Chapter 2 Rough Draft

[Jeff]

This chapter aims to define the epistemological constraints upon metaphysical theorizing (First Philosophy).

[Jeff]

Chapter 2: Metaphysical Methodology

If metaphysics is possible, the next question is, “How is it to be pursued?” However, the issue of philosophic method is as contested as that of metaphysical possibility itself. And the charges against it have a familiar ring to them. “What is the point of such speculative and idiosyncratic thinking?” “What assurance have you that any philosophic method will produce results more secure than its competitors?” These questions are often fueled by the knowledge that the history of philosophy has yielded no clear claim to a definitive method of any sort. Certainly there have been eras where a way of doing philosophy has been favored, but these have always been superceded by new methods. About all that one can say is that the historical quest for a method that produced secure and certain truths has shown us that this goal is probably unattainable. Given this history one might be tempted to suggest that it is little more than speculative introversion that compels some philosophers to continue to focus on method at all.

E. A. Burtt (1946) offers an analogy designed to thwart this sort of dismissal: Try to imagine a scientist who wouldn’t be troubled if she discovered that she had been following a laboratory procedure which she couldn’t trust to issue in results that would convince other experts in her field. Would she not immediately begin to question her own method and purge it of its deficiency? What use would it be to go on with her experiments if there was no assurance that they would be acceptable to the scientific community? There are certainly many differences between scientific and philosophic inquiry, but if this sort of reasoning makes sense for scientific procedures, then it should make sense for philosophic ones. At the very least it should help to explain why some philosophers are troubled by methodological questions.

I propose to survey possible methodological strategies. Using a classification strategy developed by Catherine Elgin (1998), I will lay out three main types of these. Though this characterization will be highly generalized, it will nevertheless allow us to advance a critique against several of them. This paper will include two such critiques. In a later paper, I will continue my evaluation of the third methodological strategy and propose a specific theory along those lines.

I. The classification scheme
Elgin is concerned with normative epistemology, i.e., with specifying the conditions under which someone ought to believe a certain proposition. The goals of a philosophic methodology are similar. Ideally such a methodology would produce propositions that others ought to believe provided that they share the same general views concerning the standards of evidence and the amount of evidence a proposition must garner to be believed. However there are some important differences between these two concerns. A philosophic methodology is generated to deal with types of evidence that are usually mustered in support of philosophic theses. Frequently the evidence is nonempirical in nature. Additionally, philosophical propositions themselves deal with norms and values. Despite these differences, philosophical propositions remain propositions. If Elgin’s categories successfully partition the general types of epistemological theories, then it seems reasonable to believe that they serve the same function even when we limit our scope to specifically philosophical propositions and the evidence marshaled in their favor. Therefore, I will adopt Elgin’s categories and terminology in what follows with the following modification: I will continue to speak of procedures of philosophic methodology, where she spoke of epistemological ones.

Elgin has developed a unique way of looking at methodological theories based upon a characterization of the procedures they employ. Elgin’s classification is itself derived from John Rawl’s classification of procedures in his A Theory of Justice. The scheme divides procedures into three groups: perfect procedures, imperfect procedures and pure procedures.

A perfect procedure is a process that recognizes an independent standard for the successful application of its method and has an absolutely reliable test to determine if that standard has been met. Such a procedure produces verifiably acceptable results all of the time—when it produces anything at all. An example might be the production of plastic likenesses of movie characters, the kind many of us give to our children. The standard here is a child’s ability to recognize the toy’s physical resemblance to the character it depicts. Further, if a toy failed to resemble its model, it would never make it to your local toy store where children would have no reason to purchase it. So the procedure produces likenesses when it produces anything at all. Elgin’s example of a perfect procedure is a precision cake-slicing machine (p27). The independent standard is its ability to slice a cake into absolutely equal slices. It will always yield an equal division when it yields any division at all. Sometimes the cake-slicing machine might yield no product, e.g., when it is presented with an irregularly shaped cake to slice.

An imperfect procedure is a process that recognizes an independent standard for the successful application of its method but has no way to determine reliably whether that standard has been met. Because such a test does not exist, a plurality of methods are likely to arise which claim to best deliver the desired result. I think that historical reconstruction is an example of an imperfect procedure. Perhaps the reconstruction of a crime scene by police detectives or the reconstruction of ancient cultures by anthropologists is a good example. Here the standard is, of course, to get things exactly right, as they actually occurred. But there is no way to guarantee that any procedure ever does this. More personally, the reconstruction of the past by our own memories may be another example. The standard is a fully accurate account of all events in the scope of some memory, but psychological studies have consistently shown that this standard is not met. As it turns out, memory may be most reliable insomuch as it is reliably defective! Assuming that one lacks an appeal to mechanical recording devises, recall may count as an example of an imperfect procedure. Elgin’s example is of a trial by jury. The standard is to acquit the innocent and convict the guilty, but as we all know, there is no reliable way to determine if this standard has been met.

