Foundationalisms

On what should we base our beliefs? "Reason and evidence," some are inclined to reply. Sure. Fine. But what is reasonable, and what counts as evidence?

 

Biblical fundamentalists think that something being asserted somewhere in the Bible is a piece of evidence that the asserted proposition is true. Why do they treat such a thing as evidence? Because, we might say, they believe a certain story about the Bible and its origins, a story which, if true, implies that biblical assertions have enormous evidentiary weight. But why think this story about the Bible is true?

 

Most everybody thinks that sensory experiences can serve as evidence for the truth of claims about a world external to the mind, claims prompted by our sense experience. Why?  Maybe there's a