Agentive Metaphors, The Selfish Gene and Puritanism about Teleological Concepts

Agentive Metaphors, The Selfish Gene and Puritanism about Teleological Concepts

Speaking metaphorically is a janus-faced speech act. Its use recommends itself because a good or apt metaphor can be, in one familiar sense, a source of unexpected insight and it is, as Thomas Hobbes pointed out, a lust of the mind to come to understanding-seeking knowledge, a lust he famously compared with ‘the short vehemence of carnal pleasure’ (see Gopnik 1998 for an exploration of the distinctive phenomenology of explanation). Their use is therefore also hard to resist. But in another sense, their use is essentially contestable because potentially misleading when an unintended reader or audience, or even the intended reader or author herself, does not always appreciate the cognitive consequences of speaking metaphorically. The case is worsened when literal and metaphorical use of words and concepts are mixed up in one complex explanatory narrative. A perspective or way of seeing things that offers insight may thus unintentionally result in a massive, persistent and glaring ‘un-insight’. The lust of the mind to come to understand things and to explain has a cognitive price.

That is, I suggest, what happened when Richard Dawkins, in The Selfish Gene, held that genes are selfish, and added, in one fell swoop, that ‘we should try to reach generosity and altruism, because we are born selfish’ (Dawkins 1976, p. 3). The full context in which the statement occurs is less known:

The argument of this book is that we, and all other animals, are machines created by our genes. Like successful Chicago gangsters, our genes have survived, in some cases for millions of years, in a highly competitive world. This entitles us to expect certain qualities in our genes. I shall argue that a predominant quality to be expected in a successful gene is ruthless selfishness…. If you wish … to build a society in which individuals cooperate generously towards a common goal, you can expect little help from biological nature. Let us try to teach generosity and altruism, because we are born selfish. (Dawkins 1976, pp. 2-3)

But it is hard to see, in view of what he writes about the selfish gene in the rest of the book, how teaching generosity and altruism could be an effective antidote against our being born selfish, given that the selfish nature of genes turns out to be a biochemical property of genes – they tend to faithfully reduplicate themselves, and it is at that level where the biological explanatory work is done. As James (2011, p. 32) put it: the idea is that if you want to understand the workings of a big scandal, you follow the money. In biology, when you want to understand the workings of evolution, you follow the genes. But how could moral education (‘teaching generosity and altruism’, writes Dawkins) undo the effects of a distinctive biochemical property of genes – to more or less faithfully reproduce themselves? Something must have gone wrong here.
I explain the source of misunderstanding created by Dawkins’ metaphor with the help of insights from theories of metaphors developed in cognitive linguistics. In section I the context of Dawkins’s book is briefly presented. In section II we sketch a model of how metaphors work and connect the issues with the role of intentional and teleological concepts in evolutionary explanations and the heuristics of evolutionary theory – roughly, the way we proceed when we study the function of a trait F within a population. In Section III we tentatively defend puritanism with respect to the deployment of teleological concepts and agentive metaphors in biology. Liberalists like Michael Ruse and Daniel Dennett often suggest that the agentive metaphors are practically indispensible, given their important heuristic value in contexts of discovery. I formulate some doubts in this paper. What Ruse (in a review of Dawkins’ book) called the ‘most brilliant metaphor of the 20th century’ may also have been the most misleading one. I draw on recent work by Peter Godfrey-Smith to underwrite puritanism in evolutionary thinking.