Finally, a pure procedure is a process that recognizes no independent standard for the successful application of its method. This means that the completion of the process produces, automatically as it were, a sanctioned result. Elgin’s example is a tournament. Sometimes the best player does not win; nevertheless, the simple fact that a player has won a tournament makes that player a champion. Another example of a pure procedure might be a democratic election. Again, it is the process of voting itself that determines the winner of an election; there is no other standard.


A perfect procedural methodology will demand conclusive evidence and will guarantee the permanent acceptability of the propositions it sanctions. Thus, if a proposition is a product of such a methodology, then it is certain; its truth, once certified by the procedure, is incorrigible. Further, perfect methodologies recognize that there are facts independent of the procedure itself. These are the basis on which the standard of acceptability is determined to have been met or not. Because the procedure issues only in truths, this original collection of facts can be systematically enlarged by further products of the procedure itself. Hence, the architecture of knowledge one this view is foundationalist. However, this characteristic also means that no nonliteral truths can be sanctioned, e.g., since metaphors are literally false, they are never the products of such a procedure. Obviously, a perfect methodology goes a long way toward our epistemic goal of preventing false beliefs from entering our worldview; however, since many truths may lack conclusive support, they may be lost to such a methodology, perhaps irretrievably. Nevertheless, a perfect methodology produces utterly secure knowledge, so perhaps the loss here has an offsetting epistemic value.

An imperfect methodology will demand merely compelling reasons and will offer only the provisional acceptability of its sanctioned propositions. Such propositions may be shown to be false at a later date or from a more general perspective. Although imperfect methodologies recognize that there are facts independent of the procedure itself, they cannot claim ever to have arrived at a verifiably adequate account of them. So, there is a realist bias to these methodologies, though this may be of a most minimal kind. Since imperfect methodologies are fallible, they tolerate a certain amount of risk of falsity. Thus, some nonliteral truths can be sanctioned, as can less than utterly reliable methods of deriving further believes from a stock of sanctioned beliefs. This means that justification within these systems is complex. Because such methodologies are fallible, the very methods by which they sanction propositions are fluid. Though certain theoretical virtues might be constant among them.

The evidence that supports a proposition sanctioned by a pure procedural methodology is constitutive of the meaning of the proposition itself. For such a methodology there is no test for reliability beyond the procedure itself. This means that the practice of the procedure is set up in such a way as to produce agreement among the qualified judges of a proposition concerning its normative value. This type of methodology is profoundly social and contextual. For the normative value of every proposition is generated by some arbitrating community. Thus, pure procedural methodologies entail relativism in one form or another.

II. Against Perfect Procedural Methodology
At the beginning of this paper I remarked that if there is one thing that the history of epistemology teaches us it’s that our epistemic predicament is such that certainty is not in the offing. This historical argument is the best reason we have for abandoning the quest for a perfect methodology. The history of skepticism, especially in the modern period, shows that rationalist responses to skepticism founder on the lack of a guarantee of self-evidency as a criterion for truth (Musgrave, p190), while empiricist responses seem to lead directly to the skepticism we wished to avoid initially (Musgrave, 99-106). Philosophy has yet to produce a sound perfect procedural methodology—though not for lack of trying, and all of the plausible candidates for conclusive standards of truth have been found wanting.

It is true that new types of perfect procedural methodologies may be produced in the future. But there is very little hope that they could be adequate guides to philosophical theorizing. For it is characteristic of methodologies of this type that they fail to yield a wide range of truths, having error avoidance as their chief theoretical virtue. As both Blaise Pascal and William James clearly saw, there are some propositions that we ought to believe because the epistemic gain that accrues if they turn out to be true outweighs a certain degree of risk of their falsity. Because it is sometimes rational to risk error, any methodology that fails to do so is to some degree irrational.

Nevertheless perfect procedural methodologies are attractive. Certainty, if only it could be had, should be our epistemic goal. The high price that we are willing to pay for it, i.e., refraining from believing even propositions that are very likely true, is probably justified. We really should be forced to give up such a laudable goal, and unfortunately, we are. The question we must now ask is this: How far we should be pushed from this original epistemic stance by the knowledge that certainty is not possible?

