Losing Our Minds
"If the experimenter would not be prepared to use a human infant, then his readiness to use non-human animals reveals an unjustified form of discrimination on the basis of species, since adult apes, monkeys, dogs, cats, rats and other mammals are more aware of what is happening to them, more self-directing and, so far as we can tell, at least as sensitive to pain as a human infant."
What are the ingredients of conscious mentality? Do unconscious minds matter? Why is consciousness ethically important?
In common with many philosophical treatments of the nature of mind, the neglect of a well worked out theory of perception leads DeGrazia to omit the greater part of organic life's mental furniture altogether. For the default option of classical perceptual realism doesn't just amount to a false theory of the world. It also leads to an impoverished conception of mind. By way of contrast - and in defiance of our ingrained direct realist intuitions - if the existence of the mind-independent world can only ever be inferred from our individual virtual world models, and not directly apprehended or "perceived", then the properties which one normally ascribes to classical macroscopic objects are inherently mental and covertly autobiographical. This is so even when one is "awake". The "awakened" condition is a mysterious mode of consciousness in which minds/brains are popularly supposed to attain a state of direct self-transcendence that they lack while dreamfully asleep.
More soberly, "out there" and "in here" don't mean what they seem. Natural selection has indeed harnessed fields of neurally organised forms of consciousness and worked them into simulating something else. It has modelled the fields of macroscopic, "medium-sized dry objects" whose patterns occupy quasi-classical branches of the quantum mechanical Multiverse; and whose (partly) causally covarying simulations stretch beyond the somato-sensory homunculus into the wider reaches of the neocortex as a whole. Yet each simulation is still ineradicably mental and indexical to its creator. So it's still an adaptive con-job. If classical realism is indeed false, and quantum mechanics is the sovereign theory of the whole cosmos, then mental life is presumably biologically ancient; even pre-Cambrian. This is on the assumption that there's an unreflective "perceptual" mentality which runs (at least) a long way "down" the phylogenetic tree. Most notably, in creatures with central nervous systems, this mentality takes the guise of species-specific virtual worlds. These are vivid "experiential manifolds" centred around a somato-sensory body-image.
Naturally, the billions of noisy, colourful, refractory, virtual worlds churned out by evolution through self-replicating DNA are not recognised and categorised by their host vehicles as [more-or-less] functionally organised modes of consciousness. For the most part, recognition of their mind-dependence would be functionally irrelevant to the organism. The cognitive skills which recognition entails demand a form of meta-representational capacity which would be neither energy-efficient nor cost-effective in genetic terms. For the functional role of most aspects of animal (and human) mind is to do duty for the non-mental as efficiently as possible. Natural selection, blindly as ever, spawns the machinery for generating dynamic simulations of the local environment. Most intimately, simulating this local environment involves running toy egocentric models of throwaway DNA vehicles (aka living organisms) themselves. Neurally-active genes code (in vertebrates) for egocentric "somato-sensory" simulations of the host organism's body. And if any intelligent organism is ever tempted to wonder why the whole world seems to be centred on itself, then it's worth asking where else, if anywhere, in a world-model would it be more advantageous for purely selfish genes to travesty one's significance in the great scheme of things.
The virtual world of a gazelle or a chimpanzee may not be as complex as that of a mature adult human. Yet it is still subjectively immense and vastly complex. Awake or dreaming, the greater part of each experiential manifold in which such simulations consist will often seem harshly or indifferently mind-independent. Each virtual world dwarfs the egocentric body-image at its centre. This sort of self-alienation is hugely adaptive for the host vehicle whose mind/neural network is running the simulations. It doesn't make the modelling process anything less than a genetically predisposed charade. Perhaps it will take the routine accessibility and long-term habitability of "artificial" virtual worlds - immersive, multi-modal and generated by (non-organic) VR-optimised computers - for the nature of this hard-wired perceptual realist illusion to sink in. Or perhaps, for many of us, its illusoriness never will; inferential realism is a philosophical position to be contemplated, not a form of life to be lived.
DeGrazia discusses animal minds minus their virtual worlds at some length and considerable depth. With his "coherence-based" approach to ethics, he appraises the moral status of animals, not in terms of the straightforward foundations of a pleasure-pain calculus, but in the light of the surprising richness and unsuspected diversity of mental life as defined even within the confines of an orthodox perceptual realist conception of mentality. Again, though fascinating and informative, DeGrazia's discussion is - I think - fundamentally skewed. This is because of his implicit reliance on an untenable realist theory of perception and his consequent diminished conception of the realm of consciousness. The more mental properties which get palmed off onto the rest of the world, the less there is to occupy one's mind - or non-human animal minds either.
