists? They argue the older school misses the
group selective component of the spread of a trait. But by marking nearly
everything as group selection, Sober and Wilson’s language acts to blur more
distinctions.

The proposed reform of language also has a more subtle side effect. The
present linguistic concentration on individual-level selection has created a
scientific culture in which it is normal to hypothesise about how various
features of organisms (eyes, wings and so on) might be “good for the organism”.
Sober and Wilson wish to create a culture in which it is normal to hypothesise
about how features might be “good for the group”.

But the critical question, and the one that is papered over, is how a rare
trait can spread. A group with many altruists may do better than one with few.
But how did such a group come about? If the population starts off with everyone
being nice, a new selfish type can increase in frequency, even when there are
few to begin with. However, if the population starts off with everyone being
selfish, then the rare altruists carry a double burden. They not only suffer
costs through scratching others’ backs, they remove the fleas from their selfish
competitors.

The authors devote a single, brief paragraph to the issue. Taking as their
example the poorly understood phenomenon of predator inspection behaviour in
fish, they posit that the trait varies continuously within the fish population.
Continuity of variation, they assert, allows a rare trait to spread, although
they make no attempt to explain why. But they themselves undermine this line of
reasoning when they argue elsewhere that it is easy to set up systems to
“police” cooperative behaviour, as it is something you either do or do not do,
hence making it easily recognisable. When it suits their argument, then,
altruism is a discontinuous trait. Remarkably, they also cite fish predator
inspection as one example.

But Unto Others is not all semantic tussling over what we mean by
“group selection”. Half the book is a semantic tussle about how to label
psychological theories of altruism.

What if being good makes you feel good? Is this altruism? If you derive a
psychological benefit for your selflessness, the sharp division between the two
traits becomes blurred. The authors argue that an evolutionary viewpoint can
help to clarify the issue. I found this section very heavy going and too
philosophical for my taste, though the evolutionary psychologists for whom this
book is intended might look on it more kindly.

I did find some ideas helpful. For example, when it comes to human altruism,
the authors advise concentrating on the problem of “social norms”. In some
cultures, people will not pick up money from the pavement. Why not? Not because
a relative is following and will get it instead. Not because they expect to get
money back in return. No, they are following a social norm: they believe it will
bring bad luck through divine retribution.

Sober and Wilson go on to assert that human groups with “fitter” social norms
will replace those with less fit norms and posit this as an adequate explanation
of human altruism. I suspect, however, that group replacement has more to do
with differences in weaponry and diseases than social norms. Every culture has
numerous social norms and I doubt that group replacement happens often enough to
do a decent job of sifting the fit from the unfit. I see no reason to suppose
that extant social norms are necessarily those that act for the good of the
group.

More critically, this line of reasoning about society again fails to address
the key question. By what mechanism, did an unselfish notion held by only a few
become the social norm?

Evolution’s rise to respectability is a puzzle. The science is not much
different from what it was some twenty years ago. Evolution doesn’t have
anything like a new space telescope with which to see more clearly. So why all
the hubbub?

What is new is the extent to which selectionist thinking is extending its
tendrils into every aspect of human life, from medicine to morality. The rise of
this Darwinisation of all things human is, however, possibly more to do with the
media than with science. Human evolution is intrinsically sexy. There is no
culture without a creation myth and Darwinian evolution is ours. By myth I do
not mean it is untrue (it isn’t), I just mean that there is some gap that we
need to fill in our minds and Darwinian evolution is an excellent plug.
Selectionist explanations also have a logical soundness that is most
seductive.

Faced with shelves groaning with popular literature on evolution, what should
you look out for? The biggest problem is that Darwinism is too often used
polemically. Stephen Jay Gould and Richard Lewontin are unusual in being upfront
about their Marxist inclinations. Creationists usually prefer to remain
cryptic.

One downside of this usurping of the field by anyone with an axe to grind is
that too often the books are written by unqualified (and biased) people. You
wouldn’t read a book on astrophysics by a biologist, so why read evolution
written by physicists?

And this leads to the nub of the problem: you need training to know what is
kosher, but how can you learn if the popular accounts are prejudiced? The
potential prejudices that writers on evolution might have are many and varied,
but they can usually be put into one of three camps.

On one side we have the pro-selectionists: Richard Dawkins, Daniel Dennett,
George Williams, Edward. O. Wilson, the Ridleys, Steven Pinker, Jared Diamond
and John Maynard Smith. This camp is usually concerned with telling you how
amazing natural selection is and how we are all products of blind but
deterministic forces acting on our genes. Were I ever to write a popular book, I
would be among them.

On the opposing side, there are those who would like you to believe that
natural selection is vastly overrated and that our current model of evolution is
missing something big—something they alone understand is the usual
subtext. Here we find anti-selectionists such as Michael Behe, Stuart Kauffman,
Stephen Jay Gould, Brian Goodwin, Niles Eldredge and Steven Rose.

