Encouraged by superficial notions of evolution,
Which becomes, in the popular mind, a means of disowning the past.
T. S. Eliot, ‘The Dry Salvages’ (1941)
While the promise appears of an evolutionary approach to the problem of
structure and agency, there is continuing resistance to the intrusion of evolutionary
or biological ideas in social theory. On the other hand, there is a
minority of enthusiasts who seem to believe that biology can explain everything
human. As Erkki Kilpinen (2000, p. 33) rightly remarks: ‘Modern
thought (post-modern, too, for that matter) tends to reduce action either to
biology or to anything-but-biology: socio-biology and deconstructionism
are today’s Scylla and Charybdis.’
One extreme position is the notion that social and economic behaviour
can be explained largely in terms of biological characteristics, such as
genes. Such a sociobiological view is rejected here, for several reasons
given in this book. Even in biological terms it is widely criticized. The other
extreme position is that biology is largely irrelevant to the study of human
society, and that social scientists may forever consider economies or societies
as if they were separated from the ecosystem and the biotic world. This
view is widespread among many sociologists and economists.
This book steers between these two extremes. Both extreme views neglect
the interactions between humanity and nature; they constrain scientific
enquiry by limiting investigation into the complex causal interactions
between nature and human society. One view overrides the causal influence
of the social environment. The other neglects that human beings, like
other organisms, have evolved from other species, and that human capacities
must be partly explained in evolutionary terms.
For much of the twentieth century, discussion by social scientists of biology
or evolution was taboo. The horrifying, racist pseudo-biology of the
Nazis helped to reinforce the academic view that biological ideas should
be entirely separated from the social sciences. Consequently, since the
3 Objections and
explanations
1920s, a large number of psychologists and social scientists have proclaimed
that human behaviour is entirely a product of human culture or
the social environment. But even if all relevant human behaviour could be
explained entirely in cultural or environmental terms, this would not justify
such a Berlin Wall between biology and the social sciences. There are
many reasons for this, and we may consider just a few.
First, whether they are largely determined by nature or by nurture, human
mental and physical capacities are bounded, and the explanation of
their limits is a matter for the science of human evolution. Human beings
are not infinitely malleable but rather special. An explanation of human
abilities and limitations must involve an understanding of the human psyche,
and how that has evolved.
Second, human beings interact with the ecosystem and other organisms.
An understanding of how humans have evolved in interaction with
their environment is not only an important topic for scientific enquiry in its
own right, but also a vital means of preserving a sustainable ecosystem and
limiting damage to the natural environment.
Third, science develops by the transfer of ideas and metaphors from one
sphere to another. Just as the economist Thomas Robert Malthus inspired
Charles Darwin, the social sciences can gain ideas and inspiration from biology.
This does not mean that any or every idea in one sphere is useful in
another. Neither does it mean that social and biological mechanisms are
similar. The whole point about metaphors is that they are inexact. It simply
means that some such ideas and metaphors can be inspirational, and for
this reason the sciences should not be sealed off from one another.
Fourth, since the famous launch of sociobiology by Edward O. Wilson
in 1975, a growing group of researchers have attempted to explain many
human capacities and behaviours in evolutionary terms. This diverse and
rapidly growing literature varies both in quality and in line of argument.
Some researchers attempt to explain human and social phenomena entirely
and crudely in terms of genes. This line of research neglects emergent
properties and causal powers at the cultural and social level. Others
see the evolution of human society partly in terms of units of information
or ‘memes’ (Dawkins, 1976). This concept also has its problems, including
a persistent vagueness concerning its meaning and its mechanisms of replication.
Characteristically, these literatures have been impoverished by
limited contact and conversation with social scientists. Fault lies on both
sides. But by rejecting biology outright, many social theorists have been
the more insular. Relatively few social scientists provide evidence that
they understand much biology or evolutionary theory.
Partly as a result of inadequate dialogue from sociologists and social
theorists, sociobiology has been led by amateurs in social theory. As
interest in the interaction between the biotic and the social has grown
outside mainstream social theory, we have seen the emergence of new
extra-disciplinary research groups and networks, some of which claim no
42 Introduction
allegiance to sociobiology. There is a now almost an alternative social science,
separate from the traditional disciplines of sociology and economics,
and with little knowledge of its predecessors. As long as the more traditional
social scientists remain with their heads in the sand, they will be ill
equipped to deal with this emerging set of rival doctrines. Continuing insularity
could mean both the impoverishment of the social sciences and the
further development of erroneous or naïve versions of evolutionary social
theory. To avoid these outcomes, it is essential that social theorists have an
improved and less reactionary understanding of evolutionary theory, and
that the new social evolutionists appreciate and understand the rich heritage
of the social sciences.
It is appropriate at this point to address some additional misconceptions
that have helped to divert social theorists from addressing some relevant
insights from evolutionary theory. We confine ourselves to some prominent
misconceptions and objections, with brief responses and counterarguments.
Later chapters of the present work will expand on some of
these points.
Darwinism means neither optimization nor reductionism
Even in economics, where evolutionary ideas have become common, several
misconceptions persist. The recent interest in evolutionary theories in
economics derives largely from the works of Richard Nelson and Sidney
Winter (1982), Kenneth Boulding (1981), Friedrich Hayek (1988) and a few
others. There was a brief former flurry of interest in evolutionary themes,
including works by Armen Alchian (1950) and Milton Friedman (1953).
Still earlier, as discussed below in the present book, evolutionary ideas
were found in economics in the 1880–1930 period.
We consider the interventions of Alchian and Friedman first, partly because
they have been highly influential among mainstream economists. In
different ways, both Alchian and Friedman used evolutionary theory to
conflate explanations of economic processes onto one level of analysis.
Alchian (1950) proposed that even if firms never actually attempted to
maximize profits, ‘evolutionary’ processes of selection would ensure the
survival of the more profitable enterprises. Friedman (1953) amended this,
by seeing ‘natural selection’ as grounds for assuming that agents act ‘as if’
they maximize, whether or not they consciously do so. Both of these arguments
have been subjected to detailed criticism. They rely on an overly
simplistic concept of competition and fail to demonstrate that some kind of
optimal behaviour does indeed result.1
Objections and explanations 43
1 E. Penrose (1952) pointed out that Alchian implausibly assumed that firms cannot know
the conditions of survival but economists can. Friedman’s evolutionary argument was
criticized in Winter (1964), Boyd and Richerson (1980), Schaffer (1989), Hodgson (1994,
1999b) and Dutta and Radner (1999).
It is a serious misconception to see evolution as always leading to either
static or optimal outcomes (Veblen, 1899a; Hodgson, 1993; Cohen and
Stewart, 1994; J. Potts, 2000). Such results occur under restricted conditions
only. In an open system, equilibria are always temporary. Some contrasting
outcomes involve positive feedbacks and are highly sensitive to initial
conditions. Possibly suboptimal phenomena such as lock-in and path dependence
are now widely acknowledged.2
More generally, evolution does not drive towards some goal or destination.
Instead it carries the baggage of its past, in a typically haphazard, ongoing
process of adaptation and selection. It is important to dispense with
all mistaken notions of evolution as an optimizing, goal-driven or necessarily
progressive process. Darwin (1871, vol. 1, pp. 166–77) himself emphasized:
‘we are apt to look at progress as the normal rule in human
society; but history refutes this. … We must remember that progress is no
invariable rule.’
But that is not the only prominent misconception. In addition, many social
scientists react to the mention of ‘evolution’ or ‘Darwinism’ as an indication
that the explanation of social phenomena is about to proceed in
purely biological terms. They mistake these labels as inevitable indicators
of an unavoidable biological reductionism in which social phenomena –
from social culture to economic performance – are purportedly explained
largely in terms of human genes.
In fact, there is an ongoing discussion within biology itself as to the possibility
of other additional units and levels of selection. Charles Darwin
(1859, pp. 235–42) himself explained that sterile insects (such as worker
ants) had evolved because integrated family communities (with sterile and
non-sterile members) had themselves become whole units of selection.
