In this part an epistemological viewpoint will be expressed, in a general way, which will provide the paradigm for the subsequent analysis of value as a matter of knowledge. A fundamental feature of that viewpoint is the Platonic thesis that knowledge, properly so called, requires a necessary ground. Thus, the theory of necessity here, which will result in the description of eight modes of necessity, is one of the most important aspects of the treatment of knowledge.
Since the epistemological viewpoint is in many respects Kantian, I will begin with a constructive criticism of Kant to show what has been retained from his familiar doctrines and the reason for the changes that have been made in them. Following that will be discussion of the salient distinctive doctrines here, following by a presentation concentrating on the theory of necessity.
One important conception in Kant that may be adopted with little change is his recognition of the unity of consciousness. The emphasis that is placed here on the implications of that conception, however, will be slightly different.
The fundamental truth about individual consciousness and everything in it is that it is a unified whole of manifold contents--in short that it is individual and self contained. Perception unfolds before each of us as a seamless but complex and varied fabric, and it is because we each possess a unique whole of perceptions that can be called by each "my" experience that we say that consciousness is unified. The "I" in its barest and emptiest meaning merely is the formal sign of the connection that the various contents of experience have to each other. Without that connection to the whole of experience, a new experience could not share in the focus of the "I" and would have to be, as Kant says, the same as nothing to me.
For the purpose here it is especially important to note that the unity of consciousness is discovered through reflection. In the presentation of external objects, we are not inclined to view experience as a unified whole since it is purely accidental that some objects are present in our experience while others are not. This suggests that the unity of consciousness is an attribute of the perceptions which are the means through which the external objects were known and thus can only be discovered when we reflect on the act of perception--i.e. when we take dual natured representation only in its internal aspect.
The unity of consciousness does not originate in and does not solely characterise self-consciousness or mediate or conceptual knowledge. Just because it is only in reflection that the unity is recognized does not mean that the unity is created by reflection. This is an important point because of the ambiguities in Kant's treatment of the subject: whether it is perception that is originally unified or only conception, where perception involves no conceptual determination whatsoever, no beliefs, no statements. Viewed with the proper generality, it must be admitted that any mental contents, perceptual or conceptual, that do not exist as part of a unified whole simply stand in no relation to the "I" and are "as nothing" to that "I."
Having recognized the unity of the contents of consciousness, the question then to ask is why the unity should hold among such manifold contents. Consciousness cannot simply be viewed as a box into which things can be placed. The unity of consciousness is not a separate substantial entity, like the Cartesian soul, distinct from the contents of consciousness which have been unified. It is obvious that the contents come and go; perceptions occur and then fade; memories come to mind and then pass away. Consciousness is thus dependent on no particular material content, and this variation enables us to abstractly conceive of the material of consciousness independent of consciousness. That material is generally to be called "sensation," but the fundamental limitation on the notion of sensation is that in principle we cannot say what it really is like in separation from consciousness. The best we can do is look at the material of perception, the color conveyed by light, the tone conveyed by sound, etc., and say that this material is sensation once 1) we conceive it as a mental content in separation from the external objects we perceive by it, and 2) we conceive it in separation from consciousness, free from the fabric and structures that bestow a relation to the "I." These conceptions, of course, leave us with nothing definite to indicate as being "pure sensation." No "patch of blue" will quite qualify for our inspection when we require that the "I" inspect it outside the experience and consciousness of the "I."
Kant approached the question of the unity of the contents of consciousness through his theory of synthesis. While this theory turns out to be too speculative to be considered true, it is important and helpful enough heuristically to consider in detail. It is helpful to imagine that consciousness is the product of synthesis, even if it isn't, and explore the consequences of that assumption.
Briefly, in the theory of synthesis we imagine that the mind actively generates consciousness by "taking up" raw sensation and ordering it according to certain innate rules. Producing or reproducing the material according to the rules means that the unity of consciousness is due to the unity imposed by the set of rules. Such unity of "rule directed activities" is described by Robert Paul Wolff. Kant's own example is counting, where the rule is to compile decades and decades of decades, structuring thereby even very large numbers that otherwise would totally escape the capacities of thought and perception.
A fundamental ambiguity in Kant concerns just what it is that synthesis generates, a structured set of concepts describing the world or our immediate perceptions. The aspect of Kant of greatest interest here is given in this passage:
What is first given to us is appearance. When combined with consciousness, it is called perception. (Save through its relation to a consciousness that is at least possible, appearance could never be for us an object of knowledge, and so would be nothing to us; and since it has in itself no objective reality, but exists only in being known, it would be nothing at all.)
The crucial notion here is that of appearance, meaning raw sensation, being "combined" with consciousness and so yielding perception. The critical question then concerns how these things are "combined," and the theory of synthesis is supposed to explain that. There are a number of interesting currents in this passage that will merit recall later. For one, the term "appearance" has been used where the reader of Kant might expect "intuition," and this is part of Kant's retrenchment of the concept of intuition that will be discussed at length later. It is perhaps also possible reading this passage to think of "perception" as meaning "judgments about perception"--but the problem with that is that somewhere in setting out Kant's system some space must be provided for phenomenal objects, which, like appearance, exist only in being known but at the same time are concrete entities external to us in space that are the perceptual focus of our sensations. Indeed, Kant has complicated things for himself here by saying that appearance is itself "known" and is an "object of knowledge," when we must prefer to say that a phenomenal object is what is known and is the object of knowledge and that the appearance is that by which that object is known. For Kant appearance as a version of "intuition" originally meant the perceptual presentation of phenomenal objects in space and time. If those objects are not then to be "nothing at all" to us, clearly we must begin to see the process of combination with consciousness as that which provides the object corresponding to the appearance.
