Printer Friendly

Rationality at Risk: Three Reasons Why Humanists Should Avoid Naturalism.

In the course of discussing the weakness of human testimony as evidence for the occurrence of miracles, David Hume (1711-1776) pointed out that "in our reasonings concerning matter of fact, there are all imaginable degrees of assurance, from the highest certainty to the lowest species of moral evidence. A wise man, therefore, proportions his belief to the evidence" (An Enquiry Concerning Human Understanding [1748], Chapter X, 'Of Miracles,' paragraphs 3,4). He overtly distinguished here the strength of people's belief from the strength of the available evidence and recommended that the former should match the latter. He lamented the tendency for people's belief in miracles to swell in strength out of all proportion to the strength of the evidence that supported such belief.

Humanists who sympathize with Hume on this point should therefore resist Hume's own attempt to naturalize the notions of evidence and proof, that is, to explain them in terms of the natural causes of mental processes such as inference and prediction. Here are three reasons why.

1. Naturalism psychologizes the concepts of reason, evidence, and proof

To naturalize these concepts is to psychologize them, which inverts and relativizes Hume's maxim: evidence-to-someone becomes whatever it is to which that person proportions his or her belief. Consequently, there can be no failure of proportion between belief and evidence, and Hume can relax. Although the strength that I assign to the evidence for your belief may not justify the strength with which you hold that belief, yet the strength that you assign to the evidence necessarily does.

The motivation for this naturalistic inversion is not hard to find. At different times and places people have counted as evidence or proof many things that would amaze or amuse us. Our ancestors once regarded trial by ordeal and confessions extracted under torture as providing conclusive proof of guilt or innocence; we do not even regard them as evidence. Naturalists regard this fact as proof, or at least as evidence, that our standards of proof and evidence are derived from our psychological nature or our social nurture, not from any awareness of the abstract Platonic 'forms' of proof or evidence; indeed, the supposition that such absolutes exist is, to them, entirely gratuitous.

Naturalism acknowledges the existence of only those things that stand in cause-and-effect relations to each other; it therefore interprets the concepts of evidence and proof in terms of our psychological responses to sensory and cognitive stimuli. We regard as evidence whatever tends to convince us, and we regard as proof whatever convinces us so thoroughly as to remove all doubt from our minds. Of course, what tends to convince me may not tend to convince you, and what removes all doubt from your mind may not remove all doubt from mine. Thus, what counts as evidence to me may not count as evidence to you, and what proves something to you need not prove anything to me.

Evidence and proof, when naturalized in terms of psychology, have the same status that many people accord to beauty and humour. We regard as beautiful anything that it gives us a certain kind of pleasure to behold, and we count as humorous anything that it amuses us to contemplate. This tempts us to define these notions in terms of our psychological responses. Whatever justifies psychologizing beauty and humour will equally well justify psychologizing justice, importance, truth, probability, evidence, and proof. That is why we have emotivist, prescriptivist, and other non-cognitivist theories of ethics. Our aesthetic taste is merely our susceptibility to being pleased in a certain way by observing various things. Our sense of humour is not a detection device that alerts us to the presence of humour; it is merely a susceptibility to being amused by different things. Similarly, our sense of evidence is not an ability to recognize evidence when we confront it but only a tendency to be convinced by various kinds of considerations.

Naturalists, therefore, side with Protagoras and the sophists against Plato. Whatever convinces me is evidence or proof to me, and whatever convinces you is evidence or proof to you; there is no such thing as what really is evidence or proof apart from its power to convince; nothing is evidence or proof unless it convinces someone; and even then, it is proof only to those whom it convinces, not to anyone else. When certain facts convince us, we call them proof; but they do not convince us because they are proof any more than jokes amuse us because they are funny or a vision pleases us because it is beautiful. It would not change matters if there were certain considerations that convinced everyone. That would only show that we all had the same sense of proof, not that proof was an inherent characteristic of the considerations that convinced us, a characteristic that they possessed independently of our ability to recognize it. Clearly, even if we all had the same sense of humour, and laughed at the same things, that would not show that humour was an inherent property of the things that amused us, a property that these objects possessed independently of our ability to appreciate it.

Naturalists believe that they have proof or at least evidence that supports their naturalism; but, given their treatment of 'support' and 'evidence,' they must regard this belief as merely a curious fact about their sense of evidence. I do not tend to be convinced by what they put forward, and to them this must be an equally curious fact about my own sense of evidence. Statistics is supposed to be the mathematical theory of evidence, and logic the theory of deductive validity. However, the eternal and unchanging realms of logic and mathematics cannot cause anything, since no events occur in them, and therefore they are removed from the world of nature. From the naturalistic perspective, they count as nothing more than Platonic myths. Naturalizing these sciences turns them into empirical studies of 'the laws of thought,' psychological investigations into our inferential propensities. These investigations yield theories for which naturalists adduce facts that they call supporting evidence, though they admit that calling these facts evidence or support merely reflects their own tendency to be convinced by them and says nothing true or false about the facts themselves.

