Three Pound Brain

No bells, just whistling in the dark…

Month: October, 2015

Graziano, the Attention Schema Theory, and the Neuroscientific Explananda Problem

by rsbakker

Along with Taylor Webb, Michael Graziano has published an updated version of what used to be his Attention Schema Theory of Consciousness, but is now called the Attention Schema Theory of Subjective Awareness. For me, it epitomizes the kinds of theoretical difficulties neuroscientists face in their attempts to define their explananda, why, as Gary Marcus and Jeremy Freeman note in their Preface to The Future of the Brain, “[a]t present, neuroscience is a collection of facts, still awaiting an overarching theory” (xi).

On Blind Brain Theory, the ‘neuroscientific explananda problem’ is at least twofold. For one, the behavioural nature of cognitive functions raises a panoply of interpretative issues. Ask any sociologist: finding consensus commanding descriptions of human behaviour is far more difficult than finding consensus commanding descriptions of, say, organ behaviour. For another, the low-dimensional nature of  conscious experience raises a myriad of interpretative conundrums in addition to the problems of interpretative underdetermination facing behaviour.  Ask any psychologist: finding consensus commanding descriptions of conscious phenomena has hitherto proven impossible. As William Uttal notes in The New Phrenology: “There is probably nothing that divides psychologists of all stripes, more than the inadequacies and ambiguities of our efforts to define mind, consciousness, and the enormous variety of mental events and phenomena” (90). At least with behaviour, publicity allows us to anchor our theoretical interpretations in revisable data; experience, however, famously affords us no such luxury. So where the problem of behavioural underdetermination seems potentially soluble given enough elbow grease (one can imagine continued research honing canonical categorizations of behavioural functions as more and more information is accumulated), the problem of experiential underdetermination out and out baffles. We scarce know where to begin. Some see conscious experience as a natural phenomena possessing properties that do not square with our present scientific understanding of nature. Others, like myself, see conscious experience as a natural phenomena that only seems to possess properties that do not square with our nature. Michael Graziano belongs to this camp also. The great virtue of belonging to this deflationary pole of the experiential explananda debate is that it spares you the task of explaining inexplicable entities, or the indignity of finding rhetorical ways to transform manifest theoretical vices (like analytic opacity) into virtues (like ‘irreducibility’). In other words, it lets you drastically simplify the explanatory landscape. Despite this, Graziano’s latest presentation of his theory of consciousness (coauthored with Taylor Webb), “The attention schema theory: a mechanistic account of subjective awareness,” seems to be deeply–perhaps even fatally–mired in the neuroscientific explananda problem.

Very little in Webb and Graziano’s introduction to AST indicates the degree to which the theory has changed since the 2013 publication of Consciousness and The Social Brain. The core insight of Attention Schema Theory is presented in the same terms, the notion that subjective awareness, far from being a property perceived, is actually a neural construct, a tool the human brain uses to understand and manipulate both other brains and itself.  They write:

This view that the problem of subjective experience consists only in explaining why and how the brain concludes that it contains an apparently non-physical property, has been proposed before (Dennett, 1991). The attention schema theory goes beyond this idea in providing a specific functional use for the brain to compute that type of information. The heart of the attention schema theory is that there is an adaptive value for a brain to build the construct of awareness: it serves as a model of attention. 2

They provide the example of visual attention upon an apple, how the brain requires, as a means to conclude it was ‘subjectively aware’ of the apple, information regarding itself and its means of relating to the apple. This ‘means of relating’ happens to be the machinery of attention, resulting in the attention schema, a low-dimensional representation of the high-dimensional complexities comprising things like visual attention upon an apple. And this, Graziano maintains, is what ‘subjective awareness’ ultimately amounts to: “the brain’s internal model of the process of attention” (1).

And this is where the confusion begins, as much for Webb and Graziano as for myself. For one, ‘consciousness’ has vanished from the title of the theory, replaced by the equally overdetermined ‘subjective awareness.’ For another, the bald claims that consciousness is simply a delusion have all but vanished. As recently as last year, Graziano wrote:

How does the brain go beyond processing information to become subjectively aware of information? The answer is: It doesn’t. The brain has arrived at a conclusion that is not correct. When we introspect and seem to find that ghostly thing — awareness, consciousness, the way green looks or pain feels — our cognitive machinery is accessing internal models and those models are providing information that is wrong. The machinery is computing an elaborate story about a magical-seeming property. And there is no way for the brain to determine through introspection that the story is wrong, because introspection always accesses the same incorrect information. “Are We Really Conscious,” The New York Times Sunday Review.

Here there simply is no such thing as subjective awareness: it’s a kind of cognitive illusion foisted on the brain by the low-dimensionality of the attention schema. Now, however, the status of subjective awareness is far less clear. Webb and Graziano provide the same blind brain explanation (down to the metaphors, no less) for the peculiar properties apparently characterizing subjective awareness: since the brain has no use for high-dimensional information, the “model would be more like a cartoon sketch that depicts the most important, and useful aspects of attention, without representing any of the mechanistic details that make attention actually happen” (2). As a result of this opportunistic simplification, it makes sense that a brain:

“would conclude that it possesses a phenomenon with all of the most salient aspects of attention – the ability to take mental possession of an object, focus one’s resources on it, and, ultimately, act on it – but without any of the mechanisms that make this process physically possible. It would conclude that it possesses a magical, non-physical essence, but one which can nevertheless act and exert causal control over behavior, a mysterious conclusion indeed.” 2

This is a passage that would strike any long time followers of TPB as a canonical expression of Blind Brain Theory, but there are some key distinctions dividing the two pictures, which I’ll turn to in a moment. For the nonce, it’s worth noting that it’s not so much subjective awareness (consciousness) that now stands charged with deception, as the kinds of impossible properties that attributed to it. Given that subjective awareness is the explicit explanandum, there’s a pretty important ambiguity here between subjective awareness as attention schema and subjective awareness as impossible construct. Even though the latter is clearly a cognitive illusion, the former is real insofar as the attention schema is real.

For its part, Blind Brain Theory is a theory, not of consciousness, but of the appearance of consciousness. It provides a principled way to detect, diagnose and even circumvent the kinds of cognitive illusions the limits of deliberative metacognition inflict upon reflection. It only explains why, given the kind of metacognitive resources our brains actually possess, the problem of consciousness constitutes a ‘crash space,’ a domain where we continually run afoul the heuristic limitations of our tools. So when I reflect upon my sensorium, for instance, even though I am unencumbered by supernatural characterizations of phenomenology—subjective awareness—something very mysterious remains to be explained, it’s just nowhere near so mysterious as someone like, Chalmers, for instance, is inclined to think.

Graziano, on the other hand, thinks he possesses a bona fide theory of consciousness. The attention schema, on his account, is awareness. So when he reflects upon his sensorium, he’s convinced he’s reflecting upon his ‘attention schema,’ that this is the root of what consciousness consists in—somehow.

I say ‘somehow,’ because in no way is it clear why the attention schema, out of all the innumerable schematisms the brain uses to overcome the ‘curse of dimensionality,’ should be the one possessing (the propensity to be duped by?) subjective awareness. In other words, AST basically suffers the same problem all neural identity theories suffer: explaining what makes one set of neural mechanisms ‘aware’ while others remain ‘dark.’ Our brains run afoul their cognitive limitations all the time, turn on countless heuristic schema: why is the attention schema prone to elicit sensoriums and the like?

Note that he has no way of answering, ‘Because that’s how attention is modelled,’ without begging the question. We want to know what makes modelling attention so special as to result in what, mistaken or not, we seem to be enjoying this very moment now. Even though he bills Attention Schema Theory as a ‘mechanistic account of subjective awareness,’ there’s a real sense in which consciousness, or ‘subjective awareness,’ is left entirely unexplained. Why should a neurobiologically instantiated schema of the mechanisms of attention result in this mad hall of mirrors we are sharing (or not) this very moment?

Graziano and Webb have no more clue than anyone. AST provides a limited way to understand the peculiarities of experience, but it really has no way whatsoever of explaining the fact of experience.

He had no such problem with the earlier versions of AST simply because he could write off consciousness as an illusion entirely, as a ‘squirrel in the head.’ Once he had dispatched with the peculiarities of experience, he could slap his pants and go home. But of course, this stranded him with the absurd position of denying the existence of conscious experience altogether.

Now he acknowledges that consciousness exists, going so far as to suggest that AST is consistent with and extends beyond global workspace and information integration accounts.

“The attention schema theory is consistent with these previous proposals, but also goes beyond them. In the attention schema theory, awareness does not arise just because the brain integrates information or settles into a network state, anymore than the perceptual model of color arises just because information in the visual system becomes integrated or settles into a state. Specific information about color must be constructed by the visual system and integrated with other visual information. Just so, in the case of awareness, the construct of awareness must be computed. Then it can be integrated with other information. Then the brain has sufficient information to conclude and report not only, “thing X is red,” or, “thing X is round,” but also, “I am aware of thing X.” 3

If this is the case, then subjective awareness has to be far more than the mere product of neural fiat, a verbal reporting system uttering the terms, “I am aware of X.” And it also has to be far more than simply paying attention to the model of attention. If AST extends beyond global workspace and information integration accounts, then the phenomenon of consciousness exceeds the explanatory scope of AST. Before subjective awareness was a metacognitive figment, the judgment, “I am aware of thing X” exhausted the phenomenology of experiencing X. Now subjective awareness is a matter of integrating the ‘construct of awareness’ (the attention schema) with ‘other information’ to produce the brain’s conclusion of phenomenology.

At the very least, the explanatory target of AST needs to be clarified. Just what is the explanandum of the Attention Schema Theory? And more importantly, how does the account amount to anything more than certain correlations between a vague model and the vague phenomena(lity) it purports to explain?

I actually think it’s quite clear that Graziano has conflated what are ultimately two incompatible insights into the nature of consciousness. The one is simply that consciousness and attention are intimately linked, and the other is that metacognition is necessarily heuristic. Given this conflation, he has confused the explanatory power of the latter as warrant for reducing subjective awareness to the attention schema. The explanatory power of the latter, of course, is simply the explanatory power of Blind Brain Theory, the way heuristic neglect allows us to understand a wide number of impossible properties typically attributed to intentional phenomena. Unlike the original formulation of AST, Blind Brain Theory has always been consilient with global workspace and information integration accounts simply because heuristic neglect says nothing about what consciousness consists in, only the kinds of straits the limits of the human brain impose upon the human brain’s capacity to cognize its own functions. It says a great deal about why we find ourselves still, after thousands of years of reflection and debate, completely stumped by our own nature. It depends on the integrative function of consciousness to be able to explain the kinds of ‘identity effects’ it uses to diagnose various metacognitive illusions, but beyond this, BBT remains agnostic on the nature of consciousness (even as it makes hash of the consciousness we like to think we have).

But even though BBT is consilient with global workspace and information integration accounts the same as AST, it is not consilient with AST. Unpacking the reasons for this incompatibility makes the nature of the conflation underwriting AST quite clear.

Graziano takes a great number of things for granted in his account, not the least of which is metacognition. Theory is all about taking things for granted, of course, but only the right things. AST, as it turns out, is not only a theory of subjective awareness, it’s also a theory of metacognition. Subjective awareness, on Graziano’s account, is a metacognitive tool. The primary function of the attention schema is to enable executive control of attentional mechanisms. As they write, “[i]n this perspective, awareness is an internal model of attention useful for the control of attention” (5). Consciousness is a metacognitive device, a heuristic the brain uses to direct and allocate attentional (cognitive) resources.

We know that it’s heuristic because, even though Webb and Graziano nowhere reference the research of fast and frugal heuristics, they cover the characteristics essential to them. The attention schema, we are told, provides only the information the brain requires to manage attention and nothing more. In other words, the attention schema possesses what Gerd Gigerenzer and his fellow researchers at the Adaptive Behaviour and Cognition Research Institute call a particular ‘problem ecology,’ one that determines what information gets neglected and what information gets used (see, Ecological Rationality). This heuristic neglect in turn explains why, on Webb and Graziano’s account, subjective awareness seems to possess the peculiar properties it does. When we attend to our attention, the neglect of natural (neurobiological) information cues the intuition that something not natural is going on. Heuristic misapplications, as Wimsatt has long argued, lead to systematic errors.

But of course the feasibility of solving any problem turns on the combination of the information available and the cognitive capacity possessed. Social cognition, for instance, allows us to predict, explain, and manipulate our fellows on the basis of so little information that ‘computational intractibility’ remains a cornerstone of mindreading debates.  In other words, the absence of neurobiological information in the attention schema only explains the apparently supernatural status of subjective awareness given certain metacognitive capacities. Graziano’s attention schema may be a metacognitive tool, a way to manage cognitive resources, but it is the ‘object’ of metacognition as well.

For me, this is where the whole theory simply falls apart—and obviously so. The problem is that the more cognitive neuroscience learns about metacognition, the more fractionate and specialized it appears to be. Each of these ‘kluges’ represents adaptations to certain high impact, environmental problems. The information subjective awareness provides leverages many different solutions to many different kinds of dilemmas, allowing us to bite our tongues at Thanksgiving dinner, ponder our feelings toward so-and-so, recognize our mistakes, compulsively ruminate upon relationships, and so on, while at the same time systematically confounding our attempts to deduce the nature of our souls. The fact is, the information selected, stabilized, and broadcast via consciousness, enables far, far more than simply the ability to manage attention.

But if subjective awareness provides solutions to a myriad of problems given the haphazard collection of metacognitive capacities we possess, then in what sense does it count as a ‘representation of’ the brain’s attentional processes? Is it the case that a heuristic (evolutionarily opportunistic) model of that machinery holds the solution to all problems? Prima facie, at least, the prospects of such a hypothesis seem dim. When trying to gauge our feelings about a romantic partner, is it a ‘representation’ of our brain’s attentional processes that we need, or is it metacognitive access to our affects?

Perhaps sensing this easy exit, Webb and Graziano raise some hasty barricades:

“According to the attention schema theory, the brain constructs a simplified model of the complex process of attention. If the theory is correct, then the attention schema, the construct of awareness, is relevant to any type of information to which the brain can pay attention. The relevant domain covers all vision, audition, touch, indeed any sense, as well as internal thoughts, emotions, and ideas. The brain can allocate attention to all of these types of information. Therefore awareness, the internal representation of attention, should apply to the same range of information.” 9

So even though a representation of the brain’s attentional resources is not what we need when we inspect our feelings regarding another, it remains ‘applicable’ to such an inspection. If we accept that awareness of our feelings is required to inspect our feelings, does this mean that awareness somehow arises on the basis of the ‘applicability’ of the attention schema, or does it mean that the attention schema somehow mediates all such metacognitive activities?

Awareness of our feelings is required to inspect our feelings. This means the attention schema underwrites our ability to inspect our feelings, as should come as no surprise, given that the attention schema underwrites all conscious metacognition. But if the attention schema underwrites all conscious metacognition, it also underwrites all conscious metacognitive functions. And if the attention schema underwrites all conscious metacognitive functions, then, certainly, it models far, far more than mere attention.

The dissociation between subjective awareness and the attention schema seems pretty clear. Consciousness is bigger than attention, and heuristic neglect applies to far more than our attempts to understand the ‘attention schema’—granted there is such a thing.

But what about the post facto ‘predictions’ that Webb and Graziano present as evidence for AST?

Given that consciousness is the attention schema and the primary function of the attention schema is the control of attention, we should expect divergences between attention and awareness, and we should expect convergences between awareness and attentional control. Webb and Graziano adduce experimental evidence of both, subsequently arguing that AST is the best explanation, even though the sheer generality of the theory makes it hard to see the explanatory gain. As it turns out, awareness correlates with attentional control because awareness is an attentional control mechanism, and awareness uncouples with attention because awareness, as a representation of attention, is something different than attention. If you ask me, this kind of ’empirical evidence’ only serves to underscore the problems with the account more generally.

Ultimately, I just really don’t see how AST amounts to a workable theory of consciousness. It could be applied, perhaps, as a workable theory for the appearance of consciousness, but then only as a local application of the far more comprehensive picture of heuristic neglect Blind Brain Theory provides. These limits become especially clear when one considers the social dimensions of AST, where Graziano sees it discharging some of the functions Dennett attributes to the ‘intentional stance.’ But since AST possesses no account of intentionality whatsoever (indeed, Graziano doesn’t seem to be aware of the problems posed by aboutness or content), it completely neglects the intentional dimensions of social cognition. Since social cognition is intentional cognition, it’s hard to understand how AST does much more than substitute a conceptually naïve notion of ‘attention’ for intentionality more broadly construed.

Goosing the Rumour Mill

by rsbakker

Just got back to find there’s been some developments! I’d resolved to say nothing anticipating anything–it just makes me feel foolish anymore. Until we have all the details hammered out, there’s not much I can say except that Overlook’s July 2016 date is tentative. I fear I can’t comment on their press release, either. Things seem to be close, though.

I know it’s been a preposterously long haul, folks, but hold on just a bit longer. The laws of physics are bound to kick in at some point, after which I can start delivering some more reliable predictions.

The Mental as Rule of Thumb

by rsbakker

What are mental functions? According to Blind Brain Theory, they are quasimechanical posits explaining the transformations between regimented inputs and observed outputs in ways that seem to admit generalization. We know the evidence is correlative, but we utilize mechanical cognition nonetheless, producing a form of correlatively anchored ‘quasi-causal explanation.’ There often seems to be some gain in understanding, and thus are ‘mental functions’ born.

Mental functions famously don’t map across our growing understanding of neural mechanisms because the systematicity tracked is correlative, rather than causal. Far from ‘mechanism sketches,’ mental functions are ‘black box conceits,’ low dimensional constructs that need only solve some experimental ecology (that may or may not generalize). The explanatory apparatus of the ‘mental’ indirectly tracks the kinds of practical demands made on human cognition as much as the hidden systematicities of the brain. It possesses no high-dimensional reality—real reality—otherwise. How could it? What sense does it make to suppose that our understanding of the mental, despite being correlatively anchored, nevertheless tracks something causal within subjects? Very little. Correlations abound, to the point of obscuring causes outright. Though correlative cognition turns on actual differential relations to the actual mechanisms involved, it nevertheless neglects those relations, and therefore neglects the mechanisms as well.  To suggest that correlative posits possess some kind of inexplicable intrinsic efficacy is to simply not understand the nature of correlative cognition, which is to make due in the absence of behavioural sensitivities to the high-dimensional mechanics of our environments.

Why bother arguing for something spooky when ‘mental functions’ are so obviously heuristic conceits, ways to understand otherwise opaque systems, nothing more or less?

Of course there’s nothing wrong with heuristics, so long as they’re recognized as such, ways for other brains to cognize neural capacities short of cognizing neural mechanisms. To the extent that experimental findings generalize to real world contexts, there’s a great deal to be learned from ‘black box psychology.’ But we should not expect to find any systematic, coherent account of ‘mind’ or the ‘mental,’ simply because the correlative possibilities are potentially limitless. So long as new experimental paradigms can be improvised, new capacities/incapacities can be isolated. Each ‘discovery,’ in other words, is at once an artifact, an understanding specific (as all correlative understandings are) to some practical ecology, one which is useful to the degree it can be applied in various other practical ecologies.

And there you have it: a concise eliminativist explanation of why mental functions seem to have no extension and yet seem to provide a great deal of knowledge anyway. ‘Mental functions’ are essentially a way to utilize our mechanical problem-solving capacity in black box ecologies. The time has come to start calling them for what they are: heuristic conceits.  The ‘mind’ is a way to manage a causal system absent any behavioural sensitivity to the mechanics of that system, a way to avoid causal cognition. To suggest that it is somehow fundamentally causal nonetheless is to simply misunderstand it, to confuse, albeit in an exotic manner, correlation for causation.



The Real Problem with ‘Correlation’

by rsbakker

stick zombies

Since presuming that intentional cognition can get behind intentional cognition belongs to the correlation problem, any attempt to understand the problem requires we eschew theoretical applications of intentional idioms. Getting a clear view, in other words, requires that we ‘zombify’ human cognition, adopt a thoroughly mechanical vantage that simply ignores intentionality and intentional properties. As it so happens, this is the view that commands whatever consensus one can find regarding these issues. Though the story I’ll tell is a complicated one, it should also be a noncontroversial one, at least insofar as it appeals to nothing more than naturalistic platitudes.

I first started giving these ‘zombie interpretations’ of different issues in philosophy and cognitive science a few years back.[1] Everyone in cognitive science agrees that consciousness and cognition turn on the physical somehow. This means that purely mechanical descriptions of the activities typically communicated via intentional idioms have to be relevant somehow (so long as they are accurate, at least). The idea behind ‘zombie interpretation’ is to explain as much as possible using only the mechanistic assumptions of the biological sciences—to see how far generalizing over physical processes can take our perennial attempt to understand meaning.

Zombies are ultimately only a conceit here, a way for the reader to keep the ‘explanatory gap’ clearly in view. In the institutional literature, ‘p-zombies’ are used for a variety of purposes, most famously to anchor arguments against physicalism. If a complete physical description of the world need not include consciousness, then the brute fact of consciousness implies that physicalism is incomplete. However, since this argument itself turns on the correlation problem, it will not concern us here. The point, oddly enough, is to adhere to an explanatory domain where we all pretty much agree, to speculate using only facts and assumptions belonging to the biological sciences—the idea being, of course, that these facts and assumptions are ultimately all that’s required. Zombies allow us to do that.

Philosophy Now zombie pic

So then, devoid of intentionality, zombies lurch through life possessing only contingent, physical comportments to their environment. Far from warehousing ‘representations’ possessing inexplicable intentional properties, their brains are filled with systems that dynamically interact with their world, devices designed to isolate select signals from environmental noise. Zombies do not so much ‘represent their world’ as possess statistically reliable behavioural sensitivities to their environments.

So where ‘subjects’ possess famously inexplicable semantic relations to the world, zombies possess only contingent, empirically tractable relations to the world. Thanks to evolution and learning, they just happen to be constituted such that, when placed in certain environments, gene conserving behaviours tend to reliably happen. Where subjects are thought to be ‘agents,’ perennially upstream sources of efficacy, zombies are components, subsystems at once upstream and downstream the superordinate machinery of nature. They are astounding subsystems to be sure, but they are subsystems all the same, just more nature—machinery.

What makes them astounding lies in the way their neurobiological complexity leverages behaviour out of sensitivity. Zombies do not possess distributed bits imbued with the occult property of aboutness; they do not model or represent their worlds in any intentional sense. Rather, their constitution lets ongoing environmental contact tune their relationship to subsequent environments, gradually accumulating the covariant complexities required to drive effective zombie behaviour. Nothing more is required. Rather than possessing ‘action enabling knowledge,’ zombies possess behaviour enabling information, where ‘information’ is understood in the bald sense of systematic differences making systematic differences.

A ‘cognitive comportment,’ as I’ll use it here, refers to any complex of neural sensitivities subserving instances of zombie behaviour. It comes in at least two distinct flavours: causal comportments, where neurobiology is tuned to what generally makes what happen, and correlative comportments, where zombie neurobiology is tuned to what generally accompanies what happens. Both systems allow our zombies to predict and systematically engage their environments, but they differ in a number of crucial respects. To understand these differences we need some way of understanding what positions zombies upstream their environments–or what leverages happy zombie outcomes.

The zombie brain, much like the human brain, confronts a dilemma. Since all perceptual information consists of sensitivity to selective effects (photons striking the eye, vibrations the ear, etc.), the brain needs some way of isolating the relevant causes of those effects (a rushing tiger, say) to generate the appropriate behavioural response (trip your mother-in-law, then run). The problem, however, is that these effects are ambiguous: a great many causes could be responsible. The brain is confronted with a version of the inverse problem, what I will call the medial inverse problem for reasons that will soon be clear. Since it has nothing to go on but more effects, which are themselves ambiguous, how could it hope to isolate the causes it needs to survive?

By allowing sensitivities to discrepancies between the patterns initially cued and subsequent sensory effects to select—and ultimately shape—the patterns subsequently cued. As it turns out, zombie brains are Bayesian brains.[2] Allowing discrepancies to both drive and sculpt the pattern-matching process automatically optimizes the process, allowing the system to bootstrap wide-ranging behavioural sensitivities to environments in turn. In the intentionality laden idiom of theoretical neuroscience, the brain is a ‘prediction error minimization’ machine, continually testing occurrent signals against ‘guesses’ (priors) triggered by earlier signals. Success (discrepancy minimization) quite automatically begets success, allowing the system to continually improve its capacity to make predictions—and here’s the important thing—using only sensory signals.[3]

But isolating the entities/behaviour causing sensory effects is one thing; isolating the entities/behaviour causing those entities/behaviour is quite another. And it’s here that the chasm between causal cognition and correlative cognition yawns wide. Once our brain’s discrepancy minimization processes isolate the relevant entities/behaviours—solve the medial inverse problem—the problem of prediction simply arises anew. It’s not enough to recognize avalanches as avalanches or tigers as tigers, we have to figure out what they will do. The brain, in effect, faces a second species of inverse problem, what might be called the lateral inverse problem. And once again, it’s forced to rely on sensitivities to patterns (to trigger predictions to test against subsequent signals, and so on).[4]

Nature, of course, abounds with patterns. So the problem is one of tuning a Bayesian subsystem like the zombie brain to the patterns (such as ‘avalanche behaviour’ or ‘tiger behaviour’) it needs to engage its environments given only sensory effects. The zombie brain, in other words, needs to wring behavioural sensitivities to distal processes out of a sensitivity to proximal effects. Though they are adept at comporting themselves to what causes their sensory effects (to solving the medial inverse problem), our zombies are almost entirely insensitive to the causes behind those causes. The etiological ambiguity behind the medial inverse problem pales in comparison to the etiological ambiguity comprising the lateral inverse problem, simply because sensory effects are directly correlated to the former, and only indirectly correlated to the latter. Given the limitations of zombie cognition, in other words, zombie environments are ‘black box’ environments, effectively impenetrable to causal cognition.

Part of the problem is that zombies lack any ready means of distinguishing causality from correlation on the basis of sensory information alone. Not only are sensory effects ambiguous between causes, they are ambiguous between causes and correlations as well. Cause cannot be directly perceived. A broader, engineered signal and greater resources are required to cognize its machinations with any reliability—only zombie science can furnish zombies with ‘white box’ environments. Fortunately for their prescientific ancestors, evolution only required that zombies solve the lateral inverse problem so far. Mere correlations, despite burying the underlying signal, remain systematically linked to that signal, allowing for a quite different way of minimizing discrepancies.

Zombies, once again, are subsystems whose downstream ‘componency’ consists in sensitivities to select information. The amount of environmental signal that can be filtered from that information depends on the capacity of the brain. Now any kind of differential sensitivity to an environment serves organisms in good stead. To advert to the famous example, frogs don’t need the merest comportment to fly mechanics to catch flies. All they require is a select comportment to select information reliably related to flies and fly behaviour, not to what constitutes flies and fly behaviour. And if a frog did need as much, then it would have evolved to eat something other than flies. Simple, systematic relationships are not only all that is required to solve a great number of biological problems, they are very often the only way those problems can be solved, given evolutionary exigencies. This is especially the case with complicated systems such as those comprising life.

So zombies, for instance, have no way of causally cognizing other zombies. They likewise have no way of causally cognizing themselves, at least absent the broader signal and greater computational resources provided by zombie science. As a result, they possess at best correlative comportments both to each other and to themselves.

Idoits guide to zombies

So what does this mean? What does it mean to solve systems on basis of inexpensive correlative comportments as opposed to far more expensive causal comportments? And more specifically, what does it mean to be limited to extreme versions of such comportments when it comes to zombie social cognition and metacognition?

In answer to the first question, at least three, interrelated differences can be isolated:

Unlike causal (white box) comportments, correlative (black box) comportments are idiosyncratic. As we saw above, any number of behaviourally relevant patterns can be extracted from sensory signals. How a particular problem is solved depends on evolutionary and learning contingencies. Causal comportments, on the other hand, involve behavioural sensitivity to the driving environmental mechanics. They turn on sensitivities to upstream systems that are quite independent of the signal and its idiosyncrasies.

Unlike causal (white box) comportments, correlative (black box) comportments are parasitic, or differentially mediated. To say that correlative comportments are ‘parasitic’ is to say they depend upon occluded differential relations between the patterns extracted from sensory effects and the environmental mechanics they ultimately solve. Frogs, once again, need only a systematic sensory relation to fly behaviour, not fly mechanics, which they can neglect, even though fly mechanics drives fly behaviour. A ‘black box solution’ serves. The patterns available in the sensory effects of fly behaviour are sufficient for fly catching given the cognitive resources possessed by frogs. Correlative comportments amount to the use of ‘surface features’—sensory effects—to anticipate outcomes driven by otherwise hidden mechanisms. Causal comportments, which consist of behavioural sensitivities (also derived from sensory effects) to the actual mechanics involved, are not parasitic in this sense.

Unlike causal (white box) comportments, correlative (black box) comportments are ecological, or problem relative. Both causal comportments and correlative comportments are ‘ecological’ insofar as both generate solutions on the basis of finite information and computational capacity. But where causal comportments solve the lateral inverse problem via genuine behavioural sensitivities to the mechanics of their environments, correlative comportments (such as that belonging to our frog) solve it via behavioural sensitivities to patterns differentially related to the mechanics of their environments. Correlative comportments, as we have seen, are idiosyncratically parasitic upon the mechanics of their environments. The space of possible solutions belonging to any correlative comportment is therefore relative to the particular patterns seized upon, and their differential relationships to the actual mechanics responsible. Different patterns possessing different systematic relationships will possess different ‘problem ecologies,’ which is to say, different domains of efficacy. Since correlative comportments are themselves causal, however, causal comportments apply to all correlative domains. Thus the manifest ‘objectivity’ of causal cognition relative to the ‘subjectivity’ of correlative cognition. 

So far, so good. Correlative comportments are idiosyncratic, parasitic, and ecological in a way that causal comportments are not. In each case, what distinguishes causal comportments is an actual behavioural sensitivity to the actual mechanics of the system. Zombies are immersed in potential signals, awash in causal differences, information, that could make a reproductive difference. The difficulties attendant upon the medial and lateral inverse problems, the problems of what and what-next, render the extraction of causal signals enormously difficult, even when the systems involved are simple. The systematic nature of their environments, however, allow them to use behavioural sensitivities as ‘cues,’ signals differentially related to various systems, to behaviourally interact with those systems despite the lack of any behavioural sensitivity to their particulars. So in research on contingencies, for instance, the dependency of ‘contingency inferences’ on ‘sampling,’ the kinds of stimulus input available, has long been known, as have the kinds of biases and fallacies that result. Only recently, however, have researchers realized the difficulty of accurately making such inferences given the kinds of information available in vivo, and the degree to which we out and out depend on so-called ‘pseudocontingency heuristics’ [5]. Likewise, research into ‘spontaneous explanation’ and  ‘essentialism,’ the default attribution of intrinsic traits and capacities in everyday explanation, clearly suggests that low-dimensional opportunism is the rule when it comes to human cognition.[6] The more we learn about human cognition, in other words, the more obvious the above story becomes.

So then what is the real problem with correlation? The difficulty turns on the fact that black box cognition, solving systems via correlative cues, can itself only be cognized in black box terms.

Given their complexity, zombies are black boxes to themselves as much to others. And this is what has cued so much pain behaviour in so many zombie philosophers. As a black box, zombies cannot cognize themselves as black boxes: the correlative nature of their correlative comportments utterly escapes them (short, once again, the information provided by zombie science). Zombie metacognition is blind to the structure and dynamics of zombie metacognition, and thus prone to what might be called ‘white box illusions.’ Absent behavioural sensitivity to the especially constrained nature of their correlative comportments to themselves, insufficient data is processed in the same manner as sufficient data, thus delivering the system to ‘crash space,’ domains rendered intractable by the systematic misapplication of tools adapted to different problem ecologies. Unable to place themselves downstream their incapacity, they behave as though no such incapacity exists, suffering what amounts to a form of zombie anosognosia.

Perhaps this difficulty shouldn’t be considered all that surprising: after all, the story told here is a white box story, a causal one, and therefore one requiring extraction from the ambiguities of effects and correlations. The absence of this information effectively ‘black-boxes’ the black box nature of correlative cognition. Zombies cued to solve for that efficacy accordingly run afoul the problem of processing woefully scant data as sufficient, black boxes as white boxes, thus precluding the development of effective, behavioural sensitivities to the actual processes involved.  The real Problem of Correlation, in other words, is that correlative modes systematically confound cognition of correlative comportments. Questions regarding the nature of our correlative comportments simply do not lie within the problem space of our correlative comportments—and how could they, when they’re designed to solve absent sensitivity to what’s actually going on?

And this is why zombies not only have philosophers, they have a history of philosophy as well. White box illusions have proven especially persistent, despite the spectacular absence of systematic one-to-one correspondences between the apparent white box that zombies are disposed to report as ‘mind’ and the biological white box emerging out of zombie science. Short any genuine behavioural sensitivity to the causal structure of their correlative comportments, zombies can at most generate faux-solutions, reports anchored to the systematic nature of their conundrum, and nothing more. Like automatons, they endlessly report low-dimensional, black box posits the way they report high-dimensional environmental features—and here’s the thing—using the very same terms that humans use. Zombies constantly utter terms like ‘minds,’ ‘experiences,’ ‘norms,’ and so on. Zombies, you could say, possess a profound disposition to identify themselves and each other as humans.

Just like us.

MJ zombie



[1] See, Davidson’s Fork: An Eliminativist Radicalization of Radical Interpretation, The Blind Mechanic, The Blind Mechanic II: Reza Negarestani and the Labour of Ghosts, Zombie Interpretation: Eliminating Kriegel’s Asymmetry Argument, and Zombie Mary versus Zombie God and Jesus: Against Lawrence Bonjour’s “Against Materialism”

[2] For an overview of Bayesian approaches, see Andy Clark, “Whatever next? Predictive brains, situated agents, and the future of cognitive science.”

[3]  The following presumes an ecological (as opposed to an inferential) understanding of the Bayesian brain. See Nico Orlandi, “Bayesian perception is ecological perception.”

[4] Absent identification there is no possibility of prediction. The analogy between this distinction and the ancient distinction between being and becoming (or even the modern one between the transcendental and the empirical) is interesting to say the least.

[5] See Klaus Fiedler et al, “Pseudocontingencies: Logically Unwarranted but Smart Inferences.”

[6] See Andrei Cimpian, “The Inherence Heuristic: Generating Everyday Explanations,” or Cimpian and Salomon, “The inherence heuristic: An intuitive means of making sense of the world, and a potential precursor to psychological essentialism.”