Concerning human society, Darwin (1859, pp. 422–3; 1871, vol. 1, pp.
59–61, 106) argued that natural selection operates upon the elements of
language as well as on individuals. Darwin (1871, vol. 1, p. 166) also proposed
that tribal groups with moral and other propensities that served the
common good would be favoured by natural selection. Darwin seemed to
endorse a version of group selection, and perhaps hinted at the natural selection
of institutions, as well as the natural selection of individuals. Several
modern biologists have argued that evolutionary selection occurs at
higher levels: not simply on genes, but on individuals, groups and even
species. Darwinian anthropologists such as Robert Boyd and Peter
Richerson (1985) and William Durham (1991) have developed a two-level
theory where there is transmission at the level of both culture and the
genes.
Even Richard Dawkins, who attained fame with his hymn to genetic
reductionism in The Selfish Gene (1976), proposed an additional level of
44 Introduction
2 Gould (1980), David (1985, 1994, 2001), Dosi et al. (1988), Arthur (1989, 1994), North (1990),
Hodgson (1993), H. P. Young (1996).
selection in the last chapter of this work, with his idea of a cultural ‘meme’.
It is difficult to propose cultural ‘memes’ and remain a genetic
reductionist, because that would require an explanatory reduction of all
memes to genes. Despite his genetic reductionist starting point, Dawkins
was driven by the logic of his own argument to adopt multiple-level selection
theory (Hull, 1980, 1981). Dawkins (1983, p. 422) thus wrote: ‘It is also
arguable that today selection operates on several levels, for instance the
levels of the gene and the species or lineage, and perhaps some unit of cultural
transmission.’3
Given that evolutionary selection can operate at the group and cultural
levels, the invocation of evolutionary ideas in social science does not imply
that explanations of social phenomena have to be, or can be, reduced to
properties and changes at the level of the genes. There are other possible
units of replication. Units such as habits, routines, customs and institutions
themselves endure through time and replicate by imitation, even if their
features are not sustained with the same fidelity as the coding in the DNA.
There is variation even between similar units of these types, and the selection
environment favours the survival of some over others. Emergent
properties and causal relations at higher levels imply that the analysis of
these higher-level selection processes cannot be explained entirely in terms
of lower-level units. The processes and time-scales involved at higher levels
may be so different from those at lower levels that much of the explanation
of higher-level phenomena must be in higher-level terms.
Later chapters of this book will examine further the application of Darwinian
principles to social theory. Several authors over the last 150 years
have proposed that Darwinian evolution takes place within society, not
simply metaphorically, analogically, or by extension of lower-level phenomena.
Darwinian evolution literally takes place at the level of society itself,
operating on social as well as biological units. The necessary
conditions for Darwinian social evolution include the existence of varied
replicating entities, and some differences in their capacities to survive in
any given situation. These replicating entities might include social rules or
routines. If Darwinian social evolution were shown to work at this level, it
would not require the invocation of biological reductionism. Such a notion
of replication on multiple levels permits accounts of social evolution involving
social rather than biological elements. Darwin’s theory implies
neither global optimization nor biological reductionism (Hodgson, 1993;
Khalil, 1997).
A wide variety of views have been gathered under the term ‘Darwinism’,
including reactionary doctrines described as ‘Social Darwinism’ and
Objections and explanations 45
3 Multiple levels of evolution have also been considered by Lewontin (1970), Arnold and
Fristrup (1982), Brandon and Burian (1984), Eldredge (1985), L. Buss (1987), Goertzel
(1992), Depew and Weber (1995), Maynard Smith and Szathmáry (1995, 1999), Brandon
(1996), Sober and Wilson (1998), L. Keller (1999), Kerr and Godfrey-Smith (2002) and
Henrich (2004).
the genetic reductionism of today. It must be emphasized that the use of
the term Darwinism here should not be taken to imply any of these things.
The term ‘Darwinian’ is used here to describe the general and causal theory
of evolution, involving variation, inheritance and selection, which is at
the centre of the classic works by Darwin (1859, 1871). More details of the
Darwinian theoretical core will be elaborated at a later stage.
The limits of sociobiology
As sociology has lost its consensus over core presuppositions, some sociologists
have established formerly alien doctrines such as methodological
individualism and individual utility maximization (Coleman, 1990).
Strong antibodies would have repulsed these invaders in the 1940–80 period.
But the crisis of modern sociology is so severe that doctrines as alien
as biological reductionism have now established themselves within. The
lure of reductionism is such that several economists (Becker, 1976;
Hirshleifer, 1977, 1985; Tullock, 1979; Robson, 2001a, 2001b, 2002), and
even some modern sociologists, have proposed that much human behaviour
can be explained in terms of our genetic inheritance.
Joseph Lopreato and Timothy Crippen (1999) diagnose the crisis in
modern sociology and creditably propose an injection of Darwinism.
However, their recommendations come too close to a form of biological
reductionism. Lopreato and Crippen (1999, p. 77) propose at some ‘ultimate
level’ that: ‘Organisms tend to behave in ways that maximize their inclusive
fitness.’ But they ignore the work of Darwinian cultural evolutionists
from Veblen (1899a, 1919a) to Boyd and Richerson (1985) who argue that
when cultural transmission enters the picture such maximization of genetic
fitness will typically be diverted or overlaid by other factors. Even if
humans ‘ultimately’ behave in ways that maximize their inclusive genetic
fitness, this tells us very little about variations in human behaviour, between
different cultures or across short periods of time. It tells us something
concerning survival, sexuality and reproduction, where some of the
strongest instinctive pressures have evolved. Some very broad conclusions
may be drawn concerning gender and family relations.4
But an infinite number of feasible social structures – and patterns of
mating and childcare – could ensure the reproduction of the human species.
Social theory is concerned with the explanations of differences within
the broader set of possibilities. Furthermore, inclusive fitness theory tells
us less about non-familial social structures, particularly those that have
emerged only in the last few thousand years, and have not been subject to
the same type or longevity of evolutionary selection pressure as those
where sex and human reproduction are paramount. Human societies vary
46 Introduction
4 See the extensive evidence cited in Lopreato and Crippen (1999).
culturally and institutionally through time and space. The lonely hour of
‘ultimate’ genetic determination never comes.
The leading sociobiologist Edward O. Wilson (1978, p. 167) declared:
Can the cultural evolution of higher ethical values gain a direction and
momentum of its own and completely replace genetic evolution? I
think not. The genes hold culture on a leash.
What is wrong with this statement? No matter how much we may (rightly)
insist that cultural evolution is (extremely) important, it would be absurd
to suggest that it can ‘completely replace genetic evolution’. The final sentence
is more problematic. If it means that culture is determined by the
genes and can be explained largely in terms of the genes, then it is false. But
Wilson himself admitted that the leash is ‘very long’ (1978, p. 167). It is
more appropriate to point out in response to Wilson that analysis at the genetic
level is chronically limited as a basis for explaining detailed human
behaviour. Wilson bypasses the task of explaining many varied actual and
possible behaviours that lie within the ample limits of genetic restraint.
The laws of gravity also hold culture and behaviour on a leash, but it
would be absurd to conclude from this truism that we can largely explain
human behaviour with the laws of physics. Human genetic constraints
were established long ago in our evolution. While knowledge of these constraints
is important, and they do tell us something about human behavioural
dispositions and possibilities, on their own they explain little of our
culture or behaviour in a modern complex society.
Wilson (1978, p. 153) admitted that human ‘social evolution is obviously
more cultural than genetic’. This is why his sociobiology is at best of highly
limited use for social science. The task of social science is to explain particular
behaviours or phenomena within the very wide zone provided by genetic
and physical constraints. Social science examines further constraints
and causal mechanisms, which themselves cannot be reduced solely to biological
terms. The genes help to form the substrata of human nature. But
they do not constitute human nature as a whole. Our genes tell us something
of our fundamental human nature, as presented at our birth, but they
tell us nothing of the specific and varied cultural contexts in which vital
human dispositions are channelled and formed. In particular, through
processes of socialization and learning, we develop a cognitive structure
by which to interpret and respond to the data received by our senses. The
initial basis of this structure is genetic, but subsequent neural development
is much a result of our interactive experiences in a social and natural context.
