Review : Sentient (Part I)

The blogging binge continues with a swift return to the default topic of consciousness. Although this time from quite a different angle to the usual.

Prelude

Jackie Higgin's Sentient is an excellent read. Naturally I chose this one because the issue of animal senses has an obvious bearing the the nature of qualia, but I'm also plenty interested in wildlife purely for its own sake (433 posts here with the "Nature" tag, and counting !). I find it fascinating that animals can experience the world profoundly differently to me and yet, even when having supposedly inferior senses and intelligence, still manage feats of which I'm incapable. 

Of course, those with superior senses are even more fascinating, but those which have different senses are surely the most interesting of all. I love the idea of colours I can't see, or senses as different from sight again as sight is from touch or smell. Why do we experience the world in the way we do ? Why doesn't our brain (usually) render vision as an audio signal ?

And if animals do have such fundamentally different senses, what does that mean for their intelligence ? In my view, qualia offer a route to free will. By allowing us access to non-physical information unconstrained by material reality, determinism is circumvented. If this line of reasoning is correct*, it means that animals have access to** information beyond what's available to me. It's not necessarily that they don't think in similar ways, it's that they - might - have a fundamentally different experience of reality altogether.

* And I'm not at all sure that it is, mind you. Just because blueness itself is not physical does not mean that the experience of blueness is not perfectly constrained by physical reality; qualia could still be absolutely deterministic.
** "Access" may be misleading. Note I do not suggest that the inner mental worlds represent some other level of reality, tending as they do to be woefully inconsistent and there being no convincing evidence for ESP and the like whatever. But read on.

The Review Bit (skip if you want to know about the animals)

Higgins takes an unusual approach, which I'm glad I didn't realise initially or else I might not have read this. While she does mention a few animal senses which are fundamentally different to anything humans possess, the main goal of the book is to compare animal senses with our own. The blurb on the back cover doesn't make this at all clear, but in the main it turns out to be a very interesting exercise indeed. While nothing as grandiose as getting inside the minds of the animal subjects is much attempted, this is played to a strength, keeping it tightly controlled and focused.

I'll note that the book is fiercely accurate in its title, unlike certain Google engineers. The book is about senses, not intelligence. Very little is made as to how animals learn or improvise; again playing this focus as a strength. This is probably worth mentioning as many people seem to confuse awareness with intelligence, which are not at all the same thing.

The book is commendably clear and readable from start to finish. Higgins is an excellent writer, balancing detail, clarity and concision pretty much perfectly. Anecdotes and descriptions of personal experiences are properly set in context. And I will cite this as a truly outstanding work for understanding the little guys of research : hardly anything in this book is depicted as a Eureka-moment revolution, virtually everything emerges from the hard graft of patiently slogging it out. Higgins does an absolutely first-rate job of conveying just how damn difficult it is, of how even apparently simple questions which seem like they should be answered in an afternoon end up taking years or even decades to answer. There are no geniuses to be found in this book at all : clever people, certainly, but much more important is that they are obsessively-interested and exceptionally determined, to the point of being bloody-minded. 

Likewise, the ongoing controversies are stated without embarrassment; Higgins does not feel like she's trying to give any particular view undue preference, and manages to state both the arguments and feelings (for scientists are all ultimately only human) of both sides without judging anyone. And this comes across very clearly also in the descriptions of how we've got the results we have when at last the conclusions seems more firmly established - yes, it's a story about challenging the consensus, but it's also about how many of those challenges are simply wrong*. Science is only in part about hypothesis-testing; it is also very much about pure exploration, of going down blind alleys only to run back in terror. If you want to get an impression of science as a process rather than a result, you could do far worse than this. It's messy, subjective, provisional, mind-wrenchingly tedious, and fun.

* Mainstream scientists may be overtly hostile to new ideas more than I suppose, but this still doesn't mean that pseudoscientists with an axe to grind ever turn out to be correct.

Yes, I know I've said all that before, but a lot of people still don't get it, so it bears repeating.

