CHAPTER IV

REFERENCE

Part II spells out some of the details of a model of representation that avoids the criticisms of the previous chapter. It will attempt to explain what representation can be if it is not similarity and how such a model of representation can be used in theory of knowledge. This chapter will address the particular problem of how a representation or a concept can be about an object without resembling it, that is, how there can be reference independent of similarity. It is the thesis of this chapter that referring is an act that we perform and that we are able to perform this act independent not only of any similarity between our representation and the object, but also independent of the correctness of the representation in general, independent of its correspondence. It is this ability that allows us to represent objectively and to transcend our present state of knowledge in a way that is not totally determined by that present state. It is this ability that makes the perspectivist fallacy a fallacy. For even if nothing can be like an idea but an idea, this does not mean that ideas are about nothing but ideas. This chapter, therefore, must bear much of the burden of the defense of the thesis of this dissertation, for our ability to represent and our ability to know by representing rest upon our ability to refer.

4.1: Representing without Similarity

In order to argue that we are able to refer independent of the correctness of our representations we first need to remind ourselves of what representing without similarity is, and we need to see what correctness of representation could be if it is not iconic resemblance. In this chapter I will be giving only a preliminary sketch of what correctness consists in. An account of what knowledge is, or how it is that we come to be aware that our representations are correct, will have to wait until Chapter Six.

In Chapter One, I held that representing was a matter of re-presenting one mode of interaction with the world in terms of another. It is essentially a connection of one mode of interaction with another by a projection of one onto the other. These modes of interaction can range from unconscious physical interactions of our bodies and sense organs with the world, in the simplest of cases, to complex abstract concepts such as circularity.

Representing is not a matter of having static atomic impressions through interacting with the world. It is a process or act of connecting interactions with the world. A paradigmatic example of representing is visual perception, where we project all of the sensory properties of the object from past experience onto our visual interaction with the object and see objects with all their various properties instead of merely seeing patches of color. The mere visual impression of an object onto our eyes and nervous system is not a representation. In fact, I argued in Chapter One, following Hegel, that such an impression on one sensory modality does not even constitute an experience. If we interacted with the world only through isolated impressions of one type, we would not be conscious of it. A rock, for example, interacts with the world in essentially only one way and is changed by that interaction. Concious perception, however, is not simply a matter of having changes impressed upon us by interaction with the world. Representing, and in particular conscious experience, involves the connection of modes of interaction.

Representation is not a matter of the similarity of impressions with their object. We should not, therefore, be discouraged by arguments that our impressions cannot be similar to external objects. Even if the impression of objects on our senses produced tiny replicas of the objects in our brains, we would not be conscious of them.[1] To become conscious of them we would have to connect their appearance with the interaction through which they arose. A camera is not conscious of the images projected within it. Representation is essentially a matter of making connections between impressions or interactions with the world.

It does not matter that our individual impressions cannot be similar to the external world because of the particular perspective they embody. Correctness of representation or correspondence is not a similarity of the phenomenological character of an impression to the object. Correctness of representation is a matter of the correctness of the connections made in the act of representing. It is not the particular properties that are connected in an act of representing that correspond or fail to correspond; it is the connections. Particular properties only enter our consciousness embedded in an act of representing, connected with other properties. We are only able to isolate and give our attention to particular properties such as redness through a complex process of abstraction.[2] The objective representational content of an act of representing, therefore, rests in the connections made within it and not in the similarity of the character of particular elements within the representation to its object.

For example, the correctness of my judgment that the type on this page is black does not consist in any similarity between my experience of its blackness, i.e. its phenomenological character, and the actual line of type. Rather, what is correct is the connection I made in my judgment between the blackness and the other properties of the page of type (spatial position, shape and combination of the letters, length, etc.) that allow me to pick the type out. This connection is correct if it connects two modes of interaction (and the properties these modes of interaction make manifest) that either interact with the same thing or interact with different things which are actually connected. In both cases the act of representing connects properties that are manifestations of things that are really connected. In the first case the two modes of interaction and the properties associated with them are different manifestations of the same thing. For example the blackness and the texture of the type are most likely manifestations of the same chemical structure which has different ways of showing itself depending on the type of causal interaction involved, light or touch. In the second case the modes of interaction are involved with different things which actually happen to be connected. This is the case in my connection of the color of the type with its spatial position on the page.