III. Against Pure Procedural Methodology
Pure procedural methodologies are familiar enough to us post-moderns. In the twentieth century, philosophers like Richard Rorty began to denounced the traditional philosophic goal of objectivity in favor of solidarity. On this new view, there can be no meaningful philosophical analysis of truth or knowledge, rather philosophy is an “ethnocentric” project that cannot appeal to trans-cultural rationality (Rorty, p26). There is no truth but the truth we create together in our own communities in this post-epistemological vision. We must transcend the epistemological quest of the Enlightenment and focus instead on our communities and our conversations.

This is an extreme departure from our original philosophic goals. Not only does the pure procedural methodologist ask that we give up certainty, which we seemed forced to do, but he also asks that we give up any alternative conception of knowledge and objectivity. And what do we gain thereby? A simple dogmatic statement that certain questions are unanswerable, that inquiry is but a metaphor, and that there is but a pragmatic distinction between physics and astrology. At worst this is an abdication by philosophers of an important role in the critique of their own cultures. At best, it is the end of philosophy as we know it.

If the hyperbole in the preceding paragraph may be excused, perhaps it will serve simply to show that there is a yawning chasm between the claim that we must give up the goal of certainty and the claim that the only viable method by which we might proceed is a purely procedural one. So, I think if a middle way can be found, one that preserves as many of our initial philosophical goals as possible, then it must be fully explored before we resort to pure procedural methodologies. This is not a direct argument against this type of methodology. But just as we should be forced to abandon the quest for certainty, similarly we should be forced to abandon the development of an imperfect procedural methodology.


Perhaps there is a direct argument against methodologies of this type, however. Pure procedural epistemologies deny fundamental constraints upon theory-building by insisting that all such constraints are internal to the methodological procedure itself, that there is no independent test—not even an unverifiable one.

A.N. Whitehead proposed four fundamental constraints upon theory building, namely: consistency, coherence, applicability, and adequacy. The first two are internal and rationalistic, consistency being logical consistency and coherence being a measure of how well our concepts hang together. (I leave this vague for now.) The last two are external to a theory and empiricistic, both referring to different ways the world is interpretable in terms of a theory. Applicability stresses the requirement that our theories be about the appropriate subjects. If theories do not have appropriate subjects, then odd—even crazy—consequences follow. For example, a theory whose terms all deal with the refraction of light might be thought to apply to the frequency of divorce among children of the clergy (Ferré, 3). Surely we must admit that our theories must be applicable to subjects that are external to them, else we begin to loose the ability to even advance theoretical interpretations. Adequacy amounts a standard of evidential completeness. It requires that all relevant evidence be considered before any conclusions are reached. The question, of course, is whether evidential completeness can be completely cashed out internally in our theories. I think that it cannot be. For there is a tension in theory-building between coherence and adequacy, which is only one of the tensions between rationalism and empiricism that theory-builders must strive to balance. The tension that I speak of is that between the coherence requirement that the meaningful terms of a theory hang together and the adequacy requirement that our theories take into account all available evidence. Facile theories need to be—and are—broadened to include ever deeper levels of experience. But why should such a tension exist if both coherence and adequacy are internal to a theory? In short, for theories to advance and grow, two conflicting tasks are constantly pursued: the task of reweaving our web of belief in light of new experience and the task of empirical reinterpretation of the world. Importantly, the latter activity seems to have an external or independent standard, albeit an imperfect one.

IV. Philosophical Aspiration
However it is judged today, Descartes’ method of doubt is rightly regarded as one of the great turning points in the history of philosophy. It did much to end an era of subservience to the past and to inaugurate a grand Enlightenment. There is scarcely another doctrine that can pretend to rival its influence. Though the shortcomings of Descartes’ procedure are now well known, the view that knowledge must be based upon absolute certainty is difficult to relinquish. Even so, we have seen that it must be. For I hope to have shown that perfect procedural methodology is untenable.

We have also seen that many of those philosophers who have recognized this fact have fallen prey to the what is perhaps the most pervasive of philosophical errors: throwing the baby out with the bathwater. As Descartes had done so many years before, contemporary philosophers tried to sweep away the past. The futility of this is perhaps greater than the quest for certainty itself. However much the need for ready understanding or the existence of personal or cultural bias might explain the narrowing of relevant philosophical data, it is such narrowing that constitutes philosophy’s greatest danger (Whitehead PR, 337). As each successive wave of philosophers ignores previous insights in the course of advancing its own doctrines, philosophy runs the risk of becoming merely a fashionable science. Today, we stand at the end of a century of many new insights in philosophy, but there was also, in Putnam’s words, “an unprecedented forgetting” of the insights of previous centuries. Our task therefore is a difficult one: We must avoid both of Descartes’ errors by abandoning the quest for certainty without thereby forgetting the lessons of the history of philosophy.


