2. Deterministic Interpretations

Some of the best-known versions of evolutionary theory are widely taken to be deterministic, and give a convincing impression of being such. Let us briefly consider Richard Dawkins' book The Selfish Gene (1976). Here it is claimed that we human beings are ‘survival machines’, built by those underlying replicators that ‘go by the name of genes’ for themselves to live in. The genes now ‘swarm in huge colonies, safe inside gigantic lumbering robots … you and me’, indeed ‘they created us, body and mind’, and ‘manipulat(e) by remote control’.Footnote ⁶ Further, ‘survival machines began as passive receptacles for the genes’.Footnote ⁷ Certainly, the arrival of consciousness ‘can be thought of as the culmination of an evolutionary trend towards the emancipation of survival machines as executive decision-makers from their ultimate masters, the genes’,Footnote ⁸ but they remain machines, and their decisions and behaviour continue to be genetically manipulated. Indeed ‘all life evolves by the differential survival of replicating entities’, and ‘the gene … happens to be the replicating entity which prevails on our own planet’.Footnote ⁹ However, genetic factors are held insufficient to explain human culture, and for this further replicators are postulated, the memesFootnote ¹⁰; yet the combination of genes and memes appears to supply the determinants of human behaviour, until in the final two paragraphs of the book Dawkins declares that conscious foresight gives human beings (and them alone) the power to resist both genetic and memetic factors. ‘We are built as gene machines and cultured as meme machines, but we have the power to turn against our creators. We, alone on earth, can rebel against the tyranny of the selfish replicators.’Footnote ¹¹ But if we are truly machines, built and manipulated by factors beyond our own control and which, in most cases, preceded our births, how can we be capable of this rebellious independence? The deterministic story told up to this point seems hard to reconcile with this closing ‘note of qualified hope’,Footnote ¹² and remains the overall message of the book.

This problem becomes clearer when determinism is openly acknowledged as implying that at each moment there is only one future. Such an implication is made explicit in Michael Greening's book As It Is; indeed Greening advances the conclusion (at page ix) that given determinism (which he endorses) human self-determination is a myth.Footnote ¹³ But this conclusion would seem to undermine the claim that humanity is capable of rebellion against the determinants that create and manipulate it. (Some further (mostly scientific) problems for Dawkins' variety of Darwinism can be found in Steven Rose in Lifelines: Biology, Freedom, Determinism;Footnote ¹⁴ but they need not concern us here.) Yet another difficulty for Dawkins' variety of determinism lies in the weakness of his theory of memes, introduced to explain human culture, a theory that has come in for serious criticisms from philosophers.Footnote ¹⁵ For if the theory of memes as determinants of culture is rejected, then a wide range of possibilities opens up for human behaviours that may not be inevitable outcomes of determinants at all.

This view, however, could be resisted on the basis of an appeal to the compatibility of determinism and free will (compatibilism). (Following van Inwagen, I use the phrase ‘free will’ in the traditional sense, to convey the claim that human beings (and possibly other animals), faced with choices between two or more mutually incompatible courses of action, frequently have it in their power to (or can) carry out each of them.Footnote ¹⁶) Accordingly I turn to a brief consideration of the stance of a defender of Darwinism who explicitly espouses compatibilism, namely Michael Ruse.

In the course of discussing the relation of human freedom to the laws of nature, Ruse claims that ‘The meaningful sense of free will is not denial of determinism, but autonomy: being able to do something without external constraint’.Footnote ¹⁷ Here Ruse is developing Hume's view that determinism and recognition of human freedom are compatible, which he later endorses.Footnote ¹⁸ Along the wayFootnote ¹⁹ he cites Daniel Dennett's analogy of a Viking spacecraft, which is at once a deterministic device and able to control itself, unaffected by external agencies. Being a deterministic device would not make us unable to control ourselves.Footnote ²⁰ But this freedom from external constraint does not correspond to the freedom that (for example) underlies moral responsibility, because for deterministic devices not more than one future is open at each moment. While such devices may be programmed to be autonomous and independent of interference from external agencies, their very programming means that they lack free will in the sense introduced above; for they are not free to adopt more than one possible course of action. No one is likely to suppose otherwise in the case of spacecraft, but if people are different they should not be understood on a deterministic basis at all. As Elizabeth Anscombe has argued, compatibilist freedom seems far too limited to satisfy our ordinary understanding of agency as expressed in phrases like ‘she could have done otherwise’.Footnote ²¹

