PHILOSOPHY PATHWAYS Issue 196

I. 'Dualism and the Individuation of Cartesian Minds: Grappling with Strawson's Anti-Dualism Arguments in 'Self, Mind and Body'' by Eric DeJardin

IV. 'Limiting the Self -- Extended Cognition and Standing States' by Matthew Sims

Issue 196 of Philosophical Pathways is dedicated to the subject of consciousness and the mind.

Questions concerning both consciousness and the mind have long been focal points that have saturated philosophical thought in one form or another; from the ancient Greeks' theories on the soul, to the writings of Descartes -- to whom it may be correct to bestow the title 'the founder modern philosophy of mind', both ontological questions such as 'what is the soul and how does it differ from all else?' and epistemological questions, 'how can I know I have a soul?', 'how can I know that the world is the way I perceive it to be?' have lead philosophers in their quest to understand both the world and its inhabitants -- namely us. Although, today, the number of philosophers who openly support Cartesian dualism are few, if any, Descartes' philosophical legacy is nonetheless far and wide reaching. This being said, the sustained development of the natural sciences over the last century can be certainly recognized as a major point of departure from having little evidence that the mind is a physical thing to having a reasonable amount of evidence that the mind is not a immaterial thing. Particularly, developments in the areas of neuroscience and cognitive science have supported the notion that the mind has more in common with the computer standing before you now than with some immaterial substance whose essence that of being a 'thinking thing'.

A natural question to ask then is, with the development of these recent disciplines having the success that they have had, why is philosophy of any use to an analysis of the mind? The old adage comes to mind 'if the shoe fits, wear it.' If neuroscience and cognitive science have provided us with satisfactory explanations of many of the mental phenomena that have concerned us, then why not just leave it to them to construct a theory of the mind? Why not let them wear the shoe? If things were only that easy! The problems that we are occupied with when analyzing the mind and the notion of consciousness are more than filling some theoretical space with the best explanation, but also delimiting the very space which needs to be explained. Cognitive science may be able to produce an adequate theory of mental functions and how various mental modules are engaged, affect and are affected by one another, however, such a theory presupposes an adequate conception of the mind itself. It is here that philosophy rears its head. Namely, it must attempt to answer many of the same questions about the mind that have captured the interest of philosophers throughout history: 'What is the mind?', 'What is an adequate demarcation of consciousness?' 'How can I know that the world is at all like that which I am conscious of ?' 'Can I know that others experience the world like me?' 'Can I know that other have experiences at all?' These questions cannot be merely answered by examining the physiological substrata of the brain, nor will they be answered by theories that take them for granted. Philosophic approaches to these questions are still as relevant as they ever were, if not more so.

Considering the continued wide-spread specialization inherent in the natural sciences today, philosophy, although it too being subject to specialization, continues to be a discipline whose essence is, to use a Humean metaphor, to turn the object of enquiry over upon all of its sides to discover its nature. This often involves the philosopher turning herself over in the process -- taking nothing for granted. It is this which distinguishes the philosophical method for an analysis of mind from that of the disciplines mentioned above. This of course is not to say that it is not the unity and whole of the labour of many distinct disciplines that will eventually lead to an adequate theory of the mind. On the contrary, today we live in a very exciting time -- one in which the efforts of a combined intellectual labour consisting of many specialized disciplines (and some which are not) can be focused upon a common goal, providing one another with the resources particular to them. Philosophy is no stranger to leaning upon resources external to its own. However, let it not be mistaken that it currently has nothing to provide in an analysis of the mind. Be it the case that the reader has taken time to engage with Philosophical Pathways, assent to this very point may likely to be assumed. However, if this reader is skeptical about the philosophy's ability to further contribute to an understanding of the mind, I encourage the reader to engage with what follows. I hope that these articles will aid in illustrating that the significance of both the questions that philosophy is apt to ask and its method of enquiry upon attempting to answer them have been and will continue to occupy a significant place in both the analysis of the mind and in the phenomenon of consciousness.

In 'Dualism and the Individuation of Cartesian Minds: Grappling with Strawson's Anti-Dualism Arguments in 'Self, Mind and Body'', Eric DeJardin begins by calling our attention to the nature of the Cartesian substance dualism distinction; the mind, an essentially thinking thing, being an utterly distinct substance as that of a body, whose essence is being a spatially extended thing. The relation of the particular person to the particular human being that the dualist argues follows from the mind-body distinction is such that the former is individuated by the mind and the later, by the body. P.F. Strawson, in his article entitled 'Self, Mind, and Body' (1966) put forth a formidable problem which occurs as a result of this view, namely that if a mind is a particular thing, one which individuates a particular person, then it must be the case that it can be subject to being counted or individuated and the very notion of an immaterial substance seems prone to being nonindividuatable. Thus, Strawson argues that given that there is no criterion of individuation available for the Cartesian, the very notion of a Cartesian mind that demarcates a particular individual, one that is subject to predication, is an unintelligible notion. Without the possibility of individuation, not only could one Cartesian mind be in fact many, but dualism is shown to be false given that the concept of a mind becomes dependent upon the concept of a human being in order to for it to be the object of intelligible predication.

In his analysis of Strawson's argument, DeJardin brings forth from it three interdependent theses, the Reduction Thesis, the Dependence Thesis and The Individuation Thesis, and then methodically proceeds to illustrate how they fail to do the work which Strawson's argument requires of them. Thus, DeJardin argues that Strawson's anti-dualist argument is untenable. After presenting two of his own very thought provoking individuation criteria for Cartesian minds, DeJardin meticulously turns the anti-dualist's cannon back upon him. He argues that the anti-dualist, given his position that bodies are to act as an individuation criterion for persons, and the problems inherent in individuation of bodies, faces serious difficulties of his own with respect to the individuation of persons. It must be noted that although DeJardin argues against both Strawson's negative and positive theses, his argument itself is not pitted against materialism and manages to abstain from assent to either the dualist or the materialist position.

Catherine Nickford, in her contribution 'On the Validity of Quantum Consciousness', presents a case for understanding consciousness as quantum mechanical phenomenon. The elusive nature of consciousness which makes it difficult to pin down as a particular physical state can, according to Nickford, be explained in a sufficient manner when seen within the framework of a quantum mechanical theory. Within such a framework, macro-physical observable matter and phenomena are hypothesized as being the result of quantum micro-particle (quanta) behavior. Quanta, considered as waves of varying probability, whose individual wave-function peaks are indicative of the maximum probability of a particle's being spatio-temporally located, Nickford suggests, can be considered as 'permanent possibilities of macroscopic objects'. Using such a framework to analyze consciousness, Nickford presents a theory from Hameroff and Penrose. This theory claims that the observation of a quantum system, which subsequent to observation is not in any particular state but in a superposition, results in a collapse of that system into a specific state. This is deemed subjective reduction, given the dependency of the specific state upon subjective observation. Objective reduction, however, is a self-collapse occurring in virtue of the behavior and properties of the system itself. Conscious events, seen as neuronal microtubule state manifestations, Hameroff and Penrose claim, are correlated to this kind of objective quantum self-collapse. Nickford maintains that if a quantum theory of consciousness along these lines can be accepted, the result would be explanatory not only of why consciousness is difficult to pin down as an observable physical state, but of its non-computational, i.e. non-algorithmic, nature. This is an advantage that such a theory has over purely materialist empirical theories about consciousness, many of which explain it in terms of being an emergent property in a determinate physical system. She concludes, when setting one's ontological gaze at a micro-level of description of reality, the truth of such a theory results in a kind of substance monism and a property dualism; the former being the substance of quantum potentiality and the later being its attributes -- the mental and the physical; together composing a unified world -- one which is both observed and experienced and yet also unrealized. However, at the macro-level of description, physicalism most accurately describes our empirically observable phenomena. Thus her ontological position may be called quantum contextualism.

In 'Can Consciousness Emerge in a Machine Simulation?', Yissar Lior Israeli gives a brief overview of some of the foundational theories of consciousness within philosophy of mind. He goes on then to argue that emergentism is the best explanation for consciousness. In short, emergentism can be understood as the notion that a system's observed dynamics cannot be predicted from its substratum constitutent components. The term emergence refers to the rise of properties that cannot be reduced to the substrata nor expected from it. One example of an emergent property by John Stuart Mill that is commonly used to illustrate the phenomenon is that of the emergence of the crystalline property of frozen H2O -- particularly exemplified in snowflakes, that is not predictable from having complete knowledge of the properties of its individual elemental components, oxygen and hydrogen. Theories of emergence are, however, far reaching in their explanations of a multitude of phenomena in biology, chemistry, physics and the social sciences. Israeli, after giving a brief exposition of how consciousness can be construed as an emergent property, then goes on to argue, upon the assumption that consciousness is truly an emergent property of the brain, the possibility of an emergent conscious state emerging in a system that functionally simulates the brain should be at least contingently possible. Upon admitting to this contingent possibility, Israeli concludes his paper with an epistemological problem, namely, if the problem of other minds has been one that has at least shown us that there is no infallible method of discovering whether or not anyone other than oneself is indeed conscious, then given the absence of such a method, if consciousness did emerge in such brain simulation, how could we possibly recognize its presence? His response reminds us that, if infallible knowledge of other minds is too much to ask for in our case, then why require it of such a possible simulation? This answer leans on the fact that we can be reasonably justified when attributing consciousness to others like ourselves because it remains to be the best explanation for others' behavior. If this is the case, then the same will have to do if and when we are faced with a complex system whose behavior is best explained by its being conscious.

In the last essay of this issue, entitled, 'Limiting the Self -- Extended Cognition and Standing States', the author presents an analysis of Brie Gertler's argument against Andy Clark and David Chalmer's extended mind hypothesis (EMH). The EMH, as construed by Clark and Chalmer's (C&C), questions standard cognitive science's notion of the brain-bound mind and goes on to make the controversial claim that given how we define having some mental state or process, if there is something which just to happens to be located externally from the subject's cranium (or body for that matter) and that something happens to cognitively function for that subject in a way such that if it were contained inside the head we would find little reason to deny pronouncing it as being a cognitive constituent, then by want of avoiding arbitrary cognitive process demarcations we should accept such an external device or action as a constituent of the subject's cognition. Furthermore, and more important to Gertler's argument, C&C claim that standing states (non-occurrent beliefs and desires -- those that one has at t but which are not entertained at t) can be cognitively extended beyond the bounds of the skull if the extension plays the correct type of cognitive function role for the subject. Gertler, in her paper 'Overextending the Mind', presents a very convincing argument against this version of the EMH which attempts to illustrate that its truth as described, along with C&C's notion that the mind is comprised of both standing and conscious -- occurrent states, results in not only a violation of a commonly accepted physical law, i.e. that nothing can occupy two distinct spatial locations simultaneously, but also flaunts our common intuitions about personal identity and the self. After presenting Gertler's argument, the author introduces another argument, the Argument from Extended Accountability, in hopes of augmenting the problem of the extended self by enlisting our moral intuitions. 'How could a sleeping person be morally accountable for committing a murder that was motivated by his set of standing states but carried out by a robot in possession of these standing states?' Then, after not having found a necessary condition additional to those offered by C&C in order to constrain standing beliefs, a proposal on a condition for qualifying as some particular person's extended standing states is turned to -- what the author calls the General Condition (GC). It is argued that because the GC distinguishes extended standing states that are someone's from states which were someone's and it is only in virtue of the former which talk about sameness of extended self becomes a relevant issue, given the intuition pumps as described, that which is claimed to be an extension fails to be so. It is concluded that if our intuitions about self-identity should be that which prevents us from accepting the EMH, then because there is no issue to be had in either Gertler's argument or in the author's augmented argument, the EMH with respect to extended standing states remains a tenable position hold.

