This site is not maintained. Click here for the new website of Richard Dawkins.

The raw deal of determinism and reductionism

One thing that comes up time and again in discussions on ethics is freedom, or more specifically some dichotomy between free will and determinism. In some flavours, determinism can be partnered with reductionism, thus generating claims that we are "merely" (what does that word even add to the statement?) colonies of cells, or selfish organisms, or blind physical forces, or robots, or any supposedly derogatory object that is calculated to connote worthlessness. In most of the discussions I've seen so far, it is implied that determinism, if correct, eradicates freedom and moral responsibility, while reductionism turns us into the "mere" X that is therefore incapable of any moral discussion.

The problems I see with this include an old one: Hume's fork. Either an event is caused or it isn't caused, in which case you've got either causality or randomness. Either way, metaphysical free will (the sort that posits an uncaused agent in the mind that isn't effected by causality outside) seems to rely mostly on gaps in our knowledge to support itself, and I think we've seen that trick done once too often to be credible. If you posit an uncaused thing that nevertheless deviates from randomness, it merely raises the question of where the difference is coming from. If I choose X option 70% of the time and choose Y option 30% of the time, if you find a cause for the difference (or candidates for one), then you've undermined your claim that it was uncaused. If you search extensively and don't find a cause, you have a roulette wheel that gives the illusion of direction, little different from a string of heads with no tails in sight. This last one might sound like a cop-out for some, but in that case I'd like them to consider the various fallacies many people have concerning probability. lists enough to be getting on with. Even Hume, who pointed out the problem of induction and then pointed out that causality was an assumption about the world that could be undone tomorrow for all we know, was pitting randomness against causality, not free will against causality.

However, I think a more pressing issue is that determinism and reductionism are not - indeed, cannot be - a threat to our ideas about freedom. Determinism's claims make no special mention of human agency (or agency at all, for that matter). True, we are not free to grow wings and fly away, but no one worries about that and everyone agrees that, by following predictable physics laws, we can build machines to overcome these difficulties. We are not free to spend more money than we can have access to; if I cannot get ahold of a million pounds, by credit card or theft or otherwise, I simply cannot buy something beyond that price range. We are not free to undo physiological facts; if I break my radius, it will not melt or glue itself together in a second no matter how much I try to will it to do so.

The rub comes in when anyone suggests anything that gets closer to the mind. I can change my beliefs and thoughts, or so it appears to me. Although I may be an outgoing person, I could reel myself in and settle down. No matter what social rules there are, I could conceivably ignore them and do my own thing, and if I was clever enough I could get away with theft or murder or lies or any immorality you care to name. It seems here I have some power to negate the causality the world imposes upon me. I might not be able to fly, but I might go on a plane or not. I can even defy my own genes by deliberately not doing things they would consider an imperative, like refrain from reproduction or even commit suicide. Surely, this is a contradiction of determinism?

Any attempt to examine this claim - for instance, by studying the "self" through neuroscience and psychology - seems to "reduce" us to electrical signals in a dense net of neurons. It is by reductive science - the sort that gave us gene-centric evolutionary theory and the four fundamental forces - that complex things collapse into simpler things. But then where's the ingredient that explains how it all works? How could electric flesh possibly be in any position to make judgements, never mind moral judgements? If we're caused, then surely we can't be identified as the causes of our own actions because I could point to a prior cause and foist blame on that, and so on. This approach (called reductionism) ends up looking like the inevitable admission that nothing interesting is going on.

I apologize if my partisanship has made me degrade these arguments' styles somewhat, but if you pick up an ethics book or look the terms up, I suspect arguments like these will feature at some point, and often in more academic language than I have mustered here. My purpose is not just to suggest that determinism has a strong case, but that the dichotomy between determinism and free will was erroneous to begin with.