I have argued that we should not go gently into a pure procedural methodology. If we are to see the truths, methods, norms, and objectives of philosophy as mere constructs, then it should only be because we have been forced to accept this view through our utter failure to articulate some middle way. If certainty is not in the offing, then it remains for us to adopt some new goal in its place, giving up as little as we must of our prior epistemic aspirations. In what remains of this paper, I will outline what I think that new goal should be, fleshing out a few of its implications for philosophical theorizing. My purpose in this exercise is practical: I am seeking to establish a few methodological principles to guide us in our subsequent metaphysical theorizing. I do not pretend to establish here an entire epistemological programme, neither do I hope to fully delineate all of the concepts that might eventually be deployed in such a programme. I won’t even engage much in the 20th century practice of providing necessary and sufficient conditions for the concepts that I will be using, as helpful as that can sometimes be. Rather, in broad stokes I hope to indicate a general way that metaphysics might be fruitfully pursued. Even in this restricted area, I make no claim to complete originality. The thrust of my position is that the proper philosophical method is continuous with the historical quest of philosophy itself.

We might do well to remind ourselves of just what that quest was. One of the most eloquent statements of it was given by the American pragmatist, Wilfrid Sellars.

“The aim of philosophy is to understand how things in the broadest sense of the term hang together in the broadest sense of the term. Under things in the broadest possible sense, I include such radically different items as not only cabbages and kings, but numbers and duties, possibilities and finger snaps, aesthetic experience and death. To achieve success in philosophy would be, to use a contemporary turn of phrase, to know one's way around with respect to all these things, not in that unreflective way in which the centipede of the story knew its way around before it faced the question, ‘How do I walk?’ but in that reflective way which means that no intellectual holds are barred” (Sellars, 1).

In short, the aim of philosophy is to make it all fit together and to understand how it fits together. It is to know what all the pieces are and to understand how to move them and how they are connected. One of the tragedies of 20th century philosophy was that many of those who abandoned the quest for certainty—and some of those who did not—tended to loose sight of this, the aboriginal nature of the philosophical quest itself.

I think that a clear recognition of the ultimate aspiration of philosophy will help us determine an appropriate methodological goal. Imagine that the aim of philosophy is somehow fulfilled. What would philosophers have in that case? Two rival answers spring to mind. One might say that philosophers would have a certain body of knowledge, that they would be in possession of a set of propositions that satisfactorily relate all the things in the broadest sense of the term. On the other hand, one might claim that the important thing that the philosophers would have is an understanding of their world, though this would presumably involve a set of propositions, a theory, as the instrument of their understanding. In other words, there seem to be at least two goals here, knowledge and understanding. Although it is unclear what the relation is between them—or even if they are entirely distinct goals, at least we can say that the historical aspiration of philosophy points toward understanding as the deeper and more relevant of the two values. Before we can continue our examination, however, we need to get clearer on what understanding amounts to. Since our ideas about knowledge are relatively clear, a comparison can be made immediately thereafter.

[Jeff]

V. Whitehead’s view of Human Understanding
Perhaps not surprisingly, understanding has seen little treatment in contemporary philosophical discussions. Fortunately, there is some indication that this situation may be changing. Recent works by both Linda Zagzebski and Catherine Elgin place understanding as the fundamental epistemic value. Unfortunately, neither of these philosophers goes very far in telling us what understanding really is. Further, if one surveys the history of philosophy one finds only three sources for discussions of this topic. Firstly, the tradition of Absolute Idealism inherited and refined Kant’s tripartite division of the human cognitive apparatus into sensibility, understanding, and reason. However, this tradition is notoriously obscure, and its conclusions so sweeping and so little explained that we would probably not attain the clarity we seek by explicating their ideas on understanding. More telling perhaps is that their distinction between understanding and reason is too fine for present purposes: We could be content with a characterization of understanding that encompassed both. So, we would do well to look elsewhere for a clear and concise account of understanding. Secondly, sociologists of the 19th century, reacting against Comptean positivism, argued that any scientific discipline that studies man must employ a method of interpreting human actions known as “understanding” (Franklin, 308). This use of “understanding” stresses empathic insight into the motivations of others. While this use of the word is an important one, it is not the primary sense in which we are interested. Rather, our interest is epistemologically broader: It could well include this sense, but is not confined to it. Again, we would do well to locate another source, if possible.