I am extremely grateful to all the contributors of this issue for dedicating both their time and effort in making it come to be. I am especially grateful to Dr. Geoffrey Klempner, my one time mentor, for giving me the opportunity of editing this issue and moreover, for providing me with a sustained opportunity to philosophize. His guidance and encouragement over the years has proven invaluable only second to his patience. I would also like to thank both Marco Totolo and Eric DeJardin for their insightful feedback and encouragement. Last but not least, I am indebted to Dima Sims, a constant source of inspiration.

I. 'DUALISM AND THE INDIVIDUATION OF CARTESIAN MINDS: GRAPPLING WITH STRAWSON'S ANTI-DUALISM ARGUMENTS IN 'SELF, MIND AND BODY'' BY ERIC DEJARDIN[1]

Cartesian dualism (henceforth: dualism) is the view that human beings comprise two utterly distinct substances, viz. mind and body, that are intimately united in a relation of mutual interaction. Since they are utterly distinct substances, minds and bodies possess utterly distinct properties. Bodies, for instance, are extended in space, and so are spatially locatable. But (Cartesian) minds are immaterial, and thus are not spatially locatable. Further, while the dualist is committed to the notion that the body is an essential element of the human being, it's the mind alone, he would claim, that, strictly speaking, constitutes the person. For a person is, the dualist would say, fundamentally a 'thinking thing', and it is (he would say) the mind that thinks, not the body. Talk about a particular person, therefore, as opposed to talk about a particular human being, is, for the Cartesian, talk about a particular mind.

However, the notion of the particularity of Cartesian minds raises a problem for the dualist. For 'mind' is a sortal term -- that is, it names a kind of thing, the individual instances of which can, at least in principle, be counted. Thus the content of the Cartesian concept of a mind must provide us with a means of counting individual minds. If it does not, then it's difficult to say in what sense the sortal 'Cartesian mind' names a genuine kind. For some terms that purport to name countable kinds can be revealed, upon analysis, to be in fact instances of what David Wiggins calls 'dummy sortals', i.e. to be terms that merely seem to name countable kinds (Wiggins 1967, pg. 29). For example, the term 'thing' is a dummy sortal, for although it makes grammatical sense to qualify 'thing' with various quantifier expressions (e.g. one thing, many thing, ten things, etc.), it's not possible to count the number of things within a given domain. (Consider, for instance, your computer: is it one thing, or is it as many things as it has separable parts? How would you decide between the two? And if you choose, for whatever reason, the latter answer, then what about the separable parts -- e.g. molecules, atoms, quarks, etc. -- that compose each separable computer part? Must we count each of them to determine how many things we have? And so on.) Hence, the dualist's concept of a mind, if it's a genuine sortal term, must provide us with criteria for individuating particular minds. That is, the concept must provide us with a principled way to ascertain whether what the dualist would call two qualitatively identical individual minds -- that is, two minds with identical properties -- should be judged to be numerically identical -- i.e. to be one and the same mind -- or distinct. And if it fails to satisfy this requirement, it's doubtful whether the concept it purports to denote as a kind is ultimately an intelligible one.

In section one, I shall present an argument developed by Strawson defending the conclusion that the sortal 'mind', as the dualist understands it, is unintelligible because it lacks an adequate criterion of individuation (Strawson 1966, pp. 169-177). I shall then argue that Strawson's argument for the ultimate unintelligibility of the dualist's concept of a mind fails to refute dualism. In section two, I shall suggest two ways in which the dualist might provide a criterion for individuating minds. Finally, in section three, I shall argue that the anti-dualist faces problems as serious as those that he argues the dualist faces vis-a-vis the individuation-of-minds problem.

I would argue that Strawson's anti-dualism argument can best be understood in terms of the interdependence of three theses, which I shall call the 'Reduction Thesis', the 'Dependence Thesis' and the 'Individuation Thesis'. I shall also argue that all three theses are problematic in their own way, and hence conclude that Strawson has not shown that dualism is either unintelligible or false. (Dualism may of course be, in fact, either unintelligible or false -- or both -- but the issue here is whether Strawson has shown it to be so.)

Strawson's argument begins with an observation concerning ordinary talk about human beings. (He calls it ordinary talk about 'persons' -- indeed, this is the term he uses throughout his essay. However, as we'll see below, his use of this term is ambiguous in the context of discussions of dualism. Hence, we'll use 'human beings' for now). We frequently ascribe both mental predicates like 'is thinking' or 'desires truth', and corporeal predicates like 'is tall' and 'has brown hair' to one and the same subject, e.g. the human being 'John'. But the dualist must judge such talk to be 'metaphysically misleading'. For while John's body is tall, it cannot think, and while John's mind can think, it cannot be tall (or short, large, etc.). Hence, the dualist is committed to the notion that ordinary talk with specific human beings as subject terms is, at least in principle, reducible to talk about their minds and their bodies. Further, whatever we predicate of human beings must be also be reducible, either to mental predicates that are predicated of minds alone, or to physical predicates that are predicated of bodies alone. This is the Reduction Thesis, and as we've seen it comprises both subject and predicate reductions.

The Reduction Thesis in general seems unobjectionable. For it merely requires that one's metaphysical position be expressible in precise terms if discussion of it in ordinary language is unclear or obfuscatory. And this just is one key job for philosophy in general. But I'd argue that Strawson's application of the Reduction Thesis to dualism is problematic. For by requiring that mental subjects be ascribed only mental predicates, and that corporeal subjects be ascribed only corporeal predicates, Strawson ignores the fact that (Cartesian) dualists are interactionists, not parallelists. That is, since dualists believe mind affects the body, some mental subjects will have corporeal predicates. And, since dualists believe that the body affects the mind, some corporeal subjects will have mental predicates. So while the dualist may be committed to the Reduction Thesis, he will not agree that he must attempt to effect his reduction in the way Strawson claims he must.

Strawson argues that the Reduction Thesis implies a further commitment on the part of the dualist. For if the dualist is committed to reducing talk about human beings to talk about minds and bodies, it follows, Strawson argues, that he is committed to the notion that the concept of a human being is dependent on the concept of a mind. I take it that the dependence relation that Strawson is deploying here can be elucidated as follows: for the dualist, the intelligibility of talk about human beings must depend on the intelligibility of talk about minds. That is, to understand what a human being is, we must first understand what a mind it. Thus, if we should discover that the dependence is the other way around -- i.e. that our concept of a mind depends for its intelligibility on our concept of a human being -- it follows that minds are not, as the dualist believes, distinct components of human beings. Rather, minds would then depend on human beings for their intelligibility. And in much the same way that our conceiving of surfaces depends on our prior conceiving of material objects implies that surfaces, in virtue of this conceptual dependence, cannot exist sans the existence of material objects, minds could not intelligibly be said to exist sans the existence of human beings (Strawson 1966, pg. 171). (I shall have more to say below on the relation between conceptual and ontological dependence when we discuss the Individuation Thesis.) Thus the dualist must aver that the concept of a mind is intelligible independently of the concept of a human being, and hence that the concept of a human being depends on the concept of a mind. This is the Dependency Thesis.

Now for Strawson, the term 'person' denotes a being to which both mental and corporeal predicates are applicable. That is, it denotes something akin to (though not identical to) our everyday use of the term 'human being'. However, I'd argue that because the dualist is an interactionist, he will use the general term 'person', which is sometimes used in ordinary language to refer to the human being, and other times used to refer to something more fundamental, in at least two distinct ways. That is, he will have two distinct concepts of a person (again, at least as the term is used in ordinary language). First, he will view persons as thinking things, i.e. as minds. Since, as I said in the very first paragraph of this essay, this constitutes the essence of a person for the dualist, let's call this strict (for the dualist) use of the term 'person' the 'Essential Person'. But second, he will (non-strictly) view persons as Essential Persons who are united, contingently yet intimately, with particular bodies. Let's call the unity referenced by this use of the term 'person' the 'Factual Person'. (N.B. though the 'Factual Person' is identical with the dualist's concept of a human being, it's not identical with the ordinary concept of a human being, and certainly is not identical with Strawson's concept of a human being/ person; hence, it's important to distinguish the Factual Person/ dualist's human being from the ordinary or Strawsonian notion of a human being/ person).

We can now reformulate Strawson's Dependency Thesis as follows: The concept of the Factual Person -- that is, the dualist's concept of the human being -- depends for its intelligibility on the concept of the Essential Person. And this allows us to restate the Reduction Thesis more precisely as well, viz. some talk of Factual Persons and their predicates is reducible to talk of Essential Persons and their predicates. For where statements involving interaction between the Essential Person and his body are concerned, the dualist just is talking about the Factual Person, and not about the Essential Person alone. And the dualist is certainly not talking about how the Essential Person interacts with some 'bodily person', a notion that has no purchase for the Cartesian.

This closes the distance between the dualist notion of dependency and what Strawson claims is the anti-dualist notion of dependency, according to which the concept of a mind is dependent on the concept of a person/ human being. For we can clarify the dualist's dependency relation further by recalling that for the dualist, the Essential Person can only be directly encountered by the individual himself; only I have direct access to my mind, and I only have direct access to my mind. So even if Strawson is right that the concept of a mind depends on the concept of a person, it seems to be the case that my concept of a person depends on my concept of my mind. For I suppose that your mind, which I cannot access directly, is essentially like mine; without this first person access to my mind, I'd be unable to conceive of your mind. And hence it is the dualist who ultimately gets the dependency relation right.

Strawson's primary target with the Reduction and Dependency theses, however, is not the predicate term, but the subject term. For Strawson argues that the very idea of what we have called the Essential Person, i.e. a Cartesian mind, is incoherent because there is no principle according to which Essential Persons can be individuated. To speak about minds is to predicate various properties of them. But to predicate a particular property of a particular essential person/ mind, we must be able to refer to that mind. And to refer to a particular mind, we must be able to explain how minds are individuated; otherwise, we are not justified in concluding that we have successfully picked out a single mind. For it seems as if what the dualist takes to be reference to a single mind could in fact involve multitudinous minds (Strawson 1966, pp. 173-174). And if we cannot refer to particular minds, then we cannot make intelligible statements about them.