I. The levels of selection controversy:
I begin with a brief sketch of the theoretical background of Dawkins’ book. In The Selfish Gene Richard Dawkins tackled an important and still disputed issue in evolutionary theory. Ever since Darwin composed The Origin of Species, the ‘units of selection’-question has always been one of the most fundamental ones in evolutionary biology (see Okasha 2008). In an evolutionary explanation of why the average speed in a zebra population has increased over time, or why the giraffe has a long neck, the unit of explanation, the ‘unit of selection’, if you want, is the organism; the differential survival and reproduction of individual zebras or giraffes causes evolutionary change from one generation to the next. But there are other possibilities. Since the principles of evolution can be formulated independently from what they operate on, it will operate on any entity that exhibits ‘heritable variation in fitness’ (Lewontin 1970), and many entities satisfy these conditions, including families and groups, and, looking ‘downwards’, cells and genes. The ‘Gene’s Eye Point of View’ (not that genes have a point of view!) has its roots in the work of Hamilton (1964), who was concerned with a problem that already bothered Darwin: how could altruism emerge in a system in which the ‘struggle for life’ was the motor behind evolution? Animals that behave altruistically have lower fitness than their selfish counterparts, and yet we observe altruism in many animals. Hamilton’s solution was that if altruism was directed at close relatives, as the altruism of parents to children, and of animals in general to their offspring, the fact that children and offspring share genes with their parents explains everything: we need not explain the effect of the altruism-gene(s) on the agent herself, but also on the agent’s direct relatives. Hamilton found out that the condition required for a gene to spread in a population, was that where c denotes the (biologically describable) cost incurred by the altruist, b denotes the benefit (idem) enjoyed by the recipient and r is the coefficient of relatedness between donor and recipient (Okasha 2008, p. 141). Hamilton’s law could only be precisely formulated after the discovery of the gene as the main (perhaps only) carrier of hereditary material. All the pro’s and con’s of reductionist strategies in science apply to this approach, but that is not the main target of our discussion.

Richard Dawkins further developed this gene-centric approach, stressing that the real question was not how an individual benefits from certain traits or behaviour, but how the frequency f a gene that underlies the trait is affected; from the gene’s perspective, altruistic behaviour is a strategy ‘devised’ by a ‘selfish’ gene, with a view to ‘ensure’ its future ‘propagation’. The concepts between quotes are clearly teleological ones, and that brings us to the topic of this paper: since there is no teleology in nature – evolution is a notoriously blind process – what is the function of these metaphors? Genes are not really selfish, and they do not ‘look at the future’. They even lack ‘a point of view’, and Okasha is very clear about that: the gene’s point of view is ‘a powerful heuristic for thinking about evolution’ (Okasha 2008, p. 142).

It is with these ‘powerful heuristics’ that Dawkins’ troubles began, for in describing the gene as selfish, Dawkins availed himself a concept that plays a key role in another, and related, area of Darwinian theories of evolution. According to the theory of natural selection, behaviour which serves to increase an individual’s fitness will be favored over behaviour which decreases that individual’s fitness. Given that cooperative behaviour tends to result in an individual’s fitness being lower than it would have been had it acted otherwise, how is it that cooperative behaviour is possible, and persists over time? We do not have to answer this fundamental question to immediately see how the concept of selfishness (and its counterpart: altruistic behaviour) is used in this context in a quite literal way: why are we not more selfish than we actually are? (Similar questions could be asked about the closely related question of altruism in sociobiology. Thomas Huxley famously argued for the incompatibility of cooperative behaviour and evolution.) If you tell a narrative, or (even more ambitiously) develop a scientific model that deals with this key problem, you will eventually end up telling a story about levels of selection and genes (if Dawkins’s theory of the gene as the ultimate level of selection is right). But such a narrative, involving selfish genes and selfish behaviour, becomes a hermeneutic time-bomb: why should its selfishness not be conceptually or causally connected with human, manifest selfish behaviour, or even more provocatively, be the basic enabling condition for human selfish behaviour? Reread some of Dawkins’s original quotes, and you will immediately see the hermeneutic mess created by what initially seemed to be an innocent metaphor: "Our genes made us. We animals exist for their preservation and are nothing more than their throwaway survival machines. The world of the selfish gene is one of savage competition, ruthless exploitation, and deceit."

An in the Selfish Gene (cf. supra):
“Like successful Chicago gangsters, our genes have survived, in some cases for millions of years, in a highly competitive world. This entitles us to expect certain qualities in our genes. I shall argue that a predominant quality to be expected in a successful gene is ruthless selfishness. This gene selfishness will usually give rise to selfishness in individual behaviour.”

