Advantageous superstitions

1. Henry Fuseli

The Swiss artist Henry Fuseli (Johann Heinrich Füssli, 1741-1825) made this remark:

"We are more impressed by Gothic than by Greek mythology, because the bands are not yet rent which tie us to its magic: he has a powerful hold of us, who holds us by our superstition or by a theory of honour" (Knowles, The Life and Writings of Henry Fuseli, volume 3, aphorism 105, page 102).

The word is "superstition" in the singular. This points to our disposition to believe or fear, rather than any particular beliefs or fears. There is something dubious within us, a mode of thought that it is easy to dismiss as irrational. But it is individual superstitions that may directly steer our thoughts and actions by virtue of the hold they have on us. The disposition only matters to the extent that there are individual superstitions, the hold of which it facilitates. But then it matters a great deal, as the reason why they can have a hold.

We may think that any rational person's mentality should exclude the disposition, and that we should automatically dismiss the idea that there was anything to black cats, horseshoes, or astrology.

Here we shall ask whether this is really so. We shall propose that while the only sensible response to such traditional superstitions is indeed to dismiss them out of hand, there is more to be said about beliefs that we all accept but that come to seem optional when we look closely. They are not traditional superstitions, but they share the feature that we lack adequate grounds to think that the world really is as they indicate. Moreover, it can be advantageous for them to have a hold on us, and therefore advantageous that we have a disposition that allows such a hold.

2. Traditional superstitions

There are two perfectly good reasons to dismiss traditional superstitions out of hand.

The first reason is that, given our scientific understanding of the world, there is no plausible mechanism by which they could operate. How could a black cat affect one's fortune? How could the presence or absence of a horseshoe on the wall, or which way up it was (a point on which superstitions differ), make any difference to the course of one's life? How could the position of the Sun relative to patterns of distant stars which only appear in those patterns when viewed from the vicinity of Earth make any difference to a baby's future character? And so on. A claim that there is a mechanism we have not yet understood is a reasonable speculation when connections within the same general field are proposed, for example a claim (made before tectonic plates were understood) that there could be some geological mechanism to connect the present positions of continents with earlier positions in which they would have had a jigsaw-like fit. It is a wholly unreasonable speculation when a connection between unrelated phenomena is proposed, a claim such as one that connects a passing black cat or a pattern in the stars with the immediate or long-term course of a life.

The second reason to dismiss traditional superstitions out of hand is that they have not been tested properly, and we have a strong suspicion that if they were tested, they would fail. In order to test them, we would first define the relevant variables carefully. For example, we would define what counted as good luck or bad luck. Then we would take a large sample of people and note whether a black cat did or did not cross each person's path on a particular day. We would measure whether each person had good, bad, or indifferent luck in the week following that day. Once we took into account the negatives (no black cat) and all types of luck, we might very well find that the supposed connection between black cats and luck vanished. We cannot be sure that this would be the result, but given the first reason, the lack of plausible mechanism, it would not be surprising. And given that first reason, the onus must be on proponents of the superstition to get such a test performed if they want the rest of us to take any notice of their claim.

This does not mean that we should entirely ignore traditional superstitions. We should be aware of their psychological hold on people, even while regretting that hold. We also need to understand them and their hold in order to understand much of our art and literature. This includes modern work that draws on the hold that long-standing superstitions have, for example the writings of Angela Carter and the paintings of Leonora Carrington. But that is all.

3. Advantageous superstitions

In this section we shall give some examples of superstitions which are so useful that it is advantageous that they should have a hold on us, at least in everyday life. Their advantages will be set out in later sections.

3.1 Scientific superstitions

3.1.1 The idea of a scientific superstition

The idea of a scientific superstition may seem bizarre. Over recent millennia, science has gradually displaced superstition. And the last survivors are exposed to the challenge noted above that from a scientific point of view, there is no plausible mechanism by which portents and actions could make the results that superstitions predict any more likely than they would have been in the absence of those portents or actions.

We can however find candidates if we broaden our notion of superstition to cover beliefs that underpin our ways of grasping the world, so that they have a powerful hold on us, but that are not forced upon us by the evidence as it would appear if not viewed through the lens of those beliefs. This is not to say that contrary beliefs would be permitted by the evidence, but that the beliefs we do have would, when the evidence was viewed without their influence, seem to be optional.