Against compatibilism, van Inwagen deploys the following argumentFootnote ²²): ‘If determinism is true, then our acts are the consequences of the laws of nature and events in the remote past. But it is not up to us what went on before we were born, and neither is it up to us what the laws of nature are. Therefore the consequences of these things (including our present acts) are not up to us.’ While he later develops this argument in greater detail, and the argument opens up much scope for discussion, it presents serious difficulties for compatibilism, and thus for compatibilist Darwinists such as Ruse. Besides, there are related problems for compatibilist Darwinists in supplying satisfactory accounts of morality, choice and deliberation and, granted the implausibility of memes, of human culture in general. While much more could be said about compatibilist Darwinism, the prospects for making it coherent and plausible are not encouraging.

Besides, there is a further problem for deterministic Darwinism. Alvin Plantinga has argued, in Warrant and Proper Function, that Darwinism implies that our apparently rational beliefs are all to be explained by factors and forces such as the need to survive and reproduce, and that for such an approach even our most considered beliefs need not be true or reliable, and could easily be deceptive or delusive. In one version of his argument, Darwinism appears to make the reliability of these beliefs low; in another, more cautious version, it simply makes them so uncertain that we have no idea whether to accept or reject them.Footnote ²³ (As he explains, it cannot be replied that our very survival shows that our beliefs must be both adaptive and reliable, since our survival could depend on our behaviour rather than on our beliefs, or even on a combination of false beliefs and additional factors that are accidentally beneficent. In his words: ‘There are indefinitely many belief-desire systems that fit adaptive behaviour, but where the beliefs involved are not for the most part true’.Footnote ²⁴ And in any case we cannot assume that our beliefs have a causal role in promoting either our behaviour or our survival.Footnote ²⁵ Thus Darwinism undermines its own rational acceptability.Footnote ²⁶ This argument seems to be on target where deterministic versions of Darwinism are in question; it is these that apparently have the implications mentioned, since for deterministic Darwinism our beliefs come about simply as a result of natural selection and other factors beyond our control (one of Plantinga's examples of another possible such factor is genetic drift), and they are the beliefs that we could not help adopting in the circumstances. But if so, they are no more trustworthy than quicksands or a Holmsian Grimpen.

Much the same would seem to apply to the HADD hypothesis; deterministic versions of this hypothesis appear equally vulnerable to Plantinga's argument, casting into doubt the credibility of a large range of beliefs and claims about the presence of one or another agent. But Darwinism can instead be held in indeterministic versions that allow room for the exercise of self-determination and autonomous rationality, and for which our beliefs are not simply resultants of factors beyond our control such as natural selection; and it is arguable that this kind of Darwinism (and perhaps only this kind: see below) can withstand Plantinga's criticism. Relatedly, some proponents of the HADD hypothesis recognise that it may better explain some of our reflective beliefs if it is held in a non-deterministic form.Footnote ²⁷

In view of all these difficulties, consideration should be given to non-deterministic versions of Darwinism and of the Cognitive Science of Religion. As Alan Holland has recently remarked, Darwin himself often spoke of animal choice;Footnote ²⁸ besides, as Blackburn confirms,Footnote ²⁹ he did not regard natural selection as the sole cause (let alone determinant) of biological phenomena. And this opens the door to the possibility that interpretations of Darwin that foreswear determinism could yet turn out to be viable.

Secular Non-deterministic Interpretations

As the title of his 1997 book (Lifelines: Biology, Freedom, Determinism) discloses, the biologist Steven Rose adopts an explicit stance on the relation of Darwinism to determinism.Footnote ³⁰ He rejects what he calls ‘biological determinism’ (within which he includes sociobiology), and adopts a materialist but non-deterministic view of living systems: ‘Far from being determined, or needing to invoke some non-material concept of free will to help us escape the determinist trap, it is the nature of living systems to be radically indeterminate, to continually construct their-our-own futures, albeit in circumstances not of our own choosing.’Footnote ³¹ The theme of self-construction later emerges as central. ‘The central property of all life is the capacity and necessity to build, maintain and preserve itself, a process known as autopoiesis.’Footnote ³² This capacity of all life is immediately related to human freedom: ‘This is why it is in the very nature of life and living processes themselves that we, as living organisms and specifically as humans, are free agents.’Footnote ³³