Firstly, my ability to change my beliefs, attitudes, and so on is an observation of the facts of neuroplasticity, which most definitely owes its workings to causality. Ironically, my ability to go against my gene's interests was put there by the genes for their own interests. My ability to curb my sexual interests, for instance, is simply the outcome of genes rewiring a mind for better social prospects. Even if I died a virgin, my relatives could pass on the seed and so the genes still win. In any case, genes have limitations such as time lag and the limits of embryology, and occasionally an organism fails, but this is a predicted consequence of the algorithm of natural selection, not a defiance of it. Genetic phenotypes can be curbed only by other genetic phenotypes, sometimes from genes sharing the same body. My self-restraint - what some may call my free will - is an override mechanism little different from the mechanism it's overriding.

Brains are constantly changing as they process both inputs and internal processes elsewhere in the same brain, and it's this internal causality that goes some way towards explaining the apparently self-generating nature of our minds. Brains, and the elusive "I" that exists in each one, themselves have causes - brains have to grow, and before then they have to be born, and before any individual birth has to be a series of evolutionary changes and generations of similarly born individuals, going all the way back to the replicators and the first organic molecules. The fact that I can shut out influences from outside (a fact which in itself requires me to detect them and respond to them, requirements which presuppose causality) while thinking of something does not mean that my thought processes are not themselves causing each other, or that they could not be traced back to processes that happened earlier.

Determinism is also becoming much more subtle. Chaotic determinism enables even a causality web to becomes fiendishly unpredictable, at least to our eyes. Probability, statistics, fuzzy logic, and continua have all conspired to allow causality into places where people were originally just ignorant. Of course, randomness can come in where causality breaks down, but in this case free will hasn't gained an ally. It is now up against a tag team.

It is not just that we are ignorant of a lot of causes. It's that, because of the very nature of a simulation machine with limited processing capacity - i.e. our brains - we're currently doomed to know only a fraction of the available knowledge. Virtually everybody has died in ignorance of something or other. Even the cumulative knowledge of the experts is mostly an increasing awareness of the sheer scope of our ignorance. The philosophical thought experiments that presuppose an omniscient observer knowing all the causal webs, quite apart from raising the possible infinite regress of self-knowledge (after all, if he's omniscient, then he knows he's omniscient and knows every fact about himself, and he knows that he knows that he knows...), are hampered by the fact that no such entity has been confirmed to exist. And there are some things it is impossible to know for sure - the very problem of induction makes us aware that some things can't be known. We'll never know for sure, for instance, what caused the Permian Mass Extinction. The best we can do (and this is by no means to belittle the palaeontologists' efforts) is to solve the crime after it's been committed, without witnesses of the crime to help us.

Speaking of crimes, the notion that responsibility evaporates because we can point to prior causes and therefore shift the blame is made dubious by the reductio ad absurdum that, following the chains all the way back, you come to the conclusion that the Big Bang should be blamed for all the crimes ever committed. Indeed, the Big Bang caused every earthquake, every intention to kill, every stroke of luck and so on. It's possible to blame the universe, of course, but this is mostly achieved by anthropomorphising the universe as something that can be blamed, and here lies the key to the strangeness of this argument. In any case, this argument only really works if culpability was solely about causality.

It doesn't work if you notice that the notion of culpability is also focused on individuals with intentions and desires, not on causal links alone. If I blame a person for robbing a bank, I'm not just saying that he caused the money to disappear. I'm saying that he intended to rob the bank, and therefore that his actions were a subset of causal events that could have been or may well be overridden by other causal events, such as returning the money, incarceration, or rehabilitation. The complexity of causal webs should be the salient factor, including the likelihood of re-offending, and especially when we still have an active influence on events. After all, a complex machine is more likely to be capable of self-regulation than a simple one.

The intentions are not a magic force that explain away the problem, either. A creature with intentions is more likely to re-offend than a creature or a thing without them. This is why we make the distinction between manslaughter and murder. We also bow to causality every time we acknowledge that accidents will happen - probability and statistics point out that, given enough time, accidents become inevitable, and we concede human limitations. Our intentions and emotions and motivations themselves are products of evolution, which is itself a causal force. Moral culpability, when you observe it in action, owes more to causality than most might concede.