This is clearly a case in which a metaphor that functions at the core of a philosophical model of a subject matter -– a theory about the level of selection in evolution – lies at the source of a deep misunderstanding. The problem with Dakwins’ choice of the word selfish in characterizing a biochemical property of genes – that they replicate themselves – and its literal counterparts – selfish and altruistic behaviour and actions in humans (and perhaps other higher animals) is that the former designates a natural phenomenon that can be scientifically described in terms of biochemical properties (‘selfish gene’ is a natural kind term), whereas the latter designate, as Michael Tomasello (2009) has stressed, not a single behavioural phenomenon but a pattern of styles of interaction to the benefit of others. A behaviour qualifies as helping if it is an altruistic act that provides instrumental service; informing someone involves the altruistic transmitting and sharing of information (telling the ignorant where the enemy is), and acts of sharing involve the transmission of physical goods (like food or instruments). It is a striking fact that helping, informing and sharing have different ontogenetic trajectories and rely upon different cognitive and affective systems (Warneken and Tomasello 2009), and there is no sense in which the countless manifestations of altruism (and egoism!) among humans could be seen as going ‘counter’ a biochemical phenomenon. Just as our capacity to build and fly planes does not go against the laws of gravity – in fact, it exploits those laws and countless other physical properties of our environment -- so is there no interesting sense in which varieties of egoistic and altruistic actions should somehow be seen to ‘go against’ a fundamental biochemical feature of genes.

Why was it so tempting to use the concept of selfishness, a manifest concept, drawn from common-sense psychology, to describe a biochemical property of genes (and perhaps the Kripkean essence of genes)? In the next section I’ll sketch a popular model of the cognitive function of metaphors; in section III I explain why the source domain, human agency, was an attractive choice.

II. Background model: metaphor as transfer of properties
From a cognitive-linguistic point of view, metaphor is the mental transfer of properties from one domain (the source domain) to another (the target domain) to create a new referential value: some of the term’s semantic properties are selected or abstracted and applied to another domain to designate a new entity in virtue of the properties considered shared by the two referents. The shared properties between the word’s original referent and the metaphorical referents are very abstract, but they constitute a semantic schema that is present throughout all the uses and which founds the semantic unity of the term. One speaks of metaphor when a shift takes place between one particular use, considered as fixing the primary meaning, and another, generally more abstract, through a process of selecting properties which are transferred from the primary domain to the other. However, it is not always easy to say exactly what the primary meaning was from which a schematic form was abstracted and then applied to another domain. it is probable that in certain cases the terms represent an abstract semantic schema from the start which is then applied to different domains: in this case, there would be no shift from a primary meaning to a metaphorical meaning, but from the beginning the word functions in various domains. As Lakoff & Johnson (1980) have shown, in metaphorical transfer one not only transports a ‘form’, but also the inferences linked to the properties of the form. This is an important feature of scientific vocabularies which have frequent recourse to it, notably because it makes it possible to take something as a basis for describing, naming and explaining something unknown.

Erasing the purely heuristic value of the metaphor had considerable and often unwitting consequences, linked to the transfer of inferences. Take the computer metaphor in philosophy of mind and the cognitive sciences. Because of the computer model, human thought was increasingly seen as an autonomous system based on the manipulation of formal symbols (sentences in Mentalese, as Fodor would have it) which could be described in terms of logic and algebra, and everything that did not belong to the traditional domain (like emotions, perceptions, or experiences) was (unintentionally) removed from the picture. (What does an emotion represent? Do experiences have representational content that meshes with representations contents of beliefs?) The problem of meaning in cognitive science has thus unintentionally been reduced to the domain of information processing: how does it take place, where in the brain does it occur and what are its neural correlates? Signification was then treated as a stable product (information to be transmitted), given as input and transformed in a process. The initial metaphor – or better: the initiating metaphor -- had considerable impact on our manifest image of the human mind – and the scientific model it supported (think of the enormous popularity of the ‘artificial intelligence’ research program).