Further, Strawson argues that the dualist cannot secure reference to particular minds by referring to particular human beings. For he argues that doing so concedes the conceptual primacy of the concept of the human being, while the dualist is committed to the conceptual primacy of the concept of the mind (Strawson 1966, pg. 174). And for Strawson, conceptual primacy and metaphysical primacy track one another. For he believes that talk about concepts -- especially fundamental concepts -- and their relations 'amounts to' talk about things and their relations (Strawson 1992, pp. 31-33). Hence, if the concept of a mind can only be individuated through its reliance on the concept of a person, then minds are metaphysically parasitical on persons. Hence, they cannot exist independent of persons, since they cannot be intelligibly conceived independent of persons. But if this is so, then dualism is false. This is the Individuation Thesis, and it's what Strawson calls dualism's 'central difficulty'.

I'd argue that Strawson conflates a number of key issues. First, he supposes that we can only refer to particulars if we can individuate them. But as Kripke has argued, we often seem to refer to individuals without individuating them property-wise. So, while I may only know that Feynman was a famous physicist -- which is hardly a unique property! -- it seems as if I can successfully refer to him through the use of the name 'Feynman' (Kripke 1980, pg. 81). Second, Strawson supposes that a criterion of individuation for some sortal amounts to a criterion for counting instances of it. But as Wiggins has argued, this is not necessarily so. For example, we can individuate instances of the sortal 'crown', such as the pope's crown, without being able to specify whether it should be counted as one crown or as three crowns (Wiggins 1980, pg. 73).

Further, borrowing from the terminology of personal identity, we might distinguish the criterion of individuation from evidence of individuation. The metaphysics of individuation is concerned with developing a criterion for individuating entities of a given type, while the epistemology of individuation is concerned with the evidence we adduce to individuate entities of a given type. And as with personal identity, the two notions come apart. Thus, good evidence of individuation may not satisfy criteria of individuation, and an appeal to criteria of individuation may not be necessary for evidence of individuation.

We can put the distinction between criteria and evidence to work in other ways as well. For Strawson claims that dualists cannot appeal to human beings (i.e. the Factual Person) to individuate minds. But it seems perfectly legitimate for the dualist to appeal to human beings as evidence of the individuation of minds, even if they cannot appeal to human beings, as Strawson does, as a criterion of the individuation of minds (since the dualist believes that the mind can exist without a body). And this fact too helps explain much ordinary talk about Factual Persons.

Finally, Strawson arguably confuses assertability conditions with truth conditions. For he repeatedly requires the dualist to show how he could know whether utterances that purport to pick out a particular mind are true, i.e. whether their assertability conditions are satisfied (Strawson 1966, pp. 174-176). But the intelligibility of such utterances -- which is what Strawson purports to be primarily concerned with -- depends rather on knowing what would have to obtain for such utterances to be true, i.e. on knowing their truth conditions. And since one can know the truth conditions of an utterance without being able to satisfy its assertability conditions, utterances about particular minds can be intelligible even if they fail to meet Strawson's demands. (For example, I can know perfectly well what the truth conditions of the utterance, 'I'm now speaking to President Obama by phone' are -- i.e. I can know what would have to obtain in order for the utterance to be true -- without knowing whether it's assertability conditions have been satisfied -- i.e. without knowing whether I'm in fact speaking to President Obama, as opposed to a clever impressionist.)

Having argued that Strawson's anti-dualism argument is unsuccessful, I shall now suggest two ways in which Cartesian minds might be individuated.

Descartes defined a mind as 'a thing that doubts, understands, affirms, denies is willing, is unwilling, and also imagines and has sensory perceptions' (Descartes 1985, pg. 19). Since we can conceive of something possessing these capacities while lacking a body, Cartesian minds are essentially immaterial (Descartes 1985, pg. 54). But we can conceive of two such minds with identical qualitative properties. How, then, could Cartesian minds be individuated?

Here is my first suggestion for individuating Cartesian minds. Suppose the human being Jones utters the Cogito -- 'I think, therefore I am'. A Cogito-type utterance cannot fail to refer, and it's necessarily true when uttered. Now let's grant Strawson's point that we cannot satisfy the assertability conditions of claims concerning whether Jones's utterance was a consequence of the activity of one or more than one mind. Does it follow that we cannot supply the truth conditions of such claims? It does not, for Cartesian minds, possessing as they do various capacities -- the capacity to affirm or will or perceive -- can thus be said to possess various dispositional or modal properties. So, if a particular mind causes Jones to utter a particular sentence, the effect will be the expression of a particular proposition. And since indexicals like the 'I' of the Cogito refer to the person uttering them, it's plausible to suppose that if two qualitatively identical minds cause the human being Jones to utter the Cogito, Jones will with one utterance expresses two distinct propositions. Whereas if one mind causes Jones to utter it, he expresses only one proposition. We may thus individuate Cartesian minds via some of their effects -- that is, via the modal property that distinct minds will express distinct propositions if they utter Cogito-type sentences.

One might object that this criterion is circular insofar as only distinct minds could, on my account, express distinct Cogito-style propositions. Hence, so the objection goes, the criterion presupposes that the minds involved in expressing those propositions have already been individuated. But it's important to keep in mind that I'm referring to dispositional or modal properties. Hence, it need not be the case that any such propositions have actually been expressed. So, Strawson's claim is that there is no way to distinguish (not epistemically, but logically) two qualitatively identical Cartesian minds. But my claim is that, if we grant (a) the suppositions above about how indexicals like 'I' function, (b) the dualist's claim that we can make sense of notion that 'I' is used to refer, at least some of the time, to the Essential Person, and (c) objects possess modal or dispositional properties, then there could not in principle be two qualitatively identical minds. For any two minds will necessarily possess the distinct modal property that Cogito-style utterances will express distinct -- indeed, singular (for each particular mind) -- propositions. And if this is right, then we can distinguish -- logically, if not epistemically -- Cartesian minds by way of these modal properties.

One may also object that (b) above is false, i.e. one might claim that the indexical 'I' refers to the human being, and cannot be used to refer to the mind alone (Strawson 1966, pg. 176). But we sometimes do meaningfully use the term 'I' in ways that cannot be understood to refer to ourselves qua human beings. If I say, 'I believe that I was Napoleon in a past life', I don't thereby assert that I believe that Napoleon possessed any of my current corporeal properties. Or, if a transgender person says, 'I was born with the wrong body', the term 'I' cannot refer to the human being as such, since it directly contrasts a specific body with the person referenced by the use of 'I'.

Here is my second suggestion for individuating Cartesian minds. Let's distinguish qualitative properties, such as redness, which two particulars could possess, from non-qualitative properties, which only one particular can possess. Nothing in Strawson's argument prohibits the dualist from appealing to non-qualitative properties, like 'being identical to oneself', to individuate minds. Or nothing in Strawson's argument would prevent the dualist from appealing to the bare particular, or substrate of each mind as an individuator of Cartesian minds. This is not to say that these alternatives succeed, for they raise many problems themselves (e.g. is there such a thing as the property of being self-identical? Can we make sense of a bare particular? And so on.) But the essential point is both that these are candidates for individuators, and nothing in Strawson's argument rules them out. Yet it seems his argument must rule them out if it's to show that dualism is unintelligible because minds cannot be individuated sans a reliance on human beings.

I shall now argue that the anti-dualist faces problems with individuating minds that are as serious as those the dualist faces. For one key part of Strawson's anti-dualism argument consists in his claim that the anti-dualist doesn't face an individuation problem. But is this right? Strawson claims that the anti-dualist individuates minds by individuating human persons (Strawson 1966, pg. 174). And how does the anti-dualist individuate persons? A Strawsonian person is a body to which both corporeal and mental predicates are applicable (Strawson 1959, pg. 104). Therefore, persons are first identified by the applicability of such predicates to particular bodies, and then individuated via their bodies.

But how are bodies individuated? By their location in 'a system of spatio-temporal relations' (Strawson 1959, pg. 25). But we can conceive of such a system that contains only two qualitatively identical objects -- such as Strawsonian persons -- which stand in perfectly symmetrical spatial relations to each other, and which thus cannot be individuated via location (Black 1952, pg. 156). Further, the location-solution merely pushes the problem of individuating bodies back a step, for now we're faced with the question of how to individuate spatial location, i.e. points in space (Taliaferro 1986, pg. 267). We cannot appeal to bodies to individuate points in space without circularity, since bodies will then be individuated by points in space, and points in space by bodies (Hoffman and Rosenkrantz 1991, pg. 189) And we cannot appeal to relations involving other points in space to individuate a particular point in space without begging the question, since doing so presumes the existence of distinct spatial points, which hence must already have been individuated (Moreland 1998, pg. 252). Thus, if we individuate minds by individuating human beings, and human beings by individuating bodies, and bodies by individuating points in space, then until we have a criterion for individuating points in space, we have no criterion for individuating minds. Therefore, if Cartesian minds are unintelligible because they lack a criterion of individuation, it follows by parity of reasoning that Strawsonian minds too are unintelligible.

A further problem with individuating persons via their bodies is that persons and bodies have distinct persistence conditions (Lowe 2000, pg. 16-18). Suppose each atom of my current body is replaced by atoms taken from a cloned body, and that each 'original' atom is destroyed as soon as it's replaced. When this process is complete, my 'original' body will no longer exist, with only the reassembled clone-body remaining. But it's plausible to suppose that the same person will persist through the atomic exchange; thus, the person can plausibly survive the destruction of the body. But suppose I die shortly after the transfer is complete. The person will no longer exist, but the reassembled body will nonetheless remain, and thus the body can survive the destruction of the person. But then specific bodies and specific persons don't necessarily track one another metaphysically. Hence, it's not clear that we can, as Strawson supposes, individuate persons criteriologically by individuating bodies (though we may do so evidentially).

I have argued that Strawson has failed to refute dualism. It doesn't follow, of course, that dualism is true, but only that Strawson hasn't provided us with good grounds for rejecting it. I then argued that we can conceive of ways in which Cartesian minds might be individuated. Finally, I argued that the problem of individuation is not an issue for the dualist alone, but afflicts the anti-Cartesian as well.

Descartes, R. The Philosophical Writings of Descartes, vol.2. translated by John Cottingham, Robert Stoothoff and Dugald Murdoch, Cambridge: Cambridge University Press, 1985

Hoffman, J. and Rosenkrantz, G., 1991 'Are Souls Unintelligible?', Philosophical Perspectives, Vol. 5, Philosophy of Religion: 183-212

Kripke, S., 1980, Naming and Necessity. Cambridge, MA: Harvard University Press

Lowe, E.J., 2000 An Introduction to the Philosophy of Mind, Cambridge: Cambridge University Press

Moreland, J.P., 1998 'Theories of Individuation: A Reconsideration of Bare Particulars', Pacific Philosophical Quarterly 79: 251-263

Strawson, P.F., 1992, Analysis and Metaphysics, Oxford: Oxford University Press.

1. I would like to thank Matthew Sims for his many helpful comments on an earlier draft of this essay, and Joanne Lovesey and Richard Chappell for comments on previous, shorter essays from which much of the content of this essay was drawn.

Metaphysicians have often tended to project their ethical ideals and intuitions into the ultimate structure of reality. This has led to the positing of numerous metaphysical entities whose nature is incoherent with known universal laws, and this is the usual defense made by reductionists in their approach to mind-body dualism. However, the variants between the corporeal and the incorporeal are not necessarily explicable in reductionist terms. In this paper, I will try to analyse and compare those ways in which consciousness is irreducible to, yet by necessity linked with its ontological physical correlates, with particular regard to recent theories of quantum consciousness.

