Resolving the Conceptual Problem of Other Minds through the Identity-Based Model A of

A bstract Christopher Peacocke’s Interlocking Account offers an example of the identity-based strategy for resolving the conceptual problem of other minds. According to the Identity Model, the sameness of meaning of a mental concept across inter-subjective domains is guaranteed by the sameness of the mental states to which the concept refers. Hence, for example, the meaning of the concept “pain” is f ixed by the sameness of the sensation of pain to which the concept refers across inter-subjective f ields. As an instance of this model, the Interlocking Account draws its most fundamental strength from the claim that human beings are similar in so far as they are carriers of conscious mental states, and that similar mental concepts have similar mental contents across individuals. The implication of this is that when similar mental concepts are used to describe contents of experience by different persons, the meanings of the concepts used are f ixed by the similarity of the contents of experience to which the concepts refer. This paper argues that this identity-based strategy fails for three main reasons: (1) the identity relation it purports to establish between one’s own case and those of others is dif f icult to achieve; (2) “the sense in which the relation of one's mind and those of others exhibits that identity is not clear;” and (3) it is an argument by analogy in disguise.


B
J B I Scholars have identi ied at least two varieties of the problem of other minds: namely, the epistemological problem and the conceptual problem (Avramides 2019(Avramides , 2001Graham 1996;Hyslop 1995). The epistemological problem of other minds raises concerns about the possibility of knowing that others have minds, and if they do, whether the contents of other minds can be known from a third person's point of view. This is the traditional problem of other minds. However, recent discussions have paid more attention to the conceptual problem of other minds. In fact, Thomas Nagel has declared that the conceptual problem of other minds is the interesting problem, which centres on "how I can understand the attribution of mental states to others (1989,(19)(20). According to Akeel Bilgrami (2010, 566), The interest in the traditional philosophical problem about other minds lies in the question whether, given what would seem a natural way we have of thinking about the meaning of mental terms, we can even make sense of the sentences we use to attribute mental states to others.
In strong support of Nagel's view, Bilgrami (2010, 566) further notes that This distinguishes the problem from a less interesting sceptical problem about other minds, in which it is a purely epistemological question how we know about other minds, on a par with the question of how we know about the external world.
According to Anita Avramides (2001Avramides ( , 2019, this is perhaps due to the fact that in talking about the mind, we must be sure that we have a general concept of the latter. By this Avramides means a concept of mind that applies to both irst-and third-person perspectives: that is, "the concept that applies to others as well as mysel " (Avramides 2001, 219). The importance of the conceptual problem cannot be over-emphasized, given its centrality to discussions about the mind more generally. For instance, she notes that the epistemological problem can hardly be raised unless the conceptual problem has been properly resolved. She writes: Where questions concerning the mind of another are taken to be epistemological, there is not only an assumption that it makes sense to think about one's own mind in advance of others, there is also an assumption that the concept one comes to have in this way is entirely general. (2001,219) 29 R C P O M One of the most popular models for resolving the conceptual problem of other minds is what can be called the identity-based model. This model takes the meaning of a mental concept to be the same for others as it is in the case of one's own use of the concept. The Identity Model of the conceptual problem of other minds inds a clear expression in Cassam's (2007, 178) view that "angry" must mean the same thing in "I am angry" and "He is angry. " Hence, "to have the concept of anger, one must be able to think someone else's anger as a state of the very same type as one's own anger. " In thinking this way, one grasps a sameness or identity relation between the irst-person use of mental concepts and the third-person use of the same concepts. By implication, if the concept of anger refers to the mental state of anger, then to describe my concept of anger as having the same meaning as another person's concept of anger, it must be the case that my mental state of anger is of the very same type as that of the person to whom I ascribe anger. This suggests the conclusion that sameness of mental states presuppositionally entails sameness of meaning of mental concepts. In this paper, I critically examine a well-known example of the identity-based approach to the conceptual problem of other minds. This is Christopher Peacocke's Interlocking Account. The Interlocking Account adopts an identity-based strategy, in that it aims at establishing sameness of meaning for mental concepts on the basis of an identity relation holding between the mental states linguistically represented by those concepts. Without presupposing this kind of identity relation as obtaining between the mental states to which mental concepts refer, it would be unclear how such sameness of meaning on the part of mental concepts could be achieved, and the solution to the conceptual problem would remain elusive. In contrast to the optimistic spirit of the Interlocking Account, however, the present paper presents arguments that move in the direction of undermining the argumentative power and explanatory force of the identity-based approach to the conceptual problem of other minds. Speci ically, it argues that the Interlocking Account fails to address three issues that are fundamental to its plausibility as a theoretical expression of the identity-based strategy for resolving the conceptual problem of other minds. These are: (i) the formulation of an identity relation between the irst-and third-person meanings of mental concepts is dif icult, if not impossible, to establish, because it is somehow impossible to establish the sameness relation between irst-and third-person levels of mental states; (ii) the notion of identity operative in the Interlocking Account is, at best, not clear, and, at worst, confusing; and (iii) the Interlocking Account is a disguised argument from analogy.