It is important that the superstitions should on current evidence be seen as optional, rather than their being seen as ruled out. Traditional superstitions are ruled out by evidence, or at least probably would be if we bothered to formulate and administer rigorous tests. For example, we would very likely find that people with horseshoes fixed to their walls had lives no better than people without them, or that the disposition of constellations at the times of people's birth could not be correlated with either their characteristics or the courses of their lives. The scientific superstitions we have in mind are not so vulnerable. Scientific progress might eventually lead us to regard them as false rather than optional, but we are not there yet.

3.1.2 Causation

Our first example is causation. We observe causes and effects, and rely on causal predictions, every day. Yet as David Hume pointed out, this may be just a habit of the mind. We repeatedly see event C followed by event E, and take it that C is the cause of E. But we do not see the causal power that necessitates E once C has occurred (Enquiry Concerning Human Understanding, section 7). Bertrand Russell, whose objections were wider-ranging and on the whole more technical than Hume's, wrote that "The law of causality, I believe, like much that passes muster among philosophers, is a relic of a bygone age, surviving, like the monarchy, only because it is erroneously supposed to do no harm" ("On the Notion of Cause", page 180). Belief in the reality of causation appears to be optional, and indeed its presence in fundamental physics is at best a shadow of its presence in everyday life. It seems as though we could get by with regularities along with confidence that they were underwritten by something or other. (They would have to be underwritten by something for the Universe to be sufficiently well-behaved for us to have evolved.) And neither Hume nor Russell objected to our noting regularities and acting on the basis that they would probably recur. But we take the option of identifying causation as that which underwrites the regularities. For this reason, causation is a superstition in our extended sense.

We do here speak of causation in the sense in which most people would understand it, including some sort of necessitation that distinguishes it from accidental regularity. Philosophers have developed other theories of causation, such as conterfactualism, dispositionalism, and process theories, and some have turned the emphasis from what causation really is to what we mean when we make causal claims, while others have been ambitiously anti-realist about causation. But such theories do not make inroads into the consciousness of most of us when we recognise individual instances of causation. We naturally think in terms of causes pushing parts of the world so that the usual effects occur. Moreover, the equations that the natural sciences offer us, equations that set out how systems evolve over time, do not in themselves satisfy our natural sense of parts of the world pushing other parts, even though one might think of the direction of time as the direction of causality.

It is belief in causation that is optional, not use of the concept. In most of the sciences and in everyday life, we have to identify events as causes of other events. It is only when we are challenged that we bring our assumption of causation to consciousness and turn it into a belief. It is then that we are driven to reflect on whether our talk of causes is legitimate. Indeed it is at that point that we may reflect on and choose between the various available theories of causation. Our regular avoidance of such debates is facilitated by the fact that we tend to focus on specific events, rather than on causation as a general notion. We say "this caused that", rather than "here is an instance of causation".

3.1.3 Free will

Our second example is free will. For most of us, it is essential to our images of ourselves and our understanding of people and social relations that we assume free will in an uncomplicated sense, without getting into the formulations that compatibilists offer in response to the challenge that the physical world offers only determinism and randomness and that the mental supervenes on the physical. We have to take it that our inner sense of freedom at the moment of decision corresponds to the reality of the world. There are some people who study the neurology and the philosophy and then say "I am a determinist, free will is an illusion, and compatibilist guidance control does not even come close to amounting to the free will that I deny". But this attitude is unlikely to be easy to sustain in daily life. Most of us are much happier retaining our uncomplicated assumption of free will.

We shall therefore regard belief in free will, the belief that arises when we bring our assumption to consciousness, as a superstition in our extended sense. It is a popular belief. It also has a powerful hold, in the sense that if a philosopher tells us we are mistaken we will cling to it as long and as firmly as we can. But given the controversy, free will cannot be regarded as established fact. And the belief lies somewhere between belief in causation and traditional superstitions in its vulnerability to turning out not even to be optional but having to be discarded. Our knowledge of neurons, and the natural assumption that the mental supervenes on the physical, for lack of anything else on which it might supervene or any plausible mechanism that would allow it not to supervene on anything, together place belief in free will at specific and substantial risk.

