While my discussions elsewhere are settling down, it’s time to finally tackle one basic issue that underlies all reductionist endeavours, I need to write this down also to justify much of my claims on matters of brains and minds. In these pages I’ve frequently hinted at the notion that any given epistemological approach necessarily comes with its own idiosyncratic problems; or, if you prefer, will come with some inherent blind-spots. In this post I’ll discuss the obvious trouble that comes with reductionism*. It’s a bit sad that this seems necessary, but after lengthy discussions I’ve come to the conclusion that writing the following in clear language might be useful: the implications I will describe seem to be missed or overlooked by far too many people.
Imagine a very young me, I used to have a blue alarm clock, very much like the one shown here. One day I got hold of a set of tiny screwdrivers and carefully dismantled it to find out how it works. After several hours, having reconstructed and de-constructed the whole thing many times, I got a very good understanding of how it worked. Satisfied with my new knowledge I went to see dad, waving the clock and proudly announcing:
Dad, I’ve discovered that this isn’t an alarm clock: in reality it is just a bunch of smaller objects arranged in such a way that they look as if this was a unique object, but it really isn’t.
I was an inquisitive and philosophically inclined child! This little story is invented, but it does sound like me at a young age. A sensible reaction from dad would have been to explain me that this is how most “things” work: they are made of smaller parts, arranged in such a way that they produce coherent behaviours. This applies to man-made mechanisms, but to living organisms as well: in the case of biology the smaller parts are much, much smaller.
The hypothetical episode accounts for the “obvious” quality of my whole point: we all know the imaginary child-me reached a naïve and unhelpful conclusion. However, we do use reductionist approaches to study much more complex natural phenomena, and when we do, we frequently forget about the inevitable implications, and repeat the same error, over and over, and over and over again.
To proceed in order, the first thing to observe is that when dismantling the clock, imaginary-me was actually applying standard reductionist methodology. The aim was to “understand how the clock works”, and to do this, I dismantled it, carefully studied how its internal parts interact, and was therefore able to more or less grasp how the overall behaviour is produced by the interactions of its smaller parts. Was I wrong in saying that it isn’t an alarm clock? Clearly, but the underlying assumption: “it is not a single entity, it is actually made of many single objects” was correct. If I had concluded “it isn’t a single object” it would have been equally naïve and unhelpful, but somewhat less wrong. We can all see why: as long as it works, in everyday life makes sense to treat and consider the alarm-clock as a single object, even if we are perfectly aware that we can dismantle it into smaller components. However, this whole thought experiment uncovers a general pattern, which we do tend to forget for complex enough domains: reductionism relies on the assumption that we can dismantle our object of study into smaller parts. Thus, it requires to treat “our object of study” as something that only appears to be unitary, but can in fact be subdivided. When it comes to alarm clocks, nobody gets surprised, but when it comes to the study of minds, almost everyone is ready to jump to unwarranted conclusions.
Before diving into philosophy of mind, there is one more thing that our alarm clock example uncovers. Another conclusion I could have reached (if I was even more philosophically inclined!) is:
Dad, I’ve discovered that the trilling sound this clock makes is an emergent quality. None of the components of the clock can make that sound on their own, but when you piece them together in exactly this way, the result is able to make this remarkable sound. It’s fascinating!
The above would have been entirely correct, but probably none of us is likely to be equally fascinated by such emergence. That’s just how things work, right? Indeed, but again, when it comes to philosophy of mind (and many other equally challenging subjects) people regularly spend plenty of efforts trying to uncover the mysterious laws of emergence, reflecting on what emergence really is or other related, ultimately self-explanatory riddles. I’ve got news (and I’m sad to realise they are news to some): emergence is an artefact of reductionism. Whenever you try to understand a phenomenon by identifying the smaller parts that produce it, and describing how such smaller parts interact, you are assuming that the actual phenomenon is an emergent property of the underlying system. Emergence is implicit in the hypotheses that make reductionism possible (and tremendously useful). There is nothing mysterious about emergence: it is an entirely (and inadvertently) invented phenomenon, it is the result of the most useful strategy that we use to understand how reality unfolds. Emergence is epistemic: it’s the flip side of the explanatory power of reductionism. Just like any other epistemic approach, reductionism has its own blind spots, one is about emergence.
That’s not to say that reductionism is flawed: I’m a fanatic defender of reductionism, but my fanaticism requires me to spend some time and effort understanding the limits and blind spots of reductionism, so to compensate my own biases. So here is one: you can’t explain what emergence really is in reductionist terms, because reductionism implies emergence. Reductionism works because regular interactions between smaller parts can frequently generate coherent behaviours on a larger scale. That’s it, can we please stop wasting rivers of ink/bytes on such a non-problem?
No, we clearly can’t. Why can’t we? I don’t really know, but we still can’t.
To show that the errors/limitations I mention above are really worth discussing and clarifying, I will use the example of philosophy of mind (because I know something about it), other people might want to try the same exercise with quantum or theoretical physics, if they are so inclined.
If Free Will is an illusion, also my alarm clock is illusory.
If you have the slightest interest in minds you almost certainly have heard of the Libet experiments on free will: according to the prevalent narrative, the experiments show that free will does not exist, it’s an illusion. If the straightforward interpretation of the results is correct we must conclude that we become conscious of our own “free” decisions only after having made them. Therefore the idea that we consciously make decision has to be wrong.