B J B
In seeking to achieving this aim, the paper starts off with a brief explanation of the conceptual problem of other minds, showing why the identitybased approach becomes an attractive position to pursue in the context of attempts to resolve it. This is followed by an analysis of Peacocke's identitybased model for understanding conceptual uniformity. In pursuing this, the paper irst considers the application of Peacocke's identity principle to non-mental concepts, where it yields a remarkable measure of success. Then, turning to its application to mental concepts and noting the dif iculties encountered by the model, the paper examines the peculiaritypeculiar feature of mental concepts that makes them not readily amenable to an identity-based analysis. It considers Peacocke's approach to resolving this issue in the context of his Interlocking Account, before inally subjecting the whole gamut of identity-based arguments to critical consideration.
T C P O M The conceptual problem of other minds, according to Smith (2010, 202), "is sometimes presented as the problem regarding the unity of concepts of mental states. " When different persons use the same mental concept (e.g., pain) to describe their mental states, do they have a uni ied conception of the concept used? Mental concepts refer to mental states, which are subjective states. This implies that the meaning of mental concepts is accessed by the individual subject in a unique way. It thus appears that the meaning of a mental concept-such as, say, pain-must be what each subject comes to have from the point of view of the subjective perspective of that individual's own feeling of pain, so that it is possible that my pain is entirely different from someone else's pain in all of its phenomenological details. Here, the saying that experience is the best teacher is correct, because experience teaches each subject to know what pain is from her own case alone, this being inaccessible to others. The question arising from this is: How do we reconcile the subjective character of mental states with the general nature of mental concepts? This is the question that the conceptual problem of other minds seeks an answer to.
The foregoing way of generating the conceptual problem of other minds derives primarily from a certain conception of the nature of concepts. According to Davidson (1987, 442), "what counts as evidence for the application of a concept helps de ine the concept, or at least places constraints on its identi ication. " This suggests that part of what determines the meaning of a concept is what counts as good evidence for its successful application. Davidson (1987, 442) notes the implications of this as follows: If two concepts regularly depend for their application on different criteria or ranges of evidential support, they must be different concepts. So if what is apparently the same expression is sometimes correctly employed on the basis of a certain range of evidential support and sometimes on the basis of another range of evidential support (or none), the obvious conclusion would seem to be that the expression is ambiguous.
The ambiguity to which Davidson alludes above results from two perspectives on a concept, in which one requires evidence and the other does not. Suppose I describe my pain as excruciating, and another subject describes her pain as excruciating. There seems to be no way of settling the meaning of "excruciating" because my own "excruciating" pain is accessed noninferentially, whereas the "excruciating" pain of the other subject is only suggested to me in her (verbal or non-verbal) behaviour. Hence, Davidson thinks that the meaning of an expression such as "excruciating pain" must be ambiguous, because it is acquired on the basis of different evidential support. Reconciling these two conceptual perspectives has constituted a central chunk of the conceptual problem of other minds. It is therefore easy to see that the conceptual problem of other minds is a problem about the meaning of mental concepts. Meanings are intersubjective phenomena; they require some kind of agreement among a community of language users about concepts, terms, words and expressions. How is this intersubjective agreement about conceptual meaning possible within the context of mental concepts, given that mental concepts refer to mental states with no objective mode of identi ication? As in Avramides (2001, 263;2019), this puzzle should be seen in the light of how to reconcile the generality of mental concepts with the unity of mental states. The supposed public nature of language is being called into question here, as it pertains to the meanings of mental concepts. If mental concepts, general as they are, are used to pick out mental states, which are subjective, it is a real question whether the former can appropriately convey the latter as our ordinary usage of them often suggests.
Mental concepts are general; their nature is such that they apply equally to both irst-person and third-person perspectives. Thus, the generality of mental concepts consists in their capability of being applied across different subjective states. For example, the mental concept pain is a general concept, because if one possesses the concept, one must be conceptually capable of attributing it both to oneself and to others in whom pain can be instantiated. However, the mental states to which mental concepts refer linguistically are uni ied because there is something it is like to have them.

B J B
The statement "I have an excruciating pain" describes a particular sensation of which I alone am intrinsically aware. If such a statement is made when I am not in pain, it raises the question: how am I to understand the concept pain? The problem that this question highlights lies in our lack of access to what it is like for another subject to be in a certain mental state.
To account for the possibility of linguistic uniformity that mental concepts are expected to exhibit, there is thus a need to establish that mental states are phenomenologically uniform across subjects-especially given the conception of mental states as the referents of mental concepts. Mental concepts can only be held to carry the same meaning across diverse domains of use if what they allegedly refer to are of the same stuff. Hence, unless the argument purporting to establish sameness among tokens of a mental state (e.g., pain) felt in different subjective ields is philosophically satisfactory, Davidson's charge of ambiguity in relation to the meaning of mental concepts will indeed be a cause for worry. This may be taken to explain the appeal of the identity-based strategy in the philosophy of mind in contemporary times. In this paper, I seek to critically examine Peacocke's Interlocking Account as an instance of the identity-based strategy for resolving the conceptual problem of other minds.