A good discussion of the intriguing combination of the computer metaphor and the metaphor of information as source domain in evolutionary biology can be found in the work of Ernst Mayr (1973), who has given the most detailed treatment of the concept of a genetic program. Mayr has invoked the program analogy in connection with his attempt to legitimate biologist’s use of ‘teleological’ or even ‘intentional’ concepts and language in biology. In order to avoid the connotations of conscious purpose and design in discussion of biological systems and processes, he advocated that biologists substitute the term ‘teleonomy’ and ‘teleonomic’ for the more traditional terms ‘teleology’ and ‘teleological’, but, as Mayr himself recognized, these new terms can only be effective if they have been purged of implications of conscious purpose and design. The parent discipline from which numerous concepts are borrowed in biology and socio-biology is information theory. This allows us to say, for example, that the DNA in a mammalian cell has a capacity of about 2 X 1010 bits of information, which is equivalent to about 100 sets of the Encyclopedia Britannica. But this is, notwithstanding its familiarity and its seductive exactness, still a metaphor: it is an account of the degree of genetic causation or control not on the basis of evidence concerning lawful generalisations that implicitly invokes intentional and/or prescriptive notions: information is always information for someone, who can determine its content, interpret it and misinterpret it, but there is no-one out there to interpret genes or genetic information. That talk about information flow in genetics is inherently metaphorical is often forgotten:

‘Though we have known for many years how genetic systems worked, breaking the genetic code – the structure of the DNA molecule – allowed a rather hard-headed species to see with its own eyes how the life of an animal was controlled by facts of nature as comprehensible as a computer program. The genetic code is a program, and it is a way of transmitting a program from generation to generation.’ (Tiger & Fox 1971, p. 58)

Talk of behaviour as being ‘programmed’ by genes is distinctively metaphorical, and many critics of sociobiology have emphasized the excessive reliance that advocates of genetic determinism have placed upon metaphorical ways of speaking (Dennett (2011) defends the use of the concept of information in genetic thinking).

Back to the selfish gene. In the selfish gene-case the target domain are genes and the source domain are morally evaluable qualities of human actions. What are the deeply entrenched habits of thought and cognitive an affective dispositions that refer us back to that source domain when trying to describe and illuminate the workings of genes and their role in the explanation of the gradual evolution and diversification of life? What explains the attractiveness of the gene as an agent harbouring selfishness?

Before I explore that question, I consider a defender of the use of agentive- and design metaphors in biology in general. Michael Ruse (2003) defends the use of design metaphors:

‘We are faced (in the case of living organisms) with what we have been calling, in as neutral a way as possible, ‘organized complexity’. (…) Whatever you call it, this complexity allows for and indeed calls for understanding in human terms of intentionality, of purpose, of design. (….) for the Darwinian, the heart is made through natural selection, but we continue, metaphorically, to understand it as made by humans. (Ruse 2003, p. 265)

Call this the heuristic view of the use of metaphors (here applied to the central metaphor of design). The claim is, as Paul Sheldon Davies (2009, p. 59) puts it, that ‘absent the metaphor we would fail to ask questions or formulate hypotheses that throw light on the causes of evolution. And if we cannot help using the design stance, why throw it away? ‘Any other stance amounts to a denial of the kind of inquirer we are’ (Davies, p. 60).

Use of design metaphors has its historical, cultural and psychological origins in the design-stance, itself a component of the intentional stance. The term intentional strategy, like that of the intentional stance, was introduced by Daniel Dennett (Dennett 1979, reprinted 1992). Being the social creature we happen to be (a contingent, evolved property of humans) we make sense of each other and allow others to make sense of us in order to arrive at mutual understanding, and this with the goal of coordinating our actions, beliefs, intentions and desires when realizing shared projects:

The intentional stance is the strategy of interpreting the behavior of an entity (person, animal, artefact, whatever) by treating it as if it were a rational agent who governed its “choice” of “action” by a consideration of its “beliefs” and “desires”… The basic strategy of the intentional stance is to treat the entity in question as an agent, in order to predict — and thereby explain, in one sense — its actions or moves. (Dennett 1996, p. 27)

In making sense of each other, we attribute large patterns of contextually appropriate true beliefs and apt desires; false beliefs (we all have false beliefs, and the world can always lead us astray) are identified on a background of shared truths. Design metaphors and agentive thinking are thus not an accidental heuristic tool, but part and parcel of our natural habit to see living things (and sometimes non-living things too!) as exemplifying agency. The intentional stance yields understanding, and is therefore essentially explanatory. What happens when concepts and cognitive habits that evolved to make sense of others and to get yourself understood by others overshoots? Researchers in the cognitive science of religion have pointed out that postulating hidden people and agents is the inevitable byproduct of our chronic search for important agents in a world full of dangerous and unexpected events (Boyer 2001). The fact that this capacity often overshoots is itself quite useful (nature teaches us to be better safe than sorry). What is the phenomenology of the kind of illusion thus created?