With the birth of quantum mechanics, a notable challenge to the traditional idea of substance as mere tangible material has arisen from the Standard Model[1] of particle physics. Within the framework of quantum theory, the elemental units (quanta) that constitute matter are dissimilar to objects encountered macroscopically. When sufficiently isolated from the environment quanta may be viewed as waves. These quantum waves (wave-functions) are essentially waves of varying probability of locating a particle at some specific position, and the peak of the wave-function indicates the location with maximum probability of a particle being found at that point[2] in space-time. Given that a quantum event begins at a particular point in space-time, there is a certain point in space-time where the cycle is complete, and it is at that point the quantum entity loses its individuality within an entangled state to become some new entity. It is at that point that the quantum process recommences. The structure of matter therefore appears to endure through time by virtue of the cyclic nature of the quantum events that constitute its very existence and in this way, quantum events may therefore be seen as permanent possibilities of macroscopic objects.

Many will accept that the philosophical basis for consciousness should be explained by some form of materialism or physicalism. Classical scientific theories[3] of consciousness assume that it simply emerges from a deterministic process similar to a computational program that is presumed to take effect in the brains of living organisms. Though so far consciousness cannot be proved to be either purely physical or purely non-physical, and quantum theory, the nature of which I have briefly described, has perhaps allowed for a new concept of matter which may leave some room for better understanding consciousness. The tenets and findings in quantum theory have revealed a counter-intuitive realm in which the solidity of matter and the very distinction between viewing subject and viewed object somehow vanish. It has been suggested by Hameroff and Penrose, in their paper Orchestrated Reduction of Quantum Coherence in Brain Microtubules: a Model of Consciousness[4], that consciousness itself is a quantum phenomenon. In this paper it is proposed that the cytoskeleton[5] of a neuron contains a number of microtubules[6] that control the function of synapses, and that the state of consciousness is simply a manifestation of the quantum cytoskeletal state and its interplay between quantum and mechanical levels of activity (p476 of Mathematics and Computers in Simulation 40, Reduction of Quantum Coherence in Brain Microtubules: A Model for Consciousness, S. Hameroff, R. Penrose (1996)). The theory begins with 'subjective' and 'objective' reduction. The former occurs when an observer measures a quantity in a quantum system; that is, a system in which discrete units of energy in the realm of atomic and subatomic length scales are measured ('quantized'). The system is not in any specific state, but in a 'superposition' of possible states until it is observed, and the observation causes the system to reduce or 'collapse' to a specific state. Superpositioned states each have their own space-time geometry, and under special circumstances to which microtubules are suited, the separation of space-time geometry of these Superpositioned states[7] reaches a point where the system must choose one state. 'Objective' reduction, therefore, is a type of collapse of the wave-function that occurs when the system itself must select between significantly differing space-time geometries, owing to the tension in the fabric of space-time that is caused by mass-energy displacement between quantum states in superposition. This 'self-collapse' results in particular states that regulate neural processes, and these states can interact with neighbouring states to represent, propagate and process information. Each self-collapse corresponds to a discrete conscious event, and sequences of events then give rise to a 'stream' of consciousness. The proteins then somehow 'tune' this self-regulated objective reduction, which itself controls the operation of the brain through its effects on coherent flows inside microtubules of the cytoskeleton. Necessarily, there are two inferences to be made here. Firstly, it would appear from this proposed model that 'proto-conscious' information is encoded in space-time geometry at the fundamental Planck-scale process[8], to which our minds are able to tune in, thereby suggesting that there is an assortment of conscious states that exist in a world of their own. Secondly, it is thought that the collapse of the wave-function is what gives nature its seeming non-algorithmic essence. Penrose argues that quantum mechanisms are non-algorithmic and 'super-computational' and therefore may, if tapped into, provide a mechanism for understanding. This may be argued for as follows. There is an algorithm for determining the truth of a mathematical proposition. The algorithm must be reliable otherwise the verdict about the proposition cannot be precisely known. However, if it is consistent, the algorithm cannot by definition be applied to itself to establish whether it is accurate. The inference is that either we cannot know something is really the case, or the method employed to discover the fact cannot be verified. Penrose believes that our ability to know mathematical and indeed all truths is incontrovertible. In this manner he suggests that we know our understanding is correct and therefore we know something that cannot be known algorithmically (p. 129 of The Emperor's New Mind, R. Penrose, (1989)).

The non-computational aspect of consciousness is indeed consistent with the very nature of quantum mechanics and the profoundly twisted geometry of relativistic space-time, in which the volatile behaviour of sub-atomic particles is to be regarded as the essential reality of an intractable world; one which gives rise to all sorts of indeterminacies fundamental in nature. Nowhere else is this better underlined than by the principle of the Einstein-Podolsky-Rosen ('EPR') paradox[9]. Bohr's[10] response, in support of such a principle, was that quanta only display such effects when in contact with physical obstacles or receptors, constantly making their presence known in a way that could only be observed and recorded in the mind of the observer. Einstein found this explanation objectionable on the basis that it conflicted with the common sense assumption that sub-atomic particles are localised pieces of matter with specific, deterministic physical qualities; even if they cannot be directly observed through consciousness. If Bohr's explanation was wrong, the indeterminacy principle was wrong and the whole fabric of quantum theory would be brought into question. Yet, quantum experiments, continued to predict experimental results with great accuracy. This was indeed a paradox -- and one which, within our apparently material universe, only succeeds in highlighting the phenomenally subjective nature of consciousness as being indefinable in terms of consciously apprehended physical principles alone. Such observations are strikingly consistent with the 'Copenhagen interpretation'[11] of quantum mechanics where elementary phenomena are only phenomena if they are registered.

The Copenhagen interpretation also leads us to the problem as to how one is able to determine whether quanta, the building blocks of physical reality, exist before registering in the consciousness of an observer. Intuitively, it seems they must, but this is the same logic that led to the EPR paradox. The 'Schrodinger's Cat' paradox highlights the difficulty of this question. Schrodinger's cat is a famous illustration of the principle in quantum theory of superposition, proposed by Erwin Schrodinger in 1935. Schrodinger's cat serves to demonstrate the apparent conflict with what quantum theory tells us is true about the nature and behaviour of matter on the microscopic level and what we observe to be true about the nature and behaviour of matter on the macroscopic level -- everything visible to the unaided human eye. In this thought experiment, the cat's state (dead or alive) essentially depends upon the collapse of the wave-function to form a physical quantum from a decaying radioactive source in a box containing the cat, and a vial of poison that will be released when the quantum registers. If the cat, poison and radioactive source are all part of a quantum mechanical system, what must we say constitutes registration? The problem of when and how the quantum wave collapse occurs is even more critical if we accept Von Neumann's conclusion in his classic work The Mathematical Foundations of Quantum Mechanics[12], that no physical separation is possible between quantum systems and classical physical objects. If quanta do not exist until they register as effects on a receptor and we have no way of knowing of them until evidence of their effects is received in our consciousness via a chain of quanta and receptors, knowing whether they exist or not without the presence of consciousness is problematic. The difficulty here, both for the investigation of consciousness and for physics is that posed by conscious distinction itself, that is, the objectively physical is always outside consciousness and is always its object, thus rendering all mechanisms we posit to explain consciousness objects of consciousness.

It must be borne in mind that the orchestrated objective reduction theory does have its limitations. We may easily identify the correlation between an observed quantum process and the physical phenomenon of brain activity, and most often the macroscopic characteristics of physical objects and activity can be seen as generated from a microscopic account. However, it is so far not possible to establish how sensory awareness could be directly causally linked to physical brain activity. The primary reasons for this are twofold. Firstly, the notions of cause and effect are applicable only to observed phenomena within the spatio-temporal sphere, and secondly, scientists are simply confined to accounting for phenomena solely in terms of other phenomena, always leaving behind an indelible residue.

Given the forms of our cognition of phenomena are presupposed in all scientific enquiry, it cannot surely fall within the province of science alone to investigate their nature. Scientific concepts and manners of representation may, in fact, themselves be interpreted in a way that does not require us to expect that anything is really being spoken of apart from perceived facts and empirical abnormalities. Berkeley, concerned at all times that terms and expressions should be related to particular 'ideas' of sense, characterised purported natural laws as no more than 'rules... observed in the production of natural effects' (p. 25 of Principles of Human Knowledge (World's Classics), George Berkeley (1988)). Austrian physicist Ernst Mach denied that such laws could be spoken of as residing 'in nature'. Only individual instances or 'cases' of laws did this, the laws more appropriately being described as 'things of the thought'. (p. 12 of Mach's Principle: From Newton's Bucket to Quantum Gravity (Einstein Studies), Julian B. Barbour, H. Pfister (1995)). Likewise, Mach questioned whether we should treat the 'unobservables' as real existents, and when discussing atomic theory he described atoms as merely representing a mathematical 'model' so as to aid in the mental reproduction of the facts. For Mach it was therefore impossible to accept the principle of their actual existence since they could not be perceived by the senses.

Substance-dualist views posit that the mental and the physical are both real, insofar as matter, having the essential property of being spatially extended, is distinct from the mind, the essential property of which is having the capacity to think. Neither can be said to be assimilated into the other, and views of this nature are rooted in Descartes Meditations. Descartes' model views the material world as pure extension, which is infinitely divisible (Part II (IV) of The Principles of Philosophy, Descartes (2012)). In this model, different portions of matter move along at various speeds in relation to one to another, which is what differentiates things in the universe.

Similarly, contemporary quantum physicists also postulate a dual reality. At the ordinary level,the level at which we operate and experience events as discrete and concrete 'realities', realities that are quite sufficiently explained in terms of classical physics (inasmuch as matter does dissolve into wave in the quantum realm). But at that other level, the level of true reality, the submicroscopic level of noumenon as understood by the quantum physicists, all is waves of probability and propensity. Reality at the quantum or 'thing-in-itself' level seems to be nothing more than chaos of tendency to exist and survive patterns of probabilities.

Many attempts to extend materialism through such emergence theories usually follow the Aristotelian tradition, which is wholly distinct to that of the dualist Cartesian conception of the relation between matter and the incorporeal. For Aristotle, there was no exact science of matter and how substance behaved was essentially affected by its form. In this way, matter was perceived as determinable and made determinate by its form. The Aristotelian model thus has many forms and substances, which necessarily resulted in many levels of ontology (Part 5 of section 1 of The Categories, Aristotle).

