Thursday, 11 October 2012

Colin McGinn on the limits of explanation

McGinn: mean but fair
There has been a widespread assumption in philosophy, and other disciplines, that the problems of the world are there waiting to be solved, and that indeed they can and will be solved by human beings. The problem of consciousness, for example: though consciousness remains mysterious from a scientific perspective, philosophers have put forward a range of views and arguments for both physicalism and dualism (see, for example, Chalmers on zombies) which seek to explain consciousness, or at least provide the basis for thinking that we could explain it. 

However, Colin McGinn, a philosopher currently at the University of Miami, thinks otherwise. McGinn has formulated ‘Transcendental Naturalism’: the view that due to our biological and evolutionary background, we are inherently limited to some extent in our cognitive abilities, and thus there will be problems (both scientific and philosophical) that we will never be able to solve or explain in our current status as human beings. To many, this view will at first sight appear extremely pessimistic, but I think it holds a lot of plausibility. 

McGinn describes conscious beings as each having a ‘cognitive space.’ This is analogous to our perceptual space, or our physical space: areas within which we can ‘move’, so to speak. Our perceptual space is our visual and auditory field, outside which we cannot see or hear anything. Our physical space is the space within which we can move, and outside of which we cannot. So too with our ‘cognitive space’ - the intellectual area within which we can think, reason, and form concepts; and outside which, we cannot reason. 

We have cognitive limitations just as
we have perceptual and spatial limitations
If we accept that we are cognitively limited in this sense (and this seems likely, since as McGinn states, ‘we are not Gods, cognitively speaking’), then we should expect that some of the problems of philosophy we will be unable to solve. McGinn most famously treats the problem of consciousness as such a problem, not unsolvable in principle, but unsolvable to us, due to its solution falling outside our cognitive space. After 2000 years of thought and experiment, the basic nature of consciousness continues to remain mysterious: it is a Hard problem with a capital H, and raises a significant explanatory gap. The natural explanation for this mystery, McGinn argues, is not that consciousness itself is inherently mysterious, or is made up of new ‘mental, non-physical stuff’, as Descartes thought. Consciousness is probably as grounded in the physical world as tables or chairs. However, its nature lies outside our cognitive space: our biologically limited concept-forming abilities do not extend to grasping it. We can no more understand consciousness than a slug can understand mathematics. 

McGinn also applies his transcendental naturalist view to other problems, such as the problem of Free Will (a response also shared with Noam Chomsky). On reflection, these conclusions may be extremely disappointing: after all, if we agree with McGinn that these problems are in principle unsolvable to us, then we must admit that they could only in principle be solved by a higher intellectual race, 1000s of years of evolution in the future. This is not a nice conclusion to draw, certainly if one is a philosopher. 

However, as is often pointed out, the unpleasant-ness of a view is not a reason not to hold it, especially if one has good philosophical reasons to hold it. And I think we do: why assume that we are intellectually so superior as to solve profound philosophical problems such as the problem of consciousness? Our powers must stop somewhere; and there is good reason to think that consciousness does indeed lie outside our cognitive space. It’s not all so bleak though; McGinn’s view does allow the physicalist/materialist a good response to the arguments for dualism drawn from the explanatory gap between the mental and the physical. The explanatory gap exists not because there is a gap in the world, between physical and mental stuff; it exists because there is a gap in our cognitive abilities to grasp the physical world, and the conscious states that arise from it.

McGinn's more substantial defence of his position as applied to consciousness can be found here, in his classic paper.

Sources: Colin McGinn, 'Can We Solve The Mind-Body Problem?' and Problems in Philosophy: The Limits of Inquiry


  1. Nice article!

    I'd be inclined to disagree, but I'm a scientist, so what do i know?

    as the progess of mathematics is basically the enlargement of the space of methods and structures of conceivable systems. Although no one can concieve of something which they can not concieve, the fact that this space is growing (mathematicians are quite good at their jobs) presuades me that it is possible.

    I'd lay bets that when someone figures out how concioussness works, the logic used will be ancient, and the maths will be over 50 years old.

    it's pretty much impossible to concieve of all of the things that allow this message to reach you from my brain at the same time, but its mostly 'understood' what is going on (subject to physic's terms and conditions.)

  2. Here's a response to McGinn's Mysterianism (what you call Transcendental Naturalism) copied (with citation) from The Cambridge Handbook of Consciousness. This excerpt is from one chapter in particular authored by Uriah Kriegel (page 38):

    The Case Against Mysterianism:

    "McGinn appears to assume that conscious states are caused by brain states. His argument does not go through if conscious states are simply identical to brain states. In other words, the argument does not go through unless any identity of conscious states with brain states is rejected. But such rejection amounts to dualism. McGinn is thus committed to dualism. On the view he presupposes, the conscious cannot be simply identified with the physical. Rather, there are two different kinds of states a person or organism may be in: brain states on the one hand and conscious states on the other.