To vividly illustrate this human phenomenon, consider how William James experienced the San Francisco earthquake when he visited Stanford University in 1906 (see also Cioffi 1998, p. 95ff):

I personified the earthquake as a permanent invidual entity… Animus and intent where never more present in any human action, nor did any human activity ever more definitely point back to a living agent as its source and origin. All of whom I consulted on the point agreed as to this feature in their experience, ‘It expressed intention’, ‘It was vicious’, ‘It was bent on destruction’, ‘It wanted to show its power’… For science… earthquake is simply the collective name of all the cracks and shakings and disturbances that happen. They are the earthquake. But for me the earthquake was the cause of the disturbances, and the perception of it as a living agent was irresistible. It had an overpowering dramatic convincingness.’ (William James, ‘On some mental effects of the earthquake, Memoirs and Studies (New York 1911, pp. 212-13, quoted in Cioffi 1998, p. 95).

James’ lucid self-observation shows that he is clearly aware of the real causes of the earthquake – the San Andreas Fault – and that his knowledge of the underlying geological mechanisms did not interfere with – did not prevent -- the way he personified the earthquake: his experience of hidden but real agency was in no way undermined by it, nor explained by it. There is, as philosophers would say, an explanatory gap between the tectonic causes and the feeling of agency caused by the experience of the quake. Willam James was, as Jean Piaget later observed, like all of us a ‘natural artificialist’ (Kelemen 2004). But we would definitely not make the mistake to think, as Cioffi points out, that someone who undertakes a seismological study of the San Andreas Fault ought to address himself to the question of what it is like to get caught up in an earthquake (Cioffi 1998, p. 96). Study of the geological phenomena would bear fruit independently of how or whether there was something it is like to experience an earthquake, and one would definitely be misled if one took the feeling of agency that was clearly a component of that experience – James’s sense of agency was, as it were, activated by the quake (did he also experience an agent?) – as a heuristic tool for seismological investigations. There just is no connection between the sense of agency experienced during the quake and a model of the San Andreas Fault, and the experience doesn’t function as a heuristic tool that can later be dispensed with. Needless to add that there is no contemporary discussion among geologists about the role of agentive metaphors in their discipline. There is no Journal of Theoretical Geology in which conceptual wars are fought over the usefulness of design metaphors in the study of seismological phenomena.

William James’s observations should be applied to the way we investigate, and think about, biological and evolutionary phenomena, and I recommend, contra Ruse (2003), a sense of conceptual puritanism. However deeply rooted in our own psychological make-up (Piaget’s artificialism, the false positives as results of overapplication of the intentional stance) and in human cultural history (religious beliefs about the world and its ingredients as being created by deities), there is no reason to assume that these concepts, now recast, or recovered, in metaphorical descriptions, are such that they are ‘necessary for the very existence of evolutionary theorizing’, as Ruse suggests. There is no need to believe in the practical necessity of a metaphor we know to be false, just as the sense of agency experiences by James and other was not necessary for the existence of seismological theorizing (see also Davies 2009, p. 71 for a critique of Ruse).

III. Agentive metaphors and Darwinian Paranoia
Peter Godfrey-Smith recently pointed out that the use of agentive metaphors in evolutionary agentive narratives is not an innocent manoeuvre. Remnants of the pull of the intentional strategy and agentive thinking can be discerned in replicator approaches in evolutionary thinking (as in Dawkins 1976) which are “in many of its presentations, designed to mesh with an ‘agential’ way of thinking of evolution (in which) evolution is treated as a contest between entities that have purposes, strategies and agenda’s” (Godfrey-Smith 2010, p. 10, see Dennett 2011 for critical remarks.) The key difficulty is obvious. We have already seen that the agential perspective is “an uneasy mix of the metaphorical and the literal”, and the agential perspective “engages a particular set of concepts and habits: our cognitive tools for navigating the social world” (idem, p. 10).

