Review : The Mind's Eye

Having been thoroughly impressed with both An Anthropologist On Mars and The Man Who Mistook His Wife For A Hat, I felt pretty confident that any other Oliver Sacks book would surely be well worth reading. So I picked up The Mind's Eye at my favourite local bookshop-bar some time ago, being determined to at last actually buy a book there and not just drunkenly peruse them as is my usual wont. And then it languished on my shelf for some time, I already being deep in my Tolkien binge at that point.

I'm afraid I have to say my ill-treatment of this (ultimately very satisfying) book continued. When I finally started it, I found it a slower burn than the others. The first few cases described are undeniably interesting but add nothing new to those covered previously. The way that we almost literally 'read' the world, assigning meaning to form and colour, is fascinating stuff, but simply repeating this without trying to understand it more deeply can quickly wear thin.

It didn't help that, now being on holiday, I naturally went on a book-buying spree. Prague has pretty good English bookshops but of course the selection back home is always better. So I did that rarest of things... I paused and... started another book. That's normally a death knell for anything I'm reading, the direct literary equivalent of eloping with a cheerleader.

(The analogy holds quite well, since my interest was diverted by a big, full colour, fully illustrated book on castles of the Welsh princes. Which is lovely, and you can read my review of that one here, but it certainly isn't worth blogging about.

I have no idea if this analogy makes sense to anyone else, but dammit it seems appropriate to me.)

After getting this shameful infidelity out of my system, I decided to resume TME. Fortunately from the point where I'd stopped it rapidly became more interesting. The theme is how mental imagery works, which for philosophy of mind is the equivalent of crack cocaine. And so once again, Oliver Sacks delivers. I'd give it a 7/10 overall; the early parts are not as interesting as some of his other stuff (at least if you've read his other works), but the later bits are if anything even more so. 

Face blindness

One of the first interesting cases is one of the simplest, prosopagnosia  : the poor ability to recognise faces. Of course everyone forgets faces from time to time, but this is well beyond the norm. In some extreme cases this can extend to one's nearest and dearest (or even one's own reflection !), but far more common is to simply have severe difficulties with more casual acquaintances. Sacks describes his own personal experience with this, and those of other academics. One went so far as to have a notice on his office door explaining that everyone should introduce themselves. 

Exactly how this happens is not well understood, perhaps because being bad at faces is so normal as to be barely recognisable as a disorder (Sacks compares it with dyslexia, which is about equally common but far more widely recognised), and can typically be circumvented. What I found interesting were cases where context helped a great deal, that if a "sufferer" saw person X at location Y, then they'd recognise them straight away, but have difficulties if they were in location Z. And that's very familiar; I've been caught out many times by seeing familiar faces in unfamiliar locations (who hasn't ?). Several times I've walked to or from work and the server at the local restaurant has waved at me, and only later have I realised who on Earth it was. Familiarity, Sacks notes, is not recognition.

It makes sense that the brain would have some dedicated functionality to recognising something as common and of such vital social importance as faces*. It also makes sense that something this specific could malfunction, or work sub-optimally. But the location-based aspect** reminded me of albatrosses, who completely fail to recognise their own chick if it falls even just outside the nest. This seems absurd, but clearly, prosopagnosia indicates that the human brain is not nearly so different from our feathered friends as we might presume. Our perception of reality is weird.

* And it partly explains my racist grandmother, because face recognition is learned early on. Reversing the typical western perspective : "One prosopagnosic colleague acquaintance, born and raised on China, went to Oxford as a student, and has lived for decades in the US. Nonetheless, he tells me, 'European faces are the most difficult - they all look the same to me.'"
**There seems to be also a correlation with having a bad memory for places.

Stereoscopy

The second interesting case is one of those exceedingly rare parts of Sacks which is wholly positive : the case of a woman regaining stereoscopic vision after losing it in infancy. In a previous book he describes how an operation ostensibly restored full sight to someone who had similarly lost vision completely at a very early age, but in that case, the results were almost tragic. The patient could only learn to see again with great difficulty, at first lost in a frustrating mass of meaningless colour and form. Just like learning to read, the information was there, but the meaning was not.

For the stereoscopic patient things were very different. Despite the prevailing medical belief that stereoscopic vision must be acquired before the age of two, the now middle-aged patient was essentially bestowed with fully normal depth perception. And this was nothing but beneficial, an endless delight to the patient to finally "see the space" between objects.

This particular patient was also a professor of neurobiology. In a direct analogy to Mary's Room, she'd told Sacks years before that she thought she had enough knowledge to imagine what stereoscopy must be like - but after the operation, she recanted, saying it was a wholly different experience. Stereoscopy turns out to be a qualia, something which cannot be adequately described but only experienced. Somehow this had never really occurred to me before.