In the middle sits the lone figure of Steve Jones, a man so universally
sceptical that unless he had his birth certificate he would doubt his own
existence. This scepticism is, to my mind, a necessary antidote to the hype. It
might strip the gloss off seductive theories, but so be it.

Although anti-selectionists often have little in common (except a preference
for seeing selection downgraded in the public imagination), I have noticed one
strange regularity. They all belong to what I shall call the “c” club. Entry to
this club requires a member to adopt a word beginning with “c” with which to
attack selectionists, if only by obfuscation. Evolution is constrained, chaotic,
catastrophic, contingent or complex (irreducible or otherwise). My main problem
with these books is that the ratio of assertion to fact is too high for my
taste. Worse still, assertions are often sold as facts with the merest peppering
of anecdote to support them. For the more theoretically minded, data just seem
to be a nuisance that spoils elegant maths. I also find many of the
anti-selectionist ideas too ambiguous to be helpful. Gould’s use of the words
“constraint” and “contingency” have, for example, left a mess of confusion that
has generated a sterile mini-industry in semantics to clean up after him. Of
course, it would be wrong to suggest that evolution is all sorted and that
selection is everything. I don’t think anyone believes this, although the
anti-selectionists usually see it as necessary to construct a false target to
attack.

The pro-selectionists, of course, make no such assertion. Instead, they play
a “what if . . .” game. What if wings, eyelashes, language or whatever, were the
product of selection? If so, they ask, what would that selection have been like
and what would follow? Consider a trivial case. Selection on eyes might, I shall
conjecture, be to allow animals to see. If so, we expect lenses to be clear and
cells in the eye to be responsive to light. We also expect organisms living in
pitch dark not to have eyes. If such predictions are borne out, then clearly
that particular selectionist theory has some explanatory power.

Pro-selectionists have the great advantage of a well-formed theory that, as I
mentioned, is logically beautiful—too elegant if anything. The seductive
simplicity of some ideas results in a tendency to shoehorn fat facts into slim
theories. Matt Ridley’s attempts to squeeze all of human cooperative behaviour
into the framework provided by the Prisoner’s Dilemma, is a bit like the Ugly
Sisters trying on Cinderella’s glass slipper—entertaining but unlikely to
be wholly successful. In selectionism, there is also an unspoken primacy given
to logic: too often when the data do not fit the selectionist theory, it is the
data that are questioned.

Darwinian psychology faces its own additional difficulties. When we consider
the evolution of the eye, I know what I am discussing. We can agree on what
defines an eye. But what about a tendency to cooperate? I cannot hold a
tendency, cooperative or otherwise, in my hand. If I could define such a
tendency, is it likely that there will be heritable variation in it? Even if
there is such heritability, how can I know whether my cooperative tendencies are
the result of selection or just something that brains can do incidental to
changes in other parts of the brain? After all, it is likely that brains can do
things now which they never had to do in the past. I have, for example, an
ability to recognise telephone boxes. No one is going to contend that this is
the result of selection acting on my ancestors to recogniseMesolithic telephone
boxes. So if the psychological apparatus I use is the result of selection, just
what is the thing that was being selected and what capabilities follow from it?
On what criteria can I make the distinction?

As Diamond suggests, one way out of this dilemma is to use data on other
species for a comparative analysis. This is fine for reproductive behaviour (all
species reproduce), but it’s no good for features that are uniquely human. How,
for example, can I ever know if the appreciation of music is something that was
selectively favoured? Without solid foundations, Darwinian psychology might slip
gently away, as it has in its previous incarnations.

Meanwhile, if you want to become better acquainted with evolution, where
should you start? I’d recommend Dawkins, who provides as clear an introduction
to the way selection works as any. For a consideration of the rest of evolution,
I would turn to the textbooks, although even these don’t get everything right.
My favourites are Maynard Smith’s Evolutionary Genetics (Oxford, 1989)
and Mark Ridley’s Evolution (Oxford, £10999, ISBN 0192892878),
although both are slightly eccentric in their choice of what to discuss. And
Douglas Futuyma’s Evolutionary Biology (third edition, Sinauer,
$62.95, ISBN 0878931899) succeeds admirably in making evolution seem like
dull but good science.

See also New ĐÓ°ÉÔ­´´â€™s Darwinian selection.

YOU might reasonably be forgiven for thinking that human evolutionary biology
is a second-rate science. One has only to turn on the radio or open a newspaper
to find someone claiming that this or that aspect of human behaviour might have
been the result of selection on Stone Age people for this or that reason. Often
it seems that plausibility alone is grounds for accepting a hypothesis. Where
are the data and where are the tests?

It is a mistake, however, to suppose that the apparent lack of rigour
displayed by some of Darwin’s disciples is a measure of the scientific
credentials of Darwin’s ideas or of the man himself. In The Expression of
Emotions in Man and Animals it is clear that Darwin did not only wish to
argue that something might be so: he wanted to know whether it is so.