Classical realism, however, is hopelessly at odds with what quantum mechanics, quite aside from a priori philosophical argument, tells us about the mind-transcendent world. Hilbert space and common-sense folk-physics are inconsistent; and ultimately it's common sense which has to be sacrificed if any unitary world picture is to be salvaged from the epistemological wreckage. Our quasi-hard-wired theatre of classical macroscopic objects leads most of its subjects to locate the contents of their visual fields in the spuriously accessible Outer World rather than in their world-selected organic minds. It is this mistaken theory of perception which gives rise to the intractable mind-body conundrum in the first instance.
At least we recognise there's a problem here. Philosophers in the grip of the ancient World-Knot ask: How can a cheesy grey mass of warm porridge - exquisite functional organisation notwithstanding - possibly give rise to consciousness, to this thought or that sensation? And the simple answer is that it can't. Classical brains - as apparently disclosed by the deliverances of naive realist perception augmented by hi-tech microscopy - are a mind-dependent artefact of particular sorts of QM-coherent quantum minds. Acknowledging this dependence isn't a disguised plea for Idealism or Scepticism. It's just a call for sophisticated realism-by-inference-to-the-best-explanation. [Philosopher Bryan Magee describes his thought in the school chapel that, by blinking, he effaced his entire perceptual world, as an "indescribably awful" realisation. He then succumbs to transcendental idealism. But surely it's better to treat the inferred mind-independent environment which presumably gives rise to each microcosm as a theoretical posit in good standing. It's a hypothesis with a whole lot of explanatory and predictive power.] Within the hugely more vast Multiverse, tiny mind-dependent classical worlds of "medium-sized objects" are themselves a highly adaptive facet of primordial-DNA-driven psychology. They get neurally activated in living organisms. In common with a schizophrenic's voices, a classical world as a whole is apprehended by its host's pre-frontal cortical module(s) as "out there" and not inside one's [somato-sensory-cortical] head. Unlike schizophrenic voices, however, these shifting classical constructs tend causally to co-vary pretty closely with certain gross macro-patterns in local regions of the Multiverse as a whole. This is why classical mind-worlds elsewhere - not a chimerical Classical World - have flourished.
There are natural mechanisms, not daily miracles, at work here. Classical mental macro-worlds are predisposed - but are not, strictly, genetically pre-programmed - to self-assemble if their constituent neurons get their weights and connections trained by appropriate sequences of stimuli from surface transducers. Formally, virtual mind-worlds can be described with the mathematical tools of artificial neural nets. Their behaviour mimics the implementation of powerful learning algorithms. Organic virtual worlds aren't classically programmed; their nets get "trained up" by peripheral input. And naturally they have a different ontology from that which the cognitive modules they interface with normally suppose; for we can't hop outside our models.
Such loose talk will appal anyone with a horror of "Cartesian materialist" homunculi. Are they all watching an infinite regress of mini-TV screens? There certainly is a real mystery here. But anyone who doubts the existence of little men in the head [irrespective of their theory of perception] should try a spell of lucid dreaming. They sure ain't anywhere else.
Virtual world prototypes stretch a long way back into the evolutionary past. Certainly, they extend far further than fully-fledged second-order representations, such as beliefs and desires, which function as though they were "about" the worlds on which they focus. These occurrent beliefs and desires, however, tend to serve as our foremost exemplars of mental states. It's these relatively late arrivals which act as simulated vehicles for "propositional content". Most mental life involves more mundane features than anything so exotic; although sometimes its features can be gruesome. When someone wantonly kills a mouse, for instance, the killer extinguishes an entire virtual world too, albeit a murine macro-world rather than its humanoid counterpart.
One of the reasons we have an impoverished conception of the mental life of animals, then, is that our individual egocentric visual worlds aren't construed as part of our mental life at all in everyday existence. The non-consensual and more environmentally-autonomous virtual worlds of the schizophrenic or well-frazzled acid-head are the exception, not the rule here. We readily grant that the voices etc in a schizophrenic's "external" environment are autobiographical features of the troubled individual's mental life. But the consensus-hallucinations of more typical types of virtual world enjoy no such recognised status; because much of us is constituted by these very hallucinations. If they were to let slip their true colours, the inference to their mind-dependence might interfere with their functional role in the informational economy of the organism. "Indescribably awful" realisations are unwholesome.
Perhaps the covertly mental status of what are often thought of as paradigmatically "physical" properties is most readily disclosed in lucid dreaming. Within a given dream, the huge rock-face, say, which one is climbing is the virtual rock-face of a virtual mountain. One measures kilometres; one is dealing in cubic centimeters. How the metrics of phenomenal spaces can arise embedded in crumpled-up neural minds defies our present understanding. But the mentality of putative "physical" properties is just as real if, to take a more savage example, one is all too awake on the African savannah and being mauled by a lion while a corresponding virtual lion is tearing bits off one's body-image. For if one is awake and being chased by a virtual lion [or a QM superposition of virtual lions etc], then its very probable causal covariation with [a QM superposition of] a real-world hungry predator(s) means it is highly adaptive to treat one's simulation as a mind-independent reality. Think about the difference for a second or so; and you're lunch.