The Libet experiment is clearly reductionist: it tries to isolate the different parts of the decision making process, and see how they interact (in a very coarse fashion!). By proceeding in this way, it assumes that there are different parts in the process, and that they interact. If they interact, it goes without saying that some things will happen before some others. Thus, what the results tell us is that indeed it seems that the decision-making process is made of smaller, distinguishable mechanisms. Or, it tells us that reduction can indeed help us to understand how we make decisions. We can thus infer that one day it might be possible to explain human decision-making in mechanistic terms . What would have been surprising is very different: if it turned out that we became aware of having decided something before any associated activity happened in the brain, that would have left open the possibility that something non-material was generating the decision, leaving various forms of dualism somewhat viable. So where does the “Libet demonstrates that free will is an illusion” claim come from? There are two interpretations of this claim, naturally: the weak interpretations claims that “free will and/or our sense of agency aren’t what they seem on first inspection“. To us it naturally seems that we consciously make decisions, but if we admit that some underlying brain mechanisms must make our “conscious decisions” possible, and therefore expect reductionism to be useful, we are already accepting that:
a. these underlying mechanisms are not consciously perceived (otherwise we would have tried to dissect them, not our entire sense of agency).
b. if these mechanisms do exist, they won’t all happen at the same time, and there would need to be a temporal distance between settling for one choice and perceiving the result of the choice.
Negating b. requires to add some magical extra-temporal element, not the typical ingredient of scientific and reductionist efforts. Thus, I don’t have a problem with this interpretation, except noting that Libet’s (and many others’) experiments merely confirm what we already assumed (points a. and b.): they confirm that our agency depends on brain mechanisms, which is what we wanted to verify from the very start.
The second interpretation is that “free will is an illusion, in the sense that it doesn’t exist”. This is the claim that gets on my nerves…
The differences with the clock example help understand why we tend to jump to the wrong conclusion in one case but not the other.
In the clock case, we perceive it as a unitary object, but we know from experience that mechanical objects are made of smaller moving parts. Each one of us would have seen, touched and experienced the presence of such (or analogous) smaller parts, and witnessed how mechanisms interact to produce coherent behaviours. On the other hand, when it comes to our own decision making processes, the only experience we have of them comes from introspection: from that point of view, they (almost) always look unitary. However we might try, we never get the chance to observe the inner workings of our (or someone else’s) decision making mechanisms. Thus, we always consider only their face value, and get genuinely surprised if a scientist uncovers some tiny detail of their inner workings. The sad fact is that we (assuming my readers are naturalistically inclined and don’t believe on supernatural souls or similar fantasies) should be surprised if scientists were unable to discover any of these inner-workings, not the other way round.
Once again: if we study a system in reductionist terms, we start with the assumption that it can be subdivided into smaller parts. Finding out that this assumption does hold grants us exactly the same right to declare that “free will is an illusion” and that “the alarm clock doesn’t exist in itself, it’s just a bunch of carefully assembled smaller objects”. Both conclusions are somewhat correct (in the most charitable reading), but the only thing they tell us is that there is hope that we can find interesting things by applying reductionism to the study of clocks and free will. Few people/scientists are surprised by the latter conclusion, but for some reason, most don’t see why it is equivalent to the “illusion” statement (in the weak form). For the clock, we would all react as my dad would have, and say something like “don’t be silly, of course it’s made of smaller parts, that’s why it works”. In the case of free will, because we are introspectively blind to its inner workings, discovering that it’s made of smaller-interacting parts, tends to make us conclude “it’s an illusion” (as in: it doesn’t exist). This isn’t formally wrong, but I completely fail to understand what makes the conclusion useful. Do we make decisions? Yes. Do we have some perception associated with making decisions? Yes, it feels like exercising free will. That’s it.
Thanks to Libet and many after him, we now have many reasons to believe that our ability of making decisions is the result of some mechanisms inside our brains. Thus, our free will is an emergent quality of our brains, in the same way as the trilling sound is an emergent quality of the alarm clock. Neither the clock, nor the trill, nor the brain, nor the fact that we make decisions are non-existent. They might be different from what they look like at first sight. But they still exist.
The same kind of reasoning applies to “the self”**. If we study our brains/minds in reductionist terms, we start with the assumption that someday/somehow we’ll be able to isolate the inner workings that produce our sense of self. We should be surprised if this attempt will prove to be impossible, not if we’ll find some ways to proceed. But above all, we should not conclude that “the self” is an illusion, only because reductionism is working as expected. In the same way, being surprised that the self will start looking as an emergent quality of brain/minds is equally naïve: it’s why we thought that reductionism would work in the first place!
This isn’t to say that free will, the self, or consciousness are not somewhat illusionary. I’m the first one to admit that they are not what they look like at first sight (that’s why philosophy of mind is so much fun!). However, if you use the successful results of reductionist science to declare that any of the above is an Illusion as in “it doesn’t really exist”, you are actually demonstrating that you don’t understand how reductionism works, and nothing else.
Notes and References:
*As far as I know, and to my complete surprise, the only argument that comes close to what I discuss here also comes from philosophy of mind, but applies to a slightly different issue, namely the claim that consciousness (or qualia, or phenomenal experience) are epiphenomena. Jaegwon Kim (1993) uses the example of pumps instead of an alarm clock, but the underlying logic seems the same to me. It is used to show that causality chains change according to how we (semi-arbitrarily) pick what is to be considered as the “underlying mechanisms” (or smaller objects that make up the clock), thus, showing that behaviour is caused by mental mechanisms does not imply that consciousness (or qualia, or phenomenal experience) can’t have any causal effect.
**Therefore hollowing out the standard incompatibilist view of free-will: if our decisions are the result of some mechanism, they can’t be our decisions – they just happen, whether we want it or not. Our will is therefore irrelevant/epiphenomenal. Hello? Of course my decisions are the result of some mechanism, the small detail is that these mechanisms (along with lots of others) are precisely the best possible definition of what makes me me. I am the system that hosts these mechanisms, they are part of me. So yes, I am taking my own decision, because I am (also) made of the mechanisms that make them possible (or necessary), duh.