Science and Nature

The illusion of causation


Featured image: Scottish philosopher David Hume.

By James Clark Ross

In this essay I argue that humans do not possess the ability to cause an effect. ‘Causation’ is simply a projection of conscious impressions onto external events, through which we believe we exert control over objects. I start with David Hume’s famous regularity theory as a basis of my argument, with which I transfigure Hume’s position to incorporate the feasibility of ‘impossibility’. Human experience is such that we do not truly cause anything: we simply associate ideas, which are born from experiences alone, with what we observe.

Hume’s regularity theory

In Enquiry Concerning Human Understanding (Hume 2008), published in 1748, David Hume penned his ideas on causation, which continue to have profound effect on philosophical thinking even today. Here I have outlined his theory but seek to reach a more-clarified and tenable position than the ambiguous one he left behind.

Hume was an empiricist: he believed that we are all born with tabula rasa minds (blank slates), with no innate knowledge of how to cause an effect. The ability to predict what will happen following an action, he would claim, is drawn from a fountain of personal experiences we derive from our senses, not from the ability to control the external world.

Intending for an event to occur is not sufficient to determine it. Indeed, if it does occur, we are not justified in concluding that we caused it. What is plausibly the case is that we project our experiences onto uncontrollable events. A layman—say, Donald—would ostensibly use their empirical knowledge of the world to demonstrate that we do cause things to happen. ‘Look’, they might exclaim, ‘I am going to cause your nose to bleed by punching you straight in the face’…

[Punch] (cause)

[Nose bleeds] (effect)

But the ‘cause’, C, could merely be the thing which preceded the ‘effect’, E, in time and space; the Universe could just be described by a series of connected states—cold chains of events which we have no say in.


Did someone necessarily cause that?

As a brief but relevant digression I have inserted here a stance that is in contrast to the one held by Persian philosopher Avicenna of the Islamic Golden Age. Philosophers advocating a position akin to him would contest Hume’s argument on the basis that we can possess innate notions of cause completely separate from experience. Avicenna, born 980 AD, posed the concept of The Floating Man, a man permanently suspended in air and deprived of all his senses from birth, removing the possibility of him ever experiencing anything. Yet this man, so this thought experiment goes, still knows who he is through his ‘knowledge by presence’—primordial knowledge of oneself which is not dependent on biological senses or even a physical body. (Douglas Adam’s sperm whale in The Hitchhiker’s Guide to the Galaxy (1979) did not possess such knowledge.)

Avicenna thus posited a transcendent self or ‘soul’. While a nice analogy, I reject it and agree with Hume: I argue that we are the results of our physical bodies experiencing the external world internally, not agents who cause it to change. How does The Floating Man know who he is? Is he anybody? If he was suddenly gifted with senses, would he be able to see objects (e.g. humans) like we see them from theirs shapes and their behaviours? Unlikely. Photons of light would reflect from an object’s surface into his eyes but he still has no learned perspective of what he is looking at, until now, to form a picture. Would he be able to predict what an action upon it would lead to? Even more unlikely. Experience is essential to that.


‘Then let the subject consider whether he would affirm the existence of his self. There is no doubt that he would affirm his own existence … And if it were possible for him in such a state to imagine a hand or any other organ, he would not imagine it to be a part of himself or a condition of his existence.’

We consciously experience events. We do this through internal impressions. We undergo different mental states—beliefs, desires, opinions, aspirations, moods, personality traits, and so forth. However, we are not causing any of this. Unlike Hume, I would argue that we do not possess freedom of will since our future actions are more likely to be determined by the Universe itself (Pereboom 2007). We are aware of our thoughts; that does not mean we are in control of them. We believe we make our own choices because we are programmed that way for whatever reason. Hence cause is already a naïve notion, an internal experience, before control over external objects is event asserted.