In general the correctness or objectivity of an act of representing consists in its having made connections between modes of interaction with the world that are caused by the things interacted with. That is, it involves connecting modes of interaction that have a common causal nexus in the part of the world interacted with in the act of representing. We will need to look more fully at this type of correspondence and how we can become aware of it in acts of knowing in Chapter Six, but this general account will allow us to see how our ability to refer is independent of the correctness of our act of representing.

4.2: Reference apart from Correctness. The Intuitive Case.

Even on the most cursory examination the position that reference depends upon similarity or correctness seems beset with insuperable problems. It seems to make it impossible to have false or incorrect representations. If reference is similarity or correctness, then a representation that is not similar to an object is not false of that object; it cannot refer to that object. It simply is a representation of something else.

Consider this example: A person under the influence of LSD is holding a cup, staring at it intently, and pointing at it. They declare: "This dog is a beagle." They have made an absurdly false statement. The absurd falsity of this statement is inexplicable apart from the fact that the person has retained their ability to isolate out a domain of interest and refer to an object, yet they do not have the slightest idea what it is they have succeeded in referring to. The statement is not absurd at all apart from this successful reference. If it were not clear that the person was talking about the cup, we might assume they were talking about some real beagle which was not present.

Cases such as these strongly suggest that reference depends on a set of abilities that are independent of the correctness of our representations. Hilary Putnam and Saul Kripke have suggested a theory of reference that begins to account for this separation of reference from representational correctness. Putnam's is by far the more sophisticated theory. I will concentrate exclusively on his views.

Frege was the first to make the distinction between representation and reference. He saw that sentences might have different representational or information content and yet refer to the same thing. Yet he still held that meaning or sense determined reference.[3] He still held that correctness of representational content is how reference was established. It was this claim that Putnam was concerned with refuting.

Putnam first argued, analogously to the argument above, that this claim made error impossible. (Putnam 1970) Holding reference to be determined by representational content makes statements of that representational content analytic truths. If, for example, the reference of 'lemon' is determined by the list of properties that we represent lemons as having (let these be properties L1, L2, L3), then any statement of the form 'Lemons are Ln' is analytically true. If an object does not have property Ln then, necessarily, it is not a lemon. It does not seem, then, that we can be mistaken in our ascription of these properties to lemons. Any representation that does not include these properties does not refer to lemons at all.

Our ability to refer independently of correct representation, therefore, is the basis of our ability to be in error, and most importantly, of our ability to correct our errors through further interaction with the object of the representation. We now need to see what Putnam took as the basis for this ability to refer independently of representational content.

Putnam (1975) advanced a type of theory that has come to be known as a causal theory of reference. The basic idea of a causal theory of reference is that reference is determined by a causal relationship of the appropriate type between a use of a word and the object or natural kind to which it refers. A causal chain leads back from the present use of a word to an original ostensive definition or baptismal event in which the name was originally applied to the immediately present object. This causal chain most often goes through the many uses of the term that have interceded between my uses and the original use. Reference, then, is a matter of being in the appropriate causal relationship to the object, not of having a correct representation of it.

Although Putnam's theory does involve some type of causal interaction with the object as an important component in reference, it would be a mistake to attribute the crude type of causal theory described above to him. It seem to me that there are three main characteristics of his theory that preclude this reading.[4]

First, Putnam holds that it is not necessary for there to be a causal chain between a particular use of a term and the object to which it refers. Reference can be established through the mediation of an introducing event in which a description of the object or natural kind is given. (Putnam 1973, p. 200) This is possible because, for Putnam, language is not an individual matter. There is a division of linguistic labor. We are able to refer to objects that we have never interacted with in virtue of belonging to a linguistic community that contains certain experts who have interacted with the objects and have knowledge about them that is not readily available.[5] (Putnam 1975, p. 227) Thus, while a causal interaction at some time, by someone is necessary for reference, all acts of referring need not be tied by a causal chain to the object referred to.

Second, it is more appropriate to say that the nature of the object determines the reference for Putnam than it is to say that a causal chain does. If I am introduced to the term 'water' in association with a particular instance of water that I have interacted with, I am then able to refer to innumerable bodies of water that I have never interacted with in virtue of their having the same nature as the water with which I have interacted. Thus, Putnam holds that 'water' refers to a particular substance, X, if and only if X "... bears a certain sameness relation (say, x is the same liquid as y, or x is the same_L as y) to most of the stuff I and other speakers in my linguistic community have on other occasions called 'water'." (Putnam 1975, p. 225).