There is a third tradition, which is unfortunately routinely marginalized by many contemporary philosophers. This is the process tradition of Henri Bergson, Alfred North Whitehead, Charles Hartshorne, and others. Whitehead in particular takes up the subject of understanding in his later writings. Especially interesting is Whitehead’s lecture called “Understanding” in Modes of Thought (MT). In the remainder of this section, I will lay out Whitehead’s fascinating view of human understanding and examine its adequacy. I do not offer Whitehead’s analysis as the definitive statement of what understanding is; nevertheless, in a field so wide we must begin somewhere. Whitehead’s views have the merit of capturing many of the philosophically important aspects of this concept, as we will see.

We use the word “understanding” in many different ways. A complete listing of these is beyond the scope of this paper, but it would include the following uses: We use the word to indicate a non-committal attitude toward an interpretation of events, as in, “My understanding of the accident is rather different, officer.” We sometimes use the word to indicate a tacit agreement, as in, “Do we have an understanding?” And I have already mentioned that the word is sometimes used to indicate a feeling of empathy. There are, of course, many other uses of the word.

Whitehead concentrates on understanding as a process of comprehension and suggests that this is the philosophically important use of the term. One of the most eloquent statements of his view can be found in the opening chapter of Process and Reality (PR). There we find Whitehead engaged in a stirring defense of a certain type of speculative philosophy. He outlines a method of imaginative generalization, which he also calls "imaginative penetration" (PR, 4). This "true method of discovery" is likened to the flight of an aeroplane. "It starts from the ground of particular observation; it makes a flight in the thin air of imaginative generalization; and it again lands for renewed observation rendered acute by rational interpretation" (PR, 5). Of course, this is only a metaphor. Even so, a general picture of understanding emerges. The first stage of the process of understanding is one involved with the particularity of the world. Then there is the approach toward insight. At various places, Whitehead indicates that the essence of such insight is the discernment of pattern. We will return to this idea momentarily. Finally, there is the integration of insight with the particularity of the world. The aeroplane touches down: The generalization is precisely interpreted to square with the stubborn facts of the world. And the process begins again. It may shock to think of philosophical understanding as a method that essentially involves imaginative generalization, but Whitehead is not alone in thinking that it does. In a recent history of skepticism, Alan Musgrave endorses a similar view. He writes,

“So the general picture of belief formation which emerges is as follows… [The] observer is active. Certain stimuli are selected and attention is paid to them; certain experiences are recognized or classified as similar to one another, depending on the point of view of the observer; and the observer is constantly jumping to conclusions on the basis of experiences which are important to that observer” (Musgrave, 74-5 [emphasis mine]).


It is worth noting that if the proper method of philosophy is essentially the same process as that of understanding, then philosophy must have a significant relation to the business of living. Whitehead makes this clear in MT. "Apart from detail, and apart from system, a philosophic outlook is the very foundation of thought and life" (MT, 63). And again, "Philosophy begins in wonder. The effort after the general characterization of the world around us is the romance of human thought" (MT, 127).

Whitehead treats of the understanding in more detail in MT. There he divides understanding into two successive phases: penetration and completion. This distinction is tightly interwoven with most of the topics he discusses in Lecture Three of MT. Penetration is said to be partial and incomplete. It is, itself, a process (MT, 43). It is an "advance" toward "disclosure," or complete "self-evidency." In individual experience there is "penetration to novelty" based upon one's causal-historical past and the immediate decisions from which the individual originates (MT, 51). Though humans are not given ultimate disclosure, our penetration is capable of continuous enlargement (MT, 51). This, Whitehead maintains, is the goal of all conscious experience.

This enlargement is fundamentally "the introduction of novelty of pattern into conceptual experience" (MT, 57). In the introduction of novelty, penetration leads us away from concrete fact, to the appreciation of abstract patterns. In this way it is contrasted with completion, which is focused on integrating the product of novelty with concrete processes. "Nothing is finally understood until its reference to process has been made evident" (MT, 46). Thus, completion is a component process of understanding arising subsequent to disclosure, the purpose of which is the integration of disclosure with the world's concrete process. The product of the process of understanding is the integration of insight. The practical consequences of this will vary with the nature of the insight itself.

