Response to my CRITICS

Abstract

I was intrigued to see that there was surprisingly little overlap in the responses by the three critics. It therefore seemed most sensible to treat each critic’s points separately, which is what I have done below. I hope I have managed to say something substantial on each of the points they brought up. While I harbour no illusions about my comments resolving all of the concerns raised to each critic’s satisfaction, I hope that the response helps to elucidate unclarities in the earlier exposition of my ideas and may, perhaps, provide resources for the further philosophical exploration of the views discussed. In relation to my discussion of Hohwy’s prediction-error minimization framework, Das rightly points out that nothing in this approach entails that our representations cannot carry information and hence yield knowledge about the external world. Indeed, some interpreters see this approach as a way of discovering the ‘hidden causes’ of whatever it is that brings about the changes in the outer nodes of our Markov blanket, arguing that the generative model ‘recapitulates’ the causal-probabilistic structure of the external environment that impinges on the organism’s sensory apparatus. It constantly generates, in a top-down manner, a flow of virtual or mock sensory signals that predicts the unfolding of sensory signals by external causes. 1 Similarly, some forms of Bayesianism see it as a way of discovering the true probabilities out there in the world (the ‘Lord’s prior’). What I am trying to argue is that neither appealing to the existence of some external item represented nor appealing to the accuracy of our representation (both the prediction-error minimizer and the Bayesian are likely to say that there are some structural features of the ‘hidden causes’ or the ‘true probabilities’ that our beliefs will capture) is necessary for the prediction-error minimization theory; in fact, that the theory would be in better shape if it dispensed with this idea. Considerations of parsimony aside, it is hard to make sense of the kind of connection supposed to exist between things inside and outside of the Markov blanket. On the one hand such a relation should be inside the Markov blanket for us to be able to think about it. Yet it also needs to reach beyond it to link up with the ‘hidden causes’. But if we can conceptualize a relationship and one relatum, the status of the other relatum does not appear to be quite as inaccessible as the idea of the explanatory-evidentiary boundary makes out. As such, the notion of a ‘connection’ with something outside of the Markov blanket is something we should do without.2 Das notes that ‘if the form of intentionalism sketched above is right, nothing in Hoffman’s account rules out the possibility that we can perceptually represent mind-independent properties of objects in our environment’. According to Das’s intentionalism, we represent a certain property in our environment if we are in an internal state that ‘has the biological function of indicating that property’. Certain internal states of an organism might have the function of indicating the presence of water. Does that mean that our internal state represents the mind-independent property of there being water in our environment? Not according to Hoffman’s account. What Hoffman sets out to show is that our perceptual capacities are not geared towards indicating properties the environment has no matter what, but that they have been shaped by evolution to indicate instances of fitness arising from an organism in an environment. An organism’s internal states do not have the function of indicating properties of the environment, but properties of the environment refracted through the lens of the organism’s specific interests and needs, i.e. instances of fitness. In these instances of fitness, properties of the organism and properties of the environment are so much bound up with one another that it is impossible to separate them. A realist critic might point out here that there must still be mind-independent properties of the environment, whether or not we can disentangle them from our representations or not. This move, however, pushes these properties into an epistemically inaccessible noumenal realm, and it becomes then unclear why the resulting position has any advantage over a more parsimonious rival that simply treats the mind-independent world as another part of the representational interface. In addition, an argument involving self-information entailed by non-self-locating claims is presented by Das, suggesting as an example that being in the British Library entails the existence of the British Library. I don’t think Hoffman would disagree with this, any more than he would disagree with saying that the claim that there is a blue folder on my desktop entails that there is something that is blue. However, this does not yet get us to ‘properties of mind-independent particulars’, but only to properties (and individuals) in the interface. Hoffman’s interface theory of perception does not rule out ‘the possibility of there being self-locating truths or of self-locating truths entailing non-self-locating truths’, but claims that any such truths are strictly truths about what goes on within the representational interface. In addition, Das points out that ‘as long as we accept a sufficiently externalist conception of knowledge … we can still be in a position to know that we are not in those sceptical scenarios’ irrealism raises. Externalism about mental states here means that such states ‘depend on extrinsic properties of an agent, not just intrinsic ones’. As an extrinsic property of me is one that essentially involves entities other than myself, externalism about knowledge clearly avoids the irrealist predicament. But anybody inclined