CHAPTER VIII

PUTNAM

In this chapter I will look at Hilary Putnam's solution to the problems posed by the arguments of Chapter Seven and the arguments he gives for this solution. He calls his solution internal realism, for it attempts to avoid relativism while making objectivity a matter of internal constraints. To understand Putnam's view and his arguments for it, we first need to see what his view is an alternative to and what led him to see an alternative was necessary.

8.1: Metaphysical Realism

Putnam calls the view he is attacking Metaphysical Realism. He defines this view most clearly in Reason, Truth and History:

One of these perspectives is the perspective of metaphysical realism. On this perspective, the world consists of some fixed totality of mind-independent objects. There is exactly one true and complete description of 'the way the world is'. Truth involves some sort of correspondence relation between words or thought-signs and external things and sets of things. (Putnam 1981, p. 49)

Putnam also holds that this view has an important implication, which he sometimes identifies with Metaphysical Realism:

The most important consequence of metaphysical realism is that truth is supposed to be radically non-epistemic - we might be 'brains in a vat' and so the theory that is 'ideal' from the point of view of operational utility, inner beauty and elegance, 'plausibility', simplicity, 'conservatism', etc., might be false. 'Verified' (in any operational sense) does not imply 'true', on the metaphysical realist picture, even in the ideal limit. (Putnam 1976, p. 125)

Putnam even identifies this claim with metaphysical realism at times:

I concluded that metaphysical realism - the view that truth outruns even idealized justification - is incoherent. (Putnam 1983, p. 85)

and

... for the sharp distinction between what really is the case and what one judges to be the case is precisely what constitutes metaphysical realism. (Putnam 1981, p. 71)

One should note a number of troubling things about Putnam's definition of the view he will attack. First, it is so vague as to include many possible views which it is difficult to see how his arguments against the view address. Let me examine his main definition.

His definition seems to include three separate positions: The first sentence of the definition states a view I will call ontological realism. It states that the world consists of some fixed totality of mind-independent objects. This, of course, is a metaphysical statement about the nature of what exists. It says what exists is independent of our minds and that it has sufficient structure independent of us to differentiate itself into objects. This to be distinguished from possible alternative views such as (1) Idealism, the view that there is no mind independent reality, and (2) the view that what exists is mind independent but has insufficient structure to differentiate itself into objects. I call this last view the Oatmeal theory of reality. This is the view that reality is a bowl of mush; you have to spoon it out yourself. This latter theory is the view Putnam espouses. It should be noted that it is quite difficult to see how one could support any of these views except by empirical evidence. In particular, it is unclear how any argument concerning the nature and possibility of representation could support any such view.

The second sentence states a view I will call epistemological realism. It states that there is one true and complete description of the world. This is essentially

the view that knowledge is a determinate relation between a description or representation and the extra-representational world and that we either have such knowledge or it is possible for us to have it. Epistemological realism is a form of the Representational Model of Epistemology. It holds that we know by representing and that it possible to have a true and complete representation of the world.

This view implies ontological realism, but it is not implied by it. Therefore, there will be at least two alternatives to it. One alternative will reject ontological realism, and along with it the external model of objectivity and the Representational Model of Epistemology. This is Putnam's internal realism. The other alternative will retain ontological realism and the Representational Model of Epistemology. It will hold that there is a world of mind-independent objects and that we can know it by representing. It will also hold, however, that it is impossible for us to have a true and complete representation of reality. It will hold that the very nature of representation precludes completeness.[1] This is the view espoused in this dissertation. In what follows, I will attempt to show that Putnam's arguments against Metaphysical Realism do not apply to this alternative.

The third sentence in the definition states a view I will call semantic realism. This is the view that there is a determinate reference relation between representations and the world. This view is most clearly expressed in "Realism and Reason": "Minimally, however, there has to be a determinate relation of reference between terms in L and pieces (or sets of pieces) of THE WORLD, on the metaphysical realist model... ." (Putnam 1976, p. 125) It seems, however, that widely differing views could accept this claim. The physical-visual model of representation, with its claim that representations are entities that have a determinate correspondence or similarity relation to objects is an instance of semantic realism. (And this is almost certainly the view that Putnam intended by his definition.) The view of reference given in Chapter Four also seems to be a version of semantic realism. Representing is seen as an act in which a domain of interest is isolated out and referred to. There is no preexisting relation between the representation and the world that constitutes reference, the act of representing itself makes the relation. Yet, there is a determinate reference relation between representations and parts of the world; the act of direct referring makes a determinate relationship. As we saw in Chapter Four, Putnam's arguments do not apply to this view of reference.

