In ‘Warrant and Proper Function’ Alvin Plantinga established his Evolutionary Argument Against Naturalism[1] (EAAN). Plantinga’s argument centres on his conclusion that Naturalism and the Theory of Evolution (NE) are rationally incompatible because they gives rise to significant “doubts about the reliability of our cognitive faculties” (Beilby 2002 pp. 2) which in turn form a rational ‘defeater’ – an argument which, on the balance of probability, gives us reason to significantly doubt an established assumption/conclusion. In this essay I will briefly outline the argument and then address some of its key flaws: Firstly I will examine his assessment of Naturalism’s view of the belief-desire relationship and critique his assumption that Naturalist’s only legitimate belief is that the content of beliefs do not influence behaviour. Secondly I will argue that, as philosophers William Ramsey, John Wilkins and Paul Griffiths suggest, the conclusion that evolution only tracks fitness with no account for truth/reliability crucially fails to properly consider the way that truth-tracking enhances fitness. Finally I will conclude that in the words of Wilkins and Griffiths “evolution by natural selection…. favours organisms that form true common-sense beliefs” (Maclaurin & Dawes 2013 pp.133) and that in the absence of any proof by Plantinga that false belief improves fitness at least as much as true belief, there are no grounds to question the reliability of our cognitive faculties and therefore EAAN in any of the forms it has taken over the past twenty years.
The Argument:
Plantinga’s EAAN is grounded in a ‘doubt’ expressed by Charles Darwin himself. In a letter Darwin expressed distrust of human beliefs or cognitions after contemplating whether we should trust the imaginings of a monkey’s mind; or more generally if we could trust a similarly complex and developed sentient organism in an entirely different environment to not have vastly diverging perceptions of reality or ‘convictions’ even if having those different beliefs would be useful for evolutionary success. Plantinga builds on this doubt by examining the link between beliefs or, in Darwin’s terms, ‘convictions’ organisms hold and the actions they undertake suggesting that there are just four potential relationships given NE. The first is epiphenomenalism which holds that “behaviour is not caused by beliefs” (Beilby 2002 pp. 6), the second is semantic epiphenomenalism which alternately holds that if not the belief itself then at least the actual literal and alterable content of that belief does not influence behaviour, thirdly and fourthly there is the understanding that beliefs do influence behaviour and are either maladaptive and will not change as a result of the environment[2] or that they are adaptive and can change as a result of their environment.
Plantinga regards both semantic and simple epiphenomenalism as the “received view(s)” (Beilby 2002 pp.10) for Naturalists which are “extremely hard” not to accept given that a belief in Naturalism necessarily implies that there is no way “the content of a belief could get causally involved in behaviour. If a belief just is a neural structure of some kind……. it is exceedingly hard so see how content can get involved in the causal chain leading to behaviour”. (Beilby 2002 pp.10) From this ‘burden push’ of the less defensible epiphenomenal standpoint onto Naturalists, Plantinga proceeds to utilize Bayesian probabilistic reasoning to suggest that the likelihood of our faculties being accurate (and therefore our ability to avoid scepticism) given NE would probably be incredibly low given epiphenomenalism, leading Plantinga to conclude that he has provided a sound and thorough defeater for Naturalism given that it is supposedly incompatible and illogical in the presence of Evolution.
The argument is essentially valid, if the content of belief has no bearing on behaviour then realistically there is little to suggest that that content will be accurate or that it ever needs to be. However, Plantinga’s premise that epiphenomenalism is the logical companion to a Naturalist perspective is not sound and there exist a variety of arguments which discredit epiphenomenalism as a whole or at least which see that even if epiphenomenalism is accurate the consequential impacts of the desire-belief paradigm mean that beliefs are linked with natural selection.
The Natural Selection Argument Against Epiphenomenalism:
Amongst the more compelling arguments against epiphenomenalism is one derived from natural selection itself. This argument runs as follows: 1) Consciousness is a result of natural selection. 2) A property is selected for because it alters an organism’s behaviour. 3) Therefore epiphenomenalism is false. (Stanford Encyclopedia of Philosophy 2012)
This line of thinking was expanded on by William James in 1879 in his article “Are we Automata”. James argued that if beliefs/consciousness had no effect on behaviour then there would be no reason to select for them, nor for us to believe that certain things that degrade fitness (e.g. pain/bodily harm) are unpleasant or undesirable etc. while other activities that enhance fitness (e.g. eating/reproducing) are pleasant or desirable etc. and in fact it could be the other way around. (Stanford Encyclopaedia of Philosophy 2012)
Ramsey’s Narrow Epiphenomenalism:
More recently in a collection of essay’s critiquing EAAN William Ramsey has suggested a novel variation on this Natural Selection Argument Against Epiphenomenalism in which he circumvents the epiphenomenal relationship without totally discounting its validity by distinguishing between mechanical cognitive function and its consequential impact with regards to selection.