Putnam's theory diverges on this point even more from the crude version of the causal theory of reference sketched above. For Putnam argues that the sameness relationship that determines the set of objects referred to by a representation is interest relative. (Putnam 1975, pp. 238-239) That is, whether an object is the same as the set of objects you usually refer to with a term, and, hence a part of the reference of that term, depends upon what is important in that particular context. For example, whether D₂O (heavy water) will be considered water will depend on whether one is interested in drinking it or using it in a chemical experiment. Therefore the set of objects isolated as a domain of interest in an act of referring is not determined by any unique causal relationship between the act and the set of objects, although some type of interaction with members of the set is necessary. The domain isolated out and referred to depends upon our interests and the context of our act of referring.

Third, in Putnam's theory the representational content sometimes does play a role in referring. When we are introduced to a term in an introducing event, the description that is used to introduce us to the term must be correct enough to lead us (the linguistic community) to an interaction with the appropriate object or type of object. (Putnam 1973, p. 201) For example, if the term 'unicorn' was introduced to a community in a fleeting encounter with a rhinoceros, your description of the animal (as a horned horse) would be so faulty that it would not allow you to isolate the animal out for interaction in the future. So, even though you were causally related to the rhinoceros in what seems to be the appropriate way, you did not succeed in referring to it because of problems with your representation of it.[6] (I will argue in the next section that this is not an exception to our thesis that reference is independent of correctness of representation.)

While it is clear from the preceding discussion that Putnam's theory is not a crude version of the causal theory of reference, it is not quite clear what type of causal theory of reference one is left with after making Putnam's concessions. Nor is it clear exactly how it is that we refer if it is not in virtue of some special causal relationship between our representations and their objects. In the next section I will attempt to answer these remaining questions within the framework of the model of representation that we have discussed so far.

4.3: Representing and Referring

It seems to me that the term 'reference', as it is generally used (by normal people, not philosophers), is a general term encompassing three more particular types of reference: direct reference, Intentionality[7], and semantic reference. The relationship of these three types of reference is simple: All Intentionality is derivative from direct reference, and all semantic reference is derivative from Intentionality. All reference, therefore, is derivative from direct reference. These terms are explained below.

Direct reference is accomplished in acts of representing that connect two or more modes of interaction both of which are actively engaged in interaction with the external world. This is not true of all acts of representing; it is only true of representations of the first type distinguished at the end of Chapter One.[8] At any rate, acts of representing that are so actively engaged make a connection between two modes of interaction as giving rise to properties that spring from a common causal nexus in the world. In doing so they isolate out a domain of interest or an object as the causal nexus of the two sets of properties connected.

The act of representing does this irrespective of the correctness of the connection made within. The person on LSD who says of the coffee mug "This dog is a beagle." isolates an object as the common causal nexus or ground of the two properties just as effectively as I do when I say "This cup is a coffee mug." We know where to go to prove them wrong just as precisely as we know where to go to prove me right. Even though they completely misconstrue the information that arises from their interaction with the cup, they do succeed in isolating the cup as the causal nexus of the properties they connect. Acts of representing in which both modes of interaction are actively engaged in interaction with the world in this way are also acts of direct referring. They pick out an object or part of the world as the common nexus in virtue of which the two modes of interaction are connected.

Such an act of representing directly refers even if the properties or modes of interaction connected in the act really do not spring from the object as their common causal nexus, as was the case with the person on LSD. The two modes of interaction need not even interact with a single unified object. The simple act of connecting the two modes of interaction isolates out what the two modes interact with as a domain of interest whether in fact it forms a common causal nexus for the modes or not, i.e., whether the connection made is correct or not.[9]

The reference established in an act of direct referring does not outlast the act itself. This was the case in the brief glimpse of the unicorn/rhinoceros. It would also probably be the case with our friend on LSD. Once he is distracted from the cup he will probably have no memory of his earlier claim and will be bewildered if you ask him "Could I have a drink from that beagle?" The type of reference that we are most interested in, therefore, is not the type established in acts of direct referring, since this type of reference does not outlast the act. We are interested in Intentionality. Intentionality, like all forms of reference, is derivative from direct referring, however. The reference established in acts of direct referring is passed on only in those cases where the act leads to the formation of concepts with Intentionality. We now need to see how this can occur.

Intentionality is a property of concepts. It is also a property, derivatively, of acts of representing that do not directly refer if those acts of representing are the expression or application of Intentional concepts. (Recall that on this view concepts are merely the active structures that do the representing, so it is natural that the activity of Intentional concepts should also be Intentional.)