There are a few minor points that irked me though. Some sections feel like an example of bait-and-switch. In particular the chapter on the vision of the peacock mantis shrimp, first describing it as having unimaginably superior colour vision to a human only to swiftly refute itself in a clumsy and abrupt way - we don't get anything about why the shrimp has such an undeniably complex visual system to accomplish an apparently unimpressive end result, nor what it's doing with the faculty of sensing polarisation (that it does at least possess, unlike humans). The chapter on the vampire bat has an analogy to the human sense of pleasure and pain which is extremely forced at best, and comparing the sense of balance in bipedal humans to quadrupedal cheetahs just left me wondering why birds wouldn't have been a more illuminating comparison.

These points are, I stress, minor. The book is rife with interesting findings that help better understand both the animal and human experience, and I'm happy to give this an overall rating of 8/10.

On to the content. This one is quite difficult to organise. Practically every chapter contains something that's interesting in its own right, covering a huge range of areas. As I read it, two major themes emerge : 1) that we often under-estimate the strength of the senses of which we're acutely aware; 2) we may have unconscious senses that influence us without our knowing. The boundary between the two is not sharp, but I found there was so much interesting stuff here I simply had to split this post in twain. 

Super sensitive senses : sight, sound, smell and stuff

Smell

All of our senses, Higgins claims, are a lot more sensitive than we generally give them credit for. We like to point out how much sharper the vision of a falcon, how much more powerful the nose of a dog (or indeed most animals), how much better the hearing of a bat or an owl. Similarly we point out how puny we are compared to an elephant, how slow compared to a cheetah... none of this is without foundation. But it does seem to be enormously exaggerated, a humblebrag to underline how truly mighty our intellect must be in order to overcome our pathetic physical frailties in our rise to dominance :

Broca wrote [in 1879] : 'This [frontal] lobe enlarged at the expense of the others, grabbed the central hegemony; the intellectual life is centred there; it is no longer the sense of smell that guides the animal: it is simply intelligence enlightened by all the senses.' He believed that... our sense of smell had been exchanged for free will.

To be fair the idea that a bigger section of the brain means it can do its job more intensely is understandable, but we now know it isn't true : not only for information processing in general, but also for the senses in particular - elsewhere Higgins describes that those with a better sense of balance actually have smaller part of the brain dedicated to this*. Analogies of the brain to a computer are deeply flawed, arguably fatally so. But this bias against smell is even worse :

* Might the brain be an example of how technological advancement would work in a less capitalist society ? We invent labour-saving devices and then go on to just do different labour instead, often apparently increasing the amount of labour we actually perform. The brain, by comparison, thinks the point of labour-saving devices is to do less work.

Freud argued that smell was only of interest to children, perverts and neurotics. He stated, 'The organic sublimation of the sense of smell is a factor of civilisation.' Its loss supposedly raised us above other animals, effectively civilising and humanising us. Our species was relegated to the status of micro-smats : 'tiny smellers' to the dogs macrosmatic supremacy.

Even leaving aside the bizarre claim that losing the sense of smell was somehow an asset, let alone a factor in the creation of civilisation itself, this is flat-out wrong. Higgins demonstrated that our own nasal sensitivity is far from shoddy. It depends on what we're being asked to sniff - some animals are better than us at some smells, but the reverse is also true. There does not seem to be a nose which is simply better or worse across the board.

Laska has quantified the smallest amounts of scent detectable to an animal. Time after time, comparing the results to human studies, he has been surprised. 'I have found that human subjects have lower olfactory detection thresholds. That is a high sensitivity to the majority of odorants detected so far, compared to most mammal species tested so far.' Humans, it seems, can do better than monkeys, macaques, sea otters, fruit bats, vampire bats, mice, rats and shrews. According to him, 'Human olfactory inferiority is a myth.'

The dog is the single species that Laska has found to flout this pattern. 'Only fifteen scents have been tests on both humans and dogs... but it is interesting to note that humans even outperform the dog with five of the fifteen odorants.'

What of the claim that a dog's nose is so many thousands of times better than ours ? Well, there are differences between the sensory apparatus and brain structures, but quantifiably assessing how that translates into experience is nightmarishly complex. In brief, the slightly poorer nasal hardware may be more than compensated for by a brain which is far more efficient and flexible; it's more of a software issue than a hardware problem.