Ramsey admits that belief’s may not present any specific, unique or identifiable “neurophysiological properties” being little more than impulses and chemical reactions along our synaptic pathways whose content may not influence behaviour. It is however important to note that the “truth value” of a given belief does depend on what that belief is about, “truthfulness can vary while the neurophysiological properties remain the same. Hence, a belief’s truth value, though partly dependent upon what is going on inside the head, is not reducible to neurological matter.” (Beilby 2002 pp.16) This leads Ramsey to conclude that “in a narrow and somewhat mechanistic sense, the truth value of a belief is indeed epiphenomenal”( Beilby 2002 pp.18) because belief’s can vary without causally influencing cognitive behaviour-generators. However in a much broader sense “it clearly is not epiphenomenal, since variance in a belief’s truth value can alter the broader consequences of the behaviour, making it either successful or unsuccessful” from an evolutionary perspective. As he surmises “Truth doesn’t make the behaviour, it makes behaviour better” (Beilby 2002 pp.18) thus allowing for truth-value to enter into the mix of characteristics for which there can be “selection pressure”. Overall this means that, even given epiphenomenalism, the fact that truth-value inescapably alters the success or failure of an organisms actions (like hunting or foraging) necessarily means that it must be “(1) a causally relevant feature of belief’s and (2) causally relevant in a way that enhances reproductive fitness” (Beilby 2002 pp.18). Thus a cognitive mechanism that generates more true beliefs in most instances will be selected for over one that does not, ultimately suggesting the whilst the ‘mechanics’ of belief may be epiphenomenal, the actual manifestations or consequences of those beliefs render them adaptive.
Adaptive Belief And The Truth-Fitness Dichotomy:
Having established that epiphenomenalism either must be rejected out of hand or, as per Ramsey, be seen to ultimately describe an adaptive relationship, I will now turn to Plantinga’s criticism of the adaptive desire-belief relationship.
For Plantinga adaptive belief is unreliable because false belief can produce the same benefits as true belief with regards to fitness. To illustrate this he gives the example of someone fulfilling the fitness-desire to survive by ‘running away from a tiger’ as a result of a misapprehension that this is the best way to play with the tiger or simply because they believe that tigers are apparitions that subconsciously signal a need to practice long-distance running. (Beilby 2002 pp.8) In this way he attempts to suggests not only that false-belief can enhance fitness just as true-belief can but also to demonstrate various “systematic ways in which… beliefs could be false but still adaptive” (Beilby 2002 pp.9). Essentially his criticism reduces to:
(P1) Cognitive faculties are designed by evolution to enhance fitness.
(P2) False belief can produce the same fitness-enhancing behaviour as true belief.
(C3) Therefore truth-value is irrelevant to fitness.
(C4) Therefore our cognitive faculties are unreliable and EAAN is true.
The key flaw with this argument is that the conclusion in (C3) is not implied by its premises and is therefore invalid. Plantinga’s contention in (P2) is that false belief can produce the same fitness-enhancing behaviour as true belief. The key word is ‘can’. He in no way demonstrates that false belief will be more likely than not to enhance fitness. He is therefore wrong to assume that truth-value is irrelevant to fitness because he has not considered the empirical question of whether true belief is more beneficial in most instances than false belief. This is because Plantinga operates on a false dichotomy between fitness and truth.
As Griffiths and Wilkins state in their ‘Milvian Bridge’ argument selection inherently enhances fitness – that is its definition. To contend, as Plantinga does, that selecting for fitness and selecting for truth are mutually exclusive processes represents a fundamental “misunderstanding of natural selection”. (Maclaurin & Dawes 2013 pp.133) Selecting for truth by enhancing an organism to be better at seeing the true state of things is a process of natural selection, a way in which the ultimate goal of increased fitness is achieved. It is only selected for because of its benefits to survival. The only hope for Plantinga’s argument is if he can alter (P2) to read:
(P2.1) “False belief enhances fitness at least as much as true belief”.