T I B M U C U Peacocke opines that for a wide range of concepts, a thinker's understanding of what it is for a thing to fall under the concept plausibly involves knowledge of identity. Speci ically, it involves tacit knowledge to the effect that the thing has to have the same property as is exempli ied in an instantiation of the concept in some distinguished, basic instance (Peacocke 2008, 168). To put the identity-based model to work in cases other than mental concepts, Peacocke considers the application of the model to some nonmental concepts. Two of these will suf ice for our purposes in this section. In the irst example, Peacocke considers an observational concept, oval. To possess this concept, according to this model, is to have tacit knowledge that for an unperceived thing to be oval is for it to be of the same shape as things one perceives as oval (Peacocke 2008, 164). This implies that understanding the concept oval, when the thing so described is not in visible range, requires prior knowledge of oval-shaped things. That prior knowledge forms the basic requirement or conceptual apparatus for understanding the concept. It is in the light of this prior knowledge that one understands all oval-shaped things, whether they be seen or unseen. This model is true of both nonconceptual and conceptual theories of content. Non-conceptual theories of content agree that the content of experience should not be thought of along linguistic lines, where it is seen as a product of the subject's conceptpossessing abilities, but instead should be understood by appeal to certain types of non-conceptual representation independent of language (Musholt 2013;Bermúdez 2001;Hurley 1997;Vosgerau 2009). Conceptual theories of content, on the other hand, maintain that the analysis of content must proceed via analysis of language (Dummett 1973). Peacocke believes that this model of understanding the concept oval will be available no matter which of these positions one adheres to.
In his second example, Peacocke applies the model to our predications with respect to non-present places and times. Suppose a weather reporter for a global television network forecasts a sunny day in Paris on a certain day of the week. Peacocke's position is that we get to understand this piece of information even though we are neither currently in the said location nor at the point of having experienced the said day of the week, because of our tacit knowledge of what it means to be sunny, based on cases we have either experienced in the past or are currently experiencing. Hence, "our understanding of what it is to be sunny at some arbitrary place-time consists in our tacit knowledge that the place-time has to have the same property as our current place-time has to have for it to be sunny here" (Peacocke 2008, 164). This is like basing our understanding of sunny predicated of other places and times on the model of our knowledge of the property of being sunny acquired through past experiences of sunny days. What equips us with this conceptual capacity is, one can say, the experience of past instances of sunny days, in the light of which we are able to understand the concept sunny, even when it is not being currently experienced.
One question arising from the foregoing is how the relation of identity is to be understood in the light of the two examples above. In the irst example, one is able to understand the property-concept oval even though the object so described is not within visible range, because one already possesses the concept through exposure to oval-shaped things in the past. Here, "concept possession" means being identical with the shape-type whose tacit knowledge one already has. Hence, an unperceived object is oval if its shape is identical with things one perceives to be oval. The same analysis applies to the second example. To grasp the concept sunny when predicated of a place and time in which one is not currently present, is to understand it in the light of the tacit knowledge acquired through past instances of sunny days. A non-present sunny day is therefore a day that has the same property as a day one perceived to be sunny earlier. In the two cases, what is really important is the property a thing must have in B J B order to be described as so-and-so (in the present case, that of being oval and sunny). The point Peacocke is making is that this property, when ascribed to any object, ixes the content of understanding. In both cases, it is the sameness of the property (e.g., being oval and sunny) that de ines the identity relation between our past knowledge of a thing and any arbitrary object with the same property. Spelt out in the above manner, the identity-involving model of grasping a concept has some attractive points. One important one, among others, is that it helps to explain how it is possible for several things to fall under a particular concept, even though each of them still retains its distinct ontological status. (Note that the different objects falling under a particular concept are those possessing a common property.) In other words, it supplies an explanation of the generality or uniformity of concepts and meanings across occurrences in different thoughts or sentences. Peacocke writes: "It is an immediate consequence of the identity-involving explanation of grasp of a concept that one and the same property is predicated both in the distinguished case and in the case understanding of which is explained by grasp of the identity" (Peacocke 2008, 165).
T P M C Nevertheless, this attractive feature only applies to observational and other spatial concepts. For when the identity-based model is applied to mental concepts, such as pain, certain problems begin to surface that pose great obstacles to the operational validity of the model. An explanation of this may be that the contents of mental concepts are not things about which we see being furnished public references, or empirical evidence, of the relevant sort. The problem encountered does not arise at the level of the individual. Intuitively, each person knows what her pain feels like from the irst-person perspective. Only when attempts are made to extend the very same meaning of the mental concept that one applies in the irst-person instance to the third-person instance does the problem of ambiguity rear its head. The problem consists in the possibility of differences in the content of the mental state to which a particular mental concept refers between the irst-person case and the third-person one. It has been argued that if a mental concept refers to different mental states in different subjects, then its meaning cannot be uni ied or the same: hence the problem of ambiguity (Smith 2010, 203-5;Skirke 2014, 233-4).