Conversely, Locke provided a wholesale rejection of the Aristotelian concept of matter and notions of substance in particular, by replacing the notion of substance with corpuscularism. He had argued in his Essay Concerning Human Understanding that the mind, observing through its ideas that an amalgam of qualities occur in regular sequences, naturally tends to explain these sequences in terms of the intrinsic 'properties' that objects have as part of their real essences. By considering the sources of human knowledge, Locke contended that we must take into account the ways in which such factors as the operations of our sense-organs contribute towards the character of our perceptual experience, and, in reducing mental and cognitive processes to mere sensations, he contended that although things may appear to us in a certain way, the characteristics that we attribute to them do not really belong to them unless they are primary in nature (such as size or shape) (viii 9, Part II, An Essay Concerning Human Understanding, Locke). That is to say, the primary qualities of objects produce ideas that 'resemble' the corresponding qualities of the objects that caused us to have those ideas. The secondary qualities of objects instead produce ideas that do not resemble the corresponding qualities in the objects that gave rise to those ideas. We may take colour as a prime example. Colour perception is not created by a chromatically discriminative optical system, for when chromatic data is accessible to the cognitive processes it can be referenced against an innate typical matrix of subjective imagery, allowing the perception of colour to affect the experiential visual process. Similarly, as a result of the operations of the mind as an agency rather than a mere recipient of discrete sense-impressions, we perceive objects as having spatio-temporal qualities. The human mind, then, is ordered in determinate ways that we use to organise our perceptions.

But it must be remembered that references to objects and causal connections are only legitimate in the context of empirical or phenomenal reality as serving to define the phenomena of the world of sense. How is it possible to make such references in order to connect the phenomenal sphere itself with what lies outside it?

The quantum world is not a world of actual events like our own but a world full of unrealised potentia. What contemporary quantum physics seems to indicate is that far below the microscopic level, the world is not a world of spatially and temporally located particles of matter. Particle 'reality' is a subjective imposition that enters the picture only when an 'observer' enters. Unobserved, true reality, at its deepest level, consists only of possibilities and tendencies and what we take to be the attributes of 'matter' are really events that occur in the acts of observing and perceiving.

It is a combination of space, time, and causality that provides the natural structure for organising our experience. These abstract categories allow us to think relationally in the most rudimentary sense; understanding simple events require one to represent the spatial relations among objects, the relative durations of actions or movements, and links between causes and effects. It is by virtue of space, time and causality that reality can manifest itself to us as a world inhabited by a variety of individual objects. Our awareness of empirical reality, then, consists in the apprehension of ideas and representations which have as their basis the data provided by the senses that are structured in accordance with the universal framework of space, time and causality imposed by us as perceiving subjects. In this manner, it may be inferred that intellect and matter may be perceived as one and the same thing, seen at some transcendent, or profounder level of 'true' reality that is unfettered by time, space or causality. Mind and matter may ultimately be conceived as attributes of what is essentially one and the same 'substance', a substance that, together with mind and matter, makes up the totality of existence.

To expound, existence may be best characterised by three realms as opposed to a mere dualistic entity; namely by the physical realm of determinism, the mental realm of actuality and the realm of potentiality. It is the mental which governs itself in the image of the former and the latter, since it is itself non-computational. The quantum aspects of brain functioning to produce the mind may be applied through physics, and from the realm of potentiality we can infer the universal Platonic Idea which governs the classification of man's place in relation to nature. The mental provides the reference frame between both of these, which takes the form of 'knowledge' of the physical and the Platonic in its own perceptions. By revisiting the Penrose-Hameroff model of consciousness, it is apparent that the electrical firing in synapses is discontinuous and discrete whereas consciousness itself is self-evidently sequential and continuous. Thus, the basis of consciousness must lie in an external, constant reference frame which is receiving these electro-chemical impressions within the space-time surrounding their occurrence. One that governs, yet perfectly coincides with, the system of operation for the aggregate organisation of massive bodies as a self-referencing, closed unit.

Alexander, Peter, Ideas Qualities and Corpuscles, Cambridge University Press Ed 1985.

Aristotle, Prior Analytics, Hugh Tredennick (trans.), pp. 181-531 in Aristotle, Volume 1, Loeb Classical Library, William Heinemann, London Ed 1938.

Atmanspacher, H., and Fuchs, C., (eds.), The Pauli-Jung Conjecture and Its Impact Today, Exeter: Imprint Academic Ed 2014

Banks, Erik C. Machian Elements and Psychophysical Relations in S. Mori, et al. (eds.), Proceedings of the 23rd Int'l Conference in Psychophysics: International Society of Psychophysics Ed 2007.

Chalmers, David J., The Conscious Mind: In Search of a Fundamental Theory, New York: Oxford University Press Ed 1997.

Descartes, Meditations on First Philosophy, translated by John Cottingham, Cambridge: Cambridge University Press Ed 1996.

Hameroff, S.R., and Penrose, R., Conscious Events as Orchestrated Space-time Selections, Journal of Consciousness Studies, 3(1): 36-53 Ed 1996

Julian B. Barbour & H. Pfister, Mach's Principle: From Newton's Bucket to Quantum Gravity (Einstein Studies) Ed 1995.

S. Hameroff, R. Penrose, Reduction of Quantum Coherence in Brain Microtubules: A Model for Consciousness, Mathematics and Computers in Simulation 40 Ed 1996.

1. This is the name given to the theory of fundamental particles and how they interact with one another.

2. Precisely where a particle is and how it is moving when observed is still 'indeterminate' and results in random measured values.

8. A process whereby the sub-atomic particle interactions in terms of quantum field theory break down and become inadequate; concepts of size and distance vanish

9. This thought experiment was designed to exhibit the failure of the 'indeterminacy principle' characteristic of quantum mechanics, in the case of the creation of a pair of twin particles and subsequent discovery of certain physical characteristics of those particles at some distance from the point of their creation. The principle behind this thought experiment was that if the particles were physical entities the indeterminacy principle failed.

10. Niels Bohr, physicist (1885-1962), was the first to apply the quantum concept, which restricts the energy of a system to certain discrete values, to the problem of atomic and molecular structure. An account of the debates between Einstein and Niels Bohr have been written by Bohr in an article titled 'Discussions with Einstein on Epistemological Problems in Atomic Physics' (From Albert Einstein: Philosopher-Scientist (1949), publ. Cambridge University Press).

11. The 'Copenhagen interpretation' holds that quantum mechanics does not yield a description of an objective reality but only deals with probabilities of observance and measurement. Assuming wave-functions are not real, wave-function collapse is interpreted subjectively. The moment one observer measures the spin of one particle, he knows the spin of the other. However, another observer cannot benefit until the results of that measurement have been relayed to him, at less than or equal to the speed of light.

12. Von Neumann, J., The Mathematical Foundations of Quantum Mechanics (1995), Princeton University Press.

There are many sci-fi books, movies and TV series depicting conscious machines; artificial entities that are aware of themselves and their surroundings; machines that both think and emote. Could such conscious machines ever be actualized? If so, it seems that an explanation is required as of how consciousness could exist in an inorganic matter. Even before that, to appreciate what such possibility would involve, we have to deal with some fundamental questions: What is consciousness? How could we come to know that a machine is conscious?

It should be obvious by now that there is a relation between brain and mind. The most straightforward evidence of this is that people with damaged brain tissue, due to injuries or sickness, behave and experience the world differently than before. For example, those suffering from Alzheimer's disease, resulting from a degeneration of the brain, not only lose their memory but also exhibit mild to dramatic behavioural changes and loss of self-identity.

This being the case, most philosophers believe that something like the possession of a brain is a necessary but not sufficient condition for consciousness, i.e. without a brain, consciousness is not possible, but having a brain is not enough; there is an extra something that enables consciousness. The main discussion in philosophy of mind is what is exactly this 'extra' and how it can explain the relation of the brain to consciousness. While avoiding the trap of supplying an explicit definition of consciousness, a term which often proves elusive, I will say that when philosophers use the term consciousness the following two criteria are commonly referred to: having a first-person point of view and having subjective experience in the sense of 'what it is like' for me to be in this state. Over the course of what follows, I hope to clarify and expand upon these criteria by presenting some different theories of mind that do try to explain what consciousness is and how it can be accounted for. In the first section of this essay I will briefly cover some of the main theories in philosophy of mind regarding consciousness for this purpose. Following this, I will present an argument in favour of accepting consciousness as an emergent phenomenon. Lastly, I'll argue that if consciousness is an emergent property, then there is no principled reason why it should not emerge from a machine simulation.

Discussions of what consciousness is can be found in early Greek and Hindu philosophy, however, Descartes may have been the first in the Western philosophical tradition to frame the mind-body problem in a clear and reasoned way. Descartes' view is that consciousness (mind) is immaterial and somehow interacts with the matter (brain) and thus, the material brain and immaterial mind are two ontologically distinct substances. This thesis is what has become known in the literature as Cartesian dualism. Today, only few philosophers or scientists support Cartesian substance dualism.

In contrast with dualism, physicalism, a form of monism, holds that there is nothing beyond physical matter, and hence consciousness is a result of brain activity. We may not yet have the correct scientific explanation of how consciousness arises from the brain but such an explanation involves nothing beyond the world of physics. Physicalists usually explain consciousness via supervenience -- the idea that psychological states supervene on physical states; if two persons are indistinguishable in all of their physical properties, they must also be indistinguishable in all of their mental properties. The same goes for computer running software, if my computer runs solitaire, then it is impossible for any other computer with exactly the same electronic states not to be running solitaire.

Thomas Nagel (1974) introduced the following argument against physicalism: physical information doesn't tell you: what it is like to be a bat. How the world is from another point of view. Nagel: 'An organism has conscious mental states if and only if there is something it is like to be that organism -- something it is like for the organism.'[1] If we'll consider pain vs. C-fibre firing, according to Nagel, there is no such thing as 'pain in itself' (C-fibre firing) as an objective experience, there's only how pain strikes me -- a subjective experience.

Another strong argument against physicalism is the knowledge argument, where Frank Jackson (1982, 1986) introduces Mary -- a brilliant scientist who learns all the physical truths about the world, vision and colour from a black and white room[2]. The argument posits that when Mary leaves her B&W room and sees a red tomato, she learns a new fact -- what it is like to see red. This occurs as a result of her having a new subjective experience by which she becomes familiar with qualia, the intrinsic qualities of phenomenal experience; in this case redness. The conclusion of the argument is that physicalism is false. While the gist of Nagel's argument is that one can have all the physical facts without having knowledge of the other persons' view, the point of Jackson's is that one can have all the physical facts without having knowledge of qualia.

Several other arguments that support the existence of qualia include the conceivability argument, inverted qualia, philosophical zombies and the explanatory argument. Chalmers (2003) says that these arguments are part of a general problem, exemplified in what has become known as the explanatory gap; Chalmers named this the 'hard problem' of consciousness -- the problem of explaining how qualia or subjective experience emerges in our minds[3].

Physicalists have not remained silent to these objections and have in response suggested several counter arguments. One such counter argument, the ability hypothesis (Nemirow & Lewis) argues that rather than having learned new facts or truths, Mary acquires new abilities; She gains know-how not knowing-that[4]. Another response emphasizes the possibility of Mary acquiring 'Knowledge by acquaintance'.[5] On this view, it is held that Mary becomes directly acquainted with the phenomenal character of colour experience, in the way that one can become acquainted with a city by visiting it.