Godfrey-Smith points out that Darwin’s achievement replaced older ways of thinking, which were based on (i) an essentialist model of organisms, (ii) the idea that species can be taken to be collectives of individuals that express a type, and (iii) a teleological outlook on biological phenomena. We focus on the third feature. For starters, it was key to the Darwinian view that at least in biology an alternative style of thinking could be developed. Darwinian explanations do not come naturally to us because we have a long cultural history of practical interacting with plants and animals and using them for our purposes, and perhaps also because there is, as Godfrey-Smith points out, “a premium on compact schemata and models with which we can impose order on (nature)” (p. 13). But simple explanations, although cognitively attractive, often create illusions of understanding. Even in the absence of a role for an intelligent designer, a teleological mode of thinking seems unavoidable for our understanding of the mindless principles of evolution by random mutation and natural selection
(Thompson 1995). But even staunch Darwinists tend to present their theories using agentive metaphors, and they often had to face misunderstandings of their key metaphors. The Dawkins/Midgley controversy, for example, was to a large extent due to the fact that Dawkins, at strategic points in The Selfish Gene (the first and last pages of the book) connected the concept of the selfish gene – intended to be a technical term – with issues pertaining to egoism and altruism, thereby developing a narrative strategy that could work only if the very same signifier was given a reading or meaning that made it sensible to think of us, humans as fundamentally selfish. The notorious remark ‘Let us try to teach generosity and altruism, because we are born selfish’ (Dawkins 1976, p. 3) suggested a non-technical, manifest reading of ‘selfish’, while the rest of the book explores genes as the ultimate level of selection; and in that theory, the gene’s selfishness indicates that they only ‘care’ about replication (note how difficult it is to even describe this idea without using agential notions). Richard Dawkins defended his use of the word ‘selfish’ with a Humpty-Dumpty Argument (‘My words have the meaning I intend them to have’):

“Philosophers may object that this kind of definition loses most of the spirit of what is ordinarily meant by altruism, but philosophers, of all people, know that words may be redefined in special ways for technical purposes. In effect I am saying: ‘Provided I define selfishness in a particular way an oak tree, or a gene, may legitimately be described as selfish’. Now a philosopher could reasonably say: ‘I don’t like your definition, but given that you adopt it I can see what you mean when you call a gene selfish’. But no reasonable philosopher would say: ‘I don’t like your definition, therefore I shall interpret your statement as though you were using my definition of selfishness; by my definition your concept of the selfish gene is nonsense, therefore it is nonsense’. This is, in effect, what Midgley has done: ‘Genes cannot be selfish or unselfish, any more than atoms can be jealous, elephants abstract or biscuits teleological’ (p. 439). Why didn’t she add to this witty little list, for the benefit of quantum physicists, that fundamental particles cannot have charm?”