Whitehead notes that neither penetration nor completion is by itself sufficient for understanding. "To feel the completion without any sense of growth, is in fact to fail in understanding. For it is a failure to sense dimly the unexplored relationships with things beyond. To feel penetration without any sense of completion, is also to fail in understanding. The penetration itself is then deficient in meaning. It lacks achievement" (MT, 48). But not even the presence of both penetration and completion is sufficient for understanding. For a self-evident disclosure is also required. In an important way, "self-evidence is understanding" (MT, 47, 50).

In Lecture Three of MT, self-evidence is associated with disclosure, which is itself associated with intuition. The complete interpretation of this relationship is beyond the scope of this paper, but we may conclude a few things about Whitehead’s notion of self-evidency. First, it is the heart of the process of understanding. It is the disclosure of insight. Second, this insight is essentially the disclosure of a pattern. Finally, the idealized process of understanding is a succession of the processes of penetration, disclosure, and completion. Ideally, the process of penetration leads to disclosure, from which springs the process of completion. Understanding, then, is the process that includes all of these subprocesses. So as a first shot at a concise Whiteheadian definition of understanding, we might advance something like this: Understanding is the process of disclosing a pattern. This isn’t a very satisfying analysis. Fortunately, Whitehead fills it out a bit more.

“Understanding always involves the notion of composition. This notion can enter in one of two ways. If the thing understood be composite, the understanding of it can be in reference to its factors, and to their ways of interweaving so as to form that total thing. This mode of comprehension makes evident why the thing is what it is.

The second mode of understanding is to treat the thing as a unity, whether or not it is capable of analysis, and to obtain evidence as to its capacity for affecting its environment. The first mode may be called the internal understanding, and the second mode is the external understanding” (MT, 46).


These two modes of understanding are really related. Whereas the first conceives of the thing as an outcome of its constituents, the second conceives of the thing causally, as a component of something greater. What must be added to our definition is this notion of the composition of the pattern, with its two related modes. Perhaps, there is a single way of conceiving composition that will allow us to capture both. Whichever mode of composition a pattern may take, we inevitably relate a component to something larger. For internal understanding this is obvious. The paradigm would be something like understanding the cogs in a gearbox (Franklin, 310). But it also holds for external understanding. There we are assimilating a unified thing to an environment, as if the cog itself were our subject instead of the gearbox. So, the idea of composition is the idea of the assimilation of component to a totality whether this be explained in terms of the component or the totality. Hence, understanding is the process of disclosing a pattern of assimilation between a component and some more inclusive kind of existence.


This is as far as we will go in our analysis. I am sure that it could be carried further, but it is enough for present purposes. Instead, let us answer some of the many questions that surround this definition. First, it might be objected that this notion of “pattern” is too vague or fuzzy to be of much use. Fortunately, patterns have been the subject of rigorous study in recent years, so a more formal definition can be given from Information Theory and Systems Theory:

“A pattern may be defined in general as an ordering among elements of a set such that when the arrangement of a subset is given, the arrangement of the remainder is indicated at a probability greater than chance occurrence in proportion to the size of the given subset” (Sayre, 152).

This definition will hold for either type of composition. In the case of internal composition, the set is given and the pattern of relating its elements is disclosed in imaginative generalization, while in the case of external composition, an element is given and the pattern of its relation to the set is disclosed by the same process.

One might also wonder whether our definition is psychologically adequate, i.e., whether it fits with our current knowledge of psychological processes. It is beyond the scope of this paper to layout supporting psychological evidence. However, recent research in cognitive science lends support to our characterization. In a series of works, most recently Philosophy in the Flesh (1999), cognitive scientists George Lakoff and Mark Johnson have argued that the fundamental human cognitive processes involve the assimilation of our experiences, which are body-mediated in philosophically important ways, to certain unconscious cognitive schema. Certainly our definition is wide enough to capture this idea, though it may be a bit too wide as it presently stands.

One might also wonder if our definition is compatible with Sellars’ view of the historical philosophical aspiration mentioned earlier (see section IV). But there is no particular reason why it shouldn’t be. Whitehead himself had a similar view of the purpose of philosophy: "The useful function of philosophy is to promote the most general systematization of civilized thought" (PR, 17). Indeed he saw philosophizing as an important part of everyday life, and thought the goal of understanding was of primary philosophical importance.