As with causation, it is only belief in free will that is optional. Its assumption in daily life, an assumption that does not generally involve explicit assent to its existence, is all but inevitable.

3.2 Ethical superstition

We tend to regard ethical rules as at least provisionally binding, and not as mere suggestions that may be ignored whenever that would happen to suit our preferences. We could say that the rules that commanded general support were the ones that were binding on us individually, with the preferences of the majority being the ground of their binding nature. But that might not do. It would not allow for a whole society to be collectively ethically misguided. We feel most comfortable with a largely unspoken assumption that ethical rules are grounded in some facts about the world that hold independently of what we currently happen to think. That is, we incline to some form of ethical objectivism. (This is not the same as objectivism in the sense associated with Ayn Rand, although we might see Rand's objectivism as an example of the more general objectivism that we have in mind.)

It is when we are challenged on our assumption of ethical objectivism that the assumption turns into a conscious belief. Then we find that there is an unresolved debate. There are plenty of philosophers who regard the belief as correct, although it takes different forms in the views of different thinkers. But there are also plenty of philosophers who think the belief is incorrect, and who offer a wide range of objections.

The existence of serious challenges to ethical objectivism means that we cannot regard it as established fact. There might not be any ethical rules that were binding for reasons other than our individual or shared preferences. And yet the idea that there are such rules has a powerful hold on us. We do not stop to think that there might not be any objective rules when we routinely praise some conduct and condemn other conduct. So the belief that there are such rules is a superstition in our extended sense. Quasi-realists might say that it was a superstition that was indeed mistaken, rather than one that might be mistaken, but they would in their own way endorse recognition of the strong hold that the superstition had on us.

As to vulnerability, this superstition is too far in its nature from the natural sciences to be exposed to scientific testing of the sort that might force its abandonment. It cannot be located on the scale of vulnerability that runs from belief in causation, through belief in free will, to traditional superstitions. Nonetheless, we cannot be confident that those who reject ethical objectivism are mistaken.

It is also not easy to say that while belief in ethical objectivism is optional, its assumption in the daily life in which it is not made explicit is all but inevitable. There does seem to be real scope for a day-to-day attitude that ethical rules are to be followed without thinking of them as objective. We could be conscious prescriptivists without disturbing our pattern of life.

4. Limits to optionality

4.1 Scientific superstitions

The world appears to us as if causation and free will are real, and we act accordingly. We have to dig some way into our scientific understanding to raise doubts. And even then, we are not forced to deny their reality. The arguments against their reality may be strong, but they are not logically conclusive, and we could always try the escape route of saying that our scientific understanding might be a misunderstanding.

The upshot is that we cannot regard causation and free will as definitely real. Belief in them is optional in that while the world appears as it would if they were real, that appearance might arise without their being real.

This is however a limited form of optionality. We are not driven to belief in the reality of causation and free will. We are however driven to acceptance of their reality in everyday life, and for causation in a great deal of science, and for free will in the routine psychological understanding of people. There is no suggestion that it would be practical to get back to the underlying detail of reality and work with that in order to eliminate use of the concept of causation, or use of broad-brush psychological concepts like the concept of voluntary decision.

4.2 Ethical superstition

Most of us live as if ethical rules had more objectivity than our individual or collective preferences, taking that objectivist attitude both to guide our own conduct (or at least to feel bad about conduct that breaks the rules) and to comment on other people's conduct. If we dive into philosophical arguments, we become less sure of this. Yet we can see that even if we were to reject objectivism we could go on conducting ourselves in the same ways, so long as we did not at moments of decision think that the rules only represented preferences and were therefore trivial to ignore. Thus the belief in objectivity is optional, but ridding ourselves of the unconscious assumption of objectivity would be mentally inconvenient. It would also be risky unless an unconscious dedication to prescriptivism were substituted.