Note that this problem does not arise for observational concepts. In the observational case, the oval-shaped (physical) object is the cause of the ovalshaped non-conceptual mental content. Hence, the subject's perceptual content is a mental replica of the object outside the subject, an object in the external world. In the case of the property-concept oval, there is an object in the world that possesses that property. This fact makes it possible to know what must obtain in order to have one's belief about the presence of an oval-shaped object con irmed as true or false. 1 To put it differently, the actual presence of an oval-shaped object enters into the identity-based analysis of the concept oval because of its role in settling any dispute that may result from an inability to differentiate between objects with, and objects without, the property oval. According to Peacocke: In the base case for an observational concept like oval, the thinker has an experience of something as oval. The thinker's perceptual experience has a content that is itself given in part by reference to a spatial type itself-what I called scenario content in earlier work. The intentional, non-conceptual content already concerns objective, and consequently public, properties and states of affairs. The identity account speci ies that for something else to be oval is for it to be of the same shape as is employed in specifying the content of the thinker's experience in the base case. This condition concerns a spatial property itself, and, modulo vagueness, there is nothing indeterminate or ambiguous about whether some object meets that spatial condition at a given time. (2008,170) Given that no problem arises with the application of the identity-based model to observational or other spatial concepts, one naturally becomes concerned as to why the same explanatory strategy fails in the case of mental concepts. Why is there determinacy in the observational case, and ambiguity in the mental case? Why is the subjective case different? In what does the difference consist? What does it mean for someone other B J B than me to fall under the concept pain? In the case of the observational concept oval, what it means for an object other than one being currently experienced to fall under it is that the object would have the same shape as oval-shaped objects do, were the relevant object to be actually present. But this fails in the case of the mental concept pain, as unlike in the case of observational concepts, the state to which this mental concept refers is not one that can be cross-checked in public terms. What is the relevant object in the case of a mental concept? Peacocke asks: "Can we say that the thinker's identity-involving understanding of the general concept pain is that for something to fall under it is for some event involving an arbitrary body, and bodily location, at some arbitrary time, to have the same property as is picked out by the unstructured local concept that corresponds in the way indicated to pain-experienced-by-me-now" (Peacocke 2008, 172). This is a very dif icult question to answer, because it fails to differentiate between two crucial cases in which there can be talk about a body feeling pain. Peacocke differentiates the two cases thus: The irst is the case in which you have pain in another body, at that bodily location, at another time. This irst case subdivides into that in which that body is your only body; and that in which, if it is possible, you have your actual body, but are also capable, by some setup, of feeling pain in another body too. In either of these subdivisions, the pain is yours. The second case is that in which someone else has pain in that body, at that bodily location, at the time in question. (2008,172) It can be observed that the identity model is only consistent with the irst case. This is because in the two subdivisions of that one, there is a portrayal of different instantiations of pain-at a different body, bodily location and time. However, since the body is either your only body or another's body whose pain you share, there really can be no question about whether the pain is yours: it is. The model is inconsistent with the second case, however, because there is no epistemic or conceptual link between your own case and the other person's case that could warrant the assertion that the latter is an instantiation of the concept pain in another body. Hence, the identity is rendered useless in the second case. Peacocke (2008, 173) concludes: "Because that exposition is consistent with only the former case being the proposed condition for an arbitrary event to be a pain-event, a case in which the pain is still yours, the account is wrong. " The difference being identi ied here between observational and other spatial concepts, on the one hand, and mental concepts, on the other, suggests R C P O M that if the identity-based model is to work in the latter case as in the former, then a variant of the model will have to be developed for the latter that takes account of this difference from the former. If the model works for the explanation of observational concepts, then there must be something the understanding of which makes its working possible. To formulate this as a question: What is it about observational and other spatial concepts that makes them such that the identity-based model works unproblematically for them without further quali ications? This is a demand for some underlying assumption or central feature to be identi ied, that must obtain in this conceptual region, such that it is possible to explain the relevant concepts without the problem of unity rearing its head. With the satisfaction of this demand, one may then ind out if the same demand can be placed on the mental conceptual region in order for it to be amenable to the sort of identity-based model whose application to observational concepts raises no qualms. In order to meet this demand, according to Peacocke (2008, 174), "we need to step back and consider what makes identity-explanations of concepts work in the cases in which they do work. " Peacocke identi ies this central feature of observational concepts-which makes them readily amenable to the identity-based explanatory model-as being the uniformity of regions of space. As he puts it, "it is a single, uniied space which makes intelligible the idea of something elsewhere being the same shape as something you currently perceive" (Peacocke 2008, 174). This assumption is the basic requirement that must be met for the identity model to work when it comes to the explanation of observational concepts. Without it-that is, if different regions of space could conceivably be held to be differently constituted, so that no two regions of space were taken to be the same-observational concepts would be subject to precisely the kind of ambiguity identi ied for mental concepts. As Peacocke writes: It is a single, uni ied space that prevents there being any ambiguity or indeterminacy of a fatal sort in the identity-account of concepts like oval. There is no such thing as a thinker perceiving something to be in an oval-shaped region of space independently of that region's being part of a single, uni ied larger space. What makes the content of the thinker's experiences spatial is that their content can contribute to his conception of the layout of things and events in the larger space of which the region he perceives is part. (2008,174) Peacocke's primary aim here is to emphasize the central role played by the uniformity of a region of space in the thinker's conception of things and events in the world. Intuition suggests that it is likely that two indiscernibly B J B identical objects will lose their identity if placed in two non-uniform regions of space. This means that without the assumption of uniformity with respect to the region of space, it will be impossible to conceive of an object not currently within visible range having the same shape as one being currently observed. Hence, given this uni ied constitution of space, "there is no such thing as a concept of oval in my perceived space that picks out a different shape from oval (and of the right size) period" (Peacocke 2008, 174). This assumption may have in luenced Wittgenstein's remarks in Philosophical Investigations, where he writes: You surely know what "It's 5 o'clock here" means; so you also know what "It's 5 o'clock on the sun" means. It means simply that it is just the same time there as "it is here when it is 5 o'clock. "-The explanation by means of identity does not work here. For I know well enough that one can call 5 o'clock here and 5 o'clock there "the same time, " but what I do not know is in what cases one is to speak of its being the same time here and there. (1958, §350) Suppose by "here" Wittgenstein means "at a particular place on Earth, " while by "there" he means "on the sun. " Suppose, further, that at 5 o'clock here, the sun usually takes on a particular angle, so that whenever the sun is observed to take up this angle, one readily knows that it is 5 o'clock, even without checking what time it is. Since the earth is under the sun, it will be impossible to imagine what the time will be like on the sun itself, where time is not determined by the position of the sun. In this case, the identity model fails, because the spaces of "here" and "there" are not uniform. Hence, although the same property "being 5 o'clock" is still instantiated, the time does not apply to both of them with the same meaning.