A cruel but striking experimental procedure is instructive here. Vivisectors sometimes surgically abolish an animal's capacity for muscular atony. This state of effective paralysis normally stops our bodies acting out our dreams. Permanent surgical ablation of the region responsible for the functional decoupling of the bodily musculature from its neural command centers during dreams, on the other hand, ensures that the dreaming organism unwittingly enacts its inner psychodramas as it sleeps; just as, controversially, we all do quasi-veridically when awake. Thus the cat doesn't just have simple beliefs and desires about the virtual mouse it chases. Its virtual body-image image chases the virtual mouse within the vast virtual spaces of a feline dreamworld. Its beliefs have both 'narrow' mental content and narrow so-called 'perceptual' content too. Sweeping aside lots of complications, they are beliefs about first-order representations expressed under another description; though since the cat's fleeting murine representations are neither "transparent" or "projectible", post-classical AI sometimes drops the "representational" tag altogether. Common-sense distinguishes, within each experiential manifold/virtual world, between experience and the object of experience. This supposed object of experience is something non-experiential to which we fancy we've got shared direct access and with which we are mysteriously 'presented'. In fact, our solid, refractory chairs and tables, sticks and stones - and squishy grey brains - are themselves distinctive modes of experience. They can be neuronally fired up by electrodes, psychedelic drugs, dreams; or selected from the psychoneural menu while one is awake by peripheral input from the mind-independent environment. Yet that environment is only inferred as the best possible explanation of the experiential evidence.
Hence a cat does not have simple beliefs and desires about a mind-independent mouse. Instead, a crude feline mouse-simulation is taking place; together with a relatively undeveloped non-verbal system of second-order representations. First and second-order representations interact and partially interpenetrate as each simulation dynamically evolves in feline neural nets. When the cat is awake, key features of its world-simulation tend to causally covary with a vastly more complicated creature - the living mouse. Yet mystical feats of feline self-transcendence are no more feasible than human clairvoyance. It can be known from the evanescence of dreams that Nature can conjure up and destroy whole CNS immensities of virtual worlds in milliseconds. These feats of destruction and creative world-making happen whenever we are awake and blink. Animal minds are no less gappy but equally real.
So what happens to the virtual world of the dreaming cat which chases phantom mice? Does it disappear and get replaced by the real world when the cat wakes up? No. But the virtual world now gets tightly sculpted, and its shifting contents neurally selected, by peripheral input. A catworld is a quite simple toy world compared to human virtual worlds. But it is still an intensely sentient mental microcosm in its own right; and, tragically, it is a killer-world with intensely sentient victims.
Why does this matter ethically?
It wouldn't do so at all, if it weren't for the fact that virtual worlds and the extended cortical minds they embody have been "emotionally encephalised" thanks to natural selection. The limbic system insinuates its processes into the furthest reaches of cortical mind. What happens in virtual worlds inherently matters because they're shot through with limbic-driven emotional meaning and significance. The encephalisation of emotion has extended not just to the cortical regions playing host to second-order representations typified by the occurrent belief-episodes of folk psychology. Our limbic processes, most notably those of the monoaminergic neurons, also infiltrate each egocentric virtual world and its vast cortical arrays too - in man and mouse alike. This infiltration accounts for the circumstance that neither we nor other animals merely "project" our feelings and values onto the [virtual] world. For many of each world's most striking features don't just seem inherently terrifying, delightful, beautiful, desirable, nasty, etc. They are inherently terrifying, delightful, beautiful, desirable nasty, etc. Thanks to the outgrowths of of our limbic emotional powerhouse, that girl [warthog/hippopotamus,gazelle etc], for instance, does not just seem sexy. She possesses the inherent property of sexiness as part of her very essence [or, more precisely, as part of her fleeting psychochemical excitation]. This identification is possible only because virtual worlds are strictly mental: in the realm of phenomenology the difference between seeming and reality dissolves. Just so long as the relevant causal covariation with the mind-independent world is retained, the emotional saturation of a [virtual] world tends also to be highly adaptive. Our genes have outrageously biased what matters to their neural creations - us - so as to differentially further their own reproductive prospects. None of this would be possible if classical perceptual realism were true; but then its intellectual sell-by date has now passed, even though the hard-wired illusion remains.
The philosophical and scientific story of mind-making is much more complicated than the simplistic outline offered here. Yet it's abundantly clear that natural selection has ensured that many organisms have horrible minds, live in horrible virtual worlds, and suffer horrible deaths. Is this an immutable law of Nature? Probably not.