It may be helpful to define the position argued for in this dissertation in comparison to Metaphysical Realism. The view espoused here is very similar to Metaphysical Realism. It differs from it in only two respects: (1) It does not hold epistemological realism, but holds a closely related version of the Representational Model of Epistemology, since it holds that there cannot be a true complete representation. (2) It holds a version of semantic realism (see Chapter Four) that Putnam did not appear to be aware of. The very fact that there is an alternative to Metaphysical Realism that is not internal realism weakens Putnam's position considerably, since most of the argument he gives for his position takes the form of objections to Metaphysical Realism.

Another problem with Putnam's definition is that it is difficult to see how it implies the consequence that truth is radically non-epistemic, a consequence he sometimes identifies with Metaphysical Realism. It seems that this implication requires an additional assumption: that justification of a representation involves stepping outside of that representation to another which represents the similarity of the representation to its object. On this view justification always involves further representation, and, hence, must always be justified itself. Truth, or the actual relationship between the representation and the object, will always outrun the chain of justifying representations; the last justifying representation in the chain might always be false.

If, however, we have a view of representation in which there are self-justifying representations in the way explained in Chapter Six, then the very relation or act in which a representation becomes justified, the very relationship in which we become aware that it is true, is the very same relationship in which its truth consists. Truth becomes an epistemic notion. It would be surprising if it were not. After all, truth is a property of judgments, one that we become aware of in knowledge through justification. The world, of course, will always outrun our knowledge - our knowledge will always be incomplete - but partial truth will not outrun knowledge or justification as seen on this view, and partial truth is all we have. So, again, we should keep in mind that Putnam's definition of Metaphysical Realism allows alternatives that are not directly addressed by his arguments.

Let us now see what his arguments against metaphysical realism are.

8.2: Putnam's Arguments

We have already seen Putnam's main argument against Metaphysical Realism, the model theoretic argument, in Chapter Seven. It is essentially an argument against semantic realism. It attempts to show that representations cannot determinately refer to extra-representational objects. Recall that it showed this by showing how there could be alternative interpretations or mappings of any representational system onto the world. This constitutes a critique of Metaphysical Realism, because it would allow more than one true mapping of a language onto the world.

This argument rests on a view of representations as objects whose relation to their object is arbitrary. They are seen as ideas, words, sentences, or linguistic systems. Given that we have no direct access to the world there is no way for us to single out a unique correspondence between us and the world. The problem is not that there can be no similarity, but that there are too many. Putnam says:[2]

The trouble with this suggestion is not that correspondences between words or concepts and other entities don't exist, but that too many correspondences exist. To pick out just one correspondence between words or mental signs and mind-independent things we would already have to have referential access to the mind-independent things. ... you cannot single out a correspondence between our concepts and the supposed noumenal objects without access to the noumenal objects. (Putnam 1981, pp. 72-73)

The premise that we are cut off from the noumenal world by a veil of representations is essential to Putnam's argument. He says this after giving an account of his use of the Lowenheim-Skolem Theorem: "This simply states in mathematical language the intuitive fact that to single out a correspondence between two domains one needs some independent access to both domains." (Putnam 1981, p. 74) Elsewhere he sets out the problem in a similar way:

Early philosophical psychologists - for example, Hume - pointed out that we do not literally have the object in our minds. The mind never compares an image or word with an object, but only with other images, words, beliefs, judgments, etc. The idea of a comparison of words or mental representations with objects is a senseless one. So how can a determinate correspondence between words or mental representations and external objects ever be singled out? (Putnam 1983 p. viii)

This premise is essential to both his model theoretic argument and the internalist conclusions he draws from it.