It is a major task of the social scientist to understand the implications
of the cultural processes of socialization and learning for human behaviour
and potential.
We have to pay much more attention to levels of replication or selection
above that of the gene. Human genetic evolution has taken place in a
Objections and explanations 47
human cultural environment. One consequence is that social selection can often
override the pressures of genetic selection. For example, the European physiognomy
is much less suited than that of Aborignines to the Australian climate.
European Australians suffer a much higher rate of skin cancer, for
example. Yet European-type institutions dominate the continent. These institutions
have been selected despite the unsuitedness of some European
genes. The genetic advantage of the Aborigines in relation to the natural
environment has been overridden by selection at the institutional level.
The importance of additional levels of selection above that of the gene
means that genes cannot tell the whole story.
These additional levels of replication or selection at the social level are not
trivial. Taking Darwinism seriously at the cultural and institutional level
means much more than ‘memes’ such as catch-phrases or pop songs. If there
are replicators of information in the social domain, then they will be structured
entities, acknowledging the structural and institutional nature of
knowledge itself (Langlois, 2001). Darwinism at the social level involves the
selection of different types of institutional structures, upon which the survival
or prosperity of nations or populations (along with their genes) can
sometimes depend. Evolution at the social level involves social relations and
structures, and more than mere information of a social kind.
Wilson’s highly limited concept of culture is symptomatic of the kind of
problem that has appeared when biologists have entered the social domain,
but with insufficient appreciation of social theory. For Wilson (1978,
p. 78) ‘cultural change is the statistical product of the separate behavioural
responses of large numbers of human beings’. On the contrary, culture is
not merely the average and variance of individual characteristics; it is also
a system involving structured, interactive relationships between individuals.
Consequently, culture is not reducible to the statistical properties of a
mere aggregate of individuals; culture involves relations between individuals
and emergent properties that are not reducible to individuals alone.
That is what gives rise to higher levels of selection, above genes and the individual.
Just as our genes loosely affect cultural possibilities, our culture
has had an effect on our genetic endowment (Boyd and Richerson, 1985;
Durham, 1991).
For much of the twentieth century, sociology rightly emphasized culture
but wrongly ignored its biological and genetic limits and preconditions.
At the other extreme, a small minority of social scientists have
surrendered to sociobiology. In both cases, a careful articulation of the interactive
relationship between the cultural and genetic aspects of humanity
has been lacking. At both extremes, an adequate discourse on emergent
properties and ontological levels has been absent. A major purpose of this
book is to outline the achievements of Veblenian institutionalists, evolutionary
theorists and emergentist philosophers in establishing a multiplelevel
evolutionary theory, where both social culture and human instincts
are acknowledged.
48 Introduction
Biology is not a means of legitimating universal market competition.
Markets themselves are neither the universal solution to the problem of
scarcity nor the exclusive domain of human competition (Hodgson,
2001c). Furthermore, competition itself is not universal in the biological
sphere (Kropotkin, 1902; Lewontin, 1978; Augros and Stanciu, 1987; Sober
and Wilson, 1998). Also some psychologists see evolution as a challenge to
some versions of individual rationality (Cosmides and Tooby, 1994a,
1994b; Plotkin, 1994, 1997; Cummins and Allen, 1998).
While some social scientists relapse into reductionism, many sociologists
resist any suggestion that the social sciences can learn from biology.
They have been trained to believe that biology and psychology simply cannot
be used to help explain social phenomena. Some will follow Durkheim
(1982, pp. 32–3) in his Rules of Sociological Method of 1901 and ‘separate the
psychological domain from the social one’ while separating the social from
the biological as well. Many will react against past abuses of biology in social
theory – when it has been called upon to support nationalism, imperialism,
racism and sexism – and draw the mistaken conclusion that all
biological explanations and metaphors have thus to be cast out of the discipline.
They will ignore the fact that no reputable theory in biology gives
any support whatsoever to any form of discrimination. Neither can we reasonably
conclude from biology that our genes largely determine our fate.
Even genetic reductionism has to admit that the social environment is an
important determining factor, including in individual development.
The general resistance of sociologists to biology is both unjustified and
deleterious. As a handful of authors have suggested, part of the disabling
crisis of modern sociology is its inability to overcome its own compartmentalization
from biology and other sciences. Instead of the language of
compartmentalization and rejection, the pressing task is to give psychology
and biology some appropriate and limited explanatory scope, restricted
by the recognition of emergent properties at the social level, and
enhanced by a revived and inclusive framework of social theory.5
Is self-organization an alternative to Darwinian
evolution?
John Foster (1997) and Ulrich Witt (1997) have proposed that the theory of
self-organization provides a basis for evolutionary thinking in economics.
Objections and explanations 49
5 Van den Berghe (1990), Halton (1995) and Lopreato and Crippen (1999) have addressed
this crisis in sociological thought. Modern sociologists who have embraced evolutionary
or even Darwinian ideas include Blute (1979, 1997), Runciman (1989, 1998), Kontopoulos
(1993), Aldrich (1999) and Chattoe (2002). In particular, Hirst and Woolley (1982), Lenski
and Lenski (1987) and Maryanski and Turner (1992) have argued that sociological studies
should also take account of the physical environment and the biological inheritance of
human beings. Influential organizational studies such as McKelvey (1982), G. Carroll
(1984), Hannan and Freeman (1989), Singh (1990) and Hannan and Carroll (1992) have
also been inspired by evolutionary biology.
Foster took the more extreme position, to argue that self-organization is a
superior alternative to any ‘biological analogy’. Foster (1997, p. 444) wrote:
Once we abandon biological analogy in favour of an economic self-organization
approach … then we are no longer interested in the microscopic
details of selection mechanisms, but in the endogenous
tendency for acquired knowledge and skills to interact to create increases
in economic organization and complexity.
Foster has alluded to modern versions of such an approach, as in the works
of Ilya Prigogine and Isabelle Stengers (1984), Daniel Brooks and E. O.
Wiley (1988), David Depew and Bruce Weber (1995), Stuart Kauffman
(1993, 1995), Weber and Depew (1996), Weber et al. (1989) and Jeffrey
Wicken (1987).
Clearly, several of the above authors were critical of some presentations
of Darwinian theory. For example, Depew and Weber (1995) noted the
changing agendas and shifting emphases of Darwinian enquiry over the
years. Others like Brooks and Wiley (1988), B. Weber et al. (1989) and
Wicken (1987) were keen to generalize evolutionary explanations and integrate
insights from thermodynamics such as the entropy law. Kauffman
(1993) made a powerful argument that natural selection alone cannot explain
the origin of complex organisms. Systems involving non-linear interactions
involve a large number of possible states, most of which would
have little survival value. Kauffman argued that processes of self-organization
channel systems into more restrictive possibilities, some of which
can have evolutionary benefits.
However, in contrast to Foster (1997), none of these modern authors
cited above saw his argument as an alternative to Darwinian theory.
Wicken (1987) wrote of ‘extending the Darwinian paradigm’, not exterminating
it. Depew and Weber (1995) considered ‘Darwinism evolving’, not
Darwinism abandoned. Weber and Depew (1996, p. 51) wrote:
the very concept of natural selection should be reconceived in terms
that bring out its dynamical relationships with chance and self-organization.
In our view, Kauffman’s recent work, as expressed in The Origins
of Order, does just this.
Note here that what is involved is a revision of natural selection theory, not
its negation. Kauffman (1995, p. 8) himself called for a ‘revision of the Darwinian
worldview’ not its abandonment. As Kauffman (1993, p. 644) also
related:
I have tried to take steps toward characterizing the interaction of selection
and self-organization. … Evolution is not just ‘chance caught on
50 Introduction
the wing’. It is not just a tinkering of the ad hoc, of bricolage, of contraption.