Concepts are complexes of dispositions to make connections between modes of interaction in acts of representing. Some concepts (most of them) have Intentionality; they have this property in virtue of including dispositions that lead to the performance of acts of direct referring upon specific domains or types of domains of interest. Concepts are about things in virtue of leading to direct interactions with those things. In these interactions the things are isolated out as domains of interest in an act of direct referring.

Most of the modes of interaction that are involved in representing the world are channeled through our sense organs, the outer surface of our body, and the proprioceptive receptors that allow us to perceive the position of our body. We possess a number of basic abilities to orient our sense organs, our body, and the relative position of parts of our body with respect to objects and to manipulate objects with respect to their orientation to our sense organs and body. The appropriate exercise and combination of these abilities lead to acts of direct referring.

For example, I have the ability to direct and focus my eyes on the cup in front of me, and I also have the ability to reach out and touch it with my hand. I can also combine these activities in an act of direct referring. My concept of the cup is Intentional and has as its object this cup because it includes these dispositions to channel these two modes of interaction into the same domain of interest. (My concept of the cup, of course, also has a great many other such dispositions that could lead to direct reference to the cup apart from the particular ones exercised in this example.) Thus, concepts have Intentionality in virtue of including dispositions to orient the body and sense organs in such a way as to bring about the connection of modes of interaction in an act of direct referring.

Concepts make connections between modes of interaction; they are the active structures whose expressions are acts of representing. The role of reference, or Intentionality in concepts, is, as the etymology of the word 'reference' would suggest, to lead the application of the concept back to the domain of interest from which the connection made by the concept arose. The Intentionality or reference of a concept to a domain of interest, then, is a matter of that concept being formed in interaction with that domain in an act of direct reference in a way that allows and leads the concept to be reapplied to the same domain.

Semantic reference is a property that signs have. Signs are physical objects or activities, such as statues, maps, and language. As we saw, such objects cannot represent by themselves; they require the interpretation of an agent with certain concepts. Signs have meaning and reference in virtue of being conventionally or naturally associated with certain Intentional concepts which in turn are dispositions that lead to particular acts of representing. Signs, therefore, are not themselves representations. They, along with their associations to concepts, are ways of getting us to reproduce particular acts of representing. They are immensely useful for the communication and storage of information, but they are not themselves representations, and they do not themselves have reference. Semantic reference is wholly derivative from connection with Intentional concepts, whose reference in turn is derivative from the acts of direct referring from which they spring and to which they lead back. (As we shall see in Chapter Seven, seeing language as not being itself a representation but as a conventional device for reproducing acts of representing by accessing particular concepts solves many of the puzzles that have troubled recent philosophy of language.)

We now need to see how reference described in these terms is independent of correctness of representation. In one sense this in non-problematic. Direct reference, as we saw, is clearly independent of the correctness of the connection made in the act of isolating the domain of interest. Insofar as all reference is derivative from direct reference, then, it is independent of correctness.

While Intentionality ultimately depends on direct reference, the abilities that lead the concept to acts of direct reference and account for their Intentionality do not themselves depend on direct reference. Therefore, we also need to see if these abilities are in any way independent of correctness.

For one class of Intentional concepts this also seems unproblematic. Many concepts are led back to interaction with particular domains of interest simply by the inclusion of basic abilities to reorient the body and sense organs the way they were originally. The person on LSD may be able to refer to the cup again even though they think it is a dog by simply reaching out to the same area of space in which they had found it before.

Often, however, concepts are led back to interaction with the same domain with the help of cues that are contained in the content of the act of representing that the concept expresses. For example, I can refer to a building or location, even if I have never interacted with it, if my concept of it includes the ability to represent its address or exact spatial coordinates. In this sense, as Putnam pointed out, sometimes the description plays an important role in fixing reference. Sometimes, the description plays an essential role in leading the concept back to interaction with the object.

This, however, does not imply that Intentionality is dependent on correctness of representation. First, even though the content plays an essential role in in fixing reference, this role does not require correctness of that content. It only requires that it lead back to the original domain; this does not require correctness. For example if your description of a location is 'one block north of the biggest oak tree you've ever seen next to the McDonald's restaurant', it may lead you to the location even though the large oak tree you saw is clearly not the largest you have ever seen and the restaurant one block north of it is a Burger King. If someone tells you to find Charlie's wife, that description may lead you to the intended person even though it turns out that Charlie is not married, although he does have a lover. Thus the representational content need not be correct as long as it leads you to the right place. Second, even if the part of the representational content that leads back to the object is correct, this does not rule out the remainder of the content being blatantly false. Therefore, at least in the sense that it leads back to a domain with reference to which it can be verified or falsified, Intentionality is also independent of correctness.