When perception is fully brought into the process of synthesis, this creates for the theory the difficulty that it postulates a process only one stage of which can possibly be examined. The product, consciousness, we can inspect, but the preexisting manifold of sensation is not available to us separately. An extreme simplification of the theory would be to imagine an assembly-line analogy of 1) accepting the raw material, 2) carrying out the operations according to the rules, and then 3) presenting the finished product. The first of those three stages is definitely hidden from us, and if we accept consciousness as the "product" in the third step, then the unfinished laboring in the second step must be hidden also. Just as bad, the third step might create a problem as well, for if merely being synthesized admitted sensations into consciousness, then every act of synthesis would add new and permanent contents to consciousness, just as sewing sections onto a quilt leaves us, indeed, with a complete quilt. If consciousness is not to be an ever expanding quilt of sensations, then the synthetic unity of the conscious products must dissolve as quickly as it is put in place.
It is tempting to say that as contents roll off the assembly line of synthesis they leave consciousness, so that contents occur in consciousness only during the second step as the activity of synthesis is occurring. That leaves us with only one of the three parts of the analogy accessible to consciousness, the part in which the action of the mind is directly working up sensible material. That is rather difficult to grasp, however, when we still cannot distinguish sensible material from the contents of consciousness that the mental activity is working them into. The assembly-line analogy fails altogether because consciousness as present to us doesn't work at all like an assembly-line: there are no step by step operations that we carry out on sensations; instead we seem to carry out all the operations all at once, so that phenomenal objects come up spontaneously to consciousness.
Kant's initial thought apparently was to regard synthesis as a function of thought, or at least of reproductive imagination--both of which are mediate functions subsequent to conscious perception. The problems of synthesis as assembly or processing are not very acute in the mediate context. When perception is drawn into the argument, however, the processes cannot work at all in the same way; and perception is inevitably drawn in as the force of the unity of consciousness is sufficiently appreciated. If we cannot appeal to the familiar mental processes of carrying out rule directed operations on some given and producing some result, then the real processes that the theory of synthesis is supposed to describe must be radically different from anything that Kant's theory allows.
Kant's very idea of synthesis as a mental activity is what must be most forcefully called into question. With the activity there is naturally also a corresponding passivity; and it is these two notions that must draw our fire first of all in refounding Kant's theory of synthesis. Kant never completely broke up his theory and redid it, but he did have the courage to follow through to some conclusions in his arguments that really contradicted what he actually had wished to say. We can begin with Kant using the most naive form of the theory of synthesis, and then gradually delve deeper until we realize that we are dealing with processes that cannot possibly belong to consciousness. At that point, to continue with Kant's courage, something so fundamental must be discarded that we would need to recommend that Wolff's book (Kant's Theory of Mental Activity) be retitled.
If consciousness were purely active with respect to the objects of knowledge, then it would simply bring them into existence. That is not an ability we have, however; instead we must be given the objects of our knowledge through perception. We are passive with respect to the existence of spatiotemporal substance; the objects act on us. Our passive datum is sensation, which the mind can then take up in its activity. Kant thus says that we possess "sensible intuition" rather than the active "intellectual intuition" which creates its objects. The problem running through the tradition is the ambiguity in the notion of "objects being given," namely that phenomenal objects as such are too easily seen as effects produced in our perception along with the sensible material, while the objects causing the effects are hidden and unknowable things in themselves. The status of the object of representation was a problem that Kant recognized and to which he devoted serious thought, but his doubts that phenomenal objects could simply be given along with sensation ended up leading him off in the wrong direction.
Since Descartes it has been a serious dilemma why a representation caused by an external object need bear any resemblance to the object or tell us anything about it. Any cause is only sufficient to its effect, and sensations as effects conceivably could have any number of possible causes, including God, the deceiving demon, etc. Kant sought to circumvent this problem by proposing that the forms of objectivity of external objects are not conveyed to us causally from without but are actually imposed by the subject from within. This "Copernican Revolution" stood the traditional relation on its head. Kant thus takes the understanding as injecting into intuition certain forms (concepts or rules) that are not to be found there originally and cannot have been borrowed from experience in any way.
Kant's move, while brilliant and of continuing importance for us, immediately must raise a most important question: What prevents us from organizing the objects of the world just any way that we like? In our ordinary application of concepts we tend to think that the world determines whether it is appropriate or inappropriate to apply a certain concept. If I have the concepts of "dog" and "cat" and I am presented with a cat, there is no constraint on me calling it a dog if I like, but it is within the capacity of my understanding to grasp the falseness of such an attribution. In order words, any concept I have can be applied rightly or wrongly. With Kant's theory the sensation, which is purely subjective, hardly seems in a position to determine whether the pure concepts of objectivity in the understanding are applied rightly or wrongly. If we follow something like a common sense view and imagine that something in sensation forms the basis for recognition in understanding that some concepts should be applied rather than other, this destroys the whole point of the theory, for we must then ask by what right sensation is determining how we are to see the structure of the world--it would have to already contain forms of objectivity that were not put in it by the understanding. This brings back the original dilemma in unchanged form.
One way to help Kant would be simply to say that the objective structure of the world is arbitrary, and one version of this might be a notion that a certain form for the world can be dictated by the structure of one's language or some other essentially arbitrary conventional set of cultural rules. On relatively superficial levels of generality this is quite true, but in any fundamental sense it is absolutely unacceptable, as Kant himself also would certainly have regarded it. To suppose that all structures of objectivity are at root arbitrary and then hold that there are objective truths about languages, cultural conventions, etc., is at least very paradoxical and in truth is hopelessly self-contradictory and question begging--reserving a special truth and universality for the theorist's statements that those very statements deny. I take this to be another case of Protagorean relativism subject to the Platonic refutation above.
From trying to do it ourselves with computers, we should know now that any notion of an absolutely amorphous and subjective input determining a specific application of a general set of rules is completely meaningless and impossible. General rules awaiting an input presuppose that specific objective structures exist in the input to be recognized.
To resolve these problems in Kant as well as to provide the definitive basis for the theory here, a new characteristic principle may be formulated: There is no causal connection between subject and object. Neither is active or passive. This is really a version of the principle of the dual nature of representation, or a clarification of it; but it is important at this point to explicitly distinguish the causal relation that must hold between external objects and our sensory apparatus in the generation of sensation and the non-causal relation of intentionality that holds between objects and the representation of objects in the subject. In this section such a distinction will be explored both with respect to Kant and in terms of a contemporary theory of knowledge, that of D.M. Armstrong in his Belief, Truth and Knowledge.