It is emergent order honored and honed by selection.
The last sentence is worthy of reflection and emphasis. Kauffman did not
conceive of his theory as an alternative to Darwinism (R. Lewin, 1992, pp.
42–3). On the contrary, once self-organized systems and subsystems
emerge, natural selection does its work by sorting the more adapted from
the less. Kauffman explained this in detail. Natural selection acts upon
these self-organized structures once they emerge. Far from being an alternative
to natural selection, self-organization requires it: in order to determine
which self-organized units have survival value. As Gary Cziko (1995,
p. 323) argued:
the laws of physics acting on nonliving entities can lead to spontaneous
complexity, but nothing in these laws can guarantee adapted complexity
of the type seen in living organisms … Of all the complex
systems and structures that may self-organize due to the forces of nature,
there can be no assurance that all or any of them will be of use for
the survival and reproduction of living organisms.
In a sense, Witt (1997, p. 489) was correct in his assertion that self-organization
‘provides an abstract, general description of evolutionary processes’
but natural selection is no less abstract, nor less general. Indeed,
self-organization involves an ontogenetic evolutionary process, in that it
addresses the development of a particular organism or structure. (In biology,
ontogeny refers to the growth and development of single organisms,
where the genetic material is given.) This does not rule out the possibility
that ontogeny can also involve the natural selection of entities within the
organism. For example, the growth of many organisms involves the natural
selection of immunities, neural patterns and (often beneficial) bacteria
in their gut (Edelman, 1987; Plotkin, 1994). Likewise, the growth of a firm
may involve the internal selection of habits or routines (Nelson and Winter,
1982). Hence some descriptions of self-organizing processes involve
some (phylogenetic) selection of constituent components of the emerging
structure.
However, accounts of self-organization or ontogeny do not necessarily
involve selection or phylogeny. By definition, phylogeny means the existence
and evolution of a whole population, within which selection occurs.
Hence natural selection is always phylogenetic as well as ontogenetic, in
that it addresses the evolution of whole populations of organisms or structures,
as well as the development of individual organisms. In general, ontogeny
may incorporate phylogeny but does not necessarily do so
(consider the examples in the preceding paragraph); but phylogeny always
incorporates ontogeny.
Objections and explanations 51
Furthermore, from the point of view of the overall evolutionary process,
complete evolutionary descriptions require a phylogenetic account of the
selection of ontogenetically developing units. Hence while self-organization
is important (and perhaps essential), it cannot provide a complete evolutionary
description. This must involve phylogeny as well as ontogeny. If
we are confined to ontogeny then our description of the overall evolutionary
process is incomplete; it does not address the differential survival and
fecundity of different (self-organized) structures or organisms. Consequently,
self-organization may be an important part of evolution and
ontogenetic development, but it cannot replace natural selection.
Self-organization theorists have shown how complex structures can
emerge without design, but these structures are themselves subject to evolutionary
selection. Some will survive longer and be more influential than
others: selection will operate. We have every reason to see these issues as
relevant to economic evolution. Conscious choices, competitive pressures,
market forces or environmental constraints operate on technologies, institutions,
regions and even whole economies. All of these contain self-organized
structures, but this neither precludes nor demotes the role of
evolutionary selection.
Yngve Ramstad (1994) also argued that biological analogies are inappropriate
for economics. One of his reasons is based on the argument of
John R. Commons that institutional evolution involves ‘artificial’ rather
than ‘natural’ selection. This is critically discussed in Chapter 13 below,
where it is established that artificial and natural selection are not mutually
exclusive, and the former always relies on the latter; so we do not have to
deal with this objection here.
The Lamarckian confusion
Many social scientists have described social evolution as ‘Lamarckian’.
6
In
fact, the relationship between Darwinism and Lamarckism is more complicated
than many have assumed. One of the most important ideas in
‘Lamarckism’ – although it pre-dates the 1809 work of Jean-Baptiste de
Lamarck by centuries – is the admission of the possibility of the inheritance
of acquired characters. This idea was popularized by Lamarck and influentially
endorsed by Auguste Comte and Herbert Spencer. The nineteenthcentury
Darwinians did not rule out the possibility of the inheritance of acquired
characters. Even in the first edition of Origin of Species, Darwin
(1859, pp. 82, 137, 209) himself endorsed this idea. Darwin never denied a
52 Introduction
6 Among those that have made the claim that social evolution is Lamarckian are Popper
(1972a), Hirshleifer (1977), Gould (1980), Simon (1981), McKelvey (1982), Medawar (1982),
Nelson and Winter (1982), Gray (1984), Boyd and Richerson (1985), Hayek (1988), C.
Freeman (1992), Metcalfe (1993) and Hodgson (1993). However, Boyd and Richerson
(1985), Metcalfe (1998) and Hodgson (2001b) uphold that social evolution is both
Darwinian and Lamarckian.
limited role for the inheritance of acquired characters and in his later life he
gave it increasing rather than decreasing attention and approval. Hence
Lamarckism (in this sense) and Darwin’s doctrine are not necessarily mutually
exclusive. We now know that the possibility of the inheritance of acquired
characters is non-existent (or highly limited) at the level of genetic
evolution. In contrast, it has been argued by many that acquired characters
can be passed on and inherited in the social domain.7
Just as there are differences between Keynesianism and the doctrines of
Keynes, and Marxism and the ideas of Marx, we must draw a distinction
between Lamarckism and Lamarck’s own views. Lamarck (like Darwin)
was a philosophical materialist and saw intention or volition as rooted in
material causes (Boesiger, 1974; Lamarck, 1984). Hence Lamarck did not
see will or purpose as ultimate drivers of evolution. It was not Lamarck
himself but later ‘Lamarckians’ that made unexplained will or purpose so
central to a depiction of evolutionary change.8
But Lamarck and the Lamarckians had something important in common:
they all believed in the inheritance of acquired characters. Hence there are
grounds to define Lamarckism primarily in terms of the inheritance of acquired
characters. Three working definitions of Darwinism, Lamarckism
and Weismannism (or neo-Darwinism) are suggested in Table 3.1.
Two internal problems with the Lamarckian theory of the inheritance of
acquired characters – even in the social sphere – are that we further require
Objections and explanations 53
Table 3.1 Definitions of Darwinism, Lamarckism and Weismannism
Term Definition
Darwinism A causal theory of evolution in complex or organic systems,
involving the inheritance of genotypic instructions
by individual units, a variation of genotypes, and a process
of selection of the consequent phenotypes according to
their fitness in their environment.
9
Lamarckism A doctrine admitting the possibility of the (genotypic) inheritance
of acquired (phenotypic) characters by individual
organisms in evolutionary processes.
Weismannism A doctrine denying the possibility of the (genotypic)
(or neo-Darwinism) inheritance of acquired (phenotypic) characters by individual
organisms in evolutionary processes.
7 The idea that Lamarck’s theory necessarily involves organisms willing their own
adaptations probably originally emanates from a 1830 caricature of Lamarck’s views by G.
Cuvier (R. Richards, 1987, p. 63). It does not derive from Lamarck himself (Burkhardt,
1984, pp. xxx–xxxi).
8 The genotype is the genetic coding of an organism. The phenotype is the organism’s
behavioural propensities and manifest attributes. The phenotype is an outcome of the
genotype and the organism’s environment.
9 But even this proposition should be treated with extreme care. A problem is defining what
we mean by a ‘characteristic’ in the social domain and what is to be treated, by contrast, as
analogous to the gene (Hull, 1982; Hodgson, 2001b; Knudsen, 2001).