The consequences of living without it also took me by surprise. True, this is just one aspect of depth perception, with a host of other factors also at work. But some, including Sacks himself, rely on this more than others (it seems stereo depth is experienced to different degrees, with some people reporting far more pronounced depth when viewing stereo images - pairs and stereograms and the like - than others). Without it the world feels like watching a movie or a painting, a continuous series of bizarre, Escher-like illusions with perspective being almost irrelevant.

I'm surprised that the differing focus of nearby and distant objects didn't destroy this effect, but it seems it didn't. Clearly this long-term effect is something qualitatively different from what one experiences if one merely closes an eye.

Hallucinations

Next, Sacks returns to his more usual bittersweet tone as he describes the early stages of his own eye cancer. I hesitated to read this. I deeply appreciate Sack's sympathy for and empathy with his patients, but my main interest is the philosophical implications, the distinctions and overlaps between mind and body. Reading for the sake purely of an emotional journey is really, for many reasons, just not my thing at all. To be blunt, I can't handle it and I don't want to.

But it's not like that. Somehow, Sacks maintains his usual style even here, and if it's necessarily more emotive than usual, the intellectual curiosity hardly plays second fiddle to this. Not for nothing is he called the poet laurate of neuroscience.

Sacks describes how his cancer caused a series of changes in his vision. With an operation to his retina, at first things seem to have been as one might naively expect : a central patch of darkness in his right eye. But as things progressed (sometimes improving, sometimes degenerating), the effects became much stranger. It began with simple hallucinations of abstract shapes, but rapidly turned to persistence of vision : "If I have been looking at something and then close my eyes, I continue to see it so clearly that I wonder whether I have actually closed my eyes."

This effect became astonishingly pronounced, but before that there were weird effects on colour. Colour vision was all but lost in the centre of his field of view in the damaged eye, with colours from the remaining peripheral vision suffusing into the centre - and of course varying hugely depending on whether he used his good left eye, right eye, or both. Having different colours available with each eye is strange enough, but it became weirder still. The blind spot would "fill in" based on the surrounding colours in "about a second or two" :

The next day I tested this with a blue sky and found the same result. The scotoma became as blue as the sky... when a flock of birds flew by, they suddenly disappeared into my scotoma, emerging on the other side a few seconds later - as if they had been cloaked in invisibility like a Klingon warship.

Colour was only the beginning. Like modern image-generative AI, the blind spot could fill in with patterns : brickwork, clouds, wallpaper... though not faces or other highly complex shapes. 

I started to think of my visual cortex not just as a rigid duplicating device, but as an averaging device, capable of sampling what was presented to it and making a statistically plausible (if not photographically accurate) representation of it.

After gazing at the bookshelves in my bedroom for a few minutes, I closed both eyes and saw, for ten or fifteen seconds, the hundreds of books arrayed in almost perceptual detail... I had the sense that my visual cortex was now in a heightened or sensitised state, released to some extent from purely perceptual constraints... These images or hallucinations have greater clarity, are more fine-grained, than perception itself, as if my inner eye had an acuity of 20/5 rather than 20/20.

I can at least dimly conceive of what this must be like. I too (Sacks was a big fan of stereoscopic imagery) appreciate stereoscopic vision experiences, mainly through VR these days, but I've got a couple of books of stereograms back home*. Viewing them, the sense of depth becomes somehow more pronounced, more "real" than in regular vision, in way that's not at all easy to articulate. Similarly, in very deep lucid dreaming, I get an experience of generating the imagery, a sense that things are somehow impossibly detailed.

* Though more recently I also have Brian May's Mission Moon 3D, which is not well written but does have wonderful stereo pairs.

Of course in Sacks case all this comes at an enormous costs. Besides the obvious, one of the stranger elements is that seeing can be quite literally believing. Eventually losing his peripheral vision in his right eye completely, Sacks describes how this is markedly different from having an eye patch : not only is the perception itself gone, but the whole concept of vision and the visual field is lost. He describes this also in patients who became fully blind and eventually lost mental imagery, or even the memory of visual experiences, absolutely. In his own case it was more limited but still severe : if something moved into missing field of view, for him it was almost as if it didn't exist. 

Then a voice to my right - her voice - said, "What are we waiting for ?". I was dumfounded - not just that I had failed to see her to my right, but that I had even failed to imagine her being there, because "there" did not exist for me. "Out of sight, out of mind" is literally true in such a situation... I am still in a world of suddenness and discontinuity, of sudden apparitions and disappearances.