I find this reply unpersuasive. Of course, any signifier can be assigned an arbitrary meaning or be given a new definition. The problem was the metaphor of selfishness, and how it was woven into a narrative in which the manifest concept of selfishness and its counterpart, altruism, had to interact with the technical concept in order to achieve the intended rhetorical effect. It is ironical that while Darwin’s main achievement was to introduce a style of explanation that aimed at eliminating agentive thinking and theorizing about living beings, some of his most popular defenders — defenders who present themselves a staunch opponents of Intelligent Design, for example — reintroduced powerful and potentially misleading metaphors that refer back to agentive thinking. But, as Godfrey-Smith adds, “the feeling that some particular way of looking at things yields understanding should not always be taken at face value, is not the end of the matter.” (p. 13).
Godfrey-Smith locates the attractiveness of the selfish gene model in the agential narrative that surrounds it. When positing an agent in an explanatory story, two explanatory schemata can be developed. The first he calls a paternalist scheme, where the explanation works because a large, benevolent agent is postulated. The agent intends that, as Godfrey-Smith puts it, “all is ultimately for the best” (2008, p. 144). Such explanations postulate gods and spirits, but Hegel’s Weltgeist would also be a good example. The second is the paranoid scheme, where the explainer postulates small, hidden powers and agents. Examples are Freud’s psychology (the unconscious as a realm full of forces that explain manifest emotions, but also the content of dreams), demonic possession narratives, and selfish genes and memes (Farrell 1996). Godfrey-Smith points out that the psychological appeal of such hypotheses “often far outruns their empirical warrant” (idem), and the seductive narrative – seductive because it appeals to the intentional stance of the reader himself -- has a considerable psychological afterlife in the mind of its readers. We suggest that the force of the intentional strategy to create a feeling of insight and understanding (even when it overshoots) explains (but obviously doesn’t justify) this tendency. The paranoid scheme is especially attractive because human minds – a phenomenon that we intuitively regard as of greatest importance – is by nature unobservable (Guthrie 2007). Fuelled by this idea of the mind it is easy for us to imagine unseen controllers at work when we speculate about unobservable mechanisms.
Godfrey-Smith argues that there is a style of selectionist thinking in biology that does not invite paranoia — “the kind of investigation when someone asks: suppose a population was like this, and such-and such a mutation happened, what would happen to that population?” (Godfrey-Smith 2009, p. 145). Call this puritanism about agentive concepts. Puritanism does not invite the idea of hidden agents and genes as “ultimate beneficiaries” for the simple reason that the concepts employed in mathematical models of evolution are thin theoretical ones (“population,” “mutation”) that have no ancestor life in our manifest image, and statistical correlations written out in complex mathematical models are connotation-free statements.
But there is also a negative side to that approach. Theoretical concepts and mathematical equations are all we need to explain how mutations occur, but it is not difficult to appreciate that except for those who are fully immersed in the language of the theory and fully understand its models when written out as sequences of mathematical equations, such formulations do not speak to us (outsiders) in terms we understand, they do not elicit a feeling of having understood things, although the mathematical models say everything there is to say about the subject matter in the context of science. Concepts like “selfishness”, on the other hand, inevitably activate cognitive habits that are constitutive of the intentional strategy. Dawkins wanted to avoid the cold language of theory and to appeal to natural our sense of understanding by using a powerful agentive metaphor, but he obviously didn’t oversee the consequences of mixing, in one and the same narrative, the literal and metaphorical meaning of selfishness. Dennett, in a critical review of Godfrey-Smith’s book, points out that opposition to agentialism was intended to avoid support for creationism. Perhaps a better reason to avoid agentive talk is to admit that the enabling conditions of evolution are, as real evolutionist insist, anonymous biochemical processes, and nothing more than that. It would certainly make it easier to avoid reductionist strategies in morality. (Another difficulty in grasping theoretical concepts and mathematical models is that evolution is gradual, and processes that often extend over thousands of generations are extremely difficult for our kinds of mind to get a cognitive grip on.) Both factors — the connotationless concepts and formulae, and the gradual character of evolution (a important factor already noticed by Darwin as a source of resistance to the evolutionary hypothesis) — at least partly explain why evolutionary explanations freed from intentional concepts and metaphors do not easily yield the kind of epistemic satisfaction we expect from ‘really understanding’ a phenomenon. The real explanation – a bold, naked formula which, if true, is probably a necessary truth about the biological realm -- feels like a non-explanation and doesn’t yield the feeling of insight. But that is perhaps the most profound illusion.

Boyer, P (2001), Boyer, P. (2001). Religion explained: The evolutionary origins of religious thought. New York: Basic Books.

Cioffi, F. (1998), Wittgenstein on Freud and Frazer. Cambridge: Cambridge University Press

Cioffi, F. (1998), Freud and the Question of Pseudoscience. Chicago: Open Court

Davies, Paul Sheldon (2009a), ‘Conceptual Conservativeness: The Case of Normative Functions’, in U. Krohs and Peter Kroes (Eds.), Functions in Biological and Artifical Worlds, Cambridge, Mass.: MIT Press, pp. 127-146

Davies, Paul Sheldon (2009b), Subjects of the World. Chicago: Chicago University Press

Dawkins, R. (1976), The Selfish Gene. Oxford: UP

Dawkins, R. (1981), ‘In Defense of Selfish Genes’,

Dawkins, R. (1986), The Blind Watchmaker. New York: Norton

Dawkins, Richard (1998), Unweaving the Rainbow. London: Penguin 1998

Dennett, Daniel (1987), The Intentional Stance. Cambridge: MIT Press

Dennett, Daniel (1991), Consciousness Explained. Boston, MA: Little, Brown

Dennett, Daniel (1996), Kinds of Minds. Towards an Understanding of Consciousness. New York: Basic Books