Rose rapidly (and understandably) denies that human beings are free in Sartre's sense of humans having no nature but being what they will. Rather we are free, according to Rose, ‘in the older, Marxist sense of the freedom of necessity. We humans, more than any other life form on earth, make our own history.’Footnote ³⁴ Clearly, Rose rejects the determinist view that at each moment there is only one future, in favour of the view that human beings (and some other animate agents too) select and thus mould their future; and his Darwinism is later explicated correspondingly. A question-mark, however, arises over his phrase ‘the freedom of necessity’, for the use of ‘necessity’ might well be thought to take back what is asserted through his talk of ‘freedom’. Rose does not explain his reference to ‘the older, Marxist sense of the freedom of necessity’, but was possibly alluding to Friedrich Engels' conclusions that ‘Freedom of the will therefore means nothing but the capacity to make decisions with knowledge of the subject’ and that ‘Freedom therefore consists in the control over ourselves and over external nature, a control founded on knowledge of natural necessity’.Footnote ³⁵

Probably Rose means here that our acts of self-construction take place in circumstances determined by factors beyond our control (by the past, by economic circumstances, or generally by our environment), and thus (to return to the passage from page 7) ‘in circumstances not of our own choosing’. Thus it is necessity that is the mother of invention. Yet his choice of phrase appears at least to constrain seriously the scope of choice; for if freedom is possible only in circumstances which are themselves determined (whether socially, economically or in some other way), and basically consists in an understanding of what could not be otherwise, then the agent too must be liable to be moulded by those circumstances, and battered by their unrelenting sequence. Besides, rather than agents being able to bring to those circumstances settled dispositions to choose and reshape the future (since their opportunities for choice could well have been too occasional and too sporadic for such dispositions to be developed), their leverage seems too limited, the scope for choice too narrowly confined, and their window of opportunity too fleeting before circumstances close in again, even if ‘freedom of necessity’ allows of genuine choice at all. Rose would probably reject these apparent implications as misinterpretations, stressing (as he later puts matters at page 153) that: ‘Development is essentially a contructivist process; the developing organism, in its being and its becoming, in its specificity and its plasticity, constructs its own future,’ and that at most junctures the organism is already the product of previous episodes of self-construction. Yet doubts still linger about the overall stability and the coherence of his emergent position. These doubts need not be grounded in some ‘non-material concept of free-will’; they concern rather the coherence of Rose's version of non-deterministic freedom, which could apparently be revised and reinforced so as to be more clearly non-deterministic without those making this move abandoning materialism thereby.

A few years earlier, another distinguished biologist and Darwinian, Richard C. Lewontin, presented a comparable version of Darwinism in Biology as Ideology: The Doctrine of DNA.Footnote ³⁶ After praising Darwin for discerning that organisms ‘adapt or die’, he proceeds to claim that biology has become frozen at the stage of recognising the organism's alienation from its environment, adding that: ‘Modern biology has become completely committed to the view that organisms are nothing but the battle grounds between the outside forces and the inside forces. Organisms are the passive consequences of external and internal activities beyond our control.’Footnote ³⁷ But this implies that the world is beyond our control, and that all we can do is to make the best of it. Even the environmental movement, he suggests, connives at this view of the environment. (Although it is not invariably guilty of this travesty, we may comment, this charge is not far off target.)

But these deterministic views are, according to Lewontin, distortions of reality, and at the same time of familiar experience. For ‘there is no “environment” in some independent and abstract sense. … Organisms do not experience environments. They create them. They construct their own environments out of bits and pieces of the physical and biological world and they do so by their own activities.’Footnote ³⁸ Like Rose, Lewontin rejects creation by the genes in favour of self-creation; unlike Rose, he avoids talk of ‘the freedom of necessity’, appealing not to Marx's analysis of history but to Kropotkins's account of social action and its possibilities. Maybe Rose's remarks about there being material limits to the freedom of living agents, plus a greater stress on Darwin's recognition that organisms need to adapt to circumstances over which they have little control, are needed as a corrective here; but Lewontin's account more unambiguously emphasises the extent of this freedom.