We dismiss an earthquake as a natural force with no malice, and we might dismiss a lion as being incapable of acting any other way when it kills someone for food, but this doesn't make it impossible for people to prevent the lion from doing it again (perhaps by killing the lion), nor does it mean putting up anti-earthquake buildings is a waste of time. We can prevent and counter a potential threat, and if the worst has happened, we can alleviate the damage. Even if criminals were put in the same category as lions (or, unlikely as this is, of earthquakes with no malice or desires), that does not prevent us from being able to identify and tackle the causes of crimes, up to and including the criminals themselves. Culpability is about identifying individuals with motives and desires, not about establishing every causal factor. Even when multiple individuals were involved, one usually has more influence than the others, and every participant can be excused or blamed based on many personal indications such as intention and ability. Nothing about causality suggests I can't identify a murderer from an innocent.


Secondly, the reductionism argument has become confused over what reductionism is about, at least from the scientific perspective. It is an explanatory mechanism for what exists. It is not a means of proving that something does not exist. It is true that, in the words of Douglas Adams, to dissect a cat in order to see how it functions is to end up with a non-functioning cat, but where's the link that says the cat doesn't (or if you prefer, didn't) exist - indeed, that the cat never existed in the first place? Certainly, neurosurgeons could home in on those lobes of the cerebrum that coordinate laughter reflexes and executive functions, but that doesn't mean there is no executive function. The "merely" adds nothing to the statement that thoughts are electrical signals. If anything, it seems to signal the speaker's disappointment that we are not angels with no attachment to disgusting or feeble fleshy shells. The fact that reductionism is considered a "reducing" idea is certainly a suggestive observation.

Much as I suspect introspection for its unreliability, I must inevitably make an appeal to it, or at least to something like it, for the next bit. Our mental processes (thoughts, feelings, and colour perceptions, say) and thus our experiences can be considered things that self-evidently exist. You can certainly claim that you don't have them, and I cannot distinguish between your claim being caused by a deliberate lack of cooperation on your part and your claim being a genuine claim of zombie-ism, but the non-random parallels between my behaviour and others, the long historical talk of consciousness and sentience, the philosophical conundrum of my red being your green, and the Mary's tent thought experiment in which a girl is raised in a monochrome room and taught redness before being given a red ball, all seem to indicate that my own experiences are shared by others in something called intersubjectivity. I am not inventing this word out of whole cloth, and the idea of common experiences is not a total certainty, but like many scientific theories, the world makes more sense using this parsimonious model than with others. There is little parsimony in the idea that other people's emotional expressions are nevertheless deceptions that hide a lack of true emotional experience. Indeed, there is so little parsimony in ideas of deception that, even if they were true, it raises the question of how the proponent could possibly know without getting the deception backwards (what if the revealing of the "deception" was itself a deception, and the real world was really real?), or without falling into an infinite regress (is that deception lying on top of another deception?). So we have reasonable grounds for supposing subjective experiences are real enough.

What are we to make of the fact that science gets on pretty well whether we make this claim or not? The reductionism argument seems to set great store on science's ability to ignore subjective experience. The first and obvious answer is that science was never the sole claimant to the truth, though it is easily one of the best means of getting closer to a complete account of the truth. The second and much more interesting answer is that the reductionist argument falls for something like the anthropic principle. It is clear enough that there is something to explain even before we begin our investigations. Certainly, we can be surprised by some of the counterintuitive ideas we encounter. It is very counterintuitive to think that solid objects are nearly entirely empty. However, this does not make the phenomena we want to explain suddenly vanish. The wall is mostly empty, true, but the notion of solidity is still explained and the distinction is still a workable one. We lack a comprehensive explanation for things like our feelings, though we have chipped away at the problem with impressive speed, but the distinction between a sentient brain and non-sentient bones or guts is still a workable one. My experience of anger does not vanish in a puff of logic simply because you can point to the hormones and EEG readings that describe it. Science is still working on the problem, but it admits the problem exists.