5. The respectability of belief

5.1 Scientific superstitions

Belief in the reality of causation and free will comes across as perfectly respectable. These beliefs are very widely shared, either explicitly when people consider them or implicitly when they are simply assumed, and their assumption has turned out to be very useful in navigating the natural and the human world. There is no reason to worry that systematic testing at the level at which they are put to work, as distinct from a deep exploration of physics or neurology, would instantly demolish the beliefs. In relation to free will, the Libet experiments and their successors come closest to such demolition, but they relate to very circumscribed choices like pushing buttons, and their interpretation is contested.

This is not to say that these beliefs should go unchallenged. Pressure to examine them closely can be brought to bear by philosophers as well as scientists. There are differing views of the nature of causation, and differing views as to what is actually claimed if one asserts its reality. Turning to free will, debates over the extent to which people are responsible for their actions lead us to ask what kind of free will can really be attributed to human beings. But most of the time, we get by perfectly well with straightforward and unsophisticated assumption of the reality of causation and free will.

5.2 Ethical superstition

Similar things could be said about belief in the objectivity of ethical rules. The implicit assumption of that belief is very widely shared, and it plays an important role in everyday life. The belief is also not vulnerable unless we conduct serious philosophical argument. Even then, it is only placed in doubt, not refuted. And most of the time it is taken for granted, not expressed in any way that would put it up as a target for debate. The belief can and sometimes should be debated, but we can get by perfectly well without having such debates all the time.

6. The virtue of belief

6.1 Scientific superstitions

It is not merely convenient to believe in or assume the reality of causation and free will. Doing so can be a condition of certain enquiries into the nature of the world, enquiries that we can expect to be fruitful. Someone who believes in or assumes the reality of causation will ask of some phenomenon what causes it, and may find an answer that he or she would not have found otherwise. Someone who thinks that people make their own choices will be motivated to investigate the psychology of choice, and will look for humanistic rather than mechanistic answers. This will not only make the answers more comprehensible. It will also greatly improve the prospects for getting answers that are of real interest and use. If we insisted on thinking in the mechanistic terms of neurons, the prospects for getting fully explanatory answers in relation to particular decisions would be remote.

What we have just said may make it seem that the scientific superstitions that concern us are no more than heuristic tools, rendering their superstitious nature harmless because it will not infect substantive claims that are consciously made. But this is not quite so. If one starts by saying "Let us pretend that there is causation", or "Let us pretend that there is free will", one will not be able to say that one has established the cause of the phenomenon of interest, or the psychology of free human choice. If research is built on a pretence, the fact that it is a pretence will undermine the status of results that also reflect the pretence. If a researcher when presenting results speaks of causes or free choices, any sense that the words used reflect pretences will cut the ground from under the results. This problem is averted when results are presented in terms that do not reflect the pretence inherent in some heuristic tool, but with causes and free choices, the problem is likely to arise. So the researcher must believe rather than pretend, or at least put awareness of pretence out of mind.

If the assumption of scientific superstitions is a high road to results, a willingness to assume them may be regarded as an epistemic virtue of a responsibilist kind. It may sit alongside virtues such as creativity and inquisitiveness when they are taken as virtues that drive enquiry forward, virtues in the context of discovery, rather than virtues that lead one to subject claims to rigorous evaluation, virtues in the context of justification. So we can celebrate our disposition to rely on superstitions that are not forced upon us by the evidence, but that are immensely useful in advancing our understanding of the world.

We can however only celebrate this disposition so long as it is kept under control. If some scientific superstition turned out to be ruled out by a combination of our evidence and theories that were well-supported, it would have to be discarded. Any results that were formulated in terms that assumed its correctness would then have to be reformulated. Less drastically, if a scientific superstition turned out to be less useful than one might have hoped, perhaps encouraging research that took us down dead ends, it would have to be called into question given that it was not in itself forced upon us by evidence.

We must also emphasise that the virtue in question here is virtue in the pursuit of knowledge, not virtue in the giving of psychological comfort. The cheering effect of wishful thinking, or a belief that a lucky charm or a treatment that happens to be a placebo will aid recovery from a medical condition, might be thought to make such states of mind virtuous, or make the encouragement of them in other people virtuous. But that is not what we have in mind here, quite apart from the fact that it is prima facie unlikely that virtue would really be found in such states of mind rather than in a robust respect for truth.