Another view adopted by both Patricia and Paul Churchland is that when it is said that consciousness emerges from the brain there is a suggestion that something else besides the neural activity is going on, something correlated but distinct from the neural activity. But then again, what is that extra thing that happens? According to this view, all questions about consciousness can be reduced to what Chalmers calls the 'easy problems' and eventually be solved. Put otherwise, the concepts of popular psychology that we use to explain our mental states (intentions, beliefs, desires, etc.) will eventually be replaced by neurobiological models that have yet to be developed. According to Patricia Churchland, the fact that it is currently very hard for us to imagine a solution to the problem of consciousness tells us absolutely nothing about whether or not this phenomenon can actually be explained. In her view, it is too easy to conclude that a phenomenon such as consciousness is inexplicable simply because current human psychology cannot grasp it.[6]

The knowledge argument continues to inspire ongoing discussion on the nature of consciousness and its relation to the physical world. The main discussion point being, how subjective experience and mental states arise from brain states.

Sometimes a system with multiple interacting components gives rise to some surprising and unpredicted dynamics that cannot be found or predicted by looking at any of the components in isolation; such an emergent phenomena is not a priori predictable from its substrata and none of the components share the property the system at large holds. The explanation for emergent phenomena takes place at another level distinct from its substrata.

Like water (H2O) has the novel properties of wetness and liquidity that cannot be found in either Hydrogen or Oxygen, by analogy mental states arise from brain states but do not share their properties, mental states are not identical to any brain states but instead emerge from them. If the later is the case, then it offers an account as to why qualia cannot be reduced to any particular physical substratum.

Emergent properties are properties of a system and dependent on that system's components, their properties, and configurations; emergence arises when the system in question passes a critical threshold of complexity and organization. Philosophers who support emergentism, view emergence to be compatible with physicalism in the sense that the universe is made exclusively of physical entities while at the same time rejecting the reducibility of the mental to the physical. Moreover, it is important to note that the truth of emergentism is consistent with the falsity of substance dualism (in the Cartesian sense).

David Chalmers (2006) defined a distinction between weak and strong emergence in order to capture the difference usage of the term 'emergence' in science and philosophy.

Strong Emergence: 'We can say that a high-level phenomenon is strongly emergent with respect to a low-level domain when the high-level phenomenon arises from the low-level domain, but truths concerning that phenomenon are not deducible even in principle from truths in the low-level domain. Strong emergence is the notion of emergence that is most common in philosophical discussions of emergence, and is the notion invoked by the British emergentists of the 1920s'[7]

Weak Emergence: 'We can say that a high-level phenomenon is weakly emergent with respect to a low-level domain when the high-level phenomenon arises from the low-level domain, but truths concerning that phenomenon are unexpected given the principles governing the low-level domain. Weak emergence is the notion of emergence that is most common in recent scientific discussions of emergence, and is the notion that is typically invoked by proponents of emergence in complex systems theory.'[8]

While we can usually say that instances of strong emergence are also instances of weak emergence -- the phenomenon is not deducible and unexpected -- the converse, however, does not hold.

Weak emergence is common in complex system theories and other scientific fields where complexity, self-organization, functional organization and system's behaviour are paramount.

Let's look at an ant colony as an example of weak emergence. The ant colony exhibits complex behaviour -- food is gathered, tunnels are excavated, waste discarded and the colony is protected and populated -- without any centralized decision making (the queen does not give direct orders). All of these processes are successfully carried out despite the fact that each individual ant acts independently; only communicating with other ants by leaving chemical traces that are picked up and acted upon by other colony members.

An objection to strong emergence -- mainly raised by physicalists -- is that even if strong emergence is possible there are no real cases of it in the world. In response, the proponents point to mental properties. Chalmers (2006) considers consciousness as the only intrinsic example of strong emergence in nature. If consciousness can be considered generally to be quite a unique state, which it is, then its uniqueness lends to the idea that strong emergence is something that is also unique.

When considering the behaviour of an ant colony, although we can analyse the behaviour of a specific ant or we can analyse the behaviour of the entire colony, the mechanism that gives rise to the organization and behaviour of the colony as a whole eludes us. Similarly, the use of fMRI (functional magnetic resonance imaging) and other tools enable scientists to explore the behaviour of neurons and their functions, researching the neurophysiological structure of the brain. However this may be, the mechanism that gives rise to consciousness from brain matter eludes us. Emergence gives us the best explanation to date of that 'mechanism' in its claiming that multiple autonomous agents (ants, neurons) interact and communicate with other local agents or occasionally remote ones, and that this results in a system's behaviour which itself cannot be reduced to the individual agents.

In response to this, I concede that biology and neuroscience may not yet have the means to explain the mechanism(s) underlying emergence, but just because it is presently inexplicable, does not mean that emergentism is a false theory or that it will always be inexplicable.

Patricia Churchland presents a view that in the history of science many problems seemed inexplicable but eventually with the progress of science an explanation was found.[9] She makes a distinction between 'we cannot now explain' and 'we can never explain'. At this point in time our understanding of the mind is still immature and we should not speculate about consciousness as a 'hard problem'. 'When not much is known about a topic, don't take terribly seriously someone else's heartfelt conviction about what problems are scientifically tractable. Learn the science, do the science, and see what happens'.[10]

While there are several on-going projects around the world researching the brain, one in particular -- the European Human Brain project (HBP)[11] -- has brain simulation as one of its objectives: 'Simulate the brain -- Develop ICT tools to generate high-fidelity digital reconstructions and simulations of the mouse brain, and ultimately the human brain.'[12] HBP supervisor, Prof. Henri Markram, believes this can be achieved by 2023.

Suppose then that consciousness is an emergent property of a physical system (the brain). If this is the case, then would our creating an accurate simulation of a human brain in another platform, say equivalent with all the neurons, synapses and the intricate and complex activity of the brain, bring about the sufficient conditions for consciousness to emerge?

Before going on, there is an important distinction to be made here in regard to artificial intelligence and artificial consciousness. It is not the purpose of this essay to explore the question of whether or not consciousness can be created artificially in some machine (machine consciousness) or whether an artificial intelligence (AI) can exhibit conscious behaviour, but rather to explore the possibility of consciousness emerging in a machine that is a brain simulation.

With this in mind, let us approach the question above. While it seems obvious that the brain is a necessary requirement for having consciousness, is it sufficient? Some philosophers believe that aspects of the agent's body (other than the brain) are constitutive for cognitive processing. This view is termed embodied cognition and the following examples[13] are representatives:

We typically gesture when we speak to one another, and gesturing facilitates not just communication but language processing itself (McNeill 1992).

Vision is often action-guiding, and bodily movement and the feedback it generates are more tightly integrated into at least some visual processing than has been anticipated by traditional models of vision (O'Regan and Noe 2001).

There are neurons, mirror neurons that fire not only when we undertake an action, but do so when we observe others undertaking the same actions (Rizolatti and Craighero 2004).

We are often able to perform cognitive tasks, such as remembering, more effectively by using our bodies and even parts of our surrounding environments to off-load storage and simplify the nature of the cognitive processing (Donald 1991).

The brain simulation should have input and output capabilities for interacting with the world. Sally Goerner and Leslie Allan Combs write: 'Consciousness always has an object. In other words, it is always about something. We are not just conscious, we are conscious of the taste of food, the smell of the sea, a tooth ache. We are conscious of joy, of boredom, of the meaning of words on the page in front of us, of the sound of music playing in the next room, of our own thoughts, of memories. The point is that virtually all experience is experience of something... Consciousness would seem to be intimately involved with the informing of the brain and mind by objects of attention.'[14]

I would concede that input and output mechanisms are needed as causal processes -- sounds, sights, etc. -- but these mechanisms are not part of a conscious states. When we see a sunset or listen to music, a chain of events occurs; one starting with the activation of receptors and proceeding subsequently to the transduction of proximal stimuli, neurons firing in the brain, integration with memory, attachment of emotion and meaning and so on. Somewhere along this process, consciousness emerges, but the stimuli (certain wavelength in the case of the sunset and series of sounds in the case of music) are not those things within which a state occurs or those things which it is constituted of; a conscious state -- the experience of seeing a sunset or listening to music -- emerges within this process.

The question of how the brain gives rise to consciousness is the key question and science has yet to identify the mechanism that is involved. This does not mean that the explanation is not 'out there' to be discovered; as Patricia Churchland prompts us to go do the science and let the future be the judge of it. Notwithstanding, the best explanation so far is given by emergence; consciousness arises from the brain via emergent property. It follows that if a simulation of the brain can be created in a machine, it is a possibility that consciousness could emerge. I acknowledge of course that it is not as simple as that, as proponents of embodied cognition are quick to point out; being that mere brain simulation is not enough, we should simulate sense organs and the simulation should be able to interact with the world. Today we do not have the technology at our disposal to create a brain simulation of small mammal, let alone a human brain with around 100 billion neurons -- the HBP is extremely ambitious in its timeline objective. But if the technology were to become available and it would allows us to build a simulation in a machine, another difficulty arises. How would we ever know that it was conscious?

As there is no 'test for consciousness', there is no method of coming to know with certainty whether other people share the same conscious states that we experience; we have a sense of self and subjective experience and we assume that other people share this experience and we believe them when they say they do. So even if a brain simulation were eventually produced and it told us that it is aware of itself, feels and experiences 'what it's like', our knowing whether or not it would qualify as conscious would be questionable at most.

So how can we test if a machine has qualia if there is no test available for knowing of the presence of consciousness in other people?

So long is there is no formal test, I would suggest that the same practice of belief that we extend to other people should be extended to the machine simulation. If its behaviour is indistinguishable from that of (what we assume to be) a conscious person and it will report to us about its subjective experience we should believe it.

'When we understand consciousness -- when there is no more mystery -- consciousness will be different, but there will still be beauty, and more room than ever for awe.' -- Dennett, Consciousness Explained (1991)

Chalmers D., 'Facing Up to the Problem of Consciousness', Journal of Consciousness Studies, 1995, pp. 200-219

Chalmers D., 'Strong and weak emergence', in P. Davies & P. Clayton (eds.) 'The re-emergence of emergence', Oxford University Press, 2006

Churchland Patricia, 'The Hornswoggle Problem', Journal of Consciousness Studies 3, 1996, pp. 402-8

Conee E., 'Phenomenal Knowledge', Australasian Journal of Philosophy, 1994, pp. 136-150

Jackson F., 'Epiphenomenal Qualia', Philosophical Quarterly 32, 1982, pp. 127-136

Lewis D., 'What Experience Teaches', In William G. Lycan (ed.), Mind and Cognition, 1990, pp. 29--57

Nagel T., 'What is it Like to Be a Bat?', In Philosophical Review 83 (October), 1974, pp. 435-50

Wilson, Robert A. and Foglia, Lucia, 'Embodied Cognition',The Stanford Encyclopedia of Philosophy(Fall 2011 Edition), Edward N. Zalta(ed.)

Sally Goerner and Allan Combs, 'Consciousness as a self-organizing process: an ecological perspective', 1998, pp. 123-127

6. Patricia Churchland, 'Chalmers' Zombies and The Hornswoggle Problem', article, 2003

7. Chalmers, 'Strong and weak emergence', in 'The re-emergence of emergence', 2006

13. Wilson, Robert A. and Foglia, Lucia, 'Embodied Cognition',The Stanford Encyclopedia of Philosophy(Fall 2011 Edition), Edward N. Zalta (ed.)