, what is the analogue of (the uniformity o ) space in the problematic case of mental concepts? This amounts to asking what the intelligibility of there being another conscious experience other than that currently being undergone by me consists in. According to Peacocke (2008, 175), "it is the conception of multiple subjects of experience that makes intelligible the idea of there existing conscious states and events other than your own. " The existence of an experience seems to suggest the idea of a subject having the experience. This is what Galen Strawson (2010, 129) refers to as Frege's Thesis: "'An experience is impossible without an experiencer' or subject of experience. " Strawson conceives the relation between experience 39 R C P O M and subject of experience in terms of necessity. In construing the relation thus, Strawson means to show the mutual inevitability of each relative to the other. As he argues, "There cannot be experience without a subject of experience, because experience is necessarily for someone or somethingan experiencer or subject of experience. Experience necessarily involves experiential what-it's-like-ness, and what-it's-like-ness is necessarily whatit's-like-ness for someone or something" (Strawson 2010, 129). If someone utters the sentence "I am in pain" when I am not in pain, that is, when the "I" does not refer to me, then it must follow that there are other subjects of experience other than me. In order to be able to understand such an utterance as "I am in pain, " I must be able to conceive of other subjects who may possibly have pains when I am not having one. Earlier, Nagel (1989, 20) wrote that "each of us is the subject of various experiences, and to understand that there are other people in the world as well, one must be able to conceive of experiences of which one is not the subject: experiences that are not present to oneself. " Without this understanding, one lacks the conceptual resource for attributing mental states to others. Against the backdrop of the above claim to the effect that there cannot be experience without a subject, Peacocke argues that there exists some kind of metaphysical interdependence between conscious states and the subject which makes each of them incomplete without the other. The nature of this metaphysical interdependence is such that it renders both conscious states and their subjects mutually parasitic on each other, making each inconceivable without the other. An actual conscious state presupposes an owner for that state, while being a subject itself derives in some way from one's ability to have conscious states. The underlying principles governing this relation of interdependence between mental states and their subjects are identi ied by Peacocke (2008, 175) in the following terms: a. Conscious states are states such that there is something it is like to be in that state, more speci ically something it is like for the subject of that state. b. Subjects are things capable of being in a conscious state.
For the identity-based model to be applied to the troublesome concepts of the mental realm, Peacocke proposes that we "treat grasp of conscious states and grasp of a potential multiplicity of subjects in a single principle. " The resultant principle, he insists, must treat both as interlocking notions, with this interlocking being a direct consequence of the mutually supporting relation holding between the experience and its subject. Thus we arrive at the Interlocking Account of other minds. Using the concept pain B J B as a paradigm example, Peacocke explains schematically that for x (i.e., a person other than me) to be in pain is both of the following: it is "for x to be something of the same kind as me (a subject); and is also for x to be in the same state I'm in when I'm in pain" (Peacocke 2008, 175). The Interlocking Account takes it to be a fact that there is multiplicity of subjects in the world. In other words, the fact that we are capable of having conscious experiences is a property that we all have in common as subjects. By implication, this means that as subjects, we all belong to the same category of beings, making each of us a thing of the same kind as others. This is why it is possible for there to be experiences which are not currently being undergone by oneself at a particular time. However, being a thing of the same kind as others does not undermine the fact of our irstpersonhood in a way that would render our experiences indistinguishable in non-trivial terms from other experiences. On the contrary, the fact of our subjectivity separates us as individual cases of consciousness, while our ability to have conscious states uni ies us as members of the same kind. This is Peacocke's way of harmonizing the con licting requirements that self and others be at the same time similar and dissimilar. Hence, the ambiguity of which other-directed attributions of mental concepts have been suspected (Davidson 1987, 422) is made up for through the recognition that there are other subjects implied and involved in the understanding of mental concepts attributed to others. As Peacocke clari ies: It does not suf ice for the understanding-condition to deal only with subjective kinds of experience, body and bodily location. Under the Interlocking Account, we specify other-attribution as the case in which we have: same state, another subject, where the subject is thought of as a thing of the same kind as oneself. (Peacocke 2008, 176) I P ' I A The question may be raised of how we should understand the notion of sameness in the Interlocking Account. An initially plausible response would be to understand it in terms of identity, since the Interlocking Account is a variant of an identity-involving model of concepts generally. However, this leads to a further question: What sense or kind of identity? Broadly, philosophers have distinguished between two types of identity: namely, numerical and qualitative. Identity of the irst sort describes a relation between "two" objects in which the supposed objects are one and the same thing, perhaps known via different intensional properties. To put it another way, two things are numerically identical if all the properties predicated of the one are present in the other, without any of them possessing more or less than the other. In the same vein, two states are numerically identical if in referring to one, one is referring to the other. According to Tomasz Bigaj, "In this fundamental sense, when