2. If psychology-independent requirements of rationality exist, we have no idea how likely it is that nature would have given us a reasoning system that conforms to them

To naturalize our faculty of reason is to regard the human mind as an information processing system that transforms sensory and cognitive input into cognitive output. For all that we can know prior to investigation, any input could generate any output. Only the peculiar nature of the processing system can determine which input is transformed into which output. If we believe certain premisses, that will cause us to believe certain conclusions; but only the processing system determines which conclusion we would jump to from given premisses. Learning that the sun has risen every day in the past convinces us that it will rise tomorrow, but with a different internal processing system it might have convinced us that the moon was made of green cheese. Which conclusions we draw, once we believe certain premisses, depends entirely on how our minds are programmed, regardless of whether the programming comes about through God, or nature, or nurture. The programming alone determines what we regard as evidence or proof.

The pocket calculator provides an illustration. If we press the buttons '2,' '+,' '3,' '=,' the display will flash the numeral '5.' The calculator does this not because it knows that two plus three equals five, in any sense in which it knows which key strokes have been entered, but because we have programmed it to respond in this way. We could have programmed it to display a different symbol in response to the same key strokes. We have instead programmed it to respond in ways that correspond with our system of representing the truths of arithmetic. This imposes a pre-established harmony between the material realm of the calculator and the abstract realm of mathematics. Elliot Sober's selection toy exhibits the same sort of pre-established harmony. It consists of a vertical cylinder sectioned at different levels by horizontal circular disks with holes whose diameters are larger in the higher disks and smaller in the lower ones. Balls of different sizes are poured into the top of the cylinder. Balls of the same size have the same colour, and balls of different sizes have different colours. The disks therefore screen balls at different levels in conformity with their different colours but because of their different sizes. The selection toy does not know the colours of the balls in any sense in which it knows their sizes, and therefore cannot be regarded as colour sensitive, no matter how reliably it sorts balls according to colour.

If our minds worked like information processing systems, then, in the absence of pre-established harmony, what would ensure, or even make it likely, that the conclusions that our processing mechanism led us to draw harmonized with the conclusions that it was rationally appropriate to draw? Evolutionary psychology would explain our information processing system as an adaptation, a functional product of natural selection. It may be very unlikely that our species would have survived the arduous ordeal of natural selection for as long as it has, equipped with a faulty cognitive system, one that does not tend to lead us to conclusions that are rationally justified by the evidence. Creatures burdened with such a system would likely have perished before they were able to reproduce their kind. However, the world is a dangerous place, and it may still be unlikely that we would have survived this long even with a correct cognitive system, one that does tend to lead us to conclusions justified by the evidence available to us.

In any case, these probabilities are irrelevant. We do not care about the probability that humanity would have survived to the present given that we have correct cognitive systems. The question at issue concerns the converse conditional probability, namely the probability that we have correct cognitive systems given that our species has survived for this long. Natural selection may have weeded out most of the species with faulty cognitive systems, but this does not entail or even suggest that most of the species that have survived have correct cognitive systems. Even if ninety-nine percent of species with faulty systems have become extinct, and all species with correct systems have survived, the remaining one percent of species that survived in spite of their faulty systems may still contain, say, nine times as many species as those with correct systems. In that case, the probability that our species would survive, given that it had a correct cognitive system, would be a hundred percent, while the converse probability that our species has a correct cognitive system, given that it has survived, would be only ten percent. To avoid this possibility, we would have to assume that natural selection ensures that only species with correct cognitive systems will survive. This assumption is highly implausible since, after all, faulty systems that have heuristic value (like those installed in my programmable calculator) may be versatile, cost-effective, and give us good enough results in the kinds of situations that we are likely to encounter. They would therefore confer a greater evolutionary advantage than logically correct systems that have a narrower range of application, use up more of our limited resources, and confer no increased practical benefit in those situations.

3. Naturalizing our reasoning processes strips us of rationality

To be virtuous requires more than that we do the right thing; we must do it for the right reason, i.e., because it is the right thing. To be honest requires more than, say, returning lost property to its owner when that would be easy, rather than keeping it for ourselves. It requires returning the property because it belongs to someone else and not because the owner has posted a reward worth more than the property. To be prudent requires more than just doing what is in our long-term best interest. It involves doing so because it serves our best interest and not because it promises immediate pleasure. In a somewhat analogous way, to be rational requires more than that we come the right conclusion, given the evidence; we must come to the right conclusion for the right reason. What makes us come to the right conclusion must be our recognition of the strength with which the evidence supports it and not our susceptibility to an appealing but irrelevant subliminal association.

No information processing system can satisfy this requirement. Such a system could 'reason' correctly every time, in the sense that it might always respond to a given input with the appropriate output. However, it would do so for the wrong reason, not because the output was appropriate but because we had programmed the system to respond as we wanted it to. With a different program, it would just as easily respond otherwise when fed the same input. This proves that the system is insensitive to appropriateness, which is a non-natural feature, and operates on some other basis.