6.2 Ethical superstition

The virtue of belief in or the unconscious assumption of ethical objectivism is clear. We are encouraged to take our obligations seriously, so that we get along together. Constant doubt about the status of ethical rules would make us less reliable neighbours and colleagues than we actually are.

7. The grounds of usefulness

7.1 Scientific superstitions

7.1.1 Broad-brush work

Scientific superstitions most obviously gain their usefulness from the fact that it would be counterproductive to get down to the microscopic detail of the world, even assuming (which may not be the case) that there would be a microscopic level that would capture all the basic facts upon which everything else supervened and that if we had arbitrarily great computational power it would be possible to investigate all the phenomena of interest in the world by working at that level.

That is, superstitions are a way to work at a broad-brush level. They are not the only way to do so. One could work with concepts that bore a close computational relationship to what would be seen at the microscopic level. This is for example what is done in statistical mechanics. Scientific superstitions, by contrast, are set out and put to work using new concepts that are not inherent in the microscopic detail. Their applicability is at closest emergent from what would be seen at the microscopic level, and perhaps not even that.

(We note, but shall not here pursue, the thought that if it were not even possible to identify emergence, it would be arguable that the relevant scientific superstitions were not optional, but were straightforwardly correct theories of the relevant aspects of the world. The starting point for the argument would be that they were not eliminable, even hypothetically on an assumption of arbitrarily great ability to collect data and arbitrarily great computational power.)

7.1.2 Structuring reality

The concepts used to formulate superstitions allow us to structure the world at a high level of abstraction. This is an aspect of working at a broad-brush level, but the aspect of structure deserves particular attention.

The classic notion of the use of concepts to structure reality is that of Immanuel Kant, for whom the world would be unintelligible if we did not use the obvious tools of spatial and temporal ordering, the identification of causal relationships, and so on.

Kant's approach was at the heart of a complex metaphysic that put the world as it was in itself wholly and permanently outside our grasp. Even if he had seen the depths to which modern physics penetrated reality, he would have maintained that this was so. Theories that identified fundamental particles and fields, and that made computations of the structure of spacetime, would still only show us the world as it appeared to us, not the world as it was in itself.

We do not need to follow Kant in order to understand the power of scientific superstition, but our attitude to scientific superstitions does have something of the flavour of his approach.

We can see scientific superstitions as giving more structure to our image of the world than it would otherwise have. Events are related by causation. Circumstances and human actions are put in clear patterns in relation to one another by use of the concept of free choice. And when we have found structuring devices that work, such as these superstitions, our belief in their correctness is encouraged.

They work not merely by making sense of what we currently observe, but also by facilitating progress. Once we have a stable picture of the world, we can re-apply the superstitions to formulate new hypotheses and then test them. We look for what caused what, or for how a subject thought through the options before making some free choice.

7.2 Ethical superstition

A belief in or assumption of the objectivity of ethical rules has the great benefit that it will support compliance without too much consideration of alternative rules, thereby sustaining a harmonious and pleasant society.

This psychological and sociological result is enough to make the superstition useful. The ultimate ground of usefulness is that people are responsive to what they perceive as binding rules, and are happy to have their interactions with others given some structure and predictability.

On the other hand, it is a short step from belief in or assumption of the objectivity of ethical rules to refusal to countenance changes to them. We sometimes need to ease up on a devotion to ethical objectivism in order to make progress. Historically it has been those who identify a supposed conscious source of objectivity, by asserting the existence of a divine lawgiver, who have been the chief obstacles to ethical progress.

8. Maintaining belief

There is no mystery about how we come to believe in or assume causation, or free will, or the objectivity of ethical rules. These beliefs or assumptions are held by people generally, and we just absorb them in the course of growing up.

What is mysterious is that the beliefs can survive recognition that while they might be correct, they might equally well be incorrect, and indeed that there are strong although not wholly persuasive arguments against them. Causation is vestigial in serious physics. Free will is at risk from the presumably mechanical operation of neurons together with the highly plausible view that the mental supervenes on the physical. And when we start to ask how ethical rules might be objective, and what strange kinds of facts might support ethical rules, we see that a claim to their objectivity would need far more substantial defences than a mere habit of belief or assumption.