    Recall that McGinn's mysterianism is of the epistemological variety. The epistemological claim now appears to be conditional upon an ontological claim, namely dualism. So at the end of the day, as far as the ontology of consciousness is concerned, McGinn is a straightforward dualist. The plausibility of his (epistemological) mysterianism depends, to that extent, on the plausibility of (ontological) dualism."

    So before accepting McGinn's arguments we must consider the plausibility of dualism, and I do not think that THAT debate has been resolved in the least.

    "Let us raise one more difficulty for mysterianism, and in particular the notion of cognitive closure. It is, of course, undeniable that rats do not understand trigonometry. But observe that trigonometric problems do not pose themselves to rats (Dennet, 1995, pp. 381-383). Indeed, it is precisely because rats do not understand trigonometry that trigonometric problems do not pose themselves to rats. For rats to grapple with trigonometric problems, they would have to understand quite a bit of trigonometry. Arguably, it is a mark of genuine cognitive closure that certain questions do not even pose themselves to the cognitively closed. The fact that certain questions about consciousness do pose themselves to humans may therefore indicate that humans are not cognitively closed to consciousness (or more accurately the link between consciousness and the brain)."

    Cambridge Handbook of Consciousness (2007) (Eds: Zelazo, Moscovitch and Thompson)
    Chapter 3: Philosophical Theories of Consciousness: Contemporary Western Perspectives, by Uriah Kriegel

    1. This is very wrong. There are a whole host of positions in the philosophy of mind concerning consciousness, of which the identity thesis (that mental states are identical to neurophysiological states) is just one. But there is also non-reductive physicalism, as well as eliminative physicalism. So if you reject the identity thesis, you can still be a physicalist.

      Furthermore, McGinn is not at all committed to dualism. His view is naturalistic, but he places our inability to understand the mind at an epistemological, NOT an ontological level. One of the virtues of his view is that it is compatible with both physicalism and dualism.

      The Dennett criticism has more weight. I have no view on that matter.

    2. PS, 'Transcendental Naturalism' is McGinn's own expression. 'New Mysterianism' is not McGinn's term; rather it is a term that originates from his opponents who seek to criticize his view.

  3. My complaint with this view is that, while we may indeed have cognitive limitations, problems formulated within our means seem like they should have solutions that themselves lie within those same means.

    Sure, slugs can't solve mathematical theorems, but they can't conjecture them either.

    So to for us; if we can formulate the problem of consciousness, shouldn't we at least suspect we could solve it?

  4. Perhaps; but I don't see any reason why we should think that an ability for being X to formulate/express a problem should have any bearing on X's ability to solve that problem. Whats the link?

    1. Indeed. I believe Kant has properly shown that such a link does not necessarily exist, e.g., our (futile) attempts of determining the existence of God on the basis of speculative reason. McGinn's view on this matter is identical to mine.

  5. What if mind is non-algorithmic? All physicalist explanations are by their very nature algorithmic, so attempting a physical/computational/ procedural/quantitative/Boolean explanation for non-algorithmic mind would be a category error.

    Non-algorithmic phenomena (if they exist) present a profound challenge to science, because some/most/all of the basic algorithmic operations and data structures may not be relevant or applicable to any 'reductionist' explanation.

    AND may not be applicable

    NOT may not be applicable

    OR may not be applicable

    IF... THEN... (and its reverse aspect 'because' ) may not be applicable

    Numbers and logical states may not be applicable

    The great 'mysterian' difficulty in talking about non-algorithmic phenomena is that although we might be able to say in general terms what they do, it is impossible by their very nature to describe how they do it. (If we could describe in a stepwise manner what was going on, then the phenomenon would be algorithmic!).

    So what phenomena would be applicable in the non-algorithmic realms? Qualia? Intentionality? Introspection?

  6. In the western movie "The Outlaw Josie Wales" the aging but vigorous indian Lone Watie pulls out a coloured transparent crystaline chunk and says "l have this rock candy - but it is not for eating - it is for looking through"
    I should think that Mr. McGinn is correct in proposing cognitive closure - a rather obvious conclusion if we assume isolated brain/minds as the effective cognitive agents. But we do not know where the limits of our brain/mind lies - and therefore we do not know where our epistemological limits lie - so pessimism is the correct term I think - in other words an attitude adopted in the absence of compelling evidence either way.
    He may well be right that the 'hard problem' will never be solved -but I think philosophical problems can sometimes act in a similar way to artworks - or to Lone Watie's candy - to help us see things from a different perspective. Sometimes a problem is not for solving - rather for provoking.
    The observation that the 'mind body' problem has not been 'solved' in the last few thousand years holds little water for me -the problem (if we can call it that) is epistemological - about how and what we can know about consciousness. All epistemological systems have both meriological articulations and aspectual filters/generators. Once a filter/generator has been posited then the possible articulation afforded by it is in principle closed - this is the province of logic in its widest sense - the province of discovery and empirical settling - but I see no limit on the number of possible aspects that can be adopted - this is the province of creativity or to use a less coloured term - generativity. So who is to say what the next thousand years will produce - or the next ten years come to that.


Blog Directory
BLOG DIRECTORY, Submit blog free, Promote Blog, Best directory