In Kant's theory the use of causality is puzzling. Apart from the problem of whether subjective causality is going to make for an arbitrary representation of the world, Kant accepts a conclusion that this representation, even if not arbitrary, still is not going to inform us about how external objects really are in themselves. Causality is one of the objective forms that is contributed by the subject, but unfortunately to make any sense out of Kant's theory, it is really necessary to think in terms of causality acting between things in themselves. This would violate Kant's own caution of "empirical realism" and "transcendental idealism" that we have no knowledge how or whether the objective forms contributed by the subject have any relation to transcendent reality. But Kant seems always to be thinking that sensation is originally the result of causal affection by external (transcendently external) objects. The sensation is then "taken up" by the subject and through synthesis a representation is constructed in which we see objects that are actually affecting us.
The paradox in Kant's view, then, is that while it may be legitimate to apply the causal form to the objects that we see, since the causal form was indeed used to construct them, it is questionable, as Kant is elsewhere explicitly aware, that we may then look beyond this representation and apply the same form to the relation between the thing in itself and the self in itself without actually assuming they are identical to the empirical object and the body. It is not enough to say that the relation between things in themselves can be "thought" but not "known"; for the whole question is about the legitimacy of thinking certain things in the effort to understand knowledge, and if it is Kant's theory that certain things about the transcendent are unavoidably "thought," despite being divorced from science and the manifold of ordinary knowledge, then it must be clearly shown what the unavoidability is.
Kant is constantly tripped up over the two levels of reality that he is required by his theory to posit: phenomenal or empirical reality, which presupposes the subject, and transcendent reality or things in themselves, which exists independent of any subject. Kant's distinction between immanent and transcendent need not be abolished, but much of the way that Kant characterizes the transcendent--as an independent order of objects different from phenomena yet between which, as between phenomena, causal interactions can occur--must be scrapped. If causality belongs to the phenomenal realm, then our simplest rule is just to be sure that the transcendent has been so construed that even the possibility or meaningfulness of causal interaction on that level no longer holds.
In the progress of Kant's thought, with the withdrawal in the Dissertation of the sensible forms of space and time and in the Critique of the intelligible forms of the categories from the object into the subject, the sticking point in the end is our continuing sense that sensation is a causal product. But the withdrawal of the form of causality into the subject counterintuitively leaves that sensation suspended in a limbo of unintelligibility between subject and object. In our reform of Kant, we wish at the same time to preserve the meaningful causal relation between external objects and our physical sensory apparatus and also the difficult ontological relation between subject and object that Kant has inherited from Descartes. Between subject and object, however, there is only the form of intentionality, and we should say that this is entirely different in kind, function, and existence from the physical relation of causality.
Our two levels of consideration are: 1) the interaction between objects, which means only phenomenal objects and which thus also means only causal interaction; and 2) the relation between subject and object, where the subject is the actual display of knowledge and consciousness that cannot be found through an examination of the physical body of the subject. These two levels, however, obviously do not correspond to Kant's problem of interactions between phenomenal objects and things in themselves. Kant has confounded the terms peculiar to one level with the relation between the two. It is the pure relation between subject and object that gives us the distinction between immanent and transcendent here. On the level of phenomenal objects the problems of immanent and transcendent do not come into play. As Schopenhauer vigorously urged against Kant (and for that matter, Descartes), there is only one order of objects of representation, the immanent and phenomenal.
With these distinctions in mind, it should be helpful to turn to Armstrong's theory. The view there is that knowledge occurs when there is a case of true belief plus a law-like connection in nature (which I will take to mean a causal relationship) between the belief state and the state of affairs in the world that the belief is intended to represent such that for various possible believers the true belief will occur with predictable regularity (under certain other conditions specified by the theory). The derivation of this from Plato's discussion in the Theaetetus is obvious and familiar from contemporary theory of knowledge. Armstrong's formula is unobjectionable, in fact, as a description of the relation between phenomenal objects. It is based on an external perspective, and indeed he calls views such as his "Externalist." But this is just the failing of such theories as attempts to answer the Cartesian problem of knowledge, and Armstrong comes very close to a decisive formulation of that failing:
....but how can we who are, as it were, behind and locked up in our own beliefs, determine which of our beliefs are properly correlated? If such a correlation is knowledge, we may sometimes know, but we will never be in a position to know that a correlation obtains, that is, know that we know. But this is unacceptable scepticism.
We do not need any kind of scepticism to greatly strengthen this objection and to limit the scope and usefulness of Armstrong's kind of theory. The Cartesian problem of knowledge was not so much a dilemma of being locked inside our own beliefs but of being locked inside the subject, with the relation, character, and even existence of external objects called into doubt by the possible solipsism of the subject. Into that gap between subject and object Kant brought the incredible gothic mechanism of the Critical Philosophy but failed to credibly bridge it.
Here I have, in the principles of the dual nature of representation and of ontological undecidability, formulated the view that the Cartesian dilemma cannot be solved because it is a false dilemma, illegitimately giving preference to one pole of the form of intentionality when each pole lays equal claim in the given context of our ordinary experience and judgments about the world. Armstrong may therefore be said to suffer from the fallacy of an external perspective, and naturally our attention is drawn to the alternatives set out by Armstrong. The major alternative for "non-inferential" knowledge that he offers is the Cartesian self-evident belief. A less powerful version of that is just the notion of "initially credible" belief. But both these alternatives are unacceptable both on the general Platonic principle that the "regress of reasons" must end in knowledge which is different in kind from "true belief" and on the specific Friesian doctrine that, apart from analytic propositions of logic, the regress of reasons must end with immediate knowledge, which is similarly different in kind from belief states (which are mediate representations).
In this situation "immediate knowledge" may be succinctly characterized as identical to the external reality of objective states of affairs that Armstrong recognizes as the actual basis for truth and knowledge. From an external perspective it is close to a self-evident truth that the world of external objects does not exist within the cranium of the subject; but from the internal perspective, which is what we actually have as occupants of consciousness, that same world of external objects does exist as a representation within the subject. That aspect of our representation as mental content is what counts as immediate knowledge, and clearly beliefs as subjective states, concepts, judgments, etc., do not belong to either immediate knowledge or the external world.
As we consider the physiology of sensation or the neurophysiology of perception, Armstrong's description is relevant and meaningful. Even Kant's theory of synthesis may be in that context, as we imagine the brain applying its cognitive programs to the input of nerves from the various senses. But to the real philosophic problem, these all fail. Considering the central place of immediate knowledge offered here, the following section will treat evident immediate knowledge, or intuition, in detail. In general terms, mediate knowledge may be said to involve some function of memory, and this will serve to distinguish from immediate knowledge all conception, recognition, and language and may make for a better psychological criterion than saying that mediate knowledge belongs to representation as mental content but not to representation as the world of external objects.
Evident immediate intentionality is perception, what Kant called Anschauung, intuition. It is convenient to start with Kant, but it will be necessary to differ sharply with him on the nature of intuition--or at least to considerably reformulate the results of his own arguments. Using a theory of active synthesis, we would need to see perception and so intuition as really the result of synthesis, while in reading Kant it usually sounds as though intuition is the same as sensation, merely the raw material for synthesis: in the Critique, one tends to think that having finished the "Transcendental Aesthetic" perception has been taken care of and only matters of thought come next.
Here the viewpoint is that even if we imagine synthesis occurring, the result must still be present to consciousness immediately since a mediate result would mean that the state preceding synthesis, containing the raw sense data or whatever, is also present to consciousness--which is false ex hypothese that synthesizes what actually constructs consciousness (a view not evolved until well into the "Transcendental Logic"). In this way Kant's originally conceived intuition would not occur to the conscious subject--a formidably incongruous notion.
As we reason through Kant's theory of synthesis, immediacy and intuition in their naive versions as raw sensation first vanish, because they cannot possibly be contained in consciousness according to the principle that the unity of consciousness requires synthesis, but then they return as we push the argument to its logical conclusion and we reflect that synthesis as a mediate process in consciousness does not make sense when what has been thought of as "mediated" is actually not present to consciousness. The result of the "mediation" appears to come up spontaneously; and even if we take seriously the idea that some synthetic activity might correspond to the brain's information processing of sensory input, from the point of view of consciousness we must say that what appears to us as phenomenal objects is given immediately.
To say that the object, or the phenomenal object, is immanent in intuition may be taken as to do no more than affirm a common sense realist principle, that the real things (tá ontôs ónta) are the things that I actually see and touch. To be contrasted with this, as Plato may be contrasted with Aristotle, is the Cartesian conclusion, ambiguously accepted even by Kant, that the real things are transcendent, standing behind the things I see and touch, which are more or less fictitious representations merely caused by them and perhaps bearing little resemblance to them.
Kant has actually accepted Descartes even while providing the elements for his refutation; for Kant, having asked what we mean by an "object of representation," answered that we mean a certain non-arbitrary structure of spatiotemporal and causal forms, etc., in experience. If this is so, then what we mean by an object is clearly immanent in experience, and what's more, according to Kant, we have put that structure there ourselves through synthesis. The latter aspect, as I have indicated, deserves to be gravely qualified, although in retaining it the Kantian conclusion would be justified that we then do not know the real (Cartesian) objects--we have merely constructed a sort of private or intersubjective analogue. On the other hand, that conclusion could be avoided by assuming that the "we" putting the structure into experience is more than the individual self--so that Cartesian matter or the Kantian thing in itself is replaced by an Idealist meta-consciousness that ensures the objectivity of the phenomenally immanent object. This is at least as awkward, and is certainly less straightforward, than the Cartesian view.
The status of an object of representation here may be seen to involve aspects both of realism and phenomenalism. The phenomenal aspect is that the objectivity of the world is only meaningful in terms of the form of intentionality and so the relation between subject and object. At the same time, the field of phenomenal objects is commonly where we ignore the subject; and the external perspective of science moves confidently through the objects without concern about the necessary correlate of the subject. There is no error in that, but the irony of it, for the theory here at least, is that realism requires an appreciation of the subject as well as the object. The realistic aspect, then, is that the existence of object and subject are genuinely separate and independent. Heidegger's sharp distinction between the existence of things as such and their qualitative character as individual things is a help here, since the former is real and separable while the latter is phenomenal and relative to the form of intentionality. And again, only immediate knowledge constitutes the objective ground in intentionality, while many kinds of mediate intentional object may or may not correspond to factual states of affairs in phenomenal objects.
The relation of intentionality is like some holographic image where both the observer and the observed are spontaneously projected from the hologram. The space into which the holographic images of observer and observed are projected is existence, and the lack of decision on the dual nature of representation, which leaves us unhappy about that to which the hologram itself "really" belongs, leaves reality as a sort of Gestalt trick of perception, where one moment we are locked into a Cartesian solus ipse with no exit, while the next moment we move confidently through a familiar world of public objects. The "spaces" of existence in this simile, internal for observer and external for observed, are the special concern of the theory of negative transcendence. In the end, however, we say that the hologram is existence itself, and the dilemma of ontological undecidability is resolved by the theory of positive transcendence. The "spaces" become perspectives within existence, not absolute ontological alternatives that we are bound to choose between. This might be taken as a triumph of phenomenalism except that positive transcendence is neither subject nor object, cannot appear except in the continuing intentional projections of subject and object, and certainly cannot be said to exist as more one than the other.
The elaboration of these stages in the theory will of course be handled in turn. What we need note now is merely the ontological dimension of the epistemological conclusions arising from a consideration of Kant. Keeping this ontological dimension in mind is especially important in giving proper credit to the status of intuitive immediate knowledge since we commonly load much of what is subjective and mediate onto intuition through the abstract ubiquity of our judgments and the powerful Gestalt of our preconscious understanding. Pure intuition, should it be possible to recover it, would be a strange state indeed, for it would have to be without the ordinary distinctions and guideposts that regulate daily life and the activities in which we commonly engage. If action is possible in such a state, it would have to be an extreme version of the Taoist ideal, without intention, without thought, without effort, without "mind"--as though our consciousness did not even exist, or existed merely as a spectator.
Together with the underlying ontological structure of the theory presented here, there is also a special epistemic structure, namely a theory of modes of necessity. These various modes are a non-reductionistic description of the different grounds of immediate knowledge underlying object language systems of theoretical and practical knowledge. This structure of necessity is elaborated both according to ontological and logical or epistemological principles. In the theory of consciousness there are three modes that will concern us: in this section, again, the focus will be on the theory of necessity in Kant, and that will involve 1) analytic necessity, which can be largely presupposed, and 2) synthetic necessity a priori as described by Kant. In the following section I will diverge from Kant into what will be called 3) the perfect mode of material necessity. That discussion will serve to open the possibility of the further modes of necessity.
An important result for Kant of his theory of synthesis was an argument for a kind of necessity that went beyond the necessity of merely analytic propositions or, as we might say now, necessity that might be derived using no more than the laws of logic. That necessity belonged to certain synthetic, logically contingent propositions that could be made a priori, or independently of experience. The basic argument in this was that synthetic propositions a priori are justified by conditions sine qua non for the unity in plurality of experience. Given that experience is a unified whole with respect to being the content of a single consciousness, Kant says that the conditions for the possibility of this unity are the abiding forms of unity that are applied through the process of synthesis and so apply equally and universally regardless of the particular material content of experience.
Kant's perspective is profoundly insightful and important, despite the objection that from an internal viewpoint the forms of unity cannot be seen as deliberately implanted into a preexisting material by some mental action. What we are given in consciousness immediately is the form and the matter together. With reflection, we come to distinguish the "form" from the "matter" by recognizing those abstract (formal) aspects of the perceptual field that do not vary with time or position. As it happens, Kant's sine qua non argument is still effective just because consciousness still presupposes the same formal unity, even though the supposed deliberate activities of synthesis are not given to our examination.
What the forms of unity are was a big question for Kant. He thought that he could derive them from the forms of pure logic, because it seemed that those would be the forms of pure thought, independent of intuition and empirical reality and so perhaps related to the non-empirical basis of reality, things in themselves. That was pretty much the doctrine of the Dissertation, that pure thought did know things in themselves, consequently called nooúmena, "things thought." In this treatment I have taken thought to be closely related and dependent on perceptual mental functions, in which case Kant's view cannot be allowed. It does seem reasonable, however, that there is a kinship between logical forms and the forms of unity in consciousness, going in the other direction, with the former evolving out of the latter. Whether or how this is so is really not a very important question here. What should be noted is that Kant's emphasis and his notion of the direct noumenal relevance of the pure forms of logic strongly biases his regard for the forms of unity both in his attempt to derive them and in his use of them after he thinks he has discovered them.
Causality is the most important form of synthetic unity in any historical perspective on Kant, but we find Schopenhauer holding that what Kant proved, in his attempts to justify ("deduce") the category of causality, was simply that there must be temporal succession, not that there must be causal succession. Indeed, what is minimally required as a basis for a synthetic unity of consciousness is a nexus of temporal unity, in order that events of one moment of time must be considered to belong to the same consciousness as the events of a preceding or following moment. Causality is not the minimal requirement, and it is easily arguable that causality is very much ontologically secondary to time in the sense that while causal events are ubiquitous and manifold, they are all events in time, presupposing temporal succession. If there are events and natural relationships where the concept of causation is not applicable, there certainly are none where the concept of time isn't.
Although more relevant to past and present scientific debates, the law that every event must have a cause still seems reasonably to presuppose a more basic law that every event must follow and must precede other events in time. The causal law adds to the temporal the requirement that the character of an event depends on both the character of certain preceding events and the requirements of various "laws of nature" that govern causal transformations. Hume's method of considering what is conceivable is helpful in distinguishing purely temporal from causal connection a nightmare succession of totally unrelated events is an easy and clear enough conception of a world without causality. A world without time yet with causality is far and away a more difficult, and for me impossible, condition to imagine.
The basic forms of synthetic unity in the perceptual field are simply the nexi of temporal succession and spatial location. In temporal terms alone, after Leibniz, these are the relations of succession and simultaneity; but the relation of simultaneity would be empty if there were no "room" for simultaneous events to coexist. The form of simultaneity is the spatial nexus, but it should be emphasized that in the full articulation of mental function the spatial nexus may be said to relate simultaneities that do not have "room" to be spatially distinct, namely states of mind. The possibility of a non-spatial reference for the spatial nexus serves to point up the important point that the spatial nexus is not the same thing as space. And the structure of space, Euclidean or non-Euclidean, Lobachevskian or Riemannian, which has often been mistakenly taken to refute Kant's Euclidean expectations, is absolutely a separate matter from the consideration of the spatial nexus as a form of synthetic unity.
In considering the temporal nexus, a question arises that is relevant both in physical debates and for us, namely, the directionality of time, the "arrow" that heads for the future but never for the past. Where in modern physics time has to an extent been assimilated to the spatial dimensions, the question has arisen why the temporal dimension, and not, apparently, the spatial, is asymmetrical. Is the temporal nexus itself directional? If it is not, then is the temporal nexus really different from the spatial nexus? After all, it is not very tempting to say that there are three spatial nexi just because there are three spatial dimensions. It is conceivable that there is only one nexus of formal synthetic unity that both spatially and temporally binds together the field of perception. This nexus is then differentiated by the intuitive ground of space, whatever that is, into the three spatial relations and by the intuitive ground of time, whatever that may be, into the asymmetrical temporal relation. If this is the case, then the problems of the identification of the intuitive grounds of space and time is separate from the theory of formal synthetic unity. It does seem to be the case, and for time the matter will be shortly taken up in relation to perfect necessity. For space, in so far as it will concern us at all, the problem is the concern of the theory of negative transcendence.
In considering Kant's notion of necessity as a result of the formal synthetic unity of consciousness, the important conclusion is that apart from the significant epistemological admission that there is a formal synthetic necessity, there is really very little of substance to be drawn from the theory. Space, time, and causality--the flashpoints of controversy in debate over Kant--all involve some ground or complication or articulation beyond the simple formal necessity that, in the most abstract and minimal way, makes the unity of consciousness possible. Where Kant was deceived and misled, and where in a sense he begged the question, was in the fullness of meaning that he gave to his pure forms of thought and hence to the pure concepts of the understanding which he believed followed from them. Kant's own awareness of his difficulties is reflected in the extent to which he reintroduced time, after having presumably disposed of it in the "Transcendental Aesthetic," into the articulation of causality in the "Analytic of Principles." The whole gothic structure of the "Transcendental Logic" is an attempt to weave out of the thinnest possible material the most substantial conceptions. It was not effort wasted, but certainly effort left incomplete.
Having cut down Kant's category of formal necessity so thoroughly, it must now be said, in all fairness to the reader, that rather more than this will come of it in the end; for it will be considered later that certain other modes of necessity come to be conceived by us according to the principle of formal necessity--the principle of connections according to law-like regularities. These will be considered in turn.
In this section the field of necessity will be expanded from Kant's notion of synthetic a priori necessity. First will come a general consideration of material necessity, with a restriction of Kant's a priori necessity into what will be called a priore necessity. Second will come a discussion of the significance of Aristotle's theory of future contingency in On Interpretation, the use of perfect and imperfect temporal aspects in the grammar of many languages, and the way in which these lead to a recognition of our first mode of material necessity, the necessity of the perfect aspect.
The sense of the terms a priori and a posteriori originally concerned what can be known before the particular cases of experience are considered and what can be known only after experience has been received as evidence. We might say, looking at the terminology, that the distinction refers to the priority or the source of the ground of knowledge: that a priori knowledge can be based on something available to thought at any time, without the need to await some datum from experience, while a posteriori knowledge literally comes afterwards, after the consultations of thought with itself. The a priori/a posteriori distinction has tended, especially with respect to Kant, to be easily construed as a form/matter distinction, with the a priori content corresponding to innate mental forms, whether those of the analytic truths of logic or those of the synthetic truths of the forms of the synthetic unity of consciousness, while the a posteriori content is given in the sensible manifold of intuition.
As such, the a priori/a posteriori distinction does not have any immediate connection with one's thinking about necessity and contingency. The fallback position for necessity is always logical necessity, that the denial of tautologies or analytic truths results in an internal contradiction. In those terms a priori synthetic truths would not be necessary. However, we call a proposition contingent only if it is possible for it to be or become, in time, false; and if there is synthetic a priori knowledge properly so called, it is impossible that it should be discovered to be false. What can be known independently of any experience may thus be considered to be necessary knowledge by virtue of its timelessness, universality, and perhaps even the inability of the world to appear except in conformity to the a priori expectations. This carries us back to a more Platonic conception of necessity, where the necessary truths are such not because of any internal logical requirement (Plato being unaware of the theory of such things) but because an ontological fact extrinsically provides for their own fixed eternity and for the impossibility of their ever becoming false. The necessity of Kant's synthetic a priori truths is similarly extrinsic, though it is more epistemological, in the conditions for the possibility of experience, than the Platonic (though there was an epistemological dimension to Plato's necessity also in his doctrine of knowledge by recollection).
For us the articulation of the extrinsic grounds becomes the major concern. Because of this the common extension of the term "a priori" becomes inappropriate, for Kant has used that term far beyond the limits that should have been strictly imposed by the formal ground for synthetic knowledge a priori that was cited in the "Transcendental Deduction," where the principle of the possibility of experience was put to use. As Kant believed that the necessity of the moral law resulted from an unconditioned version of the very same universality, and indeed the very same forms, as expressed in the case of the synthetic categories of experience, it is reasonable that the moral law and its derivatives should have been similarly said to be true synthetically a priori. But since this ground for the moral law will be rejected here, that legitimacy for the usage will be lost.
Retaining the basic distinction of form and matter as the basis for the distinction between a priori and a posteriori, the meaning of a priori will be drastically curtailed in the usage I will now establish for the purposes of this discussion. In that the "form" can refer only to principles of logic or the synthetic nexus of the unity of consciousness, no further cases of necessary truth will fall under the traditional a priori category. To signify this change, the form of the term itself will be slightly altered: by a priore, with the proper Classical Latin ablative ending, I will refer to the restricted meaning of formal necessity. By a posteriore, similarly altered, I will now refer to all extrinsic grounds of knowledge, so long as they are what we will call material grounds in immediate knowledge (whether intuitive or nonintuitive). These may or may not be necessary grounds, depending on their own peculiar character.
In that it preceded the individual cases of its application in experience, a priori knowledge was always of the nature of a prediction of what was going to occur. The form of a priori knowledge, also, would tend to be in the form of rules or laws, on the basis of which predictions of events could be made. The form of a priori knowledge will tend to assimilate modes of necessity that I henceforth will not call a priore. There are two especially interesting examples of this. First is geometry, whose axioms were certainly considered a priori knowledge until quite recently and whose form certainly continues in precisely the same way it always has. What has changed is the nature of the supposed ground; and the geometric axioms have tended to assume the character of scientific predictions which can be falsified by the event. The second example is just scientific prediction in general, whose character will be of particular concern in the theory of negative transcendence, for which I will reserve a more detailed discussion. This sense of a priori, then, is worthy of note; and, if the occasion demands, the term in its original spelling may be used for it. Otherwise, we should just say that the formalizing character of our abstract and conceptual mediate knowledge will tend to assimilate all modes of necessity to the regulative form of a priore knowledge.
From now on it will very much be more the field of the a posteriore than the a priore that concerns us, and the thin edge of our wedge of material modes of necessity will come from a reconsideration of a very ancient issue. The question of future contingency will provide the first mode of material necessity and will also establish the distinctions upon which the subsequent theories of time, change, will, and so value itself will be established.