Despite these risks, it may be perfectly acceptable to resolve to think and act as though the superstitions were correct. They have after all not been ruled out, and if inference to the best explanation can include in its evaluation of explanations their success in everyday life, they have a lot going for them. But as already noted, anything that smacked of pretence to correctness, of an attitude that it was merely as if they were correct, would risk undermining conclusions that were worded in the same terms as the superstitions.

So on what basis could superstitions that might be correct but the correctness of which was easy to doubt be upheld in a way that was strong enough not to undermine conclusions expressed in their terms? Let us consider two options.

The first option would be some form of doxastic voluntarism. We might simply choose to accept the superstitions in a sufficiently strong form of acceptance, because they were so useful. Voluntarism is however contentious. Belief is supposed to aim at truth, so it should be driven by the state of the world and not by our desires, not even noble desires to advance our knowledge and to live in harmony.

The second option would simply to resign ourselves to the psychological inevitability of the superstitions. Their hold on us is powerful not just because we have been brought up with them, but because they play large roles in our ways of life. As David Hume pointed out, philosophical doubts are only to be sustained in one's study, and are quickly abandoned once one enters into everyday society (A Treatise of Human Nature, book 1, part 4, section 7, page 269).

In relation to the ethical superstition of objectivism, the option of doxastic voluntarism would be the one we would be likely to select when placed on the defensive, and might even seem respectable. In the context of ethics, there is a degree of willingness to see an act of voluntary commitment as intellectually respectable. There is however a delicate balance to be maintained. Objectivism requires facts, and while it is certainly possible (and indeed admirable) to make a voluntary commitment to accept the facts whatever they may turn out to be, it is dubious to make a voluntary commitment to specific supposed facts in advance of receiving sufficient evidence. It is precisely the objectivist nature of the ethical superstition that should make us doubt whether voluntary commitment would be respectable.

Turning to scientific superstitions, the option of resignation to psychological inevitability would be the one we would be more likely to select when placed on the defensive. The very idea of doxastic voluntarism is manifestly antithetical to the project of finding out about the world.

Having thus resigned ourselves, we would also resign ourselves to not doubting the respectability of conclusions that we reached and that were worded in the same terms as were used in scientific superstitions, such as causal terms or the terms of free choice. Doubts that should perhaps have been transmitted from the superstitions to our expression of the conclusions would simply find no foothold in our consciousness.

This would however not suffice to make the conclusions as expressed in ways that took scientific superstitions for granted undoubtedly respectable. They would remain under the same shadow of doubt as the superstitions. But if we are to make progress in understanding the world with our limited capacity to handle microscopic detail, we must put up with that because we have to use concepts that are not forced upon us by the microscopic detail of the world and do not capture that detail. As noted above, this is not merely a matter of averaging out detail, but of conceiving the world in different ways. And we should not assume that if our brains (or our artificial intelligence systems) had arbitrarily great computational power, this need to move away from microscopic detail would end. That need might reflect limitations in principle, rather than merely in practice, of microscopic mathematical modelling or of its power to capture the nature of the world.

So the hold of advantageous scientific superstitions on us may very well be inevitable both now and in the future, either for ever or until some fundamentally new way of grasping the nature of the world is introduced. We are undoubtedly better graspers of the nature of the world for having abandoned traditional superstitions. But we are not perfect.

References

Hume, David. A Treatise of Human Nature, second edition, edited by L. A. Selby-Bigge and revised by P. H. Nidditch. Oxford, Clarendon Press, 1978.

Hume, David. Enquiries Concerning Human Understanding and Concerning the Principles of Morals, third edition, edited by L. A. Selby-Bigge and revised by P. H. Nidditch. Oxford, Clarendon Press, 1975.

Knowles, John. The Life and Writings of Henry Fuseli, 3 volumes. London, Henry Colburn and Richard Bentley, 1831.

Russell, Bertrand. "On the Notion of Cause". Chapter 9 of Russell, Mysticism and Logic and Other Essays. London, George Allen and Unwin, 1959.