You may wonder why this appears so obviously not the case in practise. The answer seems to be that we just don't need to use our nose, so we don't. Experiments have shown that when people put their nose to the ground (which of course they very seldom do), they can follow scent trails just like a dog. As well as being socially awkward, this is physically much more difficult for us. So a dog's world may well be far more dependent on scent than ours, far more odour-based than vision-based, just not necessarily because its sense of smell is anywhere near as superior as popularly supposed. It's a case of use-it-or-loose it, and we just don't use it. Dogs do.

(Incidentally, there are many examples throughout the book of the senses becoming more acute when one is damaged - hearing especially. This does not appear to be the case for smell, perhaps because our nose is in such a difficult position for continuous use. And, citing Oliver Sacks for a strange case of nasal hallucinations, Higgins notes something that was missing from Sacks' own book : the drug-addled patient described in that particular case he later admitted was himself !)

Sound

If we have a perfectly decent sense of smell but don't appreciate it because we're just not using it much, the same cannot be said of hearing. Animals like the great grey owl appear to have a phenomenally powerful sense of sound, but this again turns out to be misleading. 

Bloodhounds have special adaptations that allow their nose to be more efficient in collecting odours. Likewise the owl's whole face works like a giant ear to focus the sound. These are advantages which we just don't have, so it would be foolish to claim that our senses are wholly equal - they aren't. But collecting the sound is just the first step in the process. The second step is to convert that sound into a signal the brain can process, and the third step is for the brain to register that signal as a sensation. If we're found wanting for step one, the same cannot be said for steps two and three.

The decibel scale for sound is a logarithmic ratio, so zero decibels does not mean there is no noise present. Total absence of sound is impossible - as long as there are some particles present, there will be some waves propagating, some atoms vibrating (there is, of course, sound in space, no matter that it's inaudible to puny humans). So while a whisper might be 20 dB, under the right conditions we can hear right down to zero decibels. In specially designed anechoic chambers, researchers have reported hearing their own circulatory systems and even the sound of "my scalp moving over my skull, which was eerie, and a strange, metallic scraping noise I couldn't explain". And a 0 dB sound wave corresponds to an absurdly tiny pressure wave, a vibration smaller than the diameter of a hydrogen atom.

That's pretty mind-boggling. It's not the only example of human senses probing limits we ordinarily think of requiring fantastically advanced instruments, a point I'll return to later. The great grey owl, however, can hear the faintest whispers under perfectly ordinary conditions, e.g. a mouse beneath the snow in a real-world forest environment - it doesn't have to hope that for some reason the mouse happens to be in an anechoic chamber. It seems that there's no evidence that the owl's brain or the noise-registering hairs in its inner ear are responsible : it's entirely due to the owl's face, shaped like a collecting parabolic dish to amplify the signals. Presumably, like a telescope, this amplifies the signal-to-noise ratio, preventing the owl from being deafened by the sound of random air molecules bashing about.

One other point - Higgins says that frequency and tone are directly equivalent. But this is demonstrably not the case for photon wavelength and colour, nor for chemicals and odour. So I have to wonder if this really is the case for sound... it would be surprising if sound alone was so uniquely objective. A quick Google search reveals the existence of the condition amusia, suggesting that probably you really can be genuinely tone deaf just as you can be colour blind.

Sight

I've been conducting an experiment of my own recently. In order to test how well I can recover the signals of real galaxies by inspecting data sets, I've injected a bunch of fake ones into real but empty data (so that the noise characteristics are the same) at different brightnesses. The results are interesting. There does come a point below which the artificial galaxies are so faint that none are detected at all. But when they're even slightly brighter, it's possible to detect them at a rate high above what would be expected from chance alone. It's a strange experience - at these very faint levels, it feels like guessing, but then it turns out I'm right.

This sounds very similar to experiments trying to check if humans are capable of perceiving single photons. After years of painstaking work of more than 30,000 experiments, it seems that they are :

We found that in an alternative forced choice setting, humans perceive single photons 51.6% of the time. The importance is not how much above 50% the results are, but that are above 50% with statistical significance and statistical power, making it highly unlikely that the result is a consequence of random measurement fluctuations.