Visual thinking

There is one interesting aspect to some of the earlier cases mentioned in the books. It's possible to forget how to read but not how to write. For some people this improves by doing the physical act of tracing the letters in the air. Another learned to trace the shapes of letters with his tongue, replacing reading with a sort of tongue-based writing.

The final section of the book returns to the more general case of this, considering how vision is integrated with thought. Sacks elaborates on the case of losing the whole concept of vision, describing how in one case this was seen as a "dark paradoxical gift", since the remaining senses were so much more heightened, in their own words to the extent of being, "a whole new order, a new mode of human being". But while it can and does happen - the idea of sensory compensation is practically a stereotype - it is not necessarily typical of actual experiences. It's also possible to go in exactly the opposite direction, not to lose mental imagery, but ironically to massively enhance it.

If Sacks hints at this with his own hallucinations, with the visual cortex having nothing else to do but make stuff up, in others this can be ruthlessly controlled. His own (sighted) mother could look at a lizard skeleton and mentally rotate it, accurately drawing it from different angles from her mind alone. But one blind psychologist used descriptions and his other senses to develop his inner eye to such a degree that it served as a full replacement for his eyes - accurately enough to replace the guttering on his roof !

"... to imagine, to visualise, for example, the inside of a differential gearbox in action as if from inside its casing. I was able to watch the cogs bite, lock and revolve, distributing the spin as required. I began to play around with this internal view in connection with mechanical and technical problems, visualising how subcomponents elate in the atom, or in the living cell."

Others experience imagery in more artistic ways, or as if shown on a screen, but all of the "visual blind" seem to have a far, far more intense and vivid experience of the inner eye than most of us do. There's also a connection to proprioception, with arms and legs being directly visualised as they move them. And this can apply to other senses as well, with a deaf lip-reader describing the experience as hearing, not seeing. Much like in this famous gif, I suppose.

Sacks also touches on aphantasia, the lack of mental imagery. Our visual imaginative capacity varies considerably between individuals, but I still find it extremely hard to grasp the concept of those who don't have it at all - especially for those with highly visual professions, like surgery and even artists (is a highly developed inner ear perhaps how Beethoven managed ?). Sacks, and some of the aphantasic themselves, suggest that they must be forming mental images somewhere, they're just not consciously accessible.

But are mental images really a sort of imagery ? It appears so. Studies suggest that summoning mental images takes effort and time and uses the same area of the brain as for visual perception. The final section of the book is very similar to Livewired, which to my shame I still haven't blogged yet*. Sacks notes that, as earlier, the brain is remarkably able to adapt, re-purposing the visual cortex in the blind to handle mental imagery more precisely, freed from the need to also process incoming signals from the eyes. This may be why it hallucinates when deprived of signals in the sighted as well as explaining how the blind can develop this to such a remarkable degree.

*Stuff keeps coming up. And unfortunately the book is so good it's hard to plan out a post that wouldn't just be reprinting the whole thing and get me sued for copyright.

In similar passages to parts of Sentient, Sacks describes sensory substitution, e.g. the blind using a cane to see the world around them, transforming touch and sound into a true mental picture (Sacks puts speech marks around some of these terms but I've decided it may be worth taking this literally). Again overlapping with Livewired, he describes tactile inputs sent to the tongue from otherwise ordinary cameras, allowing a crude but effective tongue-based vision. And it really does seem to manifest itself in the brain as vision, not just as a weird sort of touch.

Finally, at the very end he wonders if the congenitally blind could also be described to have true mental imagery. Here I think I have to disagree. He suggests that language is somehow able to overcome their sensory deficiencies and enable them to truly see with the inner eyes, albeit not understanding how this works. I think a simpler explanation would be that perhaps the brain isn't entirely live-wired; it doesn't figure everything out from pure experience, but comes with some initial draft of how to handle the basics of vision : shape, colour, distance, etc., leaving only the higher meaning as that which has to be learned. But who knows ?

I've long poo-poohed the idea that reality is "nothing like" our perception of it, all the while disparaging the foolish idea of materialism. Sacks' works make it abundantly clear that our perception is far stranger than we usually realise. 

The way I think I reconcile my opposing viewpoints - and I'm quite uncertain of my own opinion here - is as follows. Perception is that which our minds construct from our senses. Our sensory apparatus directly interacts with the world, but perception is a mental act. The brain is quite capable of making up stuff all by itself, but it's also capable of constructing it directly from sensory data, albeit into wholly arbitrary forms. Most of the time those worlds are separate, and usually the rendered internal worlds are self-consistent, but the brain is more than capable of making outright mistakes and constructing impossibilities. Hence the result is what it is : a glorious but mostly functional mess.