There remains one objection that appears much more problematic: One might question the philosophical realism that our definition seems to assume. In fact, Catherine Elgin, who pioneered the imperfect procedural approach to philosophical methodology, is not a realist. She fashions herself a Constructionalist, by which she means that we find in the world no external standard to check the adequacy of our theories. Rather, the standards by which we do this are themselves “products of cognitive practices, not external checks on them” (Elgin, 127). In her view this is not a retreat to the pure procedural methodology of a Wittgenstein or a Rorty, since standards do exist external to the theory itself. However, these standards are not objective, but given by the cultural practices of rational construction itself. This is an interesting position; nevertheless, I think that it is one we should resist given the current status of the epistemological quest. If Elgin is correct in thinking that most of the problems of contemporary epistemology have been generated either by those who refuse to relinquish philosophical certainty or by those who think there is no external check on our theories whatever, then much will depend upon the new epistemic goal that we set in our imperfect procedural methodologies. Elgin claims the goal of understanding, just as Whitehead does. Unfortunately, she offers no substantive account of what it means to understand. This constitutes some motivation for siding with Whitehead on the issue. But there are better reasons. Many of the arguments against realism are themselves products of the failed quest for certainty and the skepticism that surrounded it. Alan Musgrave has argued that the safest fallback position in the current epistemological climate is critical realism (Musgrave, 274-300). A sophisticated indirect realism about the existence of external objects seems immune to many if not all of the arguments advanced against naïve or direct realism. Such a realism accepts large parts of the skeptical critique of sense perception, and the price of doing so is a necessary fallibilism. I haven’t the space to adequately survey these arguments; however, on the assumption that critical realism is compatible with imperfect procedural methodology, it seems clear that it should be the desired stance. There is no getting past the powerful human intuition that we exist within a world of colors, sounds, and other sense objects, and that we seem to be elements of such a world in the same sense as are the other things which we perceive (Whitehead SMW, 88). If an appeal is to be made to naïve experience, any methodology that denies the existence of the external world makes a mockery of it. If one were to deny this intuition, it would be incumbent to explain exactly how a common world of thought could arise without a common world of experience. And even if this were possible, it would seem to be a gross case of begging the question to build the idea into philosophical methodology when the fate of critical realism is yet to be decided. For all these reasons, I think that Elgin is too hasty in her assumption of the Constructionalist position, though she may well be correct in her assessment of it as a style of imperfect procedural methodology.

VI. Knowledge Vs. Understanding
We are now in a position to compare the adequacy of understanding and knowledge as the proper goal of imperfect procedural methodology. The traditional account of knowledge tells us that “to know” is a success verb. Once you are sure that you know, you need never doubt again. Thus, traditional accounts of knowledge are based on the Cartesian ideal of certainty. It remains to be seen if some new account of justification can make knowledge compatible with the fallibilism of modern epistemologists. Further, traditional accounts of knowledge traffic in propositions to such a degree that any nonpropositional elements that may be present in one’s cognitive system are effectively excluded from even appearing in an account of one’s epistemic status. If the findings of the cognitive scientists are correct, then there is good reason to think that such nonpropositional content abounds in human cognitive structures. Indeed, Lakoff and Johnson argue that there is no rigid symbol to world designation captureable by propositions (99-101). As Zagzebski notes, there is also good reason to question our ability to evaluate singly any particular subject’s beliefs (Zagzebski, 44). But this operation seems to be required by traditional accounts of knowledge. If all—or even some—of these allegations are correct, then perhaps knowledge is too restrictive a goal for an imperfect procedural methodology.

Understanding, as we have characterized it here, fares much better. First, it is not wedded to the tradition of certainty. Our ability to recognize patterns, to apply and extend them, frequently comes in degrees. Though understanding can become more and more complete, it can never be completely certainty. For there is always a wider context that can radically alter the importance of patterns recognized in more limited ones. When the detective discovers that her own assistant has the best motive for the murder and no good alibi, then all of the evidence collected by the assistant takes on an entirely different significance. Whitehead fully anticipated this result: “…How shallow, puny, and imperfect are efforts to sound the depths of the nature of things. In philosophical discussion, the merest hint of dogmatic certainty as to finality of statement is an exhibition of folly” (PR, xiv). Second, there is no reason that the objects of understanding must be propositions. Spatial relations, which fit visual scenes together, and motor relations, which pertain to bodily movements, are both difficult if not impossible to capture adequately with propositions or even in formal languages. But both can easily be seen as objects of human understanding. Finally, since understanding can have nonpropositional objects, there is no reason it would force us to individuate beliefs in any nonintuitive way. So, if the quest for certainty is given up, understanding appears to be a much more reasonable goal than knowledge for imperfect procedural methodology.