9 But even this proposition should be treated with extreme care. A problem is defining what
we mean by a ‘characteristic’ in the social domain and what is to be treated, by contrast, as
analogous to the gene (Hull, 1982; Hodgson, 2001b; Knudsen, 2001).
an explanation of (a) what inhibits or prevents injuries or other disadvantageous
acquired characters from being inherited, and (b) why organisms
seek to adapt to their environment. Lamarckism simply assumes that only
advantageous acquired characters will be inherited. In addition, some
Lamarckians presume a voluntarism of will, but the origin of this will itself
remains unexplained. A causal explanation of why organisms strive for
advantage or improvement is lacking.
As Richard Dawkins (1983, 1986), Helena Cronin (1991), Henry Plotkin
(1994) and others have pointed out, these gaping holes in Lamarckian theory
have to be filled by a Darwinian or other explanation. Darwinian natural
selection helps to explain how advantageous characters are favoured.
Organisms seek to adapt to their environment in terms of the production of
variations of genotype, leading to different behaviours, some of which involve
successful adaptations. Upon these varieties, natural selection does
its work. Even if acquired characters were inherited, natural selection
would be required to ensure that the advantageous rather than the disadvantageous
characters were passed on. Even if it is valid, then Lamarckism
requires Darwinism as an explanatory crutch.
Insofar as organisms are purposeful, this capacity too has evolved
through natural selection. Darwinism thus points to an evolutionary explanation
of the very origin of will or purpose itself. Hence overall, Darwinism
is a more general and powerful theory than Lamarckism. If social
theory can be legitimately described as Lamarckism, in the sense of admitting
the possibility of inheritance of acquired characters, then this
Lamarckism must be nested within a Darwinian theory (Hodgson, 2001b;
Knudsen, 2001).
Accordingly, Lamarckism is not an alternative to Darwinism, even in
the social sphere. It is erroneous to see them as rivals because Lamarckism
depends on Darwinian natural selection to complete its explanations. Even
if we can talk of acquired characters being inherited in the social domain –
and this idea itself is far from straightforward (Hull, 1982; Hodgson 2001b)
– then this does not undermine the greater explanatory importance of Darwin’s
theory.
Darwinism does not exclude intentionality
Another frequent objection to the use of these ideas in social science is that
Darwinian evolution is ‘blind’ and ignores the conscious intentions and
plans of human individuals. Because intentionality is a vital concept for
the social sciences, this may seem the most important objection, and
thereby its rebuttal is vital. Famously, in response to Alchian (1950), Edith
Penrose (1952) argued that Darwinian theories of evolution excluded the
deliberative and calculative behaviour that was characteristic of human action
in the economic sphere. However, this hard-and-fast distinction between
humans and other organisms is difficult to reconcile with the fact
54 Introduction
that humans evolved gradually from other species. If conscious intentions
are unique to humans, then when and how in evolutionary time were these
cognitive privileges bestowed upon humanity? To avoid a religious or
mystical answer, we have to assume that these cognitive attributes themselves
evolved through time, and existed to some degree in pre-human
species.
A crucial point emerges here. It is part and parcel of Darwin’s underlying
philosophy that all intention has itself to be explained by a causal process.
This causal explanation has to show how the capacity to form
intentions has itself gradually evolved in the human species, and also how
individual intentions are formed in the psyche. For Darwin, natural selection
is part of these causal explanations. There can be no first and
‘uncaused cause’. However, the fact that intentions are somehow caused
or determined does not mean that human agency is any less substantial or
real (Vromen, 2001). Human intentions are part of social reality and social
interactions involve human expectations concerning the intentions of others.
None of these points is undermined by the recognition that intentions
themselves are caused.
From a Darwinian philosophical perspective, all outcomes have to be
explained in a linked causal process. There is no teleology or goal in nature.
Everything must submit to a causal explanation in scientific terms. In
his prescient essay on the impact of Darwinism on philosophy, John
Dewey (1910a, p. 15) wrote: ‘Interest shifts ... from an intelligence that
shaped things once for all to the particular intelligences which things are
even now shaping’. Instead of God creating everything, the Darwinian focus
is on how everything, including human intelligence and intentionality,
was created through evolution. Intentionality is still active and meaningful,
but it too has evolved over millions of years. Likewise, in their textbook
on biological evolution, Theodosius Dobzhansky et al. (1977, pp. 95–6)
wrote:
Purposefulness, or teleology, does not exist in nonliving nature. It is
universal in the living world. … The origin of organic adaptedness, or
internal teleology, is a fundamental, if not the most fundamental problem
of biology. There are essentially two alternative approaches to this
problem. One is explicitly or implicitly vitalistic. … However, … this is
a pseudo-explanation; it simply takes for granted what is to be explained.
The alternative approach is to regard teleology as a product of
evolution by natural selection.
As Daniel Dennett (1995, p. 205) reported, Darwin turned the traditional
doctrine of intentionality upside-down: ‘intentionality doesn’t come from
on high; it percolates from below, from the initially mindless and pointless
algorithmic processes that gradually acquire meaning and intelligence as
they develop’.
Objections and explanations 55
Accordingly, Penrose (1952) and others were wrong to suggest that Darwinian
theories of evolution necessarily excluded deliberative and
calculative behaviour.10
On the contrary, in the social context, deliberation
and selection coexist. Furthermore, as Darwin insisted, intentions, calculations
and preferences have themselves to be explained by the methods of
science. Darwinism invokes both a theory of natural selection and a universal
commitment to causal explanations. This brings us right back to the
aforementioned central lacuna in modern social theory – the widespread
and enduring failure to provide an adequate causal explanation of human
intentionality and human motives.
It might be objected that there is more to human purposefulness than
goal-driven behaviour. After all, ants and robots are purposeful in that
sense. A key point about social interactions is that we gauge and impute
the intentions of others, in order to understand and anticipate their behaviour.
Social action is intersubjective and reflexive. It is very much about
meanings, interpretations of meaning, and imputations of meaning to the
behaviour of others. Regrettably, some enthusiasts of Darwinism have
overlooked these issues. But there is nothing in Darwinism that rules out
their inclusion. On the contrary, if interpretations of meaning and intention
are causally efficacious, then there is a Darwinian imperative to understand
their role. Furthermore, the capacities to think, interact and
interpret have themselves evolved and must also be understood in evolutionary
terms (Bogdan, 1997, 2000).
Some theorists of social evolution believe that the ‘Lamarckian’ is preferable
to the ‘Darwinian’ label because the former preserves human
intentionality. There is a deep irony here, because Lamarck himself, as
noted above, was a philosophical materialist and saw human will as
formed by material causes. No version of ‘Lamarckism’ offers an escape
from the need to provide a causal explanation of intentionality.
Not only is there nothing in Darwinism that excludes or undermines the
reality of human purposes and intentions, but also Darwinism itself, as explained
in later chapters, promoted an emergentist tradition in philosophy
that underlined the status and reality of human intentionality. In contrast,
many so-called ‘Lamarckians’ broke from the materialism of Lamarck and
proposed an unsustainable dualist ontological position where intentions
arose mysteriously from the mind, themselves being incapable of causal
explanation.
It might be objected that the explanation of human motives denies the
reality of choice. This will be discussed further below. At this stage it is
pointed out that causal determination of choices does not imply an absence
of the subjective awareness of choice. Neither does it imply that choice is
56 Introduction
10 In conversations with the present author before her death in 1996, Penrose had revised her
opinion. She was deeply fascinated with evolutionary explanations of human
consciousness, and did not take human deliberation as given or for granted.
unreal. Very small causal influences can have big effects and can thus
cause the individual to act otherwise. Finally, the notion that choice or any
other phenomenon is uncaused is unacceptable, for reasons explored later.
It is not being argued here that every theory or explanation in the social
sciences has to include an explanation of all the motives or preferences of
the individuals involved. No theory can explain everything. For some purposes
and in some circumstances, it can be legitimate to take the purposes
or preferences of the individuals involved as given. All theories involve
abstractions. In some cases it can be legitimate to abstract from the influences
on, and changes in, individual preferences and purposes. In which
case, human intentions become the elemental forces in the particular theoretical
explanation. But even in this case, the assumptions concerning human
intentions should be consistent with what we know about human
evolution and individual development. Furthermore, the use of a simplifying
assumption in one theory does not rule out the need for another theory
to explain those elements that are taken as constant or given. Intentions
and preferences still have to be explained at some stage. Explanations of
human motives should use resources from biology, psychology and anthropology,
as well as from other social sciences.