Putnam sees that physical objects or signs cannot represent themselves; they are used by us to represent things. He says this in diagnosing the problem that leads to the model theoretic argument:

This is the fatal step. To adopt a theory of meaning according to which a language whose whole use is specified still lacks something - namely its 'interpretation' - is to accept a problem that can only have crazy solutions. To speak as if this were my problem, 'I know how to use my language, but, now, how shall I single out an interpretation?' is to speak nonsense. Either use already fixes the 'interpretation' or nothing can. ................................................. But the world doesn't pick models or interpret language. We interpret our language or nothing does. (Putnam 1977, p. 24)

Thus, Putnam sees that representations, and language in particular, do not interpret themselves. We interpret it, and our use of language is the interpretation. So, Putnam saw that it does not make sense to talk of a language or system of representations that has a program of use, yet lacks an interpretation. Even though Putnam saw that representations do not interpret themselves and that we do, he still remained strangely wedded to the model of representation that made the attempt to make representations come alive necessary. He still holds that objects or mental signs are themselves representations rather than things we use to represent. This assumption is what lies behind his premise that we must have independent access to both the representation and its object in order to determine reference. If representation were seen as an activity of interacting with the world that uses objects, signs, and mental images to re-present the world, there would be no need for independent access to the world. Representing is itself an access to the world, not a veil that keeps us from it.

We shall see shortly how Putnam uses this premise and his assumption that representations are objects which gain their reference through our use to arrive at his internalist conclusions. First, however, let us review some of the arguments that Putnam gives against causal theories of reference that seem closely allied to the view expressed here. We have considered two of these arguments in Chapter Four already, but we now need to look at these in more detail and see how they fit into his general argument for internalism.

We saw already that Putnam's first argument against the thesis that a causal theory of reference can explain how representations have a determinate reference was that this theory of reference was just more symbols which themselves had to have a determinate reference. This quote gives a good indication of what Putnam's argument is aimed at:[3]

The problem is that adding to our hypothetical formalized language a body of theory entitled 'Causal theory of reference' is just adding more theory. ... If 'refers' can be defined in terms of some causal predicate or predicates in the metalanguage of our theory, then, since each model of the object language extends in an obvious way to a corresponding model of the metalanguage, it will turn out that, in each model M, reference_M is definable in terms of causes_M; but , unless the word `causes' (or whatever the causal predicate or predicates may be) is already glued to one definite relation with metaphysical glue, this does not fix a determinate extension for 'refers' at all. (Putnam 1977, p. 18)

This argument, then, is aimed at the attempt make a system of representations have a determinate reference by adding a body of empirical sentences about the the reference relation itself. Of course, Putnam is correct to point out that such an attempt will not work. The set of sentences that spells out the reference relation does not itself refer determinately. The argument, however, simply amounts to making the trivial remark that we cannot have a theory of reference that is not a theory. The fact that talk about reference must itself refer does not make all such talk circular, unless one holds that reference must be established by representing the sign and the object and mapping the one onto the other. Putnam seems to hold such a view, as evidenced by his insistence that reference is impossible without independent access to the object. A causal theory of reference, however, is trying to avoid such a view.

As we saw in Chapter Four, however, in the interactional theory of reference presented here (and in most causal theories of reference) it is not a representation or theory of the interaction or causal connection that is supposed to determine reference; it is the interaction itself. We need have no representation of it at all.

Putnam himself was aware that this is not what the causal theorists intended:[4]

At this point in the dialogue, there is an argument that I invariably get from causal realists. This runs somewhat like this: 'You are caricaturing our position. A realist does not claim that reference is fixed by the conceptual connection (i.e., the connection in our theory) between the terms "reference", "causation", "sense impression", etc.; the realist claims that reference is fixed by the causation itself. (Putnam 1983, p. xi)

Putnam argues that this reply assumes the ability to refer or single out a unique correspondence relation between our use of the word 'cause' and the real causal relationship:

Here the philosopher is ignoring his own epistemological position. He is philosophizing as if naive realism were true of him (or, equivalently, as if he and he alone were in an absolute relation to the world). What he calls 'causation' really is causation, and of course there is a fixed, somehow singled out, correspondence between the world and one definite relation in his case. Or so he assumes. But how this can be so was just the question at issue. (Putnam 1983, p. xi)

This objection would be well taken if the causal theorist were trying to provide a demonstration to prove that we do in fact refer. But no one doubts that we in fact refer; what is wanted is an explanation of how this is possible. (The premise that we do in fact determinately refer to objects plays an important role in Putnam's arguments for internalism.) A causal theory of reference of the type provided here points to an aspect of our interaction with objects in our acts of representing and uses it in an explanation of our ability to refer. The fact that this ability is exercised in this very explanation does not make it circular any more than a talk about the anatomical structures that make speech possible is circular. Neither provides a demonstration that the activity is possible; the only proof of this would be to engage in the activity. They explain how the activity is possible. So, while Putnam's argument does show that a theory that attempts to explain how reference is possible by including a theory of reference in our symbol system will run into the same model theoretic problems, it will not apply to a theory that attempts to explain how various aspects of our interaction with the world allow us to refer.

Putnam has another argument against the causal theorist, however. Even if we allow causal relations between representations and their objects, there will be to many different relations to uniquely determine reference. This is simply an application of the model theoretic argument to the causal relation itself. A representation will have many different causal relations to many different things; how will the representation know which is the intended reference relation. Putnam says:

Given that there are many 'correspondences' between words and things, even many that satisfy our constraints, what singles out one particular correspondence R? (Putnam 1981, p. 46)

................................................

For, assuming a world of mind-independent, discourse-independent entities (this is the presumption of the view we are discussing), there are, as we have seen, many different 'correspondences' which represent possible or candidate reference relations... . (Putnam 1981, p. 47)

As we saw in Chapter Four, this objection does not apply to the interactional theory of reference offered here. A representation does not have to represent or refer to its own reference relation to refer; it simply has to enter into that relation. Interacting with the world in a way such that you isolate out a domain of interest does not require that you be able to determinately represent the domain of the reference relation beforehand. The force of Putnam's argument depends on the same assumptions as his general model theoretic argument. First there is the sound premise that reference is something we fix, that signs do not represent by themselves but only when situated in a program of use. But this is combined with the assumption that representations are still objects or signs, static entities that can only be related to their objects through our activity. If our only access to the world is through such entities that are caused in some mysterious way by interaction with the world, then our representations will form a veil between us and the world. Combine this with the first premise, and it will, indeed, be impossible to refer. As Putnam says in the quote above:

The mind never compares an image or word with an object, but only with other images, words, beliefs, judgments, etc. The idea of comparison of words or mental representations with objects is a senseless one. So how can a determinate correspondence between words or mental representations and external objects ever be singled out. (Putnam 1983, p. viii)

If representations are taken as activities of interaction with the world, then the veil of ideas falls away and becomes a bridge of ideas. Representing is a way we interact with the world, not the static result of some interaction we cannot be aware of. Thus, reference does not have to be established by an interpreter from outside of the representation by assigning it an object of which the interpreter is also conscious. Reference can be established within the representation, in the very activity of interaction with the world that is the representation.

Putnam saw that objects or symbols did not interpret themselves. Yet, he did not see that this made it impossible for such objects, themselves, to be representations. Objects even when interpreted from without by an interpreter by being assigned or matched with a meaning are not representations. We represent, not objects. Representing is an activity of conscious organisms. We sometimes use objects or symbols in this process, but they are only representations when situated in the context of such an activity. Representing is not something we do by assigning symbols an interpretation. It is the act of interpreting one thing in terms of another; it is the connection of different modes of interaction. Thus reference goes on within the activity of representing, as a part of it; it is not something we assign to the representation from outside by assigning it an interpretation.

Thus, it seems that Putnam's arguments rest upon the retention of aspects of the very model of representation that he is attacking by his realization that words do not interpret themselves. Let us now see how these assumptions function in his arguments for internalism, or what he calls internal realism.

8.3: Internal Realism

Internal realism is in essence a form of the internal model of objectivity. It reinterprets the object of knowledge as internal to the representational system and, thus, makes objectivity a matter of the internal properties of the system. Putnam's version of this model involves two theses: one concerning the ontological status of objects and one concerning the nature of truth.

As we saw in the last chapter, Putnam holds that objects are constructs within theories. He says in describing his view:

I shall refer to it as the internalist perspective, because it is characteristic of this view to hold that what objects does the world consist of? is a question that it only makes sense to ask within a theory or description. (Putnam 1981, p. 49)

and elsewhere:

'Objects' do not exist independently of conceptual schemes. We cut up the world into objects when we introduce one or another scheme of description. (Putnam 1981, p. 52)

and

If, as I maintain, 'objects' themselves are as much made as discovered, as much products of our conceptual invention as of the 'objective' factor in experience... . (Putnam 1981, p. 54)

This, of course, implies an Oatmeal theory of reality. If it only makes sense to talk of objects within a theory and if we make objects through the application of our conceptual schemes, then there must be insufficient structure in the world to differentiate itself into objects. This is the thrust of Putnam's article "Why there isn't a ready-made world". The second quote above is almost an exact statement of the Oatmeal theory: The world is a bowl of mush; you have to spoon it out yourself. So then, Putnam's internal realism involves the thesis that objects are constructs in our representational systems and that the noumenal world beyond our representations has no intrinsic structure.

The second thesis of internal realism is that truth is idealized rational acceptability. Truth can no longer be a correspondence to an extra-representational world; it is an internal property of our representational systems. Yet, it is not just what happens to be accepted in the theory at any particular time. Internal realism is not a relativist theory in which there are no external constraints on knowledge. Putnam says:

Internalism does not deny that there are experiential inputs to knowledge; knowledge is not a story without constraints except internal coherence; but it does deny that there are any inputs which are not themselves to some extent shaped by our concepts... or any inputs which admit of only one description, independent of all conceptual choices. (Putnam 1981, p. 54)

and

Just as the objective nature of the environment contributes to fixing the reference of terms, so it also contributes to fixing the objective truth conditions for sentences, although not in the metaphysical realist way. (Putnam 1983, p. 86)[5]

Thus, Putnam holds that truth cannot be acceptability within a system at any one time, but acceptability under ideal conditions. He says:

'Truth',in an internalist view, is some sort of (idealized) rational acceptability - some sort of ideal coherence of our beliefs with each other and with our experiences as those experiences are themselves represented in our belief systems... . (Putnam 1981, pp. 49-50)

and

I treat truth as an idealization of justification. ... A statement is true, in my view if it would be justified under epistemically ideal conditions... . (Putnam 1983, p. 84)

Putnam holds, however, that ideal epistemological conditions can never be achieved. (Putnam 1983, p. 84; Putnam 1976, p. 125; Putnam 1981, p. 55) It is a regulative ideal or limiting concept that allows us to bring into question things that are justified at present and even our present standards of justification.

Putnam waffles (to the point of inconsistency) as to whether theories will converge under ideal conditions, whether there is only one true theory. He says:

Many 'internalist' philosophers, though not all, hold further that there is more than one 'true' theory or description of the world. (Putnam 1981, p. 49)

and

... then incompatible theories can be true.

To an internalist this is not objectionable: why should there not sometimes be equally coherent but incompatible conceptual schemes which fit our experiential beliefs equally well. ... But the motive of the metaphysical realist is to save the notion of the God's Eye Point of View, i.e. the One True Theory. (Putnam 1981, p. 73)

Yet, he also says:[6]

...the two key ideas of the idealization theory of truth are ... (2) Truth is expected to be stable or 'convergent'; if both a statement and its negation could be 'justified', even if conditions were as ideal as one could hope to make them, there is no sense in thinking of the statement as having a truth value. (Putnam 1981, p. 56)

Thus, it is an essential part of internal realism that truth be more than just internal justification. Let us look at Putnam's arguments for this thesis before we look at those for the ontological thesis, since the arguments for the theory of truth are relatively simple. Putnam's first argument is simply that truth is supposed to be something that doesn't change. (Putnam 1981, p. 55; Putnam 1983, p. 84) What is justified within a particular system can change as the system changes, therefore, according to Putnam, truth must be some idealization of justification. Putnam also has a more subtle argument for his theory of truth. If truth is a matter simply of the conditions under which a statement can be asserted, then it seems that we can make no sense out of what assertions are other than physical acts or utterances. Putnam says:

If assertion is to be taken in a suitable 'thick' sense, however, then we have to recognize that asserting is guided by notions of correctness and incorrectness. (Putnam 1983, p. xiv)

This notion of correctness requires a conception of truth that goes beyond assertability. For to assert something is not just to utter a sentence; it is to utter the sentence as representing something that is true or correct. Without a realist theory of truth, assertion becomes just utterance, a movement of the larynx. Thus Putnam says:

If a philosopher says ... that knowing the assertability conditions is knowing all there is to know about truth, then, insofar as I understand him at all, he is denying that there is a property of truth (or a property of rightness, or correctness), not just in the realist sense, but in any sense. But this is to deny that our thoughts and assertions are thoughts and assertions. (Putnam 1983, p. xv)

I have spent this time reviewing Putnam's argument for a realist theory of truth, not because I mean to take issue with it, but to show how deeply realist intuitions and arguments still have a hold on him. The question to be considered in the next chapter is whether these intuitions can be assimilated to an internalist metaphysics and epistemology.

We now need to look at the arguments and motivations that led to Putnam's ontological position, his view that objects are constructions within theories and the Oatmeal theory of Reality that this implies. We saw in Chapter Seven that Putnam's ontological theses are a way of solving the problem posed by the model theoretic argument. Putnam diagnoses the problem in this way:

The predicament is only a predicament because we did two things: first, we gave an account of understanding the language in terms of programs and procedures for using the language (what else?); and then, secondly, we asked what the possible 'models' for the language were, thinking of the models as existing 'out there' independent of any description. At this point, something weird had already happened, had we stopped to notice. On any view, the understanding of the language must determine the reference of the terms, or, rather, must determine the reference given the context of use. If the use, even in a fixed context, doesn't determine reference, then use isn't understanding. The language, on the perspective we talked ourselves into, has a full program of use; but it still lacks an interpretation.

.................................................

We need, therefore, a standpoint which links use and reference in just the way he metaphysical realist standpoint refuses to do. The standpoint of 'non-realist semantics' is precisely that standpoint. (Putnam 1977, p. 24)

Putnam, then, is committed to a theory in which understanding is determined by use. To understand a language is to be able to use it. And to understand a language we must be able to determinately apply it; that is, we must be able to refer. So understanding must determine reference. The reference of a set of symbols must be determined by a program of use in which the symbol is assigned an extension by us. This is the first premise of Putnam's argument for the ontological thesis of internal realism. This is simply a result of the recognition that objects or symbols don't interpret themselves, that meaning cannot be a result of intrinsic similarities between representations and their objects.[7] Signs gain their reference by our interpretation; we assign them a referent.

Assigning a sign a referent is only possible if we have access to both the sign and the object to which it is to refer. We must have immediate access to each of these; that is, the access cannot be through another sign. This is the second premise of Putnam's argument. We have already seen, in Chapter Seven, ample evidence that Putnam held this view, but here is a representative quote: "This simply states in mathematical language the intuitive fact that to single out a correspondence between two domains one needs independent access to both domains." (Putnam 1981, p. 74)

The central premise of his argument, at least for our purposes, is that we never have access to the world; we only have access to representations. As we saw earlier, this seems to be a result of retaining a view of representation as something that signs or objects do, albeit only with our interpretation. But even if an object does not represent on its own, if the object is taken as a representation it will form a veil or barrier between us and reality. If representations are seen as objects or signs, even ones that only get their meaning through our interpretation, they will stand as intermediary objects between us and the world. We will never see the world, only these intermediary objects. (It should be clear by now why this is importantly wrong. On the view put forward here representing is an activity of interacting with the world; it does not form a barrier between us and the world. Objects and signs only have representative powers when we are using them to represent, when they are situated in an act of representing) We have already seen (in section two above) how important this premise is to Putnam's model theoretic argument. As an indication of this consider this quote in which Putnam explains how he bases his semantics on constructive, intuitionist mathematics and how this avoids the model theoretic problems:

'Objects' in constructive mathematics are given through descriptions. Those descriptions do not have to be mysteriously attached to those objects by some non-natural process (or by metaphysical glue). Rather the possibility of proving that a certain construction (the 'sense', so to speak, of the description of the model) has certain constructive properties is what is asserted and all that is asserted by saying the model 'exists'. In short, reference is given through sense, and sense is given through verification procedures and not through truth conditions. The 'gap' between our theory and the 'objects' simply disappears - or, rather, it never appears in the first place. (Putnam 1977, p. 21)

It is this gap between the representations and the objects that allows the model theoretic argument to have force, and it is this very same gap which Putnam's ontological thesis is a way of getting around. The gap, of course disappears in Putnam's theory because objects become constructions within the theory, to which we have immediate access. This what allows us to assign our signs a determinate reference.

Given these three premises we can now make a rough reconstruction of Putnam's argument for his ontological thesis. It has the form of a transcendental argument. It begins with the fact that we do refer determinately to objects and arrives at his ontological thesis as the only way in which this fact could be possible:

1. We do determinately refer to objects.

2. Reference must be fixed by us through our use; this involves the assignment of a referent to our signs through their interpretation.

3. We can only assign a referent to a sign if we have immediate access to both the sign and the referent.

4. We have no immediate access to noumena, things as they are in themselves apart from our representation of them. We have immediate access only to representations.

Therefore,

5. The objects to which we determinately refer must be internal to our system of representations.

Of course, it is the fourth premise above that I will take issue with. The theory of representation provided here does away with the veil of ideas making premise four false and allowing for an alternative way of explaining reference. (See Chapter Four.) In a transcendental argument such as this the presentation of an alternative explanation of the fact with which the argument starts is a serious objection, since the conclusion only follows if it is the only explanation of that fact.[8]

In Putnam's theory there is no problem of reference. The objects are constructs within the fully interpreted theory:

For an internalist like myself, the situation is quite different. In an internalist view also, signs do not intrinsically correspond to objects,independently of how those signs are employed and by whom. But a sign that is actually employed in a particular way by a particular community of users can correspond to particular objects within the conceptual scheme of those users. 'Objects' do not exist independently of conceptual schemes. We cut up the world into objects when we introduce one or another scheme of description. Since the objects and the signs are alike internal to the scheme of description, it is possible to say what matches what. (Putnam 1981, p. 53)

Since we cut up the world by describing it, the world apart from our description cannot already have structure apart from our activity. If it only makes sense to talk of objects within a scheme of description, then there are no objects in the world as it is apart from our conceptual activity. The world as it is in itself is a bowl of mush; we spoon it out into objects ourselves. What we call the 'world' and what we call 'objects', then, must be constructs in our theories and not the mush of the noumenal world. Putnam, therefore, seems committed to an Oatmeal theory of Reality. The world that we refer to is of our own making. Putnam uses as a metaphor a play in which we are both an actor and author:

Kant's image was of knowledge as a 'representation' - a kind of play. The author is me. But the author also appears as a character in the play (like a Pirandello play).

................................................

I would modify Kant's image in two ways. The authors (in the plural - my image of knowledge is social) don't just write one story: they write many versions. And the authors in the stories are the real authors. This would be 'crazy' if these stories were fictions. A fictitious character can't also be a real author. But these are true stories. (Putnam 1976, p. 138)

If the world we refer to with all its structure is a construct of ours within our theories, then the noumenal world cannot have any structure by itself. Thus, Putnam attacks the notion that there is a 'ready-made world' (Putnam 1981a), that there are self-identifying objects (Putnam 1981, pp. 53-54), and that there are essential properties (Putnam 1981a).

Before going on, in the last chapter, to a critique of internalism and a defense of the view of representation presented in Part Two, I need to first consider a passage that may seem at first to argue against the interpretation of Putnam's ontological thesis given above. Putnam considers the possibility that someone might respond to his arguments by saying that we refer in virtue of some correspondence relation to noumenal objects, but that there is no one privledged or intended reference relation, so that in different models or theories a term may have a different reference. On this view reference turns out not to be determinate.[9] This is a view Putnam calls ontological relativity. This, of course, rejects the very fact from which Putnam's argument begins. Putnam could be expected to be unable to understand a position where the properties by which we describe an object don't automatically apply to the object. Where under different theories you have the same object or part of the world characterized differently, instead of having a different object in virtue of the new theory. This is just how Putnam responds:

This doctrine, however, cannot be accepted. I cannot accept it for my own language, because to do so would turn the notion of an object into a totally metaphysical notion. I know what tables are and what cats are and what black holes are. But what am I to make of the notion of an X which is a table or a cat or a black hole (or the number three or ...)? An object that has no properties at all in itself and any property you like 'in a model' is an inconceivable Ding an sich. (Putnam 1983, p. xiii)

This should be read as arguing that the thing we call the 'world' is not the noumenal world, but the world as constructed within our theory. It should not be taken as an admission that things in themselves have structure after all. Here are some quotes that indicate this as the correct reading:

The fact is, so many properties of THE WORLD ... turn out to be 'theory relative' that THE WORLD ends up as a Kantian 'noumenal' world, a mere 'thing in itself'. If one cannot say how THE WORLD is theory-independently, then talk of all these theories as descriptions of 'the world' is empty. (Putnam 1976, p. 133)

These theories, then are not descriptions of THE WORLD, the noumenal world; they are descriptions of the constructed world within our theory. Putnam also says:

If the picture of the language user that we have thus far discussed ... leads to the metaphysical fantasy of a 'ready-made world', with self-identifying objects, 'built-in' structure, essences, or whatever, and the modified picture of the mind or brain simply accepting a whole lot of different correspondences, without trying to 'fix' any particular one as the intended correspondence between word and object, leads to the metaphysical fantasy of a 'noumenal' world, with no determinate relation to our experiential world, then the trouble with this entire discussion must lie at a deeper level. (Putnam 1983, p. xiii)

Here Putnam explicitly rejects the interpretation that his rejection of ontological relativity leads him to accept a world with built-in structure. The world that we refer to is neither a noumenal world with built in structure, nor a noumenal world with no structure, it is the world as constructed within our representational systems.

It is interesting to note the motivation of Putnam's rejection of ontological relativity, for it reveals much about the motivation of his internalism in general. Consider the argument in the quote above from "Realism and Reason" (Putnam 1976). It seems that the position that Putnam wants to retain at all costs is whatever it is that we refer to, the properties that we refer to it by are really true of it. "If one cannot say how THE WORLD is theory-independently, then talk of all these theories as descriptions of 'the world' is empty. (Putnam 1976, p. 133) Strange as it may seem, one of the main assumptions behind Putnam's internalism is something like a subject-predicate or substance-attribute metaphysics. Whatever objects are, the properties we apply to them, as we describe or represent them, really do apply to them. Objects really do have properties, the very properties we ascribe to them in our representations. The properties that we represent an object as having really are in the object. The structure of reality really does mirror the structure of our language or representations.

Thus, as Putnam says, he does hold a version of essentialism (Putnam 1981a, p. 220) and a version of self-identifying objects (Putnam 1981, pp. 53-54). (He is quick to point out that this is not in a sense of any use to the externalist, but this is only because Putnam had become convinced the versions of essentialism open to the externalist could not work because of the nature of reference and the veil of ideas. He would have liked to have accepted them, as he did before he was convinced that they didn't work.) Of course, Putnam's substance-attribute metaphysics, the fact that objects mirror the structure of our representations, is simply a result of the fact that objects are constructs within our representational systems constructed in accordance with our theories. But we should not, therefore, infer that this metaphysics is simply an implication of Putnam's internalism. Rather, it is a pre-assumption of Putnam's that leads to his Internalism. His rejection of ontological relativity shows that. He had an alternative to internalism, ontological relativity[10], yet he could not stand a theory on which we could refer to the world but not say how it was. Putnam was trying to be as good an essentialist as he could; it turned out that internal realism was the best he could do.

We should also discuss how Putnam's essentialism is caused by the retention of the physical-visual model of representation and, in particular, the perspectivist fallacy, but this leads us into a critique of Putnam's assumptions from the point of view of the theory of representation presented here. Therefore, let us continue this discussion in Chapter Nine.