Determinism and reductionism thus pull through, not because they have finally grasped the truth, but because - like our excellent brain simulations and the world they simulate - they fit much better than the alternative options and are quite up to the job of helping us decipher it. Free will - at least the metaphysical kind - has by comparison done little more than stoked human egos and misled people down unpromising avenues. Indeed, it has done what intelligent design proponents have done; offered a pseudoexplanation and worked hard to justify its intellectual laziness, a paradox if ever there was one. It is tempting to blame religion for this, but religion works with what's there, and free will would never have been so alluring if it didn't appeal to people's desires for power and to people's fear for weaknesses that could be exploited. Even now, brainwashing, and something like it if you don't like the Hollywood connotations of the word, are terrifying prospects made easier if you think the mind is something that can be predicted or shaped.

The Raw Deal

I will come across as being too harsh on free will at this point, and as being too eager to demonize it. Surely, there's something in the idea? True, but the free will you describe and defend is mostly political freedom, which is mostly about a lack of coercion. Indeed, in my Collins dictionary, the second definition of free will entails lack of coercion. This free will is a worthy political goal, but it might be better to give it a less misleading name - say, freedom from coercion, freedom of choice, or emancipation. But here, determinism is not your enemy. Politics is many things - a social game, a matter of government, people deciding who speaks the softest and who wields the biggest stick - but these require knowledge of human nature. None of them requires a suspension of the laws of causality.

I may also come across as glossing over the problem of determinism. I cannot dismiss centuries of debate over the issue without explaining what caused it. Where there's smoke, there's fire. In this case, though, I suspect (as Daniel Dennett and Sam Harris do) that determinism has been confused with fatalism, and that in this case both suffered from a family resemblance.

According to my Collins dictionary, determinism is defined thus: as the philosophical doctrine that "all acts, choices, and events are the inevitable consequences of antecedent sufficient causes", which is a fancy way of saying that in principle, everything could be predicted in advance. I hesitate to align myself with the "all" qualifier, simply because randomness could play a part too, but note that human agency is not mentioned. It makes no difference whether something has intentions or desires, because nothing about intention or desire requires acausality.

By contrast, fatalism supposes that "all events are predetermined so that man is powerless to alter his destiny." Quite apart from the fact that it is not gender-neutral, the doctrine specifically mentions humanity. Moreover, it emphasizes human powerlessness. It suggests humans cannot take an active role in events and change their course. Determinism works whether humans are active or passive - its claim is that active participation is a subset of causal events, which is certainly the direction neuroscience and psychology are going.

The dichotomy should have been between fatalism and a form of free will that emphasizes our ability to change events around us (within limits), in which case free will has an easy win. Humans clearly have effects on the rest of the world, which can be quite dramatic. True as it is that we can't influence everything, our local influences are still real enough, and all that's required is the correct qualifier. Our active influence, in a chaotic, complex, sometimes random world, is our nervous systems and our other body systems causing other things, both among themselves and outside, after having been caused themselves by a non-random process called evolution by natural selection.

The original debate may well have started off as something like this, but the metaphysical free will and determinism debate strikes me as being possibly a subsequent distortion, a problem by association in which determinism ended up getting the flak for claims made by a close relative. Determinism is not an adversary of ethics, but a clarifier that cuts out too many useless misconceptions and gets to the heart of the issue.

This may seem a naive assessment, and I freely admit I'm not a qualified philosopher, but a look into the debates around science - one of the most exacting and unforgiving of intellectual pursuits - and it seems increasingly likely that such misconceptions could abound in a discipline tackling ideas that in any case are a hair's breadth away from causing confusion. If the Big Bang, group selection, and quantum mechanics (the weirdest and yet most exacting science of recent times) can generate so much confusion, what of the debates surrounding such nebulous concepts as free will, determinism, reductionism, and causality? I can't help but think determinism, or at least a variation of it, should be getting a better deal.