we say that object a is identical with object b, we want to express the thought that a and b are one and the same object-that there is not two but one individual bearing two different labels 'a' and 'b'" (Bigaj 2012, 15;Conee and Sider 2014, 8-9). Qualitative identity, on the other hand, is the identity holding between two objects that are so similar that they are almost indistinguishable. This is why qualitative identity is sometimes also called indiscernibility. Note that qualitative identity, unlike its numerical counterpart, involves two or more separate objects, which are indistinguishably similar. Sameness in Peacocke's sense does not appear to match either of the two categories of identity identi ied above. On the one hand, no two persons' pains are numerically identical: my pain is identical only to my pain at a particular time, and yours to your pain. It suf ices to say that numerical identity, in relation to conscious states, is only possible intra-personally; there cannot be an interpersonal instance of it. On the other hand, there are some dif iculties with establishing what could be involved in cases of qualitative identity between two persons' pain. In order to determine the indiscernibility of two mental states, it is required that one be able to access the two states in an objective way. But this is dif icult, if not impossible, given that only one's own mental states can be directly accessed. This implies that the requirement for determining the indiscernibility of two mental states of which one is one's own cannot be met, because the condition under which the predicate "is in pain" is true for the other person cannot be clearly speci ied.
How, then, does Peacocke explain the identity that forms the foundation of the Interlocking Account? As far as he is concerned, the importance of the identity of states to the Interlocking Account is too obvious to be ignored. In his words, the component of the Interlocking Account that speaks of identity of state is essential to determining the property of pain as the property picked out by the concept pain, and correspondingly as the property picked out by the word "pain" in the linguistic case (Peacocke 2008, 178).
Here the concept pain refers to a property of the subject concerned. The implication is that when the predicate "is in pain" is used by two persons to describe their current sensations, the two individuals have the property B J B "being in pain" in common. As such, a relation of identity appears easily established between them. It follows, therefore, that in the genuine case of other-ascription of mental concepts (say, pain), understanding the predicate "is in pain" involves tacit knowledge that the predicate is true of another subject x distinct from oneself just in case x is in the same subjective state one is in oneself that makes it rational to accept "I'm in pain" (Peacocke 2008, 178-9). The problem with the Interlocking Account, based on the notion of identity of states, is that it is not entirely clear how establishing a relation of identity among multiple mental states could be possible. The sameness of the property of being in a mental state sounds good only if there are ways in which the mental states of the subjects involved may be put side-by-side, and this relation of sameness established in an objective way. However, in the case of the mental, no such objective way really exists. Peacocke's clari ication of the phrase "same state I'm in when I'm in pain" to mean "same conscious state" does not appear to supply what is needed for the resolution of the conceptual problem of other minds being grappled with. In accepting this interpretation of sameness, there is a need to settle the question of what criterion best suits the establishing of this relation of sameness.
Three possible criteria may be suggested. The irst is a behavioural one which seeks to establish sameness of states through sameness of bodily behavior. It fails, however, because observation of behaviour is not always necessary in one's own case, while it plays a major role in the identi ication of others' presumed states. Besides, outward reactions to diverse mental states differ from one subject to another. Hence, since two subjects may react to the same mental states differently, it follows that a behavioural criterion cannot lead to the establishing of the sameness of mental states across different subjects. The second criterion is a psychological one. It hopes to establish sameness between states by identifying a mental state in one subject with the same mental state in another subject. This seems to be what sameness of states in Peacocke's sense amounts to. However, it also fails, because talk about psychological states is often irst-personal, and cannot therefore apply to third-person-based instances.
Note that the two criteria identi ied so far are what might be called "purebreed" criteria, in the sense that they appeal to monolithic principles for establishing sameness of states across persons (i.e., irst-and third-person). The third criterion mixes the behavioural and the psychological, so that the latter takes on the irst-person perspective while the former takes on the third-person one. That is, in my own case, the concept pain refers to 43 R C P O M my own directly accessed mental state, while in the case of others, "pain" refers to bodily behaviour. When this is done, however, philosophers are quick to point out the resultant problem of ambiguity which chie ly arises from arriving at the meaning of a single mental concept on the basis of two radically different evidential criteria. Davidson contends that if what is apparently the same expression is sometimes correctly employed on the basis of a certain range of evidential support and sometimes on the basis of another range of evidential support (or none), the obvious conclusion would seem to be that the expression is ambiguous. (1987,442) As Hanna Pickard puts it, "If you understand what pain is from your own case, then pain is just that wretched experience. But then how can there be pain when there is no such experience?" (Pickard 2003, 89). Pickard's question reminds us of our ignorance regarding the meaning of mental concepts that refer to states other than ours. Given this manifest ignorance, it may be the case that each person has her own peculiar concept of pain, which is different from any other, where this ultimately leads to the problem of ambiguity in the meaning of pain.