Perhaps it is programmed to respond to the physical characteristics of the way in which we have coded or represented the input. It responds to the representation, not to the thing represented. The same thing could have different representations, and different things could have the same representation. The information processing system cannot tell what is represented. With the selection toy, the input consists of balls of different colours; the colours are represented or coded by different sizes; the machine operates on the basis of size, sending balls of different sizes to different levels; this is what sends balls of different colours to different levels. Because the selection toy cannot respond to colour, which is a non-mechanical feature, its designer must impose a pre-established harmony in order to ensure that a mechanical device that operates on the basis of size will produce a result conforming to differences in colour. Nevertheless, ensuring this harmony does not make the selection toy colour sensitive.

Rene Descartes (1596-1650) realized that not even God could make humans rational if the task were framed in these terms. In his Meditations on First Philosophy (1641) he acknowledged that if God had endowed his mind with such a nature that he found certain inferences irresistible, although he would have found contrary inferences equally irresistible if the omnipotent God had created his mind with a different nature, then before he could trust his reasoning powers, he would first have to discover whether God had given his mind the right nature (Med. III, paragraph 4); for the nature of his mind determined what he would regard as evidence or proof. The attempt to satisfy himself on this point necessarily led him into the Cartesian Circle from which, like the labyrinth of the Minotaur, no one who enters can escape. However, even if God did give our minds the right nature, thereby ensuring a pre-established harmony between what we regarded as proof and what really was proof, we would still be no better than information processing systems that had been programmed in the right way. This sort of supernatural intervention cannot make us rational. We would draw the appropriate conclusions from given premisses, not because we could recognize their appropriateness, but because we would respond by nature to other features of the input, perhaps to the way in which it was coded or represented. Thus, even if our minds had the right nature, and we reasoned as we did because of our nature, still we would no more be rational than the selection toy is colour-sensitive.


In Hume's Dialogues Concerning Natural Religion (1779), which he entrusted to his friend Adam Smith for publication after his death, his spokesman Demea presents, in Part X, a classic formulation of the Argument from Evil against the existence of an omnipotent and benevolent God.
Epicurus's old questions are yet unanswered. Is he willing to prevent
evil, but not able? then is he impotent. Is he able, but not willing?
then is he malevolent. Is he both able and willing? whence then is


His power we allow infinite: whatever he wills is executed: but neither
man nor any other animal is happy: therefore he does not will their
happiness. His wisdom is infinite: he is never mistaken in choosing the
means to any end: but the course of nature tends not to human or animal
felicity: therefore it is not established for that purpose. Through the
whole compass of human knowledge, there are no inferences more certain
and infallible than these.

My friend and mentor Plato Mamo calls the argument "utterly convincing" ('On Evil,' Humanist Perspectives, issue #206, Autumn 2018). It is indeed utterly convincing, but to whom? It convinces no theists, though it disturbs some of them. If we say that it should convince them, do we assert something true or false about the Platonic realm of logic, or do we merely give expression to our own psychological tendency to be convinced by it? What happens to the certainty and infallibility of these inferences once they pass through the fires of naturalistic purification? Certainty ceases to be a guarantee of truth and becomes merely a state of maximum confidence. Calling an inference 'infallible' no longer asserts the logical impossibility that it should lead from true premisses to a false conclusion but only reflects our psychological inability to conceive how it could.

Once we have abolished the non-natural realm of logic, there remains for investigation only the psychological limitations on our powers of conception and the rhetorical strategies of persuasion. For decades, I suffered from an inability to conceive how space could have more than three dimensions, although I understood perfectly well the abstract mathematical notion of a three-dimensional sub-space of a higher n-dimensional vector 'space.' Now that I have become a senior citizen, I have fortunately overcome that disability. I would certainly not want to assign any authority to our contingent inabilities to understand, imagine, or conceive, nor to erect them into a standard to which everyone should conform. If 'good' is not simply whatever pleases God, then 'proof is certainly not whatever convinces us.'

John Ibberson, BA, MA (Calgary), DPhil (Oxford) is the author of The Language of Decision; An Essay in Prescriptivist Ethical Theory (Macmillan, 1986) and Autonomous Reason; A Defense of Rationality Against Naturalistic Materialism (Studium Generate Press, 2017). He taught philosophy at the College of New Caledonia in Prince George, British Columbia from 1989 until his retirement in 2013. He is now an independent scholar residing in Kelowna, BC.
COPYRIGHT 2019 Canadian Humanist Publications
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2019 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Ibberson, John
Publication:Humanist Perspectives
Date:Mar 22, 2019
Previous Article:The Free Speech Crisis on Campus Is Worse than People Think.
Next Article:Gender Unicorn vs Gender Elephant vs Gender Bonobo: Oh No! Genderology Wars!

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |