CHAPTER VI

AGENCY, OBJECTIVITY, AND TRUTH

The main thesis of this dissertation is that representing is an act, something we do. The combination of topics in this chapter should not, therefore, be surprising. In order to take the thesis that representation is an action seriously, we will have to see what is involved with agency or the application of concepts. In order to see how an act of representing can be objective, we will have to see how a concept can be applied to a domain in a way that allows the representation that results from the application to be caused by the domain and not the concept. We will find that it will not do simply to apply traditional notions of agency to the new model of representing, for the assumptions of the physical-visual model of representation and the perspectivist model of objectivity are built into the traditional view of agency. Therefore we must begin by seeing what agency, or the application of concepts, can be on this view.

6.1: Agency

Traditional views of agency have been concerned with explaining how humans can act in a way that is not determined by the chance and indifferent happenings of the physical world. They have mainly taken the view that we are able to act freely in virtue of our rationality. Our rationality is seen as something transcending the physical world and, therefore, able to act freely and to direct our lives in a way that will have a value independent of the particular, transitory, and indifferent goings on in the material world.[1] Reason, by virtue of its commerce with abstract ideas, forms, and principles, defines an aspect of our nature that transcends the material world providing us with both freedom from its chains and shelter from its travails.

Agency according to this traditional view is a type of causality that our reason has with respect to our material existence. Kant says:

Will is a kind of causality belonging to living beings so far as they are rational. Freedom would then be the property this causality has of being able to work independently of determination by alien causes... .(Kant 1964, 446, p. 114)

This is only the negative definition of the freedom our rationality has from constraint by material things. One of these material things, for Kant, is our own physical constitution, so that included in the things that constrain us are our passions, emotions, sensations, and feelings. The positive conception of freedom is action in accordance with law or principle that the agent as a rational creature makes for itself.[2] Thus, agency, involves the formation of a representation of the action to be brought about. This representation then exerts a causality upon our physical being to guide its action in a way that frees it from physical determination.

Since this view of agency presupposes the ability to form determinate representations, it can hardly be used to explain acts of representing. Let us look at the shortcomings of this view of agency, then, to see if they lead us to a more promising view.

The first problem stares one in the face immediately. If reason is transcendent, the causality it is able to exert on the material world seems a mystery. Once one separates off reason from the causal effects of the material world, getting back to affect the material world becomes difficult. This problem becomes extremely pressing if the split between the material realm and the transcendent realm cuts one in half, and one is dependent for their welfare on their ability to use reason to guide their action in a way that isn't prey to every passing pathological whim or fancy. Kant's solution is to invent another type of causality, self-causality. This wouldn't make sense unless the self were split in two; causation is a concept used to relate two different things or temporal states of things. The noumenal self is seen as causing the empirical self, and not the other way around. In which case it doesn't seem to be self causality at all, but the causality of our true self upon its image in the empirical world. So, besides being an ad hoc solution to the problem, the notion seems to have internal inconsistencies, not to mention the inconsistencies with other aspects of Kant's doctrine.

It seems, then, that whatever agency is, it cannot be the causation of a transcendent reason. Neither can it be the causality of a representation either transcendent or immanent. For as we saw, such a static picture or representation cannot be wholly determinate, and will never determine a particular action in all its details. An agent that simply acted out predetermined representations would seem stupid and stiff to us. They would be unable to respond to unanticipated details and changes in the situation. They would not exhibit intelligence, the ability to creatively respond to the constraints of the situation.

This has important implications for the theory of knowledge. In making agency free from possible determination by particular situations, it is also made impervious to adjustment to constraints and learning from interaction. If reason is to be transcendent it cannot be altered by its actions on the material world. The Greek response to this, at least on the traditonal reading of Plato and Aristotle, is to make reason a faculty that deals with timeless objects and to argue that man's true nature is to be realized by making himself timeless as well by contemplation of the timeless entities. To learn from experience reason cannot be transcendent. Thus the effect of the scientific revolution on philosophy is Hume's critique of the role of a transcendent reason in morals and science.

Kant saw that understanding, or the faculty of concepts, was active in experience. Once one sees this there are two ways that one can turn. Kant's move is to keep reason transcendent, to keep it out of the fray, like the eromenos, the beloved, moving others but never moved.[3] This allows the concepts that structure experience to remain representations, but it sets the phenomenal objects constructed by the concepts up as a barrier between concepts and the world. Phenomenal objects become another level of representation for Kant. He replaces the veil of ideas with a veil of phenomenal objects. This is only natural; if reason is to be kept transcendent some barrier must be formed between the direct interaction of concepts with the world. If there are no longer passive ideas to fit the bill, the passive effects of the activity of our concepts, phenomenal objects, will have to do.

The other alternative is to bring concepts down out of their transcendent realm and make them interact directly with the world. This is the view urged here (and the one urged over two hundred years ago by Thomas Reid). Once representing becomes a way of interacting directly with the world, and concepts become dispositions to enter into such interactions, our activity becomes a way of contacting the world rather than a way of building a barrier between ourselves and the world.

Agency, then becomes a matter of the activity of concepts, as discussed in the previous chapter. It requires that concepts be dispositions that have a subjective character that actively drives the concept towards its expression. That is, concepts as dispositions to connect modes of interaction through joint activation, must be able to and tend to activate the modes of interaction they connect independent of contact from the external world. It also requires that we be able to control the activity of these concepts.

This control need not require a homunculus that oversees and controls the activity of our concepts. It can simply mean that the total network of my concepts (which after all is me) with its system of inhibitory and reinforcing connections determines which of my concepts will be active and which will not. On this view, each of us is a set of physically embodied potentials. Many of the potentials that make us what we are arise from the combination and organization of physical structures, rather than from the potentials inherent in the particular structures by themselves. Conscious organisms are organized so that their activity is primarily an expression of their own complex potentials rather than those of the environment and the simple potentials of their physical components. (They do this mainly by taking and exploiting energy from the environment.) Their freedom consists in this, and their agency consists in their activity being the expression of their own complex potentials rather than the environment or the particular potentials of their components. Concepts act in and of themselves through the expression of the potentials embodied in them. We, as a system of concepts, act, through the expression of the potentials that this system has as a whole.

The fact that we are a person simply consists in the fact that process by which one state of this total network passes on to the next has a coherence in virtue of the connections within the network and that my experience of this process has a unity in virtue of its continuity. Both of these conditions can be interrupted from without. The price of making reason vulnerable to outside influences is making the self open to dissolution by circumstances beyond our control. This should not be surprising; we are not immortal after all.

Agency, then, is the active tendency of our concepts to express themselves guided by the subjective character of their activity. This ability of our concepts to act autonomously, i.e. apart from activation from the external world, as expressions of their own potential, can be disturbed by interaction with the external world. In this lies the possiblity of our having objective representations.

6.2: Objectivity

It is the ability of the application of our concepts in acts of representing to be satisfied or frustrated that makes it possible for us to arrive at objective representations. In our application of our concepts to the world, the connections they tend to make can be reinforced or inhibited. This allows us to engage in a dialectical process of interaction with the world that allows us to arrive at representations that make connections that are caused not by our concepts, but by the world.

We are able to come to have objective representations even though the phenomenological character of our representations is determined by the perspective from which it arises. It can do this because the representational content of our acts of representing lies in the connections made in them, not in their phenomenological character. In an act of representing we connect modes of interaction and attribute them to a domain of interest as their common causal nexus. If that domain is in fact the common causal nexus of the activity of the modes of interaction, then the representation is true.

We can come to be aware that our acts of representing are true, we can know by representing, through the peculiar characteristics of our concepts and their expression in agency. Our concepts are clusters of teleologicl tendencies to connect modes of interaction in acts of representing. This means that they are not only capable of doing this upon prompting of the outside world; they tend to do it on their own. Their subjective character drives and guides them to acts of representing, just as an itch leads to a scratch. This means that they can connect modes of interaction by jointly activating them apart from their activation through interaction with the world.

The application of a concept to the world involves taking two or more modes of interaction and applying them to a domain of interest which they isolate. These modes of interaction will be being activated by the operation of the concept which tends toward their connection. It will be attempting to represent the domain as the common causal nexus of the activation of the two modes of interaction. It does this by interacting with the domain through the modes of interaction while activating them. The joint activation of the modes of interaction by the concept can be either reinforced or inhibited by the interaction with the domain.

For example, consider an attempt to find a red book on a desk. Imagine that we have two modes of interaction: one whose activation we experience as the visual shape of a book, the other's activation is experienced as the color red. Imagine also that we have a concept that includes two sets of dispositions: one set tends to activate both of the modes of interaction above, the other directs our sense organs so as to apply the pathways that are able to activate these two modes of interaction from the outside to the various objects on the desk. The concept works to attempt to have a representation of a red book by jointly activating the two modes involved. In the interaction with objects on the desk the joint activation of the two modes can be reinforced or inhibited through the pathways from the senses. If we light upon an object which activates both the book mode and the red mode, we feel this reinforcement as the satisfaction of the teleological tendencies whose expressions guide our activity.

The concept will tend to interact with the domain in ways that reinforce the joint activation of the modes of interaction. It will act to do this guided by the subjective character of the experience. If it is able to settle on a way of interacting with the domain that reinforces the activation produced by the concept, then it has arrived at an act of representing that satisfies the teleological tendencies that make up the concept. We are aware when we have arrived at such an act of representing in virtue of its subjective character. In previous discussions of this view, it has been described to me as a phenomenological glow theory.[4] This seems an accurate way of describing the satisfaction as the overall character the experience has as the fulfillment of the teleological tendency, rather than some particular quality added onto the experience. But there is no need to describe this glow in detail. We all know what it feels like; it feels good.[5]

In such an act we have represented the domain as having connections which it in fact does. The connections made in such an act are objective because they are determined by the domain interacted with and not by our concept. Only if the domain activates the same connected set of modes of interaction as the concept will the connections made by the concept be reinforced and satisfaction be felt. Such acts of representing are self-justifying in the sense that it is not necessary to go outside of them to become aware of their objectivity. They bear on themselves the mark of their objectivity, their phenomenological glow. They can be described in terms that apply to Kant's reflective judgement. They cause pleasure because we find the world in accordance with our concepts even though our concepts did not cause it to be so. In other words it is the very objectivity of the connections that makes them satisfy. The amount of joint activation provided by the concept alone can be viewed as the neutral level. The concept aims at going above that level and is frustrated by any diminution. Only a reinforcing activation caused by an interaction with a domain that is a common causal nexus for the activation of the two modes of interaction can provide satisfaction by raising the level of joint activation above the neutral level.

What makes our concepts capable of arriving at knowledge is the fact that they are both active and open to alteration in the application. If they were completely passive, we might have objective representings but we could never be aware when we were having them. We would simply take whatever connections were forced on us. If we were completely active, we could never make objective connections because they would always be due to our activity. What makes objective knowledge possible is the fact that our agency is embodied and involves active tendencies that go to the world attempting to represent it in a certain way but which may be frustrated, satisfied, or altered in the process of interaction.

It is important to note that objectivity is arrived at here not through moving out of the perspective to take up new ones from which the original perspective can be judged. This is what the perspectivist model of objectivity would expect. Single perspectives and individual acts of representing, however, can be objective. The connections that are made in a particular act of representing can be caused by the domain with which it is interacting and not the perspective. The particular properties that modes of interaction give rise to in an act of representing are determined by perspective and the type of sensory apparatus we have, but the connections between the effects of a domain on these modes of interaction can be caused by the domain and not our activity. Perspectives are not windowless rooms from which there is no escape. They are themselves windows onto the world. They give a limited and incomplete view of the world, but a view nonetheless.[6]

6.3: Truth

The view of truth implied in the previous section is a correspondence theory of truth. The act of representing that produces satisfaction co-responds with reality. That is, the domain of interest responds to interaction the same way as does the act of representing initiated by the concept; both activate the set of modes of interaction connected in the representation. The connection made in the act of representing really does have the domain as its common causal nexus.

Depending on the elements stressed in an analysis of an act of representing, each of the three traditional theories of truth can seem plausible. If we concentrate on the concept involved and whether it produces satisfaction or not, it seems as if something like a pragmatic theory is correct. Goodness of concepts is a matter of whether they work or not. Do they produce satisfaction? If so, then there is little more that can be said to recommend the concept.

If we concentrate on the particular properties or concepts that are connected in an act of representing, it seems that a coherence theory fits best. We want the connections we make in representing to form a coherent network.

If we concentrate on the act of representing as a whole and its relation to the domain with which it interacts, correspondence seems to be the best theory. The objectivity and truth of an act of representing does not consist in the satisfaction it gives rise to (although this is how we come to be aware of its truth). Nor does it consist in a coherence between elements within the act of representing. The objectivity lies in the relationship of the act to the domain with which it interacts. It is the causal force of this interaction that makes the connection made in the act a result of the object and not the medium of representation. Such an interaction results in the co-respondence of the object and the representation; both make the same connection between modes of interaction.

What this shows is that there are various types of goodness for various types of cognitive objects. It seems to me that the word truth most appropriately applies to representations; it is commonly admitted that the ordinary notion of truth is of a correspondence. The pragmatic and coherence theories may be correct in their domain, but they are not correct theories of truth. Truth is a property of representations. Neither concepts nor the modes of interaction that make up the elements of a representation are themselves representations.

This chapter, then, provides an answer to the weak version of the perspectivist fallacy. It gives an account of how an objective representation can be formed from a single perspective. Cognitive activity from within a perspective can arrive at representations that are not determined by that perspective because it involves the application of concepts that actively attempt to make connections between modes of interaction that isolate out a domain of interest and direct activity towards it that results in interaction in which the concepts can be satisfied or frustrated. Perspectival interactions with the world are interactions nonetheless, and their results will contain aspects that are due to the object interacted with and not the perspective. This is fortunate, since they are also the only type of interaction it is possible to have with the world.