With the current hype surrounding Darwinian interpretations of human
behaviour, it is easily forgotten that Darwin was more than the chief proponent
of the idea of natural selection. He was also a prime advocate of the idea that
we are descended from apes. In 1872, 13 years after the publication of
Origin, Darwin argued in Expression that the evidence that we are
just another primate is, quite literally, staring us in the face. It is now
republished with a foreword, essays on the problem of universality and on
Darwin’s photographs, and Huxley’s obituary of Darwin.

As editor Paul Ekman makes clear in the foreword, Darwin was writing in a
particular philosophical context. The anatomist Charles Bell argued, from his
1806 Anatomy and Philosophy of Expression onward, that humans were
God-made separate creations and that our expressions bore evidence of this. We
have, he argued, many more expressions than animals, and unlike animals, we can
control them.

Darwin delivered a double riposte. He argued that very many of our
expressions are in anatomical, physiological and behavioural terms the same as
those of animals. And he described evidence that we cannot exercise will over
many of them. Ironically, some of the better evidence for this goes unmentioned.
With Expression, Darwin was one of the first scientists to publish a
book with photographs. He chose examples of both posed and natural subjects.
While the individuals artificially demonstrating certain expressions look like
embarrassingly awful ham actors, the pictures from life look genuine.

Darwin was not simply concerned with rebutting Bell. He was also interested
in the question of whether all races of humanity belonged to one species or, as
was then argued, were independently derived. This was why knowing whether human
expressions are universal or culturally dependent concerned Darwin. After
consulting his friends in the colonies about how the locals expressed their
feelings, he concluded that most expressions are universal.

In an excellent essay on universality, Ekman shows that Darwin was on the
right track, but that his flawed methodology meant that his contributions could
not be considered definitive. Darwin tested his ideas by using a very large
series of anecdotes, and this presents a number of difficulties. The anecdotal
evidence was not systematic, for example, so we have no idea whether he only
presented anecdotes consistent with his case. Darwin was aware of this
difficulty and discussed examples that he considered problematic.

Ekman also notes that the means by which the anecdotes were collected add
more potential biases. Writing to colonials to ask about native peoples may have
produced misleading findings, tainted by culturally specific ways of
interpreting facial expressions. Moreover, Darwin’s letters usually posed
questions that invited a given answer. Instead of asking which emotion was
expressed by the wrinkling of the eye and upwardly curved mouth, he would ask
whether happiness was expressed by these features.

With careful scientific testing, however, all such objections have since been
overcome. If for no other reason, Ekman’s essay is to be welcomed, as it
establishes that good systematic science can be done in this field, despite
obfuscatory attempts by certain social anthropologists. Darwin’s other
conclusion, that we are descended from apes, has similarly been put through
stringent scrutiny. From our knowledge of our genes and those of primates we now
have much better evidence of our ancestry than Darwin could have dreamt of.

You might reasonably ask why anyone should read this book, apart from its
obvious interest as an historical classic. Ekman’s essay on universality is a
good reason, as is Phillip Prodger’s very closely researched discussion of the
book’s photographs. Here we learn that they are not quite what they seem. They
are, in fact, photographs of detailed etchings, copied from the original
photographs! Between the original and the etching, details changed, often on
Darwin’s instructions.

The editor also provides very helpful commentaries in the text on the current
status of Darwin’s ideas. Ekman’s scholarship extends to updating and correcting
the text to be as near to what Darwin intended as possible. Much to the editor’s
credit, the book is both scholarly and lively.

For me, however, the real wonder of Darwin’s books is the way they offer an
insight into the man. It seems clear that Bell’s argument must have bothered him
from very early on. I cannot see any other way of understanding how Darwin
related so many closely observed details of the expression of animals and of his
children accrued over such a time scale. Darwin’s ability as an observational
scientist is easily forgotten in an age when his name is synonymous with natural
selection.

More than this, it is wonderful to have a glimpse of Darwin’s critical
scientific instincts. Darwin realised that he had not only to “interpret”
findings in the light of a theory, but also to look for circumstances in which
his theory predicted one thing, but an alternative was equally plausible. To
discover whether our expressions are innate or learnt, for example, he asked
whether people born blind show the same expressions as everyone else.

In these terms, then, Darwin is not dated. This is the same method of testing
hypotheses—by relying on the notion of falsifiability—that is found
in everyday science the world over. Yet the same process can be used to show
that Darwin was wrong on many counts. For example, he theorises a great deal
about the role of acquired characteristics in the evolution of expressions. Such
Lamarckian inheritance makes predictions about the way information is
transmitted from one generation to the next. For the most part the predictions
are not borne out, and hence we can reject the ideas.

By equal measure, it is to the credit of Bell that we can consider his thesis
refuted. He laid out the facts that we would see if he were right and,
conversely, those that we would expect if he were wrong. His theory was thus
framed as falsifiable. The 46 per cent of Americans who, according to a 1991
Gallup Poll, still believe that some time in the past 10 000 years God created
us more or less in our present state, should take a leaf out of Bell’s book.
Even better, they should read Darwin.