Even if we were free agents, we cannot claim that we cause E to follow C: we can only assume it happens and live to see it transpire. Donald only speculated that my nose would bleed—or, at least, that there was high chance of it bleeding—because he saw it happen during a Hollywood action movie. That is, he learned this might happen; he was not innately programmed to predict it. Hume said we use conceptions we already have to generate (what seem to us) new ideas:

When we think of a golden mountain, we only join two consistent ideas—gold and mountain—with which we were already familiar. We can conceive a virtuous horse’—a unicorn?—‘because our own feelings enable us to conceive virtue, and we can join this with the shape of a horse, which is an animal we know’. (Hume 2008, p8)


A virtuous horse.

Knowledge, then, must exclusively stem from one’s past experiences: the ideas we have are impressions of other objects we have interpreted somewhere else in our lives. So a notion of cause does not have to be directly related to the event: it can be conceived with some originality. For example, Donald might have seen red stuff coming out of his own body before; or, possibly, he might have read about nosebleeds in a book.

These objects do not need to be real in the sense we usually take them to be in: in the realm of our conscious experiences unicorns are real, just like horses; our favourite fictional characters are real, just like our friends; and gods are real, just like forces of nature. Both kinds of objects are able to provide truth and meaning to us. One is deemed extraordinary simply because we find ourselves unable to align the thoughts we have with the events that unfold. If we were suddenly able to see a thought manifest, it would lose its mysticism.

If we had the power to move mountains or control the planets just by secretly wishing these results to occur, this wide-ranging power wouldn’t be more extraordinary or further from our understanding than the power our thoughts do have over our bodies. But if we perceived any power or energy in our own will just by being conscious of it, we would know this power, know its connection with the effect, know the secret union of soul and body, and know the nature of both these substances through which one is able to operate so often on the other.’ (Hume 2008, p32)

Humans only bear witness to life’s play. When we recognise regular successions of events we project our narratives—wills, intentions, motivations—onto the Universe.

Kinds of ‘cause’

Events are constantly unfolding around us—chains of conjoined events which are contiguous in time and space. These sequences are apparent to us because we view them subjectively. Do we necessarily cause any of them to occur? No. ‘But surely it was me who caused my car to start this morning and caused this research to be done at work this afternoon and caused this food to be cooked for my family tonight. How can anyone in their right mind—i.e. someone not studying philosophy—deny these facts?’ Such a claim would be expected from someone supporting a causal ‘necessary connection’. But the ‘necessary connection’ between events, a notional power, according to Hume, can only be located in us. I am only empirically aware that I can ‘cause’ things to happen because when I ‘will’ them to occur and I understand the results, like typing on my keyboard correctly and seeing no errors (I hope), I believe it was me who caused the effects.


Credit: NASA.

Naturally, we do not hesitate to call the first the cause and the second the effect by this associative connection—but it is our minds which create this inference. Human experience is rooted in empiricism. Events could be causally-connected through some universal rule—atoms interacting with other atoms under the governance of some consistent forces of nature or because of an omnipotent god dictated things to be this exact way with their trident; however, Hume never specifically offered anything exact for this. Nonetheless, Hume and I would agree that we are not the sources of those causes:

We learn the influence of our will from experience alone. And experience teaches us only how one event constantly follows another, without instructing us in the secret connection that binds them together and makes them inseparable.’ (Hume 2008, p32)

Take three kinds of ‘cause’ we notionally take to be real but are arguably just associations of ideas:

  • Resemblance: See something happen and assign it a cause (e.g. the fact that I believe I can cause this door to open is only a result of enacting certain movements before on similar doors).
  • Contiguity: Expect something to happen because events usually precede other events (e.g. Pavlov’s dogs were conditioned such that they salivated when they heard a metronome usually preceding food; similarly, a fear of non-white people has ingrained itself within my psyche because the only non-white person in my school was a bully).
  • Cause-effect: Positing a cause (e.g. sleep is affected by the presence of a full moon; even if people’s sleep does reduce, a full moon is an extremely-unlikely cause).

In each case there is only an idea of a true cause—projections onto events simply following other events, entailed since the beginning of the Universe (E1, E2Ex), conveying thoughts to us.