I'm naturally skeptical of results slightly above a threshold value so I'll just note the paper can be read here for those who want to dig deeper. In my own experiment, the volume of the data occupied by the fake galaxies is only ~0.2%, whereas the faintest detectable galaxies are found at rates of a few percent - so it may well be that anything above 50% is indeed disproportionately significant. I would hope that they at least did tests in which no photons were injected at all, but I'm not going to read the paper right now. Higgins is anyway careful to note that others would like to see the result replicated*. Anyway :

* Though it does not seem at all controversial to state that we can detect photons in very small numbers, tens or less for sure. One other claim I've previously heard is that the sensors in the eye might be able to respond to single photons, but ordinarily the brain applies its own filters to stop us being overwhelmed by the noise. I will attest from personal experience that the brain certainly does this for sound.

When the volunteers were asked to describe such perceptions, they agreed unanimously that, whilst difficult to put into words, their experience was far from a flash of light. One said, 'If you've ever looked at a dim star in the night sky and one second you see it, but they next you don't... it's kind of like that.' Another, 'It's more a feeling of seeing something, rather than really seeing it.'

Higgins also notes that our rods, which sense brightness and not colour, are not so dense within the central region of our eye, which helps explain why very faint objects (such as, indeed, stars) are easier to see if you don't look directly at them. A colleague of mine who also enjoys amateur astronomy also notes how important it is to train your vision to such tasks : with practise, it's possible even to see incredibly faint dwarf galaxies through a small telescope.

The comparison Higgins uses in this chapter is the spookfish. This is fascinating creature of the deep that actually uses mirrors in its eyes to focus the image. While superb at low-light vision, this is a specialist. Our eyes are much better adapted to a wider range of conditions, though Higgins notes that some people are born with super-sensitive dark vision that makes them highly vulnerable to bright sunlight.

Sensory appropriation 

Most of us are aware of echolocation : bats and dolphins alike use sonar-like tactics to hunt prey and scan their surroundings. So do some blind people - more on that in the next section. But the star-nosed mole takes things in a radically different direction. Instead of replacing vision with sound, it uses an extraordinarily touch sensor located, of all places, on its nose.

Human tactile senses are impressive, being able to detect when cells are moved by just 500th of a mm, and allowing us to detect differences in a surface separated by just 0.2 mm - though this pales in comparison to our single-vibrating-atom sense of sound or single-photon sight. What seems strange to me is that our touch resolution varies across our skin, being markedly less on the back... but it doesn't feel worse there. And that's probably something worth remembering. To an animal with a wholly different sensory experience to us, its perspective of the world undoubtedly feels like "the right one", and anything differing from its norm would be somehow wrong and less real. Look at how many people have a bias against "false colour" astronomical images...

Anyway, the mole apparently does something extraordinary. Its unique protruding nasal implement is jam-packed with sensors, giving it a sensitivity and resolution I find difficult to imagine. I can't even accept that the resolution on my back is so much lower than on my fingers, so, just as I can't imagine new colours, contemplating life with a massively higher touch resolution is not at all easy. What I find really fascinating is that the mole might be "seeing" with its snout :

The star appears to act like a retina. 'The star's division into peripheral touch and central touch seems analogous to the retina in the visual system of mammals.' The star appears to act like a retina, allowing perception of the wider landscape... it is a tactile fovea - like the auditory one of an owl or a bat - to our visual fovea. As the star on the mole's snout presses down on the soil, it transmits to the brain a three-dimensional, star-shaped view of the terrain : a digital tactile image, with a particular focus on food. 'The nose may look like a hand, but it acts like an eye.'