In the first place there is a subtle difference between what I will specifically call "Determinism" and what may be called instead "Fatalism." Fatalism arises out of an observation that the events of the past cannot now be changed and that, while from our small perspective the future seems different in kind from the past, it is conceivable that this difference is only apparent and the result of our ignorance. To God, we might think, seeing past and future equally, all events have already been fixed and known from all eternity. With this in mind, Aristotle's discussion in On Interpretation upholds the view that, while true statements may be made about the future, e.g. that there will be a naval battle tomorrow, the truth of such a statement will only be contingent--until, indeed, the event comes to pass. A Fatalist perspective would be that the future was never really open and that, whether we were in a position to know it or not, the truth of propositions about past, present, or future is equally necessary due to the eternal fixity of all events.
Determinism, on the other hand, is a more sophisticated notion in which the necessity of events is due, not to their fixity in eternity or in the eye of God, but to the manner in which natural law constrains coming-to-be. The future, then, may even be regarded as open and unrestrained possibility in its own right, unknown even to God, but the character of events that come to be in the present may be said to be absolutely determined by the operation of natural laws. God as a Fatalist would merely have to look to know; but God as a Determinist would have to calculate. Like Laplace, he would have to derive the future states from the states that are given in the present.
Here we are interested in what forms the original basis for the pessimistic Fatalist extrapolation into the future. The case of material necessity that should be considered to stand first dialectically for us is the necessity of events that have already occurred. In the sense of the necessary as that which cannot be otherwise, or of the necessary as that which, now, cannot be altered by events, it is trivially true that the character of the past is necessary. Any complication in considering this results from the perspective one takes: sub specieae ternitatis one is committed to no fixed present and so may ask whether it is proper to attribute to all events in eternity the necessity of events in the past or the contingency of events in the future. On that score contingency usually wins now (with or without the addition of Determinism), as necessity was more likely to win in the Middle Ages--obviously motivated by belief in Divine omniscience and its consequences. Such sentiments are really more a matter of Zeitgeist and cultural bias, where for us each perspective must now be restricted to its proper grounds and objects. Eternity is not actually present to us for it to become a real problem whether it is essentially, as a whole, necessary or contingent. The perspective of consciousness is not like the imagined mind of God, viewing things sub specie aeternitatis. A present moment for us, having been lost, can never be retrieved; and the real possibilities that were lost when the moment passed are forever lost.
The present is somewhat different from the neat alternatives of past and future. When we fear to see something realized, seeing it realized before us means that it is already too late to prevent it. On the other hand, in seeing conditions before us, we tend to think that we can change them. The contingency of the future means that it is possible for things to be different than they are; but our power to change things means that we change them in the present, not in the future. The present is the place of change and the place of real potentiality, so that the door being closed, it can be open. The present is a paradox.
If we are to think of the past as necessary, the future as open to various possibilities, and the present as combining both these features, it would be convenient to make a division that combines the past with the necessary aspect of the present and the future with the contingent aspect of the present. This very division lies ready at hand for us in the common practice of diverse natural languages. This will provide a ready-made, and in fact Classical, terminology for this first critical mode of material necessity and its corresponding mode of contingency. Just as many languages distinguish the past tense from the present and these from the future, many others (Japanese, Russian, Arabic) distinguish the perfect aspect from the imperfect aspect. These terms are found in Latin grammar, where perfectum originally simply meant "finished" or "complete," and not "perfect" in our present English sense. These terms were actually more appropriate for Greek:
The perfect of Greek verbs is a true perfect and never has the meaning of a simple past tense as does the Latin perfect. The perfect tense denotes an action completed so shortly before the present moment that the effect of that action may be called a present state.
Here the perfect is cited as a "tense" because in the indicative in Greek tense and aspect meaning are combined. However, in the subjunctive and imperatives moods the tense systems falls away entirely, and only aspect remains. These are the old Indo-European aspects, indicated in "strong" verbs by internal vowel changes: the perfect, imperfect (called "present," however, in Greek grammar, with "imperfect" reserved for the past imperfect of the indicative mood), and aorist, where the aorist represents the action in photographic simplicity without sense of either completion or incompletion. In English there is also a full system of aspect through the use of auxiliary verbs. In the present indicative we can say: I have run (perfect), I am running (imperfect), and I run (aorist, which, however, has acquired a sense of meaning habitual action).
The necessity that events possess merely by virtue of having occurred will therefore be called here perfect necessity. The contingency that events possess because they are occurring or have not yet occurred will be called imperfect contingency. The terms thus used are of course not to be confused with their ordinary meaning in English. The possibility of confusion might warrant modification of the terms, as they are sometimes used in grammar, into "perfective" and "imperfective." However, there is little occasion here for the ordinary meaning of the English terms, since "perfection," as say with the perfection of God, is no part of this theory. My inclination, then, is to side with Schopenhauer that an unusual term should not be used so long as a common term is at hand. I simply use them in a way more evocative of their etymology than of their current usage.
The division between perfect necessity and imperfect contingency becomes increasingly important for us for the same reason that the distinction between potentiality and actuality was important for Aristotle. The distinctions are ontologically isomorphic and similar in their consequences. Indeed, where Aristotle spoke of the form and matter that composed all things in the world, my concern is with the material of consciousness, which is to be linked with actuality and perfect necessity after the manner of the Aristotelian species, and the power manifest in consciousness, through which on the one hand the world changes without our intervention and on the other hand is altered through our own agency. The sensible material is to be considered essentially immanent, while that which transcends consciousness, existence, is to be suitably associated with the imperfect aspect and the peculiar emptiness of futurity and potentiality.
In this the intuitive ground of the temporal nexus is now suggested: the ontological junction between that which is essentially of the perfect aspect and that of the imperfect. The reality of time is within the present, and as such it is not so much a temporal nexus as the ontological relation between immanent and transcendent, as I have been unfolding the meaning of these terms here. What is present to us is primarily the "arrow" itself, the feeling of what can be done or undone and what is already locked into reality. The temporal nexus as such, the connection between successive moments in time, originates first in the abstraction and dissociation of past from future as we distinguish perfect from imperfect and then these from the present where they are so intimately conjoined. Spatialized time, the "clothesline across eternity," comes at the end of a long progression of dissociation, abstraction, and hypostatization. Time conceived as space, in other words, is an illusion: our need to simultaneously visualize two separate moments, or many, in time means that we apply the form of space where such simultaneities can be imaginatively separated.