towards irrealism is unlikely to assume that cognitive states essentially involve entities beyond the cognitive interface. The regarding impasse is then prone to reduce to a debate about where the burden of proof should lie: does the irrealist have to show that the existence of a mind-independent world is incompatible with the epistemic situation we find ourselves in (as Das seems to think), or does the realist have to show that the assumption of such a world presents an explanatory advantage that the irrealist’s representation-only view cannot match? It seems that one of the reasons why the debate between the realist and his opponent is so difficult to resolve is that the two have different ideas of what such a resolution would look like. The realist demands that his opponent ‘rules out’ the possibility of a mind-independent external world, while the opponent wants to see evidence of a unique feature of the realist conception of the world that cannot be replicated by ‘realism in the image’, that is by the idea that the mind-independent world is part of the representational interface. The irrealist regards the realist’s demand as excessive, as no amount of epistemic insulation would rule out the possibility of it just happening to be case that there is the realist’s kind of world behind the representational interface (much like some events in the real world might just happen to correspond to events in some video game). The realist regards the irrealist’s offer of ‘exact replication’ of his external world as insufficient, since his epistemic default is always the existence of something behind the representation. As long as we don’t have proof that there cannot be such a thing we are justified in assuming that there is. I agree with Das that if we accept that perceptual do not represent the world as it is of this does not that these cannot be ‘true or The world behind might be with our by the of the it might out that the structure of the just to be that of But on such a is more than on the of the the of the are presents a argument that the for a substantial structural between representation and represented are is as a theory as it to I that there must be a argument to it than the claim to the the world might just out to entities with the structure as our perceptual an in Das’s It is the claim that ‘if things appear to us to be a certain in the of those we should that things are that In the philosophical this claim was by the and by so we have another here where the between the irrealist and his some philosophical But are we to the of in a point that unclear to me is much the the of is supposed to be able to I look the world from an position it to and to most that there are real out there in the world, that the structure of is that all of my to me with that it is an whether two events happen or don’t happen the that my has the resolution and so Yet some to cognitive of perception and that these do not the things are that of all of these can of be into our but that we appear to have much of what was appealing to the claim that the world see is more or the world all things the world as it and the world as our it are what is trying to the of the theory of or cognitive are any it as if the we the to the structure of the world ‘as it is in to the to which we this on a is different from the way things appear or not this is a matter we not but it seems to be that needs to be up by some kind of It is not a we can accept some further idea of what might it. concerns the that of parsimony not While it is than as irrealism only one kind of two and has the advantage of the of from the whether are as a to things are the most than simply as in our of parsimony are not the only idea that in is the point that what we mean by ‘the is with like and so on of which essentially on to such an that it is unclear much sense we can make of a world beyond these This is of a for Das, to the claim that and so on are all in the world, and more or by our But those not to the that of the our idea of ‘the is might be no more than by the will not only that a theory that with the noumenal might be more parsimonious but are likely to be such an of properties to to our cognitive be what the we in is the most As a we the existence of any substantial structural between our representations and the mind-independent world and are to accept an noumenal as the of all the world behind the representational as part of the representational seems to the most sensible An Das’s argument on is the assumption that in the of we can simply assume the existence of an external world. do not to any evidence for While this point the argument can into the I am of the that if there is no evidence the existence of we are to the existence of for is that it would us to the existence of beyond the In this my philosophical is the of the one by any has to in explanatory entities that do not the idea of a noumenal world we cannot with in any the response should not be there are of blue behind the or there but explanatory do not have but no Das the that the irrealist’s to accept beyond the interface might be by the irrealist’s According to the irrealist cannot say what he is out to The is that the irrealist cannot the the as to something beyond the since for there is no such As such any this would be or he the as for something within the such as the representation of something beyond the interface. But the irrealist and his opponent to be about different the opponent about things beyond the he means mind-independent things with in some As such nothing the irrealist can say about such according to his disagree with what his opponent is resolve this it is to be that if irrealism is any theory that essentially on things beyond the interface needs to be this we cannot make such a theory part of our trying