While the mole's nose leaves us (for once) trailing miserably in the sensory department, it's to fellow humans we need to look to see how far this analogy can be pushed. Does the mole literally form an image, or does it sense touch just like we do but better, or perhaps something else altogether ? Higgins notes the cases of blind artists who can sculpt, but even more impressive is the case of blind artists painting with perspective. How in the world would you ever realise that something would look smaller if it was further away without actually ever seeing it ? Things don't feel smaller when you hold them close or at arms length. Intriguingly, the neural activity in these cases is strikingly similar to people with vision :

'A different point of view is an intrinsically visual thing; there is no tactile point of view. What I found striking was that he has vantage point and perspective without ever having seen. It suggests that this spatial awareness is somehow ingrained, innate.'... 'Esref is congenitally blind, but if you were to see this picture of the brain, not knowing what is going on, you would say this person is seeing.'

'It's provocative, but we're arguing that the brain may not be organised into sensory modalities at all... The striate cortex is visual only if you have vision. If you don't, it quickly takes on other sensory modalities.' What neuroscientists have called the visual cortex for the past century seems not to be devoted to the eyes... it might more accurately be defined as the area of the brain best able to discriminate spatial relationships and that it will use any relevant sensory input.

Kevin Warwick says in his book that the brain is remarkably plastic, capable of receiving and interpreting completely different sensory data than what the nerves are normally used for. It should be noted that others have attested differently, that vision is necessary for the perception of space. Why this should be different between individuals, or whether the mole does something vision-like or something else with its marvellous nose... I have no idea.

Conclusions to part one

Sensory sensitivity is astonishing, literally rivalling the finest scientific instruments available. This does not mean we record things with anything like the controllable, measurable objectivity of specialist equipment, because that's not at all what our brain is trying to do. Some argue that our perception is based around utility, not truth; I say that truth and utility are not in the least mutually exclusive and accuracy has a utility all of its own. But undeniably, we cannot count photons or state the exact decibel level or molecular count for some particular compound. For whatever reason, that's just not how the brain works.

What might be more interesting about this extreme level of sensitivity is that when you get down to the level of one photon, one particular molecule, one atom of movement, you're getting to the realm where quantum effects could potentially be important. I remember being chided for this back on Google Plus because apparently mentioning "quantum" and "consciousness" in the same sentence is an anathema to scientific inquiry. 

Well, I don't mean to suggest that quantum woo plays any role in mind at all (it might, but that's another story). What I would claim is that if the brain has really such extreme sensitivity, then things like the double-slit experiment could be important. CCD multipliers work by multiplying signals from single photons up to macroscopic levels, converting minute amounts of input energy into measurable electrical flows. So if the brain does the same, it does not seem beyond the pale to me at all to suppose that quantum effects play at least some limited role in consciousness, just as we can observe wave-particle duality effects and become puzzled by them. That doesn't strike me as woo, but mundane.

(It's often stated that the warm, wet environment of the brain is not a place where quantum effects are expected. I am not sure why this should be, unless there are large-scale field effects at work. Otherwise, I would not expect such rapid, minute processes to care at all about whether they're in a brain or a vat of liquid helium, and quantum tunnelling plays a role in keeping the Sun shining precisely because the temperature is astronomically high.)

More philosophically, how does our perception relate to our understanding of reality ? How would it be altered if we altered our perception - which reality is the "right" one ? That of a butterfly, a dog, an ant, a person ? Or is there only one reality which are all filtering differently ?

I lean strongly towards the latter. Not only does the existence of a blind artist understanding perpesctive point towards a shared but filtered reality, an innate understanding, but there are no things, only stuff. The problems lie not in reality itself, which exists all by itself, but in our descriptions of it, which don't. It's an age-old question of whether mathematics was invented or discovered; viewing mathematics as a language to me implies that it was both, in a sense. Clearly objects do obey physical numerical laws, but numbers themselves, equations and formulae, are not found in nature, nor is it necessary for cheetahs to do mathematical calculations to avoid falling over. Maths is a description of what's going on, but it is not the same as the the stuff that's happening itself. More on the importance of language for sensory perception in part two.

The sensory filters which animals have developed are astonishing. But we should not be too awestruck - our own capabilities are no so inferior. By the same token, we should probably not give ourselves too much intellectual credit either. As we discover that our senses are not so hopeless compared to animals, so we discover that animal intelligence is not so pathetic compared to our own. As we'll see next time, even things we think we understand continue to surprise us - all that patient, unrevolutionary work of unknown but ferociously determined researchers is paying dividends.