The soundness of (P2.1) is a question of empirical fact – potentially a problematic prospect for philosophers concerned with scepticism. Intuitively, however, a few arguments can be made for truth-tracking as enhancing fitness. Firstly, as Wilkins and Griffith argue, “human cognitive mechanisms” are “highly expensive” (Maclaurin & Dawes 2013 pp.137). Our brains make up around 2% of our bodies yet consume 20% of our oxygen meaning cognitive function beliefs and all, “come at a high price” (Maclaurin & Dawes 2013 pp.137) and if there were no substantial benefit to having them then increased cognitive function would be selected against. So we can conclude cognitive function does benefit fitness. How it does this is less definite. Certainly in some instances systematically false beliefs assist fitness – it is “well documented… that human beings have unrealistically positive views of themselves and of their prospects in life” (Maclaurin & Dawes 2013 pp.135) because, amongst other things, having an optimistic outlook may encourage humans to endeavour to succeed when their actual chances of success are incredibly minute. The key point to consider is that although ‘false’ optimism can systematically improve fitness it does this not because truth value is irrelevant but because “like all evolutionary processes, truth tracking operates under constraints” (Maclaurin & Dawes 2013 pp.137) and when weighed up against other fitness-enhancing factors (cognitive processing speeds, emotions etc.) the ‘fitness calculation’ works out that true-belief is traded off at the expense of another fitness-enhancing benefit. Time, energy and resources devoted to forming and verifying accurate beliefs could be better spent on other fitness-enhancing endeavours like feeding or reproducing. As Griffiths and Wilkins point out “the scientists whose work is usually cited to demonstrate how badly human being track truth have long argued that” the failings of human knowledge represents an approach “which sacrifices being right all the time for being right most of the time at a greatly reduced cost.” (Maclaurin & Dawes 2013 pp.138)
For Plantinga to succeed he would have to show that most of our beliefs are more compromised than not as a result of this truth-sacrificing trade off. The main problem with doing this is that true belief has very tangible benefits in most instances. Take the example of a small foraging mammal attempting to evade predators. The more time it spends foraging, the greater its fitness. However the more time it spends foraging the less likely it is to escape danger. Its belief’s should ideally be tunned to maximise foraging time whilst minimising danger exposure. Thus a more accurate belief about the whereabouts of a predator as opposed to a harmless larger animal will be crucial to enhancing its fitness. This type of reasoning leads Wilkins and Griffiths to the conclusion that “an organism can succeed because it is better than its rivals at tracking truth just as it can succeed because it is better at foraging or because it has a more efficient respiratory system.” (Maclaurin & Dawes 2013 pp. 137) Would we be as successful if we were unable to understand the laws of gravity or any of the other scientific laws of the universe that have allowed us to create the technologies responsible for our domination of earth? The answer ostensibly is no.
In sum Plantinga’s arguments only illustrate instances where false belief enhances fitness and to assume that any organism would be benefited in its ‘truth-trade-off’ by consistently having more false belief than true “has absolutely no plausibility from the standpoint of naturalism….” Were an advanced organism to have a more substantially false belief-oriented cognitive system “the operation of such a mechanism would be utterly mysterious – inexplicably generating unrelated false belief-desire pairs that just happen to always keep the agent out of trouble.” As Ramsey concludes “The operation of such a mechanism would be miraculous, and naturalism doesn’t allow for miraculous mechanisms.” (Beilby 2002 pp.23)
Conclusion:
So while Plantinga’s EAAN is a highly sophisticated attempt to ‘defeat’ Naturalism it ultimately fails to do this as a result of two key flaws. Firstly, it falsely assumes that Naturalism must entail epiphenomenalism when there is a strong selection based argument against this or, alternately, Ramsey’s ‘narrow-epiphenomenalism’ method of circumventing the harmful implications to NE put forward by Plantinga. Secondly its classification of true belief as an alternative to fitness instead of as a facet of it, and its failure to prove that false belief is at least as beneficial to fitness as true belief (when in reality there are common sense reasons to reject this) removes all legitimate grounds for the suggestion that our cognitive systems are unreliable, validating Naturalism and unravelling EAAN.
Bibliography:
Beilby, J.K. (ed.). 2002. Naturalism Defeated? Essays on Plantinga’s Evolutionary Argument Against Naturalism. Ithaca, NY: Cornell University Press.
Maclaurin, J. & Dawes G. (eds.) 2013. A New Science of Religion. London, UK: Routledge.
Robinson, William, “Epiphenomenalism”, The Stanford Encyclopaedia of Philosophy (Summer 2012 Edition), Edward N. Zalta (ed.). Accessed 7/09/2013 via <http://plato.stanford.edu/entries/epiphenomenalism/>.
[1] The belief that there are no supernatural forces or entities in the universe including God or the soul and that the universe can be entirely explained by scientific thought.
[2] Plantinga himself argues evolution and the maladaptive belief-behaviour relationship are incongruent as creatures with maladaptive beliefs “from the point of view of fitness…. would be better off without them.” (Beilby 2002 pp.8).
© Anthony Small and thethinkingantman, 2015. Unauthorized use and/or duplication of this material without express and written permission from this blog’s author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Anthony Small and thethinkingantman with appropriate and specific direction to the original content.
thethinkingantman 