The ambiguous bogeys of mechanism and determinism
Some social scientists may object that the argument here is ‘mechanistic’ or
‘deterministic’. The social sciences are satiated with rebuttals of ‘mechanistic’
and ‘deterministic’ doctrines, but these words are themselves rarely
and poorly defined. Others, in contrast, enthusiastically take up the idea
that human beings are ‘mere machines’, as a warrant for their version of
scientific enquiry. Their opponents see such ‘mere machines’ statements as
sufficient condemnation of the approaches involved. Yet rarely is the
‘mere machines’ idea clarified and explored further. Both enthusiasts and
critics fail to adequately define their terms.
Consider the words of Richard Dawkins (1976, pp. x, 2, 21–5), who described
humans as ‘survival machines’, ‘machines created by our genes’ or
‘robot vehicles blindly programmed’ to preserve their genes. These
phrases are designed to shock. But all shock and explanatory value is lost
when it is realized that Dawkins did not explain adequately what he meant
by a machine. He would admit that humans have consciousness and purpose,
but provided minimal exploration of the meaning of these terms. Yet
he also repeatedly ascribed ‘ruthless’ will and purposefulness to the ‘selfish
gene’. Dawkins denuded the human individual of purposefulness, but
only by repeatedly ascribing intentionality to the genes. For him, genes are
purposeful but humans are mere machines. But he failed to explain the difference.
A consequence is that the concept of intentionality is undermined.
With both enthusiasts and opponents of ‘mechanistic’ doctrines, rhetoric
triumphs over substance.
Objections and explanations 57
Among those social scientists that emphasize consciousness and choice,
things are only slightly better. Economists typically make a song and
dance about choice. But only the maverick economists reflect upon its substance
and definition (Shackle, 1961, 1976; Buchanan, 1969; Loasby, 1976;
T. Lawson, 1997). Sociologists write of human agency and self-reflection,
but the underlying presuppositions are inadequately explored.
To proceed further, some definitions must be attempted. A provisional
definition of a ‘mechanism’ is a structure involving causal connections but
lacking an adequate capacity for self-reflection and intentionality. A minimal
feature of intentionality is the capacity to prefigure a goal in conscious
thought. Leaving further important questions on one side, it is already
clear, with this rough definition of a mechanism, that the Darwinian theoretical
approach embraced here is not mechanistic. This is because Darwinism
does not deny intentionality, at least in the sense of the existence of
consciousness and prefiguration. Darwinism simply asserts that human
intentionality is itself caused, and in turn it requires some causal
explanation.
In contrast, despite its verbal emphasis on ‘choice’, much of mainstream
economics is mechanistic in the sense of lacking adequate notions of human
self-reflection, intelligence, intentionality or will. This is because human
agents are often modelled as automata, with limited cognitive or
learning capacities, reacting crudely to stimuli from their environment that
are somehow unambiguous. Furthermore, some mainstream economists
claim that the same basic model of human agency, based on utility maximization,
also applies to lower organisms, including ‘honeybees, ants and
schooling fish’ (Landa, 1999, p. 95), or even bacteria (Tullock, 1994). This
simply confirms the observation that the degree of intelligence and self-reflection
found in human beings is not encompassed by a standard and allegedly
ubiquitous utility function. To overcome the limitations of
‘mechanistic’ models of human agency, significant attention must be given
to factors such as the number and complexity of stimuli, cognitive processes,
interpretative ambiguity, and so on (Bandura, 1986; Hodgson, 1988;
Simon et al., 1992; Witt, 2000; Loasby, 2001; Vanberg, 2002).
Other writers see the term mechanistic as denoting something very different,
such as an emphasis on quantitative rather than qualitative factors
of change. Again, the approach adopted in the present work is not mechanistic,
even in this alternative sense. Both qualitative as well as quantitative
changes are emphasized here. Still other writers associate ‘mechanistic’
with an atomistic ontology, in which entities are said to possess qualities
independently of their relations with other entities. Others use the term
‘mechanistic’ to describe systems whose functional specification denies
variation or diversity in the functional parameters. But again, the approach
adopted here is not mechanistic in any of these senses.
The Darwinian ideas that every event is caused, and that even human
motivations have to be subjected to causal explanation, will provoke in
58 Introduction
some quarters the accusation of ‘determinism’. This too is misleading and
at least in some senses mistaken. The very word ‘determinism’ connotes a
confusing multiplicity of meanings. Essentially, there are at least three different
versions of ‘determinism’, as briefly described below.
1 Predictability Determinism. Determinism is sometimes defined as
the epistemological doctrine that ‘any event can be rationally
predicted, with any desired degree of precision, if we are given a
sufficiently precise description of past events, together with all the
laws of nature’ (Popper, 1982, pp. 1–2).
2 Regularity Determinism. A different definition of determinism is the
notion that any given set of circumstances and state of the world must
lead to a unique outcome: ‘given A, B must occur’ (Blanshard, 1958, p.
20). Regularity determinism involves a denial of randomness and
chance in the universe. This is an ontological rather than an
epistemological notion: it says nothing about what we may be able to
know or predict.
3 The Principle of Determinacy. Another definition of determinism is
the notion that every event has a cause (Urmston, 1989). This is again an
ontological statement about the world, otherwise known in
philosophy as ‘the principle of universal causation’ or ‘ubiquity
determinism’. As Mario Bunge (1959, p. 26) put it, the ‘principle of
determinacy’ means: ‘Everything is determined in accordance with
laws by something else’.
These crucial differences require that the ambiguous word ‘determinism’
be defined whenever it is used. The principle of determinacy is central
to Darwinism and is adopted here.11
But the other two versions are rejected.
There is nothing in Darwinism that involves any commitment to the first
two versions of determinism. Furthermore, these three versions of determinism
are logically independent: one does not flow from the other. Predictability
determinism – the dream of Laplace – is itself countered by the
realization of analytical and computational limits in the face of complexity,
and even of the limits of mathematics itself (Gödel’s Proof), and more recently
by theories of computability, chaos and complexity. There are nonlinear
systems with such a high degree of sensitivity to initial conditions
that no amount of accurate measurement of the appropriate parameter values
can provide a sufficiently accurate prediction (the Butterfly Effect).
Predictability in the human domain is also confounded by the logical problem
of predicting future knowledge or creativity. If prediction led us to
Objections and explanations 59
11 Note the difference with Hodgson (1993) where I was equivocal over the adoption of the
principle of determinacy. I was then needlessly worried that this principle might be
incompatible with free will and genuine choice. Bunge (1959), Bhaskar (1975), Earman
(1986) and Auyang (1998) have rightly argued that predictability determinism is quite
different from the principle of determinacy.
know future knowledge, then it would be present knowledge, not knowledge
confined to the future (Popper, 1960, 1982).
The principle of determinacy does not mean that the future is inevitable,
at least in the sense of unavoidable. As Dennett (2003) pointed out, knowledge
of causal determination enhances rather than diminishes the possibility
of avoiding an outcome. Determinacy does not mean inevitability.
The principle of determinacy does not rule out the possibility of statistical
determination, where effects are stochastic but with regular statistical
properties. Statistical laws are still laws. If outcomes were statistically determined,
and statistical determination was not merely apparent but real,
then the second proposition – regularity determinism – would strictly and
generally be false. But the third proposition would not be undermined.
Even if determination is not statistical but links one set of causes with
one set of effects, then there are still objections to regularity determinism.
Roy Bhaskar (1975) rightly rejected regularity determinism on the grounds
that it would work only if it were confined to a closed system, and most
systems are in fact open. The possibility of exogenous disturbances undermines
regularity determinism in specific systems.