Dennett, Daniel (2011), ‘Homunculi Rule. Reflections on Darwinian Populations and Natural Selection’ by Peter Godfrey Smith’. Philosophy and Biology (forthcoming)

Godfrey-Smith, P. (2009), Darwinian Population and Natural Selection. Oxford: UP

Godfrey-Smith, P. (2011), ‘Agents and Acacias: Replies to Dennett, Sterelny and Queller’, in Philosophy and Biology , pp. 501-515

Gopnik, Alison (1998), ‘Explanation as Orgasm’, Mind and Machines 8, pp. 101-118

Guthrie, S. (2007), ‘Anthropological Theories of Religion’, in M. Martin (Ed.), The Cambridge Companion to Atheism. Cambridge: UP

Farrell, John (1996), Freud's Paranoid Quest: Psychoanalysis and Modern Suspicion. New York: University Press, 1996

James, William (19011), ‘On some mental effects of the earthquake, Memoirs and Studies (New York 1911, pp. 212-13)

Hamilton, W.D. (1964), ‘The Genetic Evolution of Social Behaviour I and II, Journal of Theoretical Biology 7, 1-16, 17-32

Kelemen, D. (1999a). Beliefs about purpose: On the origins of teleological thought. In M. Corballis & S. Lea (Eds.), The descent of mind: Psychological perspectives on hominid evolution (pp. 278–294). Oxford, England

Kelemen, Deborah (2004), Are Children ‘inutitive theists’ ? Reasoning about purpose and Design in Nature’, Psychological Science 15, 295-301

Lakoff, G. and M. Johnson (1980), Metaphors We Live By. Chicago: Chicago University Press

Lewontin, R.C. (1970), ‘The Units of Selection’, Annual Review of Ecology and Systematics I, pp. 1-18

Mayr, Ernst (1974), ‘Behavior Programs and Evolutionairy Strategies’, American Scientist 62 (1974), 650-659.

Midgley, Mary (1985), Evolution as Religion. London: Routledge

Okasha, S. (2008), ‘The Units and Levels of Selection’, in Sarkar & Plutynski (Eds.), pp. 138-156

Robert, Stéphane, Words and their Meanings: Principles of Variation and Stabilization, in M; Vanhove (ed.), From Polysemy to Semantic Change. Studies in Language Companion Series 106, John Benjamins Publishing Company 2008, 55-92

Roughgarden, J. (1996), The Theory of Population Genetics and Evolutionary Ecology: an introduction. New York: MacMillan.

Ruse, Michael (2003), Darwin and Design: Does Evolution Have A Purpose? Cambridge (Mass.): Harvard University Press

Sarkar, S. and A. Plutynski (Eds.) (2008), A Companion to the Philosophy of Biology. Oxford: Blackwell

Shtulman, A. (2006), ‘Qualitative Differences Between Naïve and Scientific Theories of Evolution’, Cognitive Psychology 52, pp. 170-194

Searle, John (2007), What is language: Some preliminary remarks. in S.L. Tshatzidis (Ed.)n John Searle’s Philosophy of Language. Force, Meaning and Mind. Cambridge: UP, 15-49

Sober, E. (1998), ‘What is Evolutionary Altruism?’, in D.L. Hull and Michael Ruse (Eds.), The Philosophy of Biology, Oxford: UP, pp. 459-479

Thompson, M. (1995), ‘The Representation of Life’, in R. Hursthouse, G. Lawrence and W. Quinnn (Eds.), Virtues and Reasons. Oxford: OUP, pp. 247-97

Tiger, L. and R. Fox (1971), The Imperial Animal. New York: Holt, Rinehart and Winston 1971

Tomasello, M. (2009), Why We Cooperate. Cambridgde (Mass.): MIT Press

Warneken, F. & Tomasello, M. (2009). Varieties of altruism in children and chimpanzees. Trends in Cognitive Science, 13, 397-402.