Stances such as those of Rose and Lewontin disclose some of the possibilities for secular non-deterministic Darwinism. Within each there are problems of internal and external consistency, but these problems may well be surmountable, and seem more likely to be surmounted than the deep conceptual problems that beset determinist and compatibilist Darwinisms. Relatedly, the economist Herman Daly is critical of neo-Darwinism as deterministic, believing that this helps deprive its adherents of the possibility of finding meaning in life, and maintaining that only non-deterministic forms of Darwinism uphold this possibility.Footnote ³⁹ Nevertheless, further versions of non-deterministic Darwinism should also be considered, versions having a theistic character.

3. Theistic Non-deterministic Interpretations

In an essay which partly concerns and is sympathetic to the Cognitive Science of Religion, it could appear paradoxical that theistic interpretations are under consideration at all. For it is sometimes assumed that any kind of sympathy towards the findings of this new discipline precludes religious belief, at least in its traditional forms, including theistic ones. However, some of its practitioners have argued, to the contrary, that the Cognitive Science of Religion is both religiously neutral and neutral between world-views, and should not be construed as explaining religion away, or even as seeking to do so.Footnote ⁴⁰ The details of this discussion cannot be entered into here, but it seems reasonable in the circumstances to proceed on the assumption that the Cognitive Science of Religion on the one hand and theistic beliefs on the other are not incompatible. Adopting this assumption clears the way for a consideration of some versions of Darwinism that are both non-deterministic and theistic.

Two versions of theistic and non-deterministic Darwinism are considered here, those of the philosopher and theologian Keith Ward and of the biologist Kenneth Miller. While Ward's position was originally conveyed in God, Chance and Necessity (1996),Footnote ⁴¹ it is conveniently summarised in a later anthology chapter entitled ‘Theistic Evolution’.Footnote ⁴²

Ward accepts that ‘human beings have descended by a process of mutation and adaptation from other and simpler forms of organic life over millions of years’, and that natural selection is ‘the main but possibly not the only driving force of evolutionary change’.Footnote ⁴³ (This passage suffices to attest that Ward is a Darwinian, even though he is reluctant to accept that scientific explanations are sufficient as such to explain reality.) At the same time he advances the thesis that the universe was created by God, recognising, however, that some forms of this hypothesis are falsified by acceptance of evolution, and that versions claiming that God directly creates every single event are among them. Rather than determining every event, God selects a world both of natural laws, of random mutations and of ‘the generally stochastic processes of genetic modification’, and thus ‘chooses chance’.Footnote ⁴⁴

Ward's suggestion of the reason for this choice is that indeterminacy (or ‘lack of sufficient causality’) is ‘a necessary, though by no means a sufficient, condition for the emergence of free and responsible choice on the part of agents generated by the evolutionary process’Footnote ⁴⁵ (Miller also endorses this view.Footnote ⁴⁶). Here Ward implicitly argues that the importance and value of ‘libertarian freedom’ would help explain the need for nature's partial indeterminacy, and thus the creator selecting it. ‘Libertarian freedom is that property of a rational agent by which, on at least some occasions, no prior physical state, even with the addition of a set of general laws of nature, entails one specific outcome of a given situation. There are real alternative possible futures, and only the agent's uncompelled decision decides which future is realised’.Footnote ⁴⁷ Ward relates all this to his beliefs about consciousness, value and obligation, and their cosmic role (holding that the universe was created in order to realise states of intrinsic value),Footnote ⁴⁸ but the detail of his claims need not concern us here. (I have argued elsewhere that a less anthropocentric and more biocentric theory of value might render his axiological claims more cogent.Footnote ⁴⁹) There turn out to be a range of grounds for Ward's variety of theism (see ‘Theistic Evolution’, pages 266–8 and 274), but for present purposes neither such grounds nor their merits need concern us either. It should, however, be recognised that he is aware of potential problems for the coherence of his ‘theistic evolution’ (e.g. of the question of whether the creator's purposiveness can be reconciled with the randomness of the evolutionary process), and that he offers solutions to a range of such problems.Footnote ⁵⁰