Counterfactual approach: possible worlds

A ‘counterfactual analyst’ might seek to define reality in terms of ‘possible worlds’. By conceptually removing possible causes and testing whether effects would still manifest, they explore the nature of causation. For example, I might believe my swinging a cricket bat against windows causes them to smash. However, even if it does smash I cannot claim that I definitely caused it, for there was a chance the window would not have shattered. So if E1 had not occurred, E2 might not have occurred either. In a more-extreme example, I might postulate that my pen will not fall to the ground due to gravity when I release it from my grasp. The counterfactual analyst could argue that there are real versions of reality where it might not. There are no guarantees as effects that are less likely to occur are only less ‘impossible’. Thus there is always something indeterministic which pertains to conjoined events, at least in this universe.


The implement that necessarily caused the glass to shatter?

During our lifetimes, in our own experiences, we see enough of the same thing such that we begin to expect certain outcomes and we associate causes and effects. Our worldviews are built on such experiential assumptions, where we place the most confidence in future events which stem from the most likely events, in accordance with narratives conceived from our prior experiences. A scientist develops evidence in exactly the same way. Particle physicists at the Large Hadron Collider announced the existence of the Higgs boson to the world only when they saw it with 5-sigma (99.9999-%) certainty. This was used to ascertain the Standard Model theory. Scientists, like anyone else, describe things for the ideas they have constructed—connecting outputs to inputs for the sake of human understanding—in this case describing the rules behind things smashing together. Their certainty only removes levels of impossibility.

We can never be certain of an event unless we could do something infinitely-many times—but infinity is another human construct and something we can never reach. Statistically speaking, too, we can never be sure that causation is anything more than constantly-conjoined events.


For any given event many different permutations—many alternative states of affairs or possible worlds— could have actualised but with different degrees of likelihood. Do we need to stretch the realms of reason as far as abstractly possible, to infinity, to reach actual impossibility (red)? Say, could an electron be an elephant if the world was recombined?


Hume, as discussed, believed that we contrive ideas of causation from personal experience (and the will of the mind). However, his view demands that the human mind cannot possess innate knowledge for it to begin existence as a blank slate. But, as an empiricist, Hume would have been intellectually obliged to accept new evidence which suggested knowledge was innate to humans. Yet this is exactly what some studies suggest. One such study (Muentener and Carey 2010) purports to show this in infants. More than a summation of learned experiences from their surroundings, the authors claimed that the infants were able to expect certain things to happen, suggesting that there is a bodily origin to how humans make predictions. It could then be claimed that we genuinely possess causal insights into the physical world, akin to Avicenna’s ‘knowledge by presence’. Some animals, like chimpanzees, too, have been shown to use notional senses of causation to generate creative plans of action (Kohler 1925).

If these arguments are correct, Hume’s entire stance is undermined, for humans could be credited with knowledge outside of experience. Moreover, they could claim that we are capable of exerting control over objects and not limited to cognitively holding impressions of them. If, however, these experiments are instances of scientists demonstrating incorrect conclusions, Hume’s position would not be undermined.

The jury is still out on this one.


Figuring out or already know the world?


The arguments against all human causation being a necessary connection between constantly-conjoined events are compelling. While we can be quite sure an ‘effect’ will follow a ‘cause’ from experience, without ever possessing innate knowledge of what can happen, without ever having certainty, without ever having free will, we cannot truly claim to possess the power to cause anything.


Hume, David. 2008 [1748]. Enquiry Concerning Human Understanding, ed. Jonathan Bennett.

Kohler, Wolfgang. 1925. The mentality of apes. New York: Harcourt. Cited from: Passer, Michael et al. 2003 Psychology: Frontiers and Applications, First Canadian Edition, McGraw-Hill: Toronto, p267.

Muentener, Paul, Carey, Susan. 2010. Infants’ causal representations of state change events, Cognitive Psychology, 61: 63–86.

Pereboom, D. 2007. “Hard Incompatibilism” and “Response to Fischer, Kane, and Vargas”, Four Views on Free Will, Fischer, J., Kane, R., Pereboom, D. & Vargas, M, Oxford: Blackwell Publishers.

No Comments

Leave a Reply