At this point in the theory, our sense of a posteriore and imperfect modes of contingency opens the way into the dynamical transcendent and the possibility that further modes of necessity, after the manner of the a priore and the perfect, may be distilled out: the increasingly weak and restricted senses that will include matters of value. It should be kept in mind that while each sense of necessity will be secure and uncontradicted in its domain, each sense of contingency will always only be relative since it may include cases that will turn out to be necessary in terms of a weaker mode of necessity. Thus it is reasonable to speak of logical contingency, a posteriore contingency, and imperfect contingency, even though the whole point of the theory here is to deny that all logically contingent propositions are actually contingent. In the theory of absolute transcendence we will eventually consider the sense in which even analytically true propositions may be take to be contingent.
A cornerstone of Plato's thought was that knowledge, to qualify as knowledge, must be eternal and unchanging. From this conception of knowledge he inferred that the objects of knowledge must also be eternal and unchanging, thereby disqualifying the things of experience, transient and mutable as they are, from any role as ground for cognition. Aristotle, who by no means disagreed with Plato on the nature of knowledge, denied the validity of the inference to immutable objects removed from the world of experience. One of Aristotle's insights was that the regularity of change in the world provides a suitably immmutable basis for knowledge. There is no need to substantially alter this view even today, although the tradition seems to be of many minds about how to regard the matter metaphysically--namely whether the regularities in the world belong there objectively or are put there by a projection from the brain, from language, from a traditional cultural mythos, or from the scientific enterprise itself. All the subjectivist approaches, however, are little better than admissions of epistemic defeat.
More on the epistemological than on the ontological side of his theory, Plato took the ultimate foundation for knowledge to be intuitive--though reduced in the world to a memory of the pre-natal intuition. Aristotle, who was the first to clearly conceive the logical issue that not all synthetic propositions could be logically derived from others without an infinite logical regress, also regarded substantive knowledge as ultimately justified and grounded by intuition--insight by pure mind (noûs). From these beginnings philosophy made little enough progress until Kant, who was the first to see that the truth of Aristotle's synthetic "first principles of demonstration" might be grounded in some way other than through tautological or intuitive self-evidence. The Friesian School was virtually the only post-Kantian tradition to correctly address these issues and appreciate their significances.
The world is an interplay of chance and law, contingency and necessity. I have agreed with Plato that all knowledge is founded on some aspect of necessity, and so far in this treatment the three aspects of necessity have been: 1) analytic necessity, which is often the only kind allowed in modern epistemology and metaphysics; 2) synthetic a priore necessity, which is a function of the unity of consciousness after the manner of Kant's "principle of the possibility of experience"; and 3) perfect synthetic a posteriore necessity, which refers to the aspect of fixity, definiteness, and immutability of the material, immanent contents of consciousness. Although each of these, and the modes of necessity to follow, is weaker and more restricted than those preceding, perfect necessity is of special interest as the fundamental ground for all factual knowledge of the world, since the definiteness of all states of affairs in intuition is an expression of it. It may at first seem strange to think of mere fact as involving necessity, but familiarity should resolve this strangeness once we examine the meaning of necessity and its history and cease to beg the question by instantly equating the term with what is now so well understood as logical necessity.
In the Timaeus Plato seemed to think it a sufficient justification of the doctrine of Forms to say that it was equivalent to believing that there is a difference between knowledge and opinion. This is still close to the crux of the issue. Modern epistemology, as in the theory of Armstrong, is actually little able to go beyond the suggestion in the Theaetetus that knowledge is true opinion accompanied by an explanation or justification. This is, after all, not so bad in the light of the paradigm of scientific method. A scientific hypothesis is a kind of educated guess, an elaborate opinion, which gains credibility in proportion to the consistency with which observational and experimental evidence supports it--or at least fails to refute it. There is an abiding paradox in the Theaetetus conception, however, which was obviously appreciated by Plato and has been more recently treated by Nelson. All opinions, if believed, are believed to be true. Thus, a true opinion and a false opinion, taken entirely by themselves, are no different and cannot be rationally discriminated. The true opinion is discovered or proven to be true by means of its "explanation or justification"; but if the truth of the opinion is known by means of the justification, then the justification certainly is equivalent to knowledge of the matter at hand, which renders the "true opinion" aspect of the theory superfluous. Given direct knowledge, we discard opinion.
The suggestion of the Theaetetus thus is an elaboration that leaves the basic question entirely unchanged--the problem of knowledge is simply thrown back one step, to the content of the "justification." Where a scientific observation is merely a necessary, but not a sufficient, condition to the truth of the belief, this regress is obscured. And in an "externalist" theory, which describes a regular connection that is outside our inspection, the requirement of justification is transformed into something really alien to the Cartesian problem of knowledge. The Thaeaetetus can be well taken in one sense, however; for if instead of "true opinion" or "true belief" we substitute "true mediate representation," then, for cases of conceptual knowledge, the "justification" will, usually, contain something distinct from the mediate representation-although it must still count as knowledge, immediate knowledge, in itself.
Nelson's view was that if nothing can be accepted at face value as being knowledge, then we can never accept anything as being known. No criterion will ever enable us to admit that something is to be accepted as knowledge unless the criterion can be independently known to be true. Otherwise we may as well admit that we just go along picking out rules to suit our prejudices regardless of any relation to reality they might have (a view not without its appeal, to be sure). Nelson, like Plato, considered it essential to keep in mind always that basic difference between knowledge and opinion.
While the major and critical conclusions of the theory of consciousness must be epistemological, it also has obviously developed important psychological, logical, and ontological conceptions. These all provide essential materials for us in understanding the nature of will and value in both its internal aspect and its external and objective role. For both Plato and Kant theoretical knowledge was a prerequisite but transient feature on the way to the theory of value and practical reason. Here the case has been the same.