Would regularity determinism apply to the universe as a whole? Given
that the universe is interconnected and systems are open, the regularity
‘given A, B must occur’ could not be specified A corresponded to a complete
description of all the possible influences on B, from throughout the
universe. In practice, the statement ‘given A, B must occur’ will itself be indescribable
in its massive scope and complexity. Strictly, with unlimited
interconnectedness, the ‘given A, B must occur’ statement will only pertain
if A is a complete description of the state of the universe. The idea of regularity
determinism cannot apply to any limited description of the world,
and complete descriptions are unattainable. Consequently, even if regularity
determinism applied to the universe as a whole, it would offer little
epistemological guidance for science.
Having rejected or disabled the first two versions of determinism, the
(third) principle of determinacy is retained. Indeed, it is a necessary foundation
for science. A theological definition of a miracle is something that
happens without a scientifically explicable cause. If science admits the possibility
of an event without a cause, then it has abandoned its own mission.
We can retain a broad view of the nature of science, but the quest for meaning
and explanation is indispensable to any version of the scientific enterprise.
Of course, we cannot prove the unfeasibility of an uncaused cause. In
general, proofs of causality, or of its absence, are impossible. But science is
nevertheless obliged to search for causal explanations, and determinacy
must thus be assumed. In many circumstances, prediction will be impossible.
Nevertheless, the quest for some kind of causal explanation must remain.
To behold a first and uncaused cause is to issue licence at that point
to abandon the quest.
60 Introduction
Darwinism is thus incompatible with the idea of George Shackle (1976)
that human intentionality is an ‘uncaused cause’. A problem with
Shackle’s position is that it involves an investigatory closure. Once we affirm
an ‘uncaused cause’ we say that science should explain this much, but
no more. We may move so far down the causal chain, but no further. We
arrive at a causal and explanatory roadblock, policed by the adherents of
the ‘uncaused cause’. Admittedly, all ontological commitments involve
dogma in the sense that they cannot be directly verified by experience. But
the principle of determinacy is preferable to the ‘uncaused cause’ in that it
does not place dogmatic bounds on the scope of scientific enquiry and explanation.
The preferable ontological commitment is one that rules out
miracles and denies any no-go zones for science. The roadblock must be
opened, even if the road ahead is treacherous and complex.12
How can a first and ‘uncaused cause’ be compatible with the recognition
that other outcomes are caused? How is this special causal void to be
explained? Is it ubiquitous to nature, or does it lurk merely in a mammalian
neural system? Or is it unique to humans? How can evolution explain
its sudden appearance?
The uncaused cause is sometimes defended as the requirement of real
choice. If our choices are determined, how could we have acted otherwise?
Choice may be seen to lack substance if there was no alternative. But crucially,
two situations of choice are never identical in all details. Even if two
situations are very similar, we could act differently because of slight influences
with sufficiently magnified effects. For instance, the very fact that we
are reflecting upon the possibility of ‘acting otherwise’ may be sufficient
for us to make a different choice. A multiplicity of conflicting causal influences
acts upon our decisions, with complex feedbacks and interactions. In
some cases, our own deliberations can have big effects. Complexity, emergence
and sensitivity can make choice real, despite the fact that it is
determined.
Dewey (1894, pp. 338–9) notably responded to the proposition of an
uncaused ego with the insistence that ‘it becomes necessary to find a cause
for this preference of one alternative over the other’. He continued: ‘when I
am told that freedom consists in the ability of an independent ego to
choose between alternatives, and that the reference to the ego meets the scientific
demand with reference to the principle of causation, I feel as if I
were being gratuitously fooled with’. For Dewey, in full Darwinian spirit,
the need for causal explanation could not be abandoned.
Some authors argue that if our will is determined then we can hardly be
held responsible for our choices and our actions. It is alleged that as a result
Objections and explanations 61
12 My own position has changed on this issue. Hodgson (1993) failed to decisively reject the
notion of an uncaused cause. However, I also noted that chaos theory suggests that even if
the world is deterministic, it may appear as entirely spontaneous and free. I now believe
that the admission of the possibility of an uncaused cause is not only unnecessary but also
untenable, for the reasons given here.
of such ‘determinism’ there can be no basis for morality or law. Two brief
responses are appropriate here. First, if our will is determined, then moral
pressure and legal sanctions still can have an effect on our actions. Consequently,
there is no ground for abandoning morality or law. Second, even
if, on the contrary, our will was an ‘uncaused cause’, then we would be no
more responsible for the capricious and spontaneous processes that led to
our actions. The ‘uncaused cause’ adds nothing extra to the importance of
morality or law. They are important in any case.
For example, a philosophically minded murderer might claim that his
decision to pull the trigger of his gun was caused by events beyond his control.
Another might claim that her intention to murder appeared spontaneously
(or uncaused) in her mind. The first murderer is just as responsible
for the murder as the second. In both cases the prosecution would argue
that the (caused or uncaused) inclination to kill should have been resisted
by the murderer, so that the murder did not take place. The principle of determinacy
does not diminish the burden of individual responsibility.13
The position adopted here does not rule out some notions of novelty,
nor even of ‘free will’. A number of philosophers – including the Greek
scholastics, David Hume and Jean-Paul Sartre – have argued that an idea
of free will is compatible with the principle of determinacy. In philosophy,
this position is known as ‘compatibilism’. It upholds that even if our
choices are determined then that does not rule out the reality of the process
of choice. It is beyond the scope of this work to establish the possibility of
compatibilism. I simply note that the Darwinian position stated here admits
ground for the ‘compatibilist’ argument that ‘choice’ and ‘free will’
can be reconciled with the proposition that every event is determined. The
human will is a real cause, but it is a proximate rather than an ultimate
cause (Mayr, 1982).14
If novelty simply refers to unpredicted outcomes then we have no difficulty
admitting such possibilities, even if every event is caused. We now
know from chaos theory that even if every event is determined, the world
is still often unpredictable. Randomness and apparent indeterminacy remain.
Novelty may be caused, but it will often appear as entirely spontaneous
and free. Prior causes always exist, but the complexity of the system
may make them especially difficult to identify. In open, complex, non-linear
systems all sorts of novelties are possible. What are ruled out of the picture
are novel effects that do not themselves obey actual scientific laws.15
62 Introduction
13 For a discussion of related ethical themes, in the context of Darwinism, see J. Richards
(2000).
14 On compatibilism see Sterba and Kourany (1981), Dennett (1984), Honderich (1993) and
Vromen (2001).
15 On the compatibility of novelty with the principle of determinacy see Bunge (1959, ch. 8).
Vromen (2001) rightly argues that evolutionary theory cannot be rejected on the grounds
that it fails to predict novelty. If novelty involves unpredictability then it is unpredictable
by any theory.
A note on causation
Aristotle identified four types of causality: the formal, material, efficient
and final. His notion of ‘cause’ was somewhat broader than the modern
meaning. It included assertions of the origin, nature, form and material
constitution of a phenomenon. In much modern usage, a narrower sense of
the word is invoked: causes are always taken to mean the specific factors
leading to an effect. As Jochen Runde (1998, p. 154) put it: ‘a cause of an
event [includes] anything that contributes, or makes a difference, to the realization
of that event in one or more of its aspects’.
Within this narrower and more dynamic meaning of ‘cause’, two of Aristotle’s
categories remain: ‘efficient’ and ‘final’ causality. Efficient causality
is similar to the concept of causality in the modern natural sciences. The
word ‘efficient’ here does not necessarily refer to an optimal (or any other
particular type of) outcome. It simply means capable of having an effect.
Final causality, or ‘sufficient reason’, is teleological in character: it is directed
by an intention, purpose or aim. Hence, within this narrower notion
of a cause, Aristotle promoted a causal dualism. Much later, René Descartes
retained a similar division, with his dualistic separation of physical
matter from the independent, volitional and supposedly immaterial human
soul.16
This distinction persists in modern thought, where the natural sciences
embrace descriptions of cause and effect involving matter and energy, and
the social sciences find their causal fuel in human intentions, purposes or
beliefs. The compartmentalization of the natural from the social sciences
encourages a form of dualism with two different conceptions of cause.17
Here the attempt at reconciliation proceeds in materialist terms. The
modern natural sciences admit no cause that does not involve the
rearrangement or transformation of physical matter or energy. We may
call this the ‘materialist condition of causality’.18
According to this view, all
causes involve movements of matter and transfers of energy or momentum,
as a necessary but not sufficient condition of it being a cause.