to between and Das is in that for the irrealist it is to an point outside of that representational but that does not that the of irrealism on such an The we to irrealism out to be is it according to the irrealist’s The to this to be a for the irrealist has the resources to that specific are not can that there are no and so he can that the beyond the representational is not Das the that if we make the assumption that on with entities the irrealist seems to be to to entities like and since he has with them. about the such a is since it our in with the entities The between the irrealist and his opponent is here not about the as such, but about to to It is that Das of in his of the he this to to real to a relation that us and our representational states with entities however, is an assumption that will appear to the we are that there is one relation that as a between the representational inside and the outside of the the irrealist’s claim that all our resources only to on to something within the representational is Yet there is nothing the irrealist from the as long as we it to about The of as to something within the representational interface than beyond it is an It is simply a of more in the debate about the cannot in on the of our theory by in since the of like and the like is by the other The is a that within the of the not a of the interface with entities that are not part of the interface. notes that that irrealism is not a but an about the external This might appear to The that I irrealism from is in the of truths it cannot be an that there is as would have it. The of irrealism with is on the idea that we can have an towards the external world there is such a world, or an is no such does not claim that we information to this but that on the of to us the more position is to the existence of an external world. or is a view of the world is the and irrealism is not. This is where the between irrealism and concerns the the irrealist into we accept a account of perception according to which the world we in is a that us will be on the The still on such as the two which will in on the they all the way to whatever the will on the since in the world us therefore have a of in which no is up the is a then nothing is real the sense of being have that that is and it if to an it is however, we simply up in a position to where we from inside the to the outside world of the our find ourselves inside the might also be that irrealism the and the theory about The a there is no behind the can we still make sense of the between the The an there is no true can we argue that are about the way things In response to the that the only in the interface. is nothing beyond the interface that the still be a the that is and we are epistemically with I it there are no the is and we are epistemically with like are are still The irrealist is hence not his and it since there are no in of he is not saying that there are no and that it the of so would by his point is that an we might up with a this however, that so that we will it. This is not if to get but by the for better than while we might to the of as this should not be as an theory that the of As the of is to point some position does not of the In this we should also point that something behind the of the that about mental the I to argue that no sense is to be of a behind the of can I still to the idea of of there being cognitive we since they are I do not that the of our cognitive that are inaccessible to us form a for the claim that all is The is for example by things that are to or of as by the of the and the of can about such and they can in our they to us in the way and them in to things that do and we inaccessible mental to of the mental we can As for the it is that out that we are about the way things cannot be the of outside of the of in to have an look the world. any must be from within the interface. we that certain as a world behind do not make sense we have this, we have not an true view of the world, an our view is better than the not in of approach to but in to the of The discussion in the by to the raised in the the whether it still a theory any true This is an which is in an of the debate between two different of the and The between these two is a matter of but one point is that is as some philosophical while is to be a with the any position of also the with a to the whether is to that the claims an irrealist be as might be part of a true a theory that does not to or but that of whatever there and is the point of our This point is simply ‘the that our with an this is not a about the world, but a about our we carry them out according to the rules that form part of the in which we we find that we reach the that might have like has another it. I don’t think there is here a would disagree but I also find it hard to see that this is an approach according to which is to are part of and about are by to that are also part of should this about by true means be to be an The I is that for there are two different sets of epistemic the of a of and is to truths about the world is the of water and so and the is to such truths as have no however, should we accept this of our epistemic it not the kind of epistemic that we to all of truths we might them in different I would to that we whether or not the on the is real water or a we the we trying to whether or some things are are with the epistemic we and have to all with some of these is no that us to the by it from those by epistemic outside of this about the whether the irrealist any It seems as if the irrealist’s can all be as claims about the of the with his the are to a to a no claim to or to such a claim can be to be have no for an can be to be in the world all the way or and The only other is are of a virtual world. we assume is we up with a the two of a