[Jeff]

VII. Theoretical Virtues
Finally, we must consider how adopting understanding as our methodological goal will affect our philosophical theorizing. The goal of understanding makes plausible several theoretical virtues. By a theoretical virtue, I mean a quality of a theory that maximizes the likelihood of understanding. I shall argue that the four virtues mentioned in section III, namely: consistency, coherence, applicability, and adequacy, are necessary virtues of philosophical systems, though they may not be the only ones.

It is important to notice that these theoretical virtues are not quite the same thing as criteria for theory choice. Neither are they equivalent to criteria for epistemic justification. For this reason I have avoided both terms in favor of the idea of “theoretical virtues.” Theory choice has to do with the weighing of competing theories, and it emphasizes such virtues as conservatism and simplicity. These are not straightforwardly virtues of pattern recognition and generation. On the other hand, criteria for epistemic justification don’t seem to be what we are after either. Most of these theories understand their criteria to be part of an adequate account of knowledge. As I hope to have made clear in the last section, understanding is a rather different sort of epistemic goal.

What we are looking for are the limiting principles in accordance with which a theory could be constructed to stand the best chance of advancing philosophical understanding. I will take a theory to be some scheme of ideas intended to capture a pattern. “The first requisite is to proceed by the method of generalization so that certainly there is some application; and the test of some success is application beyond the immediate origin. In other words, some synoptic vision has been gained” (PR, 5). Our definition of understanding helps us to see what Whitehead has in mind here. The ideas that we use to capture a pattern must give us some probability of predicting the composition of either the totality, if the theory is based on a consideration of elements, or the elements themselves if it is based on the giveness of the totality. The standard of applicability is therefore minimal. It merely tells us that the theory applies to the pattern in one of the two ways in which they can be characterized. Thus, it is a necessary constraint upon all theory. What the elements of a pattern are is, of course, an open question. But generally patterns are important to the degree that they relate previously unrelated elements or allow us to grasp new totalities.

“The second condition for the success of imaginative construction is unflinching pursuit of the two rationalistic ideals, coherence and logical perfection” (PR, 6). The purpose of these twin goals is to explore the interpretation of experience in terms of some scheme. Mere logical consistency of its component ideas is not enough for understanding, though it is required. I will take its necessity as given; it is a requirement so long as our theories are to mean something rather than nothing. Whereas the consistency requirement can be viewed as a negative condition for philosophical theories, coherence is a positive one. Whitehead’s conception of coherence is not quite the same as that found in theories of epistemic justification. He is not referring to a property of an epistemic agent, but a property of the scheme of ideas in which patterns are expressed. Whitehead insists that it is only by giving our ideas their widest possible scope that their coherence can be explored. “In framing a philosophic scheme, each metaphysical notion should be given the widest extension of which it seems capable. It is only in this way that the true adjustment of ideas can be explored” (AI, 237). Coherence is then the adjustment of ideas within the schemes of their origin. This is not mere logical consistency, a minimal requirement for adjustment at best. Rather, coherence requires that the scope of any metaphysical principle should not be limited otherwise than by the necessity of its meaning. But why should this be a necessary property of philosophical systems? Again, our definition of understanding helps to answer this question. The ideas necessary to express a pattern are not simply read off the pattern as we find it. In cases in which patterns are incapable of rigorous philosophical statement, this is obviously so. Rather, our ideas are framed by schematizations of our imaginative generalizations, as Whitehead and others have claimed. These generalizations need not be either expressed in the simplest language or appeal to the fewest theoretical entities. However, by giving our metaphysical principles the maximal scope that their meanings can tolerate, we insure both of these goals. Coherence, then, is a principle of theoretical parsimony, akin to Ockham’s razor, which is widely held to be a necessary constraint on theorization.

Adequacy, the final theoretical virtue that I will discuss here, becomes relevant in the completion phase of understanding. Whereas applicability stressed the appropriateness of the pattern to some part of experience, adequacy stresses application of the pattern to all relevant parts of experience. Much more is required of a theory than a single important application; in principle theories demand evidential completeness. “The adequacy of the scheme over every item does not mean adequacy over such items as happen to have been considered. It means that the texture of observed experience, as illustrating the philosophical scheme, is such that all related experience must exhibit the same texture” (PR, 4). Again, why should this be a necessary feature of philosophical systems? As I argued in section III, it is the tension between adequacy and coherence that makes our theories both broad and deep. It would be easy to stop theorizing with a “foolish consistency.” But Emerson’s point is that facile theories are banal. A standard of evidential completeness coupled with a commitment to explore the maximum scope of our philosophical principles is the best way to ensure that our theories genuinely advance human understanding.