14. Sally Goerner and Allan Combs, 'Consciousness as a self-organizing process: an ecological perspective', 1998, pp. 123-127

IV. 'LIMITING THE SELF -- EXTENDED COGNITION AND STANDING STATES' BY MATTHEW SIMS

In the last decade, the hypothesis of situated cognition has received a fair amount of attention. This hypothesis, in short, claims that although the brain is necessary with respect to the obtaining of various cognitive processes, it is not sufficient for many cognitive processes to obtain. Given this superficial description, one may certainly feel, at least one should feel, it is a quite trivial hypothesis. It needn't be a point of contention for most that a cognitive process, say of visually perceiving some object, not only involves the brain but also involves the use of some sensory organs such as the eyes and the perception of some distal object. Nor would it ruffle many feathers to claim that the content of many of our beliefs are is the result a process that involves not only our sensory organs, which present distal information about the world, but the world itself as being the object of our attitudes and veridical perceptions. All superficialities aside, situated cognition is to correctly be understood as making a claim that is anything other than trivial. As presented within the framework of one strain of situated cognition, the extend mind hypothesis, EMH, the claim is a constitutive one; there are events or objects additional to the brain which figure in constitutively to the obtaining of cognitive states. Such states might be those of involved in perceiving, problem solving, language acquisition and use, memory recall, and belief and desire formation. Andy Clark and David Chalmers, two of the pioneers of EMH, constrain the types of cognitive states that can be extended to those that are standing -- those states such as 'beliefs embedded in memory' (Clark and Chalmers, p. 33); in short, those beliefs and desires which are not at t being entertained but are nonetheless a part of one's mental makeup.

The purpose of this essay is to focus on something that can equally be regarded as a vexing explanatory challenge brought about by and a difficult objection to the EMH; a difficulty which, even if one were to accept EMH, would have to be faced regardless of its being found true. It can be summed up in asking this: 'If an agent's cognition were to be extended, where does the agent's self-identity stop? Is she merely her biological body or is she her biological body and the extension?' The latter brings about the unintuitive result that she can exist in two distinct, non-overlapping spatial regions simultaneously. This difficulty is brought to light and exploited by Brie Gertler in her paper 'Overextending the Mind'. In this essay, I will present Gertler's argument, showing how it exemplifies the problem of extended self-identity. Then I will augment it in such a way as to transform it into a case where our intuitions about moral accountability are enlisted. By nudging the problem into a moral framework, I hope to sharpen our intuitions about self-identity and as a result clearly demarcate where it is that one begins and where that same person ends in extended cases. After this I will introduce a necessary condition for being a standing state, along with a further necessary condition for being an extended cognitive state. I will then argue that because the problematic cases of extended identity do satisfy the former condition but fail to satisfy the later, the unintuitive conclusions that follow from both of the presented arguments against EMH cannot be pushed through. Therefore, we are not forced to choose between the correctness of EMH or our intuitions about the delimitations of the self.

Some preliminary remarks are needed before turning to Gertler's argument. Being that her argument begins by assuming the truth of EMH as favored by Clark and Chalmers, henceforth referred to as C&C, in their paper 'The Extended Mind', it would be beneficial to understand the rationale behind C&C's argument in favor of the possibility of extended minds. In their paper C&C invite us to imagine a person Otto, who failing to have a functioning standing memory due to Alzheimer's disease, enters all of his trusted beliefs into a notebook. In circumstances that require access to his standing beliefs and desires, it is to this he refers. The force of their parity argument might be said to be this: if the extension functions in a way that is similar enough to something which we would have no problem accepting as a constituent of a cognitive state were it cranially contained, then on pains of drawing arbitrary distinctions, we should have no problem in accepting something which extends outside the head as such a constituent. Thus it argued that Otto's journal, playing the same cognitive role as a normally functioning person's standing beliefs do, counts as a part of Otto's mind -- his standing memory to be sure. Additionally C&C put forth a short list of necessary conditions for being a standing belief. They include: being easily accessible, being constantly available, and being subject to automatic endorsement.[1] From the assumption that one's standing beliefs are constituents of one's mind and that because Otto's standing beliefs are in his notebook, the conclusion is drawn that it is possible that the one's mind can extend out into the world beyond the skin.

Gertler's argument, what I will refer to henceforth as the argument from overextension (AFO), takes the form of a reductio ad absurdum. Assuming the truth of EMH with respect to standing states, the AFO goes on to show that this assumption leads to an unintuitive result; one that is not only incompatible with our common concept of the self but also conflicts directly with a widely accepted notion in physics -- nothing can occupy spatially distinct and non-overlapping locations simultaneously.

ii. If Otto's journal is as a constituent of his mind, then on pains of arbitrariness, a robot, whose 'action plan' program is informed exclusively by an input system that detects the environmental layout and all of Otto's uploaded notebook entries and who satisfies C&C's standing belief criteria should also count as being a constituent of Otto's mind.

iv. If a constituent of X's mind produced some token action, then X produced that token action. (assumption)

v. If one day, while Otto were sleeping, the robot went out and performed an action based entirely on those standing states, say go grocery shopping and thereafter bake banana bread, then because those beliefs and desires are sufficient for explaining the action and they are constituents of Otto's mind, then those actions would be Otto's actions. (iii. and iv.)

vi. If X performs a token action and Y performs a numerically identical token action, then X is identical to Y. (Principle of Token Action-Identity)

vii. Otto occupies two non-overlapping spatial locations at the same time; both sleeping in one place and say shopping in another place.

viii. No one thing can be in two distinct and non-overlapping places at one time

If this analysis represents Gertler's argument correctly, it becomes apparent that in order for it to get off the ground, premises iii. and iv. are must be accepted. On a first pass, premise iii. seems to be a kind of Humean truism about motivation.[2] For instance, say I had the fleeting occurrent belief that 'I should eat something' and my coming to believe this was the result of previously having a painful feeling in my stomach. Addition to these, say I also had an occurrent desire to eat an apple located in my backpack -- a desire that I still have but one which I am not aware of any longer. Now, given these standing beliefs and desires (and a multitude of other background beliefs and desires) is my reaching into my backpack for the apple sufficiently explained? Well C&C seem to think so. In their original argument, they claim that Otto's walking to the MoMA is sufficiently explained by merely referring to the entries in his notebook. If this is the case, then my reaching for the apple can also be sufficiently explained by my standing beliefs and dispositions. Whether or not this is really the case, I will take up in the next section. However, let's continue to analyze the argument in hope of understanding what leads to the unpalatable conclusion that Gertler draws.

Premise iv. does not require a stretch of imagination for one to admit to its plausibility. If some action is caused exclusively by a set of beliefs and desires, where beliefs and desires are understood as constituents one's mind, then the agent from which those beliefs and desires would spring, would also be the cause of that action; it is her mind responsible for the action and thus it is her action.

Continuing, premise v. follows from premises iii. and iv., while premise vii. follows from the acceptance of what I've called the Principle of Token Action-Identity stated in premise vi.; a principle which I think most would take to be obvious. If the argument is valid, the result commits us to the conclusion that Otto exists in two distinct non-overlapping regions of space simultaneously. That this result is inconsistent with the laws of physics seems to be a good enough reason to look else where within the argument to locate a faulty premise(s).[3] Gertler, herself, responds to the problem by denying the mind is constituted by standing states and argues instead that occurrent states demarcate the mental. Doing this, she is able to accept the contingent possibility that the mind may be extended in the way that C&C have formulated. The mind as she understands it, however, falls under a narrower concept. While Gertler's solution may or may not persuade for a number of reasons, taking issue with it is beyond the scope of this essay.

What should be gleaned from the absurd conclusion that follows from accepting all AFO's premises? One thing is that it reveals a potential problem with the extended hypothesis; if we allow for objects laying beyond the skin to be extended constituents of standing cognitive state, objects that are spatially distinct from those agents whom they are extensions of, then it seems as if we must also extend the notion of the self as far as those extensions reach in order to be consistent with plausible notions about motivation, action, and agency. Now whether or not this is considered to be a problem or merely a challenge is dependent upon one's pre-existing notion of the self and its peripheral notions. If one such peripheral notion becomes unintelligible as a result of a spreading out of the self, and that notion is essential to practical and normative reasoning, then what seemed a challenge is revealed to be nothing less than a problem.

To illustrate the weight of such a problem, I would like to present an argument that is based upon the AFO, what I call the argument from extended accountability (AFEA). It is by means of it, that I hope to engage our intuitions about how the acceptance of EHM seems to have intractable results concerning the notion of the morally accountable self.

i. Otto's journal is a constituent of his mind. (assumption -- the truth of EMH)

iv. If a constituent of X's mind produced some token action, then X produced that token action. (assumption)

v. One day Otto, after having seen his neighbor abuse his dog, thought to himself, 'People like that shouldn't be allowed to live.' He also for a moment felt a fleeting desire to harm his neighbor. Rather on act on it, Otto went to sleep. If, while Otto was sleeping, the robot murdered the neighbor based entirely on Otto's standing states, then because those beliefs and desires are sufficient for explaining the action and they are constituents of Otto's mind, then those actions would be Otto's actions. (iii. and iv.)

vi. If X performs a token action and Y performs a numerically identical token action, then X is identical to Y. (Principle of Token Action-Identity)

ix. Otto is morally accountable for committing the murder (vii., viii. and Leibniz's Indiscernibility of Identicals)

One question that should immediately arise in response to this argument is 'how large of a role should we allow our moral intuitions to play?' Unlike the AFO, the AFEA does not end in a metaphysical inconsistency but in a distinct kind of uneasy moral feeling that 'certainly there is something that sets Otto apart from being accountable for the murder, even if it can be explained exhaustively in virtue of his standing states.' In the following section I will make an effort to locate the reason as to why our intuitions, under the assumption that they are relevant enough to matter, are correct. If our intuitions that Otto is not morally accountable can be supported, then given the strength of this support, a plausible constraint on sameness of personal identity with respect to extended standing systems may appear in the clearing.

A proponent of EMH might waste little time in denying that our intuitions are as the conclusion says they are or she might claim that if they are so, then there is good reason to believe these intuitions to be mistaken. In process, she may ask us to consider an analogy between this kind of morally extended robot case and a case of extended moral accountability of parents with respect to their children. Imagine that someone's child steals something from a store. It would seem certain that upon getting caught in the act, the child AND the parent would be held accountable. The parent as legal guardian is responsible for her/ his child's wellbeing and wellbeing can easily be construed as referring to actively constraining the child's behavior such as to not flout the social laws as well as not allowing situations to occur which might endanger the child's safety. If this is plausible, then any intuition about Otto not being accountable for the behavior of his robot is just wrong; particularly considering that it is much like a child, who after having acquired many of the beliefs and desires of its parent, carries out some action as a result of having those very beliefs.