Another objection that could be brought against Peacocke's Interlocking Account is that it is a disguised version of the argument from analogy. The classical version of the latter takes the perceived similarity between one's own case and that of another as a warrant for the conclusion that others' mental states are not signi icantly different from one's own. Consider Stuart Hampshire's version of it-as reconstructed by Norman Malcolm: When my behaviour is such-and-such there comes from a nearby human igure the sound "He feels giddy. " And generally I do feel giddy at the time. Therefore when another human igure exhibits the same behaviour and I say "He feels giddy, " it is probable that he does feel giddy. (1958,971) It is important to point out here that others' judgment of my behaviour as a feeling of giddiness is arrived at on the basis of their having had some similar experience in their own case. It is their association of the content of their own inner feeling with their outward behaviour that puts them in a position to interpret other people's behaviour as instances of feeling giddy.
As a preface to the foregoing account of feeling, Hampshire differentiates between what he calls "autobiographical" and heterobiographical" statements. By the former he means "a statement describing someone's feelings or sensations which explicitly shows, in the actual form of its expression, B J B that the author of the statement is also its designated subject" (Hampshire 1952, 2). Autobiographical statements take the irst-person singular form; hence, they usually start with "I" used in the normal sense. An example of the use of such a statement is "I feel giddy, " where the statement describes "somebody's momentary feelings or sensations which are expressed in the irst-person singular. " On the other hand, heterobiographical statements are statements describing someone's feelings which are not expressed in the irst-person singular. "He feels giddy, " "you feel giddy, " and "they feel giddy" will then be examples of heterobiographical statements (Hampshire 1952, 2). Note that a heterobiographical statement can sometimes take on the form of an autobiographical statement. This is what happens, for instance, when I hear the sound corresponding to the utterance "I feel giddy, " but there is no experience in me that makes the statement true. My understanding of the meaning of such a statement will be an extension of the meaning that "I feel giddy" has when I am both the author of the statement and the subject of the experience. It is the case, therefore, that the predicate "feel giddy" must be capable of being applied both to irst-and third-person perspectives. Of course, given Hampshire's analogical analysis, one can only make heterobiographical statements if one can make autobiographical statements. In other words, a conceptual capability of an autobiographical kind serves as the basis for making heterobiographical statements. In a similar fashion, Peacocke distinguishes between base, local concepts, on the one hand, and non-basic or global concepts, on the other. Although he does not devote any speci ic section of his work to the clari ication of these kinds of concept, one could try to arrive at a sense of what he takes them to be by looking through some of the things he says about them at different points in his book. For instance, he writes in one place that "In the base case of an observational concept like oval, the thinker has an experience of something oval. " In another, Peacocke has this to say: In the base case, the thinker can be regarded as employing a recognitional concept C d of a spatial kind (subscripted with 'd' for the distinguished case), which intuitively is something like perceived-by-me-now-to-be-oval. The Concept C d is not really structured, of course. The hyphenation in the italics is just to indicate the fundamental condition for something to be the concept's reference. C d is true of an object at a time just in case the object is perceived by the thinker to be oval at that time.… Correct and rational application of C d registers the instantiation of a property itself. (2008,170) Against the foregoing backdrop, one may understand base concepts to be those possessed in virtue of irst-person encounters with the experiences to which they refer. Roughly, in the case of observational concepts, base cases provide templates for individual concept-possessing subjects to grasp attributions of properties to objects that are not within a perceivable range. As Peacocke (2008, 171) writes, "for something to be oval is for it to have the same shape property P as things must have if they are to fall under the local observational concept C d in the distinguished, base case. " Put in the context of the mental concept of, say, pain, a base case would be the pain that one feels or possesses an experience of. Hence, a base case of pain will be an experienced case in the light of which other-ascriptions of pain are understood.
On the other hand, non-basic or global cases are general concepts in the sense of their being subject-neutral. They belong to neither this nor that subject, but are capable of being picked out by any subject's experience. They are parasitic upon any conscious state's capacity to instantiate the relevant feeling, where the feeling is a kind and the instantiation its token. The base case bears a dual relation to the global case in that it is both an instance of, and a template for, the global one. As an instance, it is propertyidentical with the global concept, and as a template, it provides the basis for understanding the latter. This helps to make sense of the claim that there are multiple subjects of experience in the world of which one is oneself just one instance. The central claim of the identity-based model is that a correct application of a mental concept will imply the identity of the base case with the global case. That is, to understand the predicate "in pain" when attributed to others is to conceive it as an instance of a global case, not instantiated in the base case. According to Peacocke, the irst person plays a double role in the Interlocking Account, a role not played by any other way of thinking of a person or a subject. It enters the base case both for the concept of a subject and the concept of a given conscious state. The Interlocking Account gives a legitimate undergirding to the intuitive claim that one knows from one's own case what it is for someone to be in pain, or in some other given conscious state. It gives a corresponding undergirding for the intuitive claim that one knows from one's own case what it is to be a subject. (Peacocke 2008, 181) The point we have been pursuing so far has been that the Interlocking Account is a variant of the argument from analogy, and this could not be more obviously and clearly shown than it is in the above quote.