Objections and explanations 63
16 See Bunge (1980) and Stich (1983, 1996).
17 Mainstream sociology has typically taken it for granted, or even by definition, that ‘action’
is motivated by reasons based on beliefs. Others have criticized the adoption of such a
‘folk psychology’ that explains human action wholly in these terms. The critics point out
that such explanations are a mere gloss on a much more complex neurophysiological
reality. These dualistic and ‘mind-first’ explanations of human behaviour are unable to
explain adequately such phenomena as sleep, memory, learning, mental illness, or the
effects of chemicals or drugs on our perceptions or actions (Bunge, 1980; Stich, 1983; P. M.
Churchland, 1984, 1989; P. S. Churchland, 1986; A. Rosenberg, 1995, 1998; Kilpinen, 2000).
18 Aristotle’s concept of ‘material cause’ was quite different. It referred to the material makeup
of an object as part of the explanation of its nature. There is a striking lack of consensus,
even among philosophically inclined social scientists, over the terms involved here. An
alternative to imposing the ‘materialist condition of causality’ would be to use Aristotle’s
phrase ‘efficient causality’. But the intention here is to escape from the Aristotelian causal
framework. Furthermore, ‘efficient’ can easily be misunderstood, especially by
economists.
In philosophy, the precise definition and logical form of a causal statement
is highly complex and still unresolved (Sosa and Tooley, 1993;
Salmon, 1998). The ‘materialist’ condition imposed here, however, is to
some degree independent of this philosophical debate. Although, at least
at the social level, we cannot understand causality completely in terms of
identifiable material relations, all relevant causal relations involve movements
of matter and transfers of energy or momentum. In physics and elsewhere,
causes are not fully understood, but all causes satisfy this
materialist condition of matter–energy transfer.
Intentions can satisfy the materialist condition of causality if intentions
are understood as involving transfers of matter or energy, including at the
neural level. Indeed, any action or communication involves movements of
matter and transfers of energy or momentum. Notwithstanding the fact
that they are caused, intentions themselves are causes. Intentions are real
but do not require an entirely different kind of causality. The causes and effects
of intentions have to be explained, in terms that include the important
role of mental prefiguration and judgement. They are special causal mechanisms
but not an entirely separate (teleological) fundamental type of
cause.
The fact that the sciences are still saddled – well over two millennia after
Aristotle – with more than one version of causality, is rarely a subject for
discussion. The wall between the natural and the social sciences has
averted us from this question. Yet when dialogue does occur between the
biologist and the social scientist then the problem emerges.
The development of quantum physics – particularly in the so-called Copenhagen
interpretation – has sometimes prompted a rejection of causal
explanation in the terms outlined here. However, statistical determination,
as expressed in probabilities, does not imply the absence of a cause.
Charles Sanders Peirce gave the name ‘tychism’ to the doctrine of the
probabilistic nature of causation. However, even if outcomes are stochastic,
statistical determination is still involved. Statistical determination or
tychism does not mean indeterminacy. Quantum physics does not necessarily
lead to an abandonment of some standard principles of causal determination
(Bunge, 1959).
Quantum physics may be consistent with a non-statistical version of
causal determination. We may not be able to offer precise predictions of
the motion of subatomic particles simply because of our ignorance of all
the causes that bear upon them. As Bertold Brecht had a character explain
in his play Me-Ti: ‘Their movements are difficult to predict, or cannot be
predicted, only because there are too many determinations, not because
there are none.’ The same may be true of the subatomic quanta. Albert Einstein
and others were concerned about the abandonment of strict causation
by the quantum physicist for some form of stochasticity, and remarked
that he could not believe ‘in a God who plays dice’. Einstein retained faith
‘in complete law and order’. Consequently, Einstein et al. (1935) argued
64 Introduction
that explanations of quantum phenomena had to be completed by the addition
of ‘hidden variables’ (Salmon, 1998). This issue remains controversial
in physics (Cushing, 1994).
Determined, non-linear systems can simulate stochastic behaviour. Ian
Stewart (1989) has thus conjectured that chaos theory can thereby bridge
the gulf between the apparent randomness of the quanta and the operation
of causal laws. The throw of a die leads to apparently random effects, but
that does not rule out the outcome being a unique result of a specific combination
of prior circumstances and events. Similarly, random number
generators in computers use multiplication and numerical truncation to
generate a series of (pseudo-)random numbers. Apparent stochastic behaviour
may be an aggregate outcome of non-probabilistic causal processes
operating at lower, micro levels.
However, for the purposes of this book, no ruling is necessary, or will be
made, over the admissibility or otherwise of statistical or stochastic determination.
The minimum core ontological position maintained and defended
here involves the principle of determinacy or universal causation
(proposition three above), and a commitment to a singular overall type of
materialist causality that can connect diverse domains. The principle of
statistical determination is ruled neither out nor in. The rise of chaos and
complexity theory, has given a recent fillip to Einstein’s proposition that
strict rather than statistical determination is everywhere at work. But it is
not necessary to take a position on this here.
Forward to the past
The previous chapter pointed to an evolutionary approach, in which the
transformation of agents as well as structures is explained. The case is
made that this approach should be Darwinian, in that social phenomena
are also subject to principles of variation, replication and selection. This
evolutionary and Darwinian approach to social theory recognizes emergent
properties in the social domain and does not attempt to explain the social
in entirely biological terms. It does not see evolution as an optimizing
or teleological process. Evolution occurs on multiple levels, including the
social as well as the biological. It emphatically includes human
intentionality but sees its emergence as subject to a causal and evolutionary
explanation. Human agency is a cause, but it is a cause in turn that has
to be explained. Darwinism involves more than variation, replication and
selection – it invokes an unrelenting search for causal explanations.
Remarkably, such an approach to social theory was partly developed in
the 1890s, particularly in the United States, but also drawing on publications
from Germany, Britain and France. The works of a connected group
of British thinkers including Charles Darwin, Thomas Henry Huxley,
George Henry Lewes and Conwy Lloyd Morgan were of particular importance.
Within a few years, all of the ideas described in the preceding
Objections and explanations 65
paragraph were present. Yet this extraordinary episode in American social
theory is largely neglected by the social theorists of today.
One of the principal authors of this Darwinian transformation of social
theory in America was Thorstein Veblen. His work synthesized the evolutionary
theory of Darwin, the instinct–habit psychology of William James
and William McDougall, and the pragmatist philosophy of James and
Charles Sanders Peirce. Fatefully, however, Veblen did not build up his insights
into a systematic and comprehensive treatise. Nevertheless,
Veblen’s works were widely cited, and he influenced a number of other
important thinkers. The evolution and nature of this brilliant, neglected
but incomplete synthesis is explored below.
However, as explained in Parts III and IV, the intellectual environment
of American academia began to change rapidly and substantially after
1914. In philosophy, pragmatism was displaced by positivism, and instinct–habit
psychology was replaced by behaviourism. Meanwhile, the
inheritors of Veblen’s legacy failed, partly because of the less favourable
intellectual environment in the interwar period, to complete a theoretical
system for institutional economics.
Remarkably, however, the present philosophical and psychological environment
is more conducive to the revival and development of the original,
Veblenian project. This book is an attempt to contribute to this task.
The next two chapters examine some key issues that are related to
Veblenian thought. One chapter discusses the impact of Darwinian and
Spencerian evolutionary ideas on psychology and the social sciences. The
other addresses the concept of emergence and some of its implications.
Chapters 6 to 10 will focus more specifically on Veblen’s contribution.
66 Introduction