The non-deterministic component of Ward's theistic evolutionism (what he calls ‘libertarian freedom’) is available to be adopted or pressed into service by secular non-deterministic Darwinists (such as supporters of Rose or Lewontin), even if Ward's theistic metaphysic is not endorsed; it does not, for example, turn intrinsically on a ‘non-material’ definition of ‘freedom’. Or rather, it is thus available unless the term ‘libertarian’ is construed (or misconstrued) as involving a freedom that is wholly unconstrained. For, construed in this way, free action appears to amount to entirely inexplicable and thus random action, and to be far from the responsible choices that Ward regards as important and central. But, as John Lucas has argued, action that is unexplained in terms of sufficient antecedent physical conditions can be far from random where explanations in terms of the agent's reasons are concerned.Footnote ⁵¹ (And even if David Wiggins' reply to Lucas were to be accepted that a causal account of reasons for belief must still be given,Footnote ⁵² that causal account need not involve sufficient conditions, as opposed to conditions with the role of (say) the necessary conditions of the belief or action for which an explanation is sought.) Thus Ward's ‘libertarian freedom’ need carry no implications or overtones of randomness; rather it concerns decisions of agents who select, albeit in determinate contexts which are beyond their powers of present choice, between two or more genuinely open futures. Indeed it brings to the light what a consistent non-deterministic stance would need to say.

Kenneth R. Miller defends a comparable and congruous version of non-deterministic Darwinism in Finding Darwin's God.Footnote ⁵³ Miller is a biologist with a strong commitment to Darwinism, who rejects the claims of Intelligent Design theorists that aspects of the biological world display ‘irreducible complexity’, incapable of being produced by evolution, and replies accordingly (at pages 129–164) to the claims of Michael Behe made in Darwin's Black Box.Footnote ⁵⁴ But he also rejects the shared view of adherents of Intelligent Design and of materialists such as Lewontin that ‘evolution implies an absolute materialism that is not compatible with religion’,Footnote ⁵⁵ and proceeds to defend in the rest of the book a combination of Darwinism and theistic creation.

Here he expounds the discovery of quantum indeterminacy, and thus the abandonment of determinism on the part of twentieth-century physics. The world's indeterminacy, he suggests (as noted above) is a necessary condition of free will, and a creator wanting to facilitate free will would be likely to generate a world of regularities (which would thus be comprehensible and allow of purposeful action on the part of creatures) and also of indeterminacy (without which there would be no free agency). Thus the world was so fashioned as ‘to allow us the freedom and independence necessary to make our acceptance or rejection of His love a genuinely free choice’.Footnote ⁵⁶ Miller here cites with approval Schrödinger's conjecture that the randomness of mutations may correlate with the indeterminacy of quantum mechanics,Footnote ⁵⁷ adding that ‘the DNA molecule is structured in precisely a way that makes the behavior of individual atoms, even individual electrons, significant’, and that ‘events with quantum unpredictability … exert direct influences on the sequences of bases in DNA’.Footnote ⁵⁸ Thus he would reply to the possible objection that quantum variations even out at the macroscopic level and thus make no difference, that on the contrary ‘Mutations are just as unpredictable as a single photon passing through a diffraction slit’, and that, because these items can be related, ‘evolutionary history can turn on … the quantum state of a single subatomic particle’.Footnote ⁵⁹ If he is right here, then his own overall stance is clearly strengthened, although this interpretation of the implications of quantum physics is perhaps not indispensable to his case.

Miller makes no attempt to argue for God's existence, but maintains that the world is much as we might expect it to be if fashioned by a creator desirous of both universal regularities and the kind of indeterminacy necessary if responsible action and free responsiveness to the creator were to be possible. After discussing various problems for adhering to Darwinism and simultaneously to theism, Miller concludes by reaffirming his belief both in Darwinism, and in the kind of God presupposed in the final paragraph of The Origin of Species:

Ward and Miller are not the only writers to defend theistic and non-deterministic versions of Darwinism. Another such writer is John T. Haught.Footnote ⁶¹ However, enough has been said to illustrate how this pattern of beliefs can be integrated; and, while neither Ward nor Miller would probably accept all the claims and arguments of the other (for example, one is a Protestant and the other a Catholic), the contours of a theistic and non-deterministic Darwinism have sufficiently come to light.