We now need to see what kind of causal theory of reference we are left with on this view. What makes it accurate to call this view a causal theory of reference at all is that it holds that reference requires some interaction with the object, some act of direct referring. This interaction is clearly, however, not a single type of static causal relationship. First of all, referring is an act we perform, an interaction, not a static relationship between a representation and an object. As an act, referring is a process; it takes time. An essential part of isolating out a domain of interest is connecting different modes of interaction at different times and attributing them to a domain as their common causal nexus. Such an act cannot occur in isolation, it is always part of a dialectical process of representing the world according to concepts, altering the concepts in the interaction, and reapplying the concepts in a subsequent act of representing. We can analyze out particular acts of representing or experiences, but they, like the components of an act of representing, have the properties they do only in the context of a wider process. Referring is an act that requires time. It is something organisms do, not a property they have in virtue of their causal relations at any one moment.

Second, there is no one type or natural kind of causal interaction that is reference. There are as many types of interactions that can establish reference as there are acts of direct referring. Just as their is no natural kind of interaction in which putting a basketball through the hoop consists, there is no one type of interaction that is referring. Yet we seem to have no problem in achieving either of these results (nor in failing to achieve them). Neither activity is a static relation that can be characterized generally; both are acts.

4.4: Putnam contra a Putnamian Theory of Reference

Putnam, in his later work, provides a critique of the type of causal theory of reference he once held. Although I will discuss Putnam's arguments as a critique of Metaphysical Realism in great detail in Chapter Eight, it is necessary here to at least sketch out the general outline of his main argument and my responses.

Putnam's main line of argument is that causation is essentially a theoretical notion, a way in which we represent certain aspects of the world. He develops this into two main objections to causal theories of reference: (1) If a causal theory of reference is just another theory, just another representation, it cannot determine reference independently of the correctness of representation. Its ability to determine reference depends upon its ability to correctly represent the appropriate causal relationship. He says, "The problem is that adding to our hypothetical formalized language of science a body of theory entitled 'Causal theory of reference' is just adding more theory." (Putnam 1977, p.18) This does not make reference independent of correctness, for the theory of reference must be correct to refer determinately itself:

Notice that a 'causal' theory of reference is not (would not be) of any help here: for how 'causes' can uniquely refer is as much of a puzzle as how 'cat' can, on the metaphysical realist picture. (Putnam 1976, p.126)

(2) Putnam also argues that causation is essentially an explanatory notion and that it does not make sense to ascribe causal relations to things apart from our conceptualization of them. Putnam says:

... the notion the materialist really uses when he employs 'causal chain', etc., in his philosophical explications is the intuitive notion of an explanation.

But this notion is certainly not physically definable. (Putnam 1981a, p. 213)

Even if we could ascribe such obviously explanatory notions to things in themselves, there would be too many causal relations, or too many explanatory chains, to determinately fix reference. What counts as an explanation depends on the context.(Putnam 1981, p. 46; 1981a, pp. 211-214)

I will answer each of these arguments in turn. Before I begin to respond to the first argument, I should point out in passing its similarity to our favorite example of the perspectivist fallacy, Berkeley's argument that we cannot conceive of material objects because we can form no conception of them which is not a concept. So Putnam seems to argue that a theory of reference as independent of representation and theorization is impossible because we can't give a theory of reference that isn't a theory.

Putnam's first argument does have force against those who think that by adding a representation of the reference relationship to our system of representations we can make it come alive and represent apart from our interpretation. But this, of course is not the view espoused here. People without a causal theory of reference refer just as well as those of us that have heard of it. Reference is not a matter of having a representation of the relationship between ourselves and a domain. It is a matter of making a relationship in an act of representing.

One need not already have a representation of the relationship to make it, nor need they already have a representation of the domain to isolate it out in an act of referring.[10] You need not have a determinate representation of how the ball will go through the hoop to make a basket. You need not have a determinate representation of exactly the molecules of water that will be included to scoop a cup of water out of the lake, yet the result is determinate. Nor need you know exactly how many crumbs will be included to cut out a piece of cake.

To do these things requires concepts, and concepts are by their nature indeterminate. But the expressions of these concepts are indeed determinate. Concepts are not representations and, hence, action does not require determinate representation of the outcome. Determinate outcomes arise in the act of applying concepts to the world in acts of referring and representing. Representations are not mental entities that stand as a veil between us and the world and thus require a unique reference relation to allow us to gain access to the world. Representing and referring are acts by which we directly interact with the world. They need no pre-existing relation to carry them across the veil of ideas. There is no veil of ideas, and they make their own relations to the world as they go. Thus, the fact that the theory of reference provided here is itself a theory does not prevent it from explaining how reference is possible.

We now need to answer Putnam's second argument. This argument seems to rest on an ambiguity in our notion of causation. It seems that Putnam is right that our normal notion of causation is an explanatory one. When we ask for the cause of something we want an explanation of why it is the way it is. There is, however, one type of explanation that is so common that it has come to be called by the name of the more general notion, causation. This is explanation by physical interaction. The billiard ball moved because another ball hit it. In this sense, causation is not just an explanatory model or account. It is an actual interaction whose force can be felt and which can be referred to apart from any explanatory role it might play.[11]

If someone were to sneak up behind me and give me a whack on the side of the head, I would feel the force of this interaction without being able to explain it at all or using it to explain some other change. The interaction does not explain the felt force; what is felt is the interaction. Any explanatory account of the interaction, of course, would not exhaust or adequately represent the felt interaction itself.

This sense of causal interaction is not a causal explanatory chain, and Putnam is right to point out that this type of causation cannot be given a physical account (nor a mental one for that matter). Any account would pick out certain features of the interaction as felt or experienced as particularly relevant or salient for some explanatory purpose. (The interaction as felt has a force not exhausted by any explanation, which, I suppose is in some oblique way the source of the adage that what a person who is unable to grasp the force of some explanation needs is "a whack on the side of the head".)

It may be more appropriate to call the theory advanced here an interactional theory of reference to distinguish it from cause in the sense of explanatory chain. The role that the interaction plays in the theory of reference advanced here is not an explanatory or representational one. The interaction is not itself an explanation or representation of the reference. It is the act of referring and representing itself.

Of course, there are too many relationships or interactions going on in an act of referring to allow us to pick out just one as being "reference". Luckily, we do not have to pick out the interaction from the list of possible ones; we simply have to enter into it.

There are too many relationships between a ball and a basket to pick one out as scoring a basket, yet this does not hinder us in putting the ball through the hoop (though many other things might). There are too many relationships between ourselves and a cup of water to specify one as scooping out that cup of water, yet we scoop the water out just the same. We do not have to pick out which act of referring we are going to do; we just have to do it. After all, it is not the interaction that we have to refer to. The interaction allows us to refer to the object.

Thus, the causal or interactional theory of reference advanced here holds that referring is an act of isolating out a domain of interest by connecting two modes of interaction with the world and attributing them to the domain as their common causal nexus. This theory of reference is the answer to the strong version of the perspectivist fallacy. Recall that the strong version held that it is impossible from within a perspective or from a system of representation to refer to anything outside of that perspective or system of representation. This rested on the view that reference depended on correct picturing. Particular pictures are so distorted by perspective and representative medium that they cannot picture anything outside of that perspective. Nothing is like an idea but an idea.

Once we see that referring is an act that can be accomplished independent of correctness of representing, we can see that no matter how distorted our perspective it still can isolate out a domain of interest, just as spatial perspective defines a point in space. This ability is the basis of a non-perspectivist external model of objectivity. Only if we are able to direct our activity to a domain apart from the correctness of our particular perspective will interaction with that domain be able to constrain our representation of it from that same perspective.

We have also seen the beginnings of the answer to the weak version of the perspectivist fallacy in this chapter. If correctness of representation is a matter of the correctness of the connections made in an act of representing, rather than the similarity of the properties that arise from the particular modes of interaction, then we can begin to see how there can be objective representation without similarity. If the modes of interaction (and the properties they manifest) connected in the act of representing really do have that domain as their common causal nexus, then there is an aspect of the representation that is caused by the object and not by the subject. This can be so even though the particular quality of the properties that arise from our modes of interaction is completely dependent on our peculiar physical and conceptual makeup. The objective content of acts of representing lies not in the properties involved in the representation. These are only the manifestations of particular modes of interaction. Objective content lies in the connections between these properties and feeling of those connections in an act of representing.

We will see how it is that we can become aware that the connections made in an act of representing are caused by the object and not the subject in Chapter Six. But first we need to find out more about concepts. We need to know more about what they are, how they work, why they are not representations, and why they have caused so much trouble in the history of philosophy.