B J B
Other-ascription of mental concepts is impossible without the base case. The concepts one learns from these various base cases are uni ied concepts, because they are applicable to both irst-and third-person perspectives. Serrated and happy, as examples of uni ied concepts, behave like each other, on Peacocke's view, in that "serrated" allows us to understand something about the mountain range currently in view as well as about the edge of the bread-knife in our kitchen drawer at home, while "happy" allows one to understand one's own happiness as well as happiness which is not one's own. (Skirke 2014, 238) In both cases, the base case equips one with a conceptual capacity to otherascribe the concepts one ascribes to oneself. The same point is reiterated by Joel Smith (2010, 214), who opines that "A central feature of the Interlocking Account is that, for a range of conscious states, thinking of another as being in a mental state Ψ is to think of them as being in the same state that one is in when one is Ψ oneself. " An objection might be raised against the above interpretation of the Interlocking Account by invoking Peacocke's principle that "Understanding precedes Evidence. " 2 This principle privileges understanding over and above evidence in matters pertaining to other-ascriptions of conscious states. However, the argument from analogy prides itself on the observed (physical) similitude between one's own case and those of others, which seems to suggest that the allegation against the Interlocking Account to the effect that it is a disguised argument from analogy mis ires. Correct as this view is, it should be noted that the argument from analogy, as traditionally deployed by scholars such as J. S. Mill, Bertrand Russell, Stuart Hampshire and Alec Hyslop (among others), is an attempted solution to the epistemological problem of other minds. Hence, as an epistemological solution, its priority is showing that there is evidence for the existence of other minds.
However, the Interlocking Account is not meant to be a solution to the epistemological problem; rather, it is a solution to the conceptual problem. Recall that the conceptual problem has to do with the meaning of mental concepts and, as a problem about meaning, it places more of a premium on the understanding of mental concepts than it does on evidence for otherascriptions of mental states. Hence, what Peacocke's principle claiming that "Understanding precedes Evidence" really amounts to is just this: that where mental concepts are concerned, understanding comes before evidence. Yet it should be noted that this understanding derives from the base case: in the case of observational concepts such as oval, the base case is the oval-shaped object either previously experienced or currently being experienced, while in the case of a mental/conscious concept such as pain, the base case is the experience of pain either experienced on a prior occasion or now being experienced. For example, the meaning of "He is in pain" derives from the understanding of the base case "I am in pain. " Hence, given that the base case provides a standard for understanding the meaning of mental concepts used by others, and thereby provides the warrant for other-ascription of mental states, the allegation against the Interlocking Account being a disguised form of analogical inference stands.
If it is indeed the case that the Interlocking Account amounts to an instance of argument from analogy, then it follows that it is open to some of the attacks incurred by this strategy for resolving the problem. Some of these objections (see Malcolm 1958, 969-78) have been identi ied in the literature; hence, there is no need to rehearse them in the present context. However, the central weakness of all analogical arguments, which renders them inadequate when it comes to resolving the (conceptual) problem of other minds, certainly merits reiteration here: it is the problem of the theoretical appropriateness of the methodological derivation of other minds using one's own mind as a model or basis. This cannot be resolved by appeal to inductive method, as with its evidential strength resting on just one case it yields only the weakest form of inductive procedure. Neither can the problem be circumvented by an appeal to the sort of sameness exempli ied when we understand the statement "She is in pain" as an instance of the kind that is what I mean when I say "I am in pain. " Such a reading tries to establish an identity relation between my mental state and those of others, and to reduce the latter to the same conscious state I am in when the relevant mental state obtains. However, the solution is thwarted by lack of criteria for other mental states. I simply do not have the relevant point of view to know whether the mental states of others are what I take them to be. As Malcolm (1958, 971) writes: "You cannot improve my understanding of 'He has a pain' by this recourse to the notion of 'the same' unless you give me a criterion for saying that someone has the same as I have. " C This paper has been concerned with one particular version of the famous identity-based model for attempting to resolve the conceptual problem of other minds: namely, Peacocke's Interlocking Account. It was noted that the 48 B J B attractiveness of identity-based theories as far as the conceptual problem of other minds in concerned, lies in the notion that mental concepts are linguistic representations of mental states, the idea being that sameness of meaning across subjective domains with respect to mental concepts presupposes sameness of the mental states to which they refer over diverse subjective ields. On such a construal of the problem, the identity-based model inds itself required to demonstrate how the relation of sameness can be established between the meanings of mental concepts and the raw feel of the mental states to which they refer-across persons. This, as the central challenge facing all identity-based theories that seek to address the conceptual problem of other minds, is one that also confronts Peacocke's Interlocking Account.
The present paper argued that Peacocke's strategy for meeting this challenge, as enshrined in his theory, fails. Its ultimate conclusion, then, is that any attempt to argue in the manner of identity-based theories that does not dismiss the subjective character of mental states is bound to encounter the line of objection levelled here against Peacocke's Interlocking Account. On the other hand, if the subjective character of mental states is denied, the identity-based model collapses, and its attraction and explanatory force are lost, because the other side of this identity has been eradicated. To forestall this, the identity model must strive to ind an alternative way of reinforcing the sameness relation between the meanings of mental concepts and the mental states to which they refer, such that the subjective character of the latter is not tampered with and the problem of the indeterminacy of others' mental states is properly taken care of.