Dear Paul,

The problem of 'bringing perceptual parts together' is discussed in a wide variety of places from Descartes to contemporary neuroscience. The question is where to start and what aspect to home in on. The key source texts that immediately come to mind might be:

Descartes in Meditations and Leibniz in Monadology make it clear that a perceiving subject must have a single indivisible relation to the totality of what it perceives so that relation cannot be of a familiar mechanical sort. So it would seem that there cannot be any parts - just complexity, and Leibniz spends a lot of time justifying this.

William James worries about the 'combination' of what might be called microexperiences as parts of a total experience on a panpsychist background in Principles of Psychology. He concludes that there is no plausible basis for this.

The Gestalt movement also focused on the totality and indivisibility of experiences but I am not familiar with the literature.

In more recent times a number of neuroscientists have suggested that the elements or parts of a 'scene', such as the features of various objects, might be brought together as a whole through synchronisation of cellular firing. This has become a confused area because it is almost certainly true that synchronisation is used by the brain to 'triage' features to different objects during processing. However, although the champions of synchronisation often imply that this is the same as 'bringing perceptual parts together into a totality' it is not and the same mechanism cannot achieve both. Almost certainly synchronisation cannot itself achieve perceptual 'binding' for the simple reason that the synchronised signals are in different places.

A whole group of people have also tried to 'bring perceptual parts together' using the concept of unified fields within the brain. A number of the theories that venture into quantum theory involve this - Frohlich's 1968 paper, Hameroff and Penrose, and others. Others like Pribram, McFadden and Freeman deal with fields in a more general way. (The relevant references should pop up on Google fairly easily with the names.) I have recently written a paper in Journal of Consciousness Studies looking at the problems of theories invoking 'unified fields' and have suggested, rather like Descartes, that what we need is not a unified field but a unified relation between a field and some 'indivisible substance' which in modern physics would be a mode of excitation. My website (linked to by my RG page) goes into my background reasons in various ways.

Within neuroscience perhaps the most popular theories of what makes certain signals co-contribute to experience are those of Tononi (Integrated Information Theory) and of Changeux and Dehaene. To my mind there is still a locality problem for these, but C&D propose some very plausible background requirements.

So there's lots of stuff, but to my mind the most interesting thing is that we are still not agreed on what could be feasible within biophysics.

For me recently studying Sandra Cisnero's "The House On the Mango Street" in vignette form provided the literature that considers what could perhaps be thought of as perceptual mereology...

"Philosophy in the Flesh" johnson and lakoff 2008 mentions 'mereological summativity' at least once in reference to embodied cognition. I believe Husserl also involves himself with the topic in "The Crisis of the European Sciences and Transcendental Phenomenology".

Thank you Ryan, P

The mereological fallacy (within neuroscience) is discussed extensively in Bennett and Hacker's (2005) "The Philosophical Foundations of Neuroscience". Especially in Chapter 4, 'Sensation and Perception'.

-Hope this helps.

Thanks Michael.

Absolutely! I think this topic is an important one.

An important aspect to applying the mereological principle to the field of psychology is, for instance, when explaining the causal sequence of events of seeing an apple, “the upshot of the immensely complex sequence of neural events that occur [in the brain] is not that an image or quale miraculously springs into existence, either in the brain or anywhere else. It is, rather, that the person sees the apple” (Bennett and Hacker, 2005, p.306). Thus, the case in point here is a conceptual one, not an empirical one, for we cannot doubt that our brains are critical in order for us to manifest the capacity for consciousness.

Put differently -- with respect to illustrating another analogy is to think about the relation between the source of energy generated from a car.  It makes no sense whatsoever to ascribe horsepower solely to the pistons of a car since horsepower is an attribute of the entire automobile (even though pistons are a necessary component for generating horsepower). To this end, it makes no sense whatsoever for neuroscientists to “localize” consciousness to a particular part of an organism, namely the brain (even though it’s a necessary feature for it), because the brain does not experience consciousness, the person does. Thus, the source of confusion is that neuroscientists “neuroify”—a term borrowed from Raymond Tallis in Aping Mankind: Neuromania, Darwinitis and the Misrepresentation of Humanity (2011)—conscious attributes when they should be attributed, conceptually, to the person who experiences it.

Dear Michael,

With all due respect to you, I think Bennett and Hacker's proposal is fatuous, in a way only Oxford philosophy has the gall to be. Hacker is like the caterpillar in Alice in Wonderland, which may have been a caricature of an Oxford philosopher for all I know.

The ordinary man says that a person sees a cat. But the ordinary man says that the sun rises in the morning. Both may be very muddled ideas. The point of philosophy and science (which are really the same) is to show that the ordinary man (like Hacker) often has confused ideas and rather more counterintuitive ones may be better.

What is a 'person'. The engine analogy is invalid because we know what the parts and the totality are. We do not actually know what we mean by person. If it is a whole human being then that does not see. The liver is part of a human being and is completely uninvolved in seeing. If the person is some non-local dynamic construct to do with 'behaviour' then we still have to decide what it is. Person and seeing are just useless words when it comes to understanding subjective experience.

And even if 'seeing' has to apply to 'persons' then there is no justification for retaining this stricture for 'experiencing consciousness'. The ordinary man does not even think about that. What could a 'person' possibly refer to here? It refers to a belief in an entity far more confused than Descartes's soul. Descartes's analysis is extremely specific and entirely consistent with physics, just wrong in practical detail. Hacker's analysis is more like that of Galileo's inquisitors - empty belief.

There is also the point that to define mechanism in terms of behaviour is circular. If you define a screw driver as something that drives screws you learn nothing. The value of science is that it explains behaviour in terms of internal mechanism based on generalised dynamic rules. Hacker may hate science but he has nothing to tell us of interest and would do well to go back to the library. Goodness knows what Bennett thought he was up to.

Thanks for your response, Prof. Edwards -- all very good points!

The contributions from Bennett and Hacker are limited from the larger scheme, but they offer some good clarifications to the use of language in the field, in particular.

The main thrust from Bennett and Hacker's book has to do with the inaccurate language used by neuroscientists. For example, saying that the brain has experiences, believes things, interprets clues, and makes guesses, and formulates hypotheses. This language presumes oddities in their assumptions, such that neurons contain knowledge, and estimate probabilities and present arguments. This is just erroneous.

Put differently, their main project is to introduce the mereological fallacy in neuroscience, which is the ascription of attributes to parts of an animal that can only be intelligibly ascribed to the whole. These kinds of mistakes are equivalent to saying that the engine of a plane flies, or the fuse of a a clock indicates time. Likewise, it does not follow that if a car is fast, that its carburetor is also fast. It is a problem with language.

From the onset, their arguments are based on Aristotle and Wittgenstein, specifically in respect to the detailed determination of actualities and potentialities, which amounts to a dynamic metaphysics.

Interestingly enough, Bennett is a world-renowned neuroscientist.

I am aware of Bennett and Hacker's intent and claims, Michael, but they are talking nonsense. They are not clarifying the use of language. Like the caterpillar, they are insisting on a particular usage, which is not the way language works either in science or ordinary life. Usage shifts constantly with context and growth of understanding. 

The irony seems to be that although Hacker invokes Wittgenstein he is doing precisely what the lat Wittgenstein said we should not do. Moreover, the idea that it is 'persons' that are conscious or see is simply not what ordinary people say. They say 'I am conscious' or 'you see the cat' and you and I are persons. However, if asked if persons are conscious most intelligent adolescents might well say 'but what actually IS a person'. The belief in persons is something that philosophers insist on but most ordinary people are not quite so sure they know what they are going on about. And Hume famously said that he could never find his own person, so what hope of finding others?

The analogies with machines are false and misleading. An engine does not fly. However, if you ask what drives a train you can say the locomotive, or you can say the steam engine in the locomotive, or the transmission system, or maybe the steam itself. All of these work pretty well. So there is absolutely no reason why there should not be valid answers to 'what experiences cats on mats?' at a finer grain than 'person'. The extended mind people have shown neatly that 'person' is actually quite arbitrary and inappropriate as a domain of mind. Mary Midgeley talked of the magic membrane that is put around the person, which of course does not exist. Person is a very dubious domain to allot experience to. A neural domain is much more promising, as Descartes realised. There is absolutely nothing wrong with using language this way.

There are certainly problems with saying that brains see unless one qualifies the usage carefully. It is almost certainly not the brain anyway, since the olfactory bulb and hypothalamus do not see. And neuroscientists don't talk like that anyway. This is straw man stuff. But I can see absolutely no problem with saying that the brain interprets clues. The liver does not do that but the brain seems pretty certain to be where clues are interpreted. Hypotheses get formulated in brains as far as I can see. 

And for sure the brain seems to be where experiences are. I agree that the whole brain does not 'have an experience' but largely because I suspect there are millions of experiences all going on at the same time in a brain. The only reason why there is a myth that experiences are not located in specific parts of brain is that our brains do not have machinery for referencing internal events in internal space - only in time and in relation to the world. That is obviously related to the fact that there is no biological advantage in knowing that experiences are in the temporal lobe or thalamus, only if they were at teatime and in the kitchen. If experiences are part of biology they must be local like everything in science. To think otherwise is the sort of trivial mistake that occurs when muddled philosophers are too lazy to think about how their ideas relate to the real world.

Although I know a lot of neuroscientists I had actually never heard of Bennett before I heard of this book but he probably is quite well known in the field. I cannot think why he teamed up with Hacker. But a lot of neuroscientists are blind to the need for locality when it comes to experience. They allow the rules of biology to evaporate once inside Midgeley's magic membrane. It is all quite peculiar and I suppose goes to show that really rather few human beings think rigorously!