This kind of response gets something very correct and something very incorrect. The good news first is that, yes, those who do have children are likely to take responsibility for their children's actions and are often held accountable for those actions -- particularly when they reflect the desires of parent. Now for the bad news -- the argument presented makes specific use a case of murder for the purposes of pushing our moral intuitions as far as they can go. Would we hold a parent accountable for their child committing a murder? An answer to this question may begin by mentioning that accountability is gradient and that given the reality of such a case, we may not hold the parent fully accountable but only partially so. However, this is distinct from Otto's case in that we, accepting certain premises, are led to hold Otto fully accountable given the fact that the robot's standing states are exclusively Otto's standing states and it was these states that exclusively led to the murder. Such standing state exclusivity, in the case of a murderous child seem to require nothing less that the child being some kind of robot.[4] This, of course, is not to deny the influence of a parent's beliefs and desires on her/ his child. What it is denying, however, is the analogy's force given the distinct natures of the two cases. If the analogy is found wanting, then it is simply not a plausible way to disentangle ourselves from the unintuitive result of accepting all of the premises that led up to the conclusion. This is not to say that EMH is the culprit but only that until the other premise are found false, it remains to be suspect.

Why is it that we have the intuitions that we do in the case as described? A good a place as any to approach finding an answer may be questioning C&C's conditions for standing beliefs and desires. In hopes of illustrating that they are not restrictive enough, I invite the reader to imagine an individual, Bill, who has a functioning memory and thus has accessible standing beliefs, and yet Bill by some fluke of nature cannot modify or correct his beliefs when presented with new conflicting evidence. In such case, would we really be willing to grant Bill, although having states that are accessible, consistently available and subject to automatic endorsement, the possession of standing beliefs? Let's advance a negative answer to this question for sake of seeing where it may lead us in our enquiry. Thus one might reply, 'If there were such states, whatever kinds of states these would be, their function would be far from those of standing beliefs and desires!'

Before pursuing this, more needs to be said about the notion of modification. I purpose that the modification of a standing state may best be understood in respect to its membership status, i.e. whether or not it belongs to a total coherent set of beliefs and desires. My token belief that 'it's raining outside', after looking out the window and seeing no rain, is not modified so as to be the same belief negated. The new belief that 'it's not raining outside' is a different belief altogether! My previous belief is modified in the sense that it becomes a member of the set of beliefs that I don't assent to (although I may have access to). Its content, however, remains the same. Now if Bill were come to have the fleeting belief that 'its raining outside' and then minutes later look out the window to see no rain, he, unable to modify his standing belief, would still assent to it and simultaneously come to have the new belief that 'it's not raining'. The reader can surely see where this is going; his new belief, although true and justified, would be overridden by the immutable standing belief 'it's raining outside'. If this is true, Bill would always, and not as a result of his own will but by disposition alone, believe that it's raining outside. So one might argue that such a 'belief' fails to function in a way that would be beneficial to him and thus should not be considered to be a standing belief. This person may also suggest an addition to C&C's standing state conditions such as the following. Let's call this the rational coherence condition (RCC):

RCC: Where X is a belief or desire, and S is a system, for all X, if X is standing cognitive state in a system, S, then X must be subject to potential modification by S given the presence of new information and its rational coherence with the rest of S's background standing states.

When returning to C&C's original example with Otto's notebook, assuming something like the RCC is correct, we can see that Otto's notebook, although it is located beyond the skin, is subject to modification given new and more coherent information. If Otto, while underway to the MoMA, sees that the museum has moved, he is able to modify the notebook entry given this new information. Moving on to the AFO, however, it seems that the RCC is not satisfied in the case as described. When Otto falls asleep, his potential ability for altering his desires and beliefs given new information is put on hold. This in itself is not problematic, given that the constraint's 'potential' qualification leaves room for a sleeping agents standing beliefs to still qualify as such. The problem is that there are some beliefs that Otto does not have but would have if introduced to new information -- beliefs which could possibly affect the justificatory standing of his standing beliefs. However, given the robots informational input system, its set of standing states, and an action program, it fails in a way in which Otto would succeed. Were Otto at a grocery store in the banana aisle and had smelled the bananas himself, he might have had a change of mind -- be it that the smell might have been so overwhelming that his desire for anything to do with bananas is overridden by a want for fresh and odorless air. The robot, as described cannot modify its 'desires' and 'beliefs' in such a way. Its states, or whatever they are, fail to satisfy RCC and thus fail to count as standing states.

This kind of reasoning I think is mistaken for two reasons. The first reason is that RCC is not a necessary condition for having standing states such as beliefs or desires. A further example may make this clear. Imagine now that Bill's brain damage causes him to see and hear an imaginary friend, Anna and nothing anyone could say or do, no amount of inconsistent evidence or new occurrent beliefs are able to unhinge or make him give up his assent to 'Anna exists'. If, however, Bill's behavior can be best explained in terms of him having the belief that 'Anna exists', then surely it would be correct to accredit him with such a belief even if it is immutable.[5] If this were the case, then to require the RCC in addition to those of C&C's necessary conditions for something qualifying as a standing belief would be a mistake. At most, the RRC would be something that could constrain the kind of beliefs that qualify as rational beliefs, but this, I trust, is too trivial to be of any use. Although, for most, the fact that the RCC is not necessary should be sufficient evidence to justify our completely turning our gaze from it, by not abandoning it straight away, the second reason as to why a search for necessary conditions on standing states is a misplaced effort becomes clear. It is only when this mistake is clearly cast that another more plausible condition may expose itself to us. This is what I will now attempt to illustrate.

Assuming that the RCC were necessary for standing beliefs, its being so wouldn't do the work needed of it. The reason being is that given the complete set of Otto's standing beliefs and desires, the robot would be able to surely satisfy the RCC. It would simply, given that set, infer how to modify the belief or desire. If it has the standing conditional belief that 'if bananas are overly ripe, then they are not to be purchased', then this belief would override the standing desire for banana bread and the robot might simple return home empty-handed as Otto would have. This brings out a striking point about beliefs and desires. Beliefs and desires are not just about overly ripe bananas or banana bread, they are much of the time about 'those horrible smelling overripe bananas' or 'this mouthwatering smelling banana bread'. The intentional content of standing states are often determined in part by phenomenal content of occurrent states. To be sure of this, just look at the computer screen before you. To limit your experience to 'there is an illuminated whitish screen peppered with black letters' would be to pilfer the very feel of your experience for sake of perhaps easily communicating something about it. Your experience presents something present rather than descriptive or conceptual. 'THIS!' This phenomenal content is demonstrative in that it refers to particular qualities as presented by your subjective occurrent states. So, given the input of overly ripe bananas, even if the robot's set of standing states might be able to be used to inferentially deduce a response similar to one that Otto might have, unless its encoding of phenomenal content is identical to the way in which Otto would encode it, what at most could be said is that it satisfies C&C's conditions and the RCC but in way that is distinct from the way Otto would satisfy it. Moreover, the robot as described fails to have phenomenal content given that it fails the capacity to have occurrent states. This last point carries over to the AFEA also. Although Otto's robot may contain a set of beliefs and desires which were had at one point by Otto, because Otto is absent at the scene of the murder, there is no possibility for the non-occurrent robot to have exactly the same kind of new input which is necessary for identical kinds of belief and desire modification. If this the case, then it fails to say that the difference between Otto's input + phenomenal content and the robots input + action program would result in a difference in the way that the set of standing states are modified. This point can be emphasized further by looking at a quite plausible general condition for successful extension of standing states. What I'll for simplicity sake just call the GC.

GC: If an agent's, A's, standing states are cognitively extended via extension, E, then it must be the case that E represents A's motivational set such that that set will supervene on A's occurrent states.[6]

This is just to say that there is no change in the robot's M-set without a change in Otto's occurrent states. Why would a supervenience relation be relevant here? For one, in order for something to be someone's M-set, whether or not an extended someone, that person's M-set should be acknowledged as being dependent upon her/ his conscious states. To imagine a case where this does not hold is to imagine a case where one's beliefs and desires are not dependent upon one's own conscious experiences; thus it unintelligible to even call them one's own beliefs and desires. To see the GC in action, just note that in both AFO and AFEA a change in the robot's input states, might alter its set of standing beliefs and desires. It then might modify Otto's pre-existing M-set in a way that has nothing to do with Otto's consciousness. Thus, calling such a modified state a constituent of Otto's M-set would be plainly incorrect. Furthermore, as described, even if Otto were awake, his occurrent states wouldn't make a difference to the robot's M-set without an additional upload. Given this, while Otto sleeps there is no route in virtue of which the robot's M-set could supervene on Otto's occurrent states. If this is the case, then neither the AFO or the AFEA can be pushed through to their unintuitive, if not absurd, conclusions. Of importance also is to point out that in coming to this difference that makes a difference, we, in carefully tracing out our reasons for abandoning a condition that was unnecessary for standing states, were able to locate one that is plausibly necessary standing state ownership. For given the nature of the problems that the AFO and AFEA both present, what was needed was a constraint not on what qualifies as a standing belief but what qualifies as someone's extended standing belief.

If one concedes to the assumption that one's standing states count as being constituents of one's mind, then Otto's standing states fail to be his the moment it is admitted that they can alter without his occurrent states altering. As a result, Otto's mind is extended only insofar as his M-set can be affected by him. Sure, our beliefs are still our even while we sleep, but what makes them ours is that their potential for modification is tied up with our potential occurrent alterations. This conclusion differs from Gertler's. While she locates the mind narrowly in only occurrent states, I am putting forth that the mind is constituted of both standing and occurrent states but that an agent's standing states -- assuming the possibility of extension -- are dependent upon the occurrent states of that very agent. Where this dependency relation holds, so does the extended mind relation. To conclude, because Otto's mind is not in more than one place at one time, the difficulty of establishing the bounds of the self becomes superfluous; at least in the way in which these cases have been described!

Clark, Andy (2008): Supersizing the Mind: Embodiment, Action and Cognitive Extension, Oxford; OUP, 2008.

Gertler, Brie (2007): 'Overextending the Mind', in Gertler and Schapiro (Eds) Arguing about the Mind, (New York: Routledge pp. 192-206).

Hume, David Treatise of Human Nature, edited by L.A. Selby-Bigge, 2nd ed. Revised by P.H. Nidditch, Oxford: Clarendon Press, 1975.

Williams, Bernard (1981): 'Internal and external reasons' reprinted in his Moral Luck (Cambridge: Cambridge University Press, 1981).

1. C&C do mention a fourth criterion, but given their own admittance that it is arguable, I have not included it in this list.

2. David Hume, in his Treatise argued, I think quite successfully, that the having of beliefs and desires are sufficient for motivation. (David Hume, A Treatise on Human Nature, book II, section III)

3. Although quantum mechanical theory may provide a compelling reason to disregard this result as something that should be considered incorrect.

5. I am indebted to Eric DeJardin for this example and his pointing this out to me.

6. The term motivational set, or M-set, should be understood here as Bernard Williams coined it in 'Internal and external reasons'; as one's subjective set of pre-existing beliefs and desires against which one measures the coherency of new beliefs and desires in a rational manner. (Williams, pp. 102-103). Of course, if the possibility of extended standing states are to be granted to non-human animals, the notion of M-set must be relinquished for some kind of set that is not constrained by rationality.

www.geoffreyklempner.net

klempner@fastmail.net

International Society for Philosophers

Wisdom begins with wonder