theory of whatever there and of a point of our also be in of between the and his is the of all the claims by the true say that of whatever there is simply means that for the opponent up with the can show that the existence of to a Similarly, any entities to in an with can be further to an with The between this and the is that the is about the world, and to any entities still truths about the world, truths that are supposed to no matter what, or of there The to what is and is not in It is on by as the of and that certain entities cannot be in a theory of the world. It does however, a of the world that might why this is the these are then the is no whether or not the has a that of whatever there is and the point of our I that makes a case in of an But this in does not a that of whatever there is is to be Does this mean a about the world, or is it a about the we for the This it seems to the of the relation between irrealism and is as the position that I am while all other and things are a of my mental it is that the are irrealism is incompatible with the existence of any real However, we might also as the position that other are not of any about the of It is this form of that is in the discussion about the of other mental we for example in the of the two and it is also the one that we should have in in the we like we a representational or are different interface up as in a video or I the only one or are there other that do so as we about what is happening beyond our representational interface. such or is since it something the interface theory we cannot a view from outside of our interface into as it is in in to whether or not there are other we see from the The position that there are no other beyond our interface is as as the position that there is nothing beyond the since it the kind of true theory the in out to show to be The only response to the of the irrealist can is to say that from the within the interface which is the only we can there are other this does not provide us with any information about the of the irrealism and In one sense the is our view of the world the existence of other one then are they also real from the As long as an to this more than information within the interface it cannot be a from the can only be by us through our and if the interface our as a the existence of a in the the cannot be that my argument for irrealism on the of the world to the representational or as regarding it as a noumenal something behind the to which our This he such as or of the view of perception that forms the to much of the discussion in arguing that we should an account of According to such an are not representational about the they are and with the What this means on what we mean by ‘the the world is an of and things in a mind-independent in a mind-independent a realist is into this of the view of This makes it of in a discussion where the of such a view of perception is being On the other the world also be with the representational interface. As for the irrealist our clearly are and with the representational the the position is clearly true from an irrealist the does not reduce to no more than in the to since the and the realist accept the existence of an true theory of the structure of the world, while the irrealist does not. on is the irrealist would to see an argument for the about the external world by or he the as from within the representational Similarly, to argue that a account of perception us … with the objects or not be as being in with the irrealist’s It is that the here cannot be real entities out in a we would not be with a but with a realist view of But if the status of these objects is that they out to be of the representational or if they are simply with whatever our perceptual relation is to us to the irrealist’s of the representational there is no for the irrealist to to this account of While me in of a between an position and an beyond our epistemic reach I am about the the between the two of the might be to In some the can be as a to a world in it into a kind of hard to from the irrealist position in all but But of we cannot have it and must be that our response does not from one to the other to the of any is that the irrealist model be as it to account for the our appear to it seems to me that irrealism has resources to account for such for example no why of the should not be something that can within the representational nor would the claim that all perception within the of the interface that we cannot between or some according to the irrealist the and it does not link us up with entities of different in one a in the in we are to of our representational interface. But these are to other in different and it is this in structure that us to one from the in of us is likely to the of being be able to to in it will not this to the of our cognitive in a way that is quite different from the way our and this is what makes them as different of this is to point out that water cannot our but it does not from this that of water are and with the while of water are not. the are to cognitive states of one kind while the are to cognitive states of another kind The existence of these different us to and but it does not on that only the are with the also notes that we are of some features of the perceptual that are suggesting that the irrealist account for this by these features within the representational but that this would explanatory with and Yet this seems to that there is no more for such features than there is to example to in for the a with of of In the way in which the does not have to the ‘as if by an external of the of the features not be within the perceptual interface. a would the of the interface beyond It is only the features to that the perceptual interface has to from something like the features to something more