I interrupt my mythology book reviews to turn to the completely different matter of neuroscience.
David Eagleman's Livewired was one of my favourite reads of recent years. He put forward a genuinely technologically optimistic study that has absolutely sod-all to do with that stupid manifesto, and felt like a much-needed counterweight to the equally stupid "all technological innovation is crap" cynicism which seems to pervade social media. He also explored things philosophically, claiming that it is a actually possible for us to experience new qualia.
I've mentioned Livewired a few times over the years and it's my shame that I never did a review of it. I think I'm going to have to re-read it to do it justice.
Anyway, his earlier, shorter book The Brain was an obvious read. The major theme of Livewired was the brain's tremendous adaptability, how it could repurpose itself to accomplish new tasks with old hardware. Eagleman did a very thorough job of describing just how far this could go, setting out both when surprising levels of flexibility were possible but also when limits would be reached.
There are plenty of obvious overlaps between these two books. Since The Brain covers these much more concisely, perhaps by covering them in brief here, I can shorten my eventual review of Livewired... which would otherwise risk just rewriting the whole damn book, 'cos it was bloody good.
The Brain is a very good, very short read. I give it 8/10 for being such an excellent but little compendium on how the brain works. It doesn't tie itself down in unnecessary caveats or tangents but doesn't skip the uncertainties either. It just gets on with things, so so will I.
Semi-permanent plasticity
Brain development, says Eagleman, continues roughly up to the age of 25 or so. Sure, the earliest years are important (in Livewired he makes a bit more of this, noting that certain very basic skills are almost impossible to develop once out of childhood, e.g. feral children will never fully integrate into society), but much happens throughout adolescence and early adulthood too. New connections are formed and discarded as they prove useful or unhelpful (I believe in Livewired he describes this as a process of babbling).
Teenagers brains, he says, are literally different from adults, just as children are. They're socially awkward risk-takers not just because of lack of experience, but also because of their neurology (though of course, presumably experience plays a large role in shaping their neural structures). They are, paradoxically, emotionally hypersensitive but also prone to seeking out highly emotional activities, and unable to control their emotional responses as much as adults are. In one experiment they got physically anxious when asked to sit in a shop window but at the same time they send each other naked pictures of themselves*. They are quite literally immature.
* Not literally at the same time, you understand.
But that's not the end of the story at all. If all the major wiring is in place by around 25, substantial redevelopment is still possible indefinitely afterwards, at least for re-using old networks in new ways. Sufficient practise at a skill can cause macroscopic changes in the brain basically at any age; in Livewired he says this can occur in a matter of hours.
This plasticity, claims Eagleman, is unique to humans. Most animals rely much more on hardwired instincts*. Human adaptability gives us the tremendous advantage of of immense versatility, which far exceeds the penalty of our very long development period. Continuous practise may even help prevent the onset of dementia such as Alzheimers, not by fortifying the connections so much as creating redundant ones as backups. As some connections fail, even the ageing brain is able to repurpose old ones to do different jobs. The degradation can't be stopped but its effects can be reduced.
* I wonder, though, how much research has been done on animal brains in this regard. After all, animal intelligence has continuously exceeded our expectations.
Decision time
Plasticity also plays a role when the brain has to make a choice. The brain acts, says Eagleman, as a series of competing networks each vying for supremacy, although I rather prefer his other analogy of a parliament. The idea that we have many voices inside of us is quite real, with different networks continuously firing until, finally, the brain acts to "crush ambiguity into choices" and makes a decision.
Exactly how this happens is still somewhat unclear. There's a reward system, in that the brain favours networks which have previously made predictions that have been validated : if one mode of thinking has successfully predicted a good outcome in the past, that network will be favoured. Conversely, those networks which don't give predictions in agreement with reality are downvoted. The brain, then, acts as a sort of prediction engine, continually checking its findings against different models*. Even abstract concepts can act as a reward or punishment, but immediate, tangible effects tend to override all these. Hence it's easy to avoid doing the things we should do in favour of something else (eating ice cream instead of studying).
* This makes it all the more mysterious that the brain is generally able to do this very well for low-level activities (few people continuously stab themselves with forks) but is often shite at higher reasoning (like believing in the Jewish Space Laser).
Eagleman's suggested self-help solution is to be like Odysseus and nail ourselves to the mast : ahead of time, ensure that the actual situation we'll find ourselves in is more tangibly rewarding/punishing than it otherwise would be. Don't give the brain the option to be distracted by temptation. Perhaps more interestingly, watching our neural responses (when we've got the option to do so) has also proven helpful in getting participants to learn self-control. Maybe one day we'll have phone-accessible EEG-hats and can cultivate our own desired responses using an app...
But exactly how the brain decides, "this network is the winner" and decides that this is the right approach is nowhere made clear. Emotions, says Eagleman, act as a heuristic for decision-making, combining all the possible effects into a simple sensation we can respond to, which explains that uncomfortable sensation while we haven't committed to anything. This, though, is still more description than anything explanatory, albeit a useful one.
Finally, it's also interesting to me that neural activity is more coherent and correlated while we're asleep than when we're awake. Is the brain better able to get on with things without that pesky conscious mind sticking its nose in ? Maybe.
The nature of reality
This will have to be either very brief or extraordinarily long, so I'm going for the former. The world we experience is not the outside world itself, says Eagleman. Not only is there a delay in processing different sensory signals, but there's even a different delay for each type of perception. Yet somehow, the brain combines all of this into a unified experiential whole.
And this unified approach appears to be crucial : to assign meaning to a sensory input, it must correlate to something else. If you only have visual data, you won't be able to see. While I had the impression from some of the case studies of Oliver Sacks that we learn to read the world around us, it's more subtle than that – interactivity is crucial, an idea developed further by Peter Godfrey-Smith (another one who's books I really must blog up sometime). Interestingly, when this multisensory approach is denied, the result is hallucinations... eerily similar to LLMs.
Perhaps one of the most compassionate parts of the book describes Schizophrenia as a sort of waking dream. In this condition, says Eagleman, suffers experience hallucinations without any kind of distinction between them and real, external sensory inputs. This naturally explains their behaviour, which may be perfectly rational but in response to a reality all of their own. I wonder if it doesn't go even further than that : in my dreams I rarely respond rationally to anything, yet at the time it feels coherent and logical. Maybe the logical reasoning centres are also impaired, without affecting the sensation that things have been done correctly... the reward networks might be all messed up.
More philosophically, Eagleman stresses that our perceptual reality is not reality itself : "the real world is not full of rich sensory events; instead our brains light up the world with their own sensuality". Perceptual time is also not experiential time (let alone real time). Experiments have shown that while high-stress situations make us feel like time has slowed, they don't affect our ability to think or perceive more quickly.
Here I think I'd need to clarify with Eagleman himself exactly what he's getting at. I agree, our perceptions don't correspond directly to reality, in that our sensory experience has no special claim on validity (compared to, say, a bat or a whale), and that we have to learn how to create our own mental worlds*. But all the same, surely our experiences do have some equivalence with external reality. There is something outside that induces an experience; redness is not totally meaningless, nor is pain or sound or heat. We can verify these sufficiently well so as to reliably identify exceptions (like the responses of Schizophrenics) as being disconnected from that external reality. So in that sense, I don't agree that our perceptions are purely fabrications; since we have access to nothing else, it doesn't feel at all useful to me to claim that perception is not reality. Perceptual reality is the only thing we will ever have any access to.
* Eagleman mentions an interesting case of synesthesia in which letters are associated with colours. I wonder how this works, given that the letter symbols themselves have no meaning until they're learned.
Finally, I also don't agree that we don't have free will. Eagleman says (as do many others) that consciousness does appear to have some uses, particularly when learning a task for the first time or when things need to be done in an especially careful, controlled way, or when big-picture thinking is required – integrating one line of thinking from one subject with something seemingly unrelated. Seems fine to me, though as to why conscious experience should be needed for this is anyone's guess. And as for those experiments where brain activity can be predicted ahead of the conscious sensation of having made a choice... nah, I've covered that umpteen times already. Far more likely, in my view, that conscious sensation actually does involve being in control rather than a weird way in which the brain constructs a narrative* of deliberation for no good reason.
* Though I do also agree that constructing a narrative is something the brain does a great deal of, as well as creating a theory of mind for us to predict and respond to other people. Interactivity plays an interesting part here, too : while those with too much Botox are hard to read, they themselves have difficulty reading others as they're not able to be so expressive.
You will all become one with the Borg
Well, maybe. The brain's adaptability is not unlimited but it's tremendously powerful all the same; Eagleman gives the case of a girl with literally half a brain who lived a perfectly normal life (and see that recent post about the notorious but misreported Phineas Gage). A key concept Eagleman develops here is the "plug and play" model of sensory inputs. Essentially, our sensory peripherals – eyes, ears etc. – can be replaced and the brain will still find a way to deal with the information. It takes a while for the brain to learn how to process it, but it works (Kevin Warwick came to much the same conclusion in "I, Cyborg").
Personally I think this is most interesting stuff in Eagleman's repertoire. I love the idea of being able to, quite literally, reject your reality and substitute my own. And this actually works. Experiments have enabled people to "see" using tactile sensors on their backs, foreheads and tongues, hooked up to cameras. They can identify objects and accurately judge distances. It works equally well if the tactile interface is connected to an audio sensor instead of a visual one. And this can be done wirelessly, letting volunteers experience sensory input from other places on the planet – and in reverse, to control mechanical arms remotely.
Real, it turns out, really is just an electrochemical reaction in your brain.
Of course, things are still limited. The possibility of digitally scanning the brain is a stupendous, unrealisable challenge. Transhumanist dreams of fully uploading our consciousness are also likely futile even in a materialistic perspective, even if we did manage to scan our entire brain and its neural patterns. If consciousness is indeed just those patterns, then we could simulate the brain inside a computer, yes... but, Eagleman says, while it might well be something (or someone) that experiences something, it still wouldn't be us. Even from his perspective of consciousness being emergent, which I don't really agree with*, a transfer just doesn't work. A copy ? Sure, if it's really the patterns that are conscious, and not something to do with the physical substance. Even then, it would have to be capable of change, of forming new memories and responding to new sensations.
* He attempts to refute the argument of Leibniz's Mill, saying that yes, you can't find perception in any single mechanical part of the brain, but maybe you could in the emergent whole. I just don't see it. To me, the gap between experience and physicality is just too large, too fundamental to ever be bridged. The only way I can see this working is if mind and matter are, somehow, of essentially the same stuff (neutral monism, idealism, etc. etc. etc., you know the drill by now). Otherwise, I have no clue how a gigantic abacus could ever be conscious, or how this would happen if we moved the balls around but only in a very particular way.
In short, a fascinating and worthwhile read. I'm going to have to try and get hold of his TV documentaries and read his other books (of which I'm glad to see there are several) at some point. One day I should try a fuller write-up against materialism, but what's really interesting is that often the disagreement with my own, somewhat more dualist perspective, is that the disagreement isn't really that stark after all.