Clearly many problems confront anyone defending the conjunction of theism and Darwinism, but this is not the place to debate them. The focus here concerns instead the coherence of combining Darwinism and indeterminism, and the works of Ward and of Miller are suggestive of how this combination can be defended in a theistic context. Some will find the corresponding secular conjunction (as in Rose and Lewontin) more digestible, particularly when fortified by Ward's definition of ‘libertarian freedom’ and perhaps by Miller's account of quantum indeterminacy and its implications. However, those who, with Ward, maintain that scientific explanations show the phenomena of biology to be scientifically explicable but themselves incomplete without metaphysical underpinning (e.g. of belief in the reality of progress, both in science and in history) may instead find the Ward/Miller synthesis more intellectually satisfying and thus in this regard less unstable. (They would not, of course, be committed to the full range of (say) Ward's views on intrinsic valueFootnote ⁶² or Miller's views on the unacceptability of deism.Footnote ⁶³

4. Non-deterministic Darwinism and HADD

It remains to explain how non-deterministic Darwinism of the kind favoured by Ward and Miller escapes Plantinga's argument. As was mentioned earlier, where this kind of Darwinism is concerned, our beliefs are not all simply resultants of factors beyond our control such as natural selection, for they are frequently mediated by autonomous rationality and reflection. The worry embedded in Plantinga's argument (Plantinga calls it ‘Darwin's Doubt’) is that our beliefs are thrown up by natural selection (and other evolutionary processes beyond our control) and geared to our survival and to our production of surviving offspring in a manner that does not in any way depend on or imply (let alone guarantee) their reliability or truth, and which relatedly precludes our adopting any other beliefs instead; they have no greater prospect of reliability (to use Darwin's way of expressing this worry) than the beliefs (on the same matters) of a monkey, which may well be geared to monkey survival, but need not be geared in the least to verisimilitude, let alone to truth.

Maybe Darwin was underestimating the autonomy and reflectiveness of monkeys, but this problem, in any case, relates most obviously to deterministic versions of Darwinism for which our beliefs are simply the resultants of natural selection and other evolutionary factors beyond our control (as he probably considered the beliefs of monkeys to be), and thus only apparently up to us.

Plantinga's argument, however, should not be construed as succeeding in undermining the reliability of the generality of our beliefs simply on the basis of Darwinism and evolutionary theory in general. It casts doubt on the reliability of our (apparently) rational beliefs only if a deterministic version of Darwinism and of evolutionary theory is in question, for it is such versions that imply that the adoption of no other sets of beliefs was open to us.

Non-deterministic Darwinism, by contrast, can maintain that our beliefs, despite being formed against a background of natural selection and being subject to various necessary causal conditions (and thus not the unconstrained reflections that existentialist accounts of freedom might lead us to expect), are (in many cases) the conclusions of autonomous rationality, and could have adopted a different content. (The reflectively held belief of Darwinists in Darwinism might comprise a relevant example.) As such, there is no reason not to regard them, despite all the causal influences to which we are exposed, as just as much the outcomes of autonomous thought and of active processes of reflection and/or deliberation as we normally suppose, and as capable of being reliable accordingly. (Our reflective beliefs may still sometimes fail to be reliable and turn out to miscarry, but the Darwinian factors indirectly involved in producing or facilitating them need not be regarded as casting doubt on their reliability. Besides, their very liability to intelligent error underlines the underlying possibility of their being rational and reliable.) For they are not simply the products of factors and forces beyond our control, since we could have formed other beliefs, and since our thinking could have issued in more than one set of cognitive states or outcomes.

Indeed, short of an independent argument to the contrary, it will be such non-deterministic versions of Darwinism alone that allow for our beliefs to be both autonomous and potentially reliable; for deterministic versions of Darwinism, implying as they do that no other beliefs than those actually adopted could have been adopted, preclude these possibilities. Accordingly (and in the continuing absence of such an argument) non-deterministic forms of Darwinism (unlike deterministic forms) seem to elude Plantinga's argument. The versions of Darwinism proposed by Rose and Lewontin, and perhaps more clearly by Ward and Miller, continue to explain most biological phenomena through selective pressures, without making the beliefs and actions of conscious creatures in general into irresistible reactions or ineluctable episodes.

The HADD hypothesis too, as it turns out, works better in a non-deterministic form, at least where reflective beliefs are concerned. Barrett (who invented the name) is sometimes inclined to write as if representing HADD as operating in a deterministic manner where nonreflective beliefs are concerned, as when he claims, about people who notice crop circles, that ‘Once HADD registers the crop circles as a trace of agency, it produces a nonreflective belief that the crop circles were left deliberately by someone.’Footnote ⁶⁵ However, he goes on to recognise that (after this belief has been adopted) a process of inference is relevant to the formation of either the reflective belief that the agent was a human being or the contrasting reflective conclusion that the agent was a supernatural agency, and takes his readers through the steps of the relevant reasoning.Footnote ⁶⁶ Each of these conclusions, however, is presented as resistible, as when ‘sceptics scramble for alternative explanations’.Footnote ⁶⁷ More than one cognitive outcome (by way of reflective beliefs) is thus assumed to be possible in such situations (a non-deterministic assumption), even if, at the intuitive level, the formation of at least some nonreflective beliefs is irresistible.

A little later, Barrett contrasts the situations of different observers who, while wandering through a wood, notice ‘a rustling of leaves and cracking of twigs nearby’.Footnote ⁶⁸ For those with little at stake, their HADD offers a resistible interpretation involving (animal or human) agency; but for those whose situation (of need or of fear) is urgent and desperate, their HADD would ‘shout’ or even be ‘screaming at them’ (apparently irresistibly). Yet shortly afterwards he writes (in the same connection) of endangered people who ‘have both personal and immediate contexts encouraging the easy and rapid detection of agents’;Footnote ⁶⁹ here the word ‘encouraging’ suggests that even these intuitive responses can be resisted. (For what encourages us can hardly be taken as necessitating our responses.) Thus not even the generation of nonreflective beliefs by our HADD is here being treated as deterministic. More recently Barrett has explicitly remarked: ‘Notice that these cognitive explanations are not deterministic: underlying cognitive mechanisms like HADD and folk psychology do not determine the outcome of individual belief formation.’Footnote ⁷⁰

Elsewhere Barrett apparently endorses Brelsford's view that ‘CSR's claim [sc. the claim of the Cognitive Science of Religion] that religious beliefs derive from (unconscious) intuitions arising from basic evolved structures of mind need not imply complete determinism, since intuitions need not be seen merely as fixed outcomes of structures of mind but as capable of being consciously reformed or overridden via intentional reflection and practice’.Footnote ⁷¹ Thus Barrett's considered view favours a non-deterministic interpretation of the Cognitive Science of Religion, and therewith of the HADD hypothesis.

Here, surely, his comment is well-advised, since if he adopted a deterministic interpretation of the genesis of our beliefs about agency (reflective beliefs included), Plantinga's argument would indicate that these beliefs are not in general reliable, contrary to what we ordinarily believe, and contrary to his own assumptions about his own agency and that of his readers. But with a non-deterministic interpretation, there is no need to regard our reflective beliefs about agency (whether religious or otherwise) as the inevitable outcomes of processes beyond our control, or as compulsive intuitions, let alone as epistemically contaminated goods; even when our reflective convictions are generated by the evidence, it is we that remain responsible for the convictions we form. The HADD hypothesis still makes a difference, predicting strong intuitive impulses to detect agency in response to many an ambiguous stimulus, impulses themselves due to our evolutionary history. Yet we retain a capacity not to succumb to these impulses and adopt irresistible matching beliefs, but to reflect and reach our own conclusions autonomously.

Non-deterministic versions of Darwinism, then, and non-deterministic versions of such hypotheses of the Cognitive Science of Religion as HADD, turn out to have major advantages over deterministic versions, at least if based on the kind of indeterminism presented by Ward and Miller, and implicitly supported in large measure by Rose and Lewontin. All these positions have their problems, but these non-deterministic stances make it possible to endorse Darwinism without those doing so undermining thereby the credibility or rationality of their own beliefs, whether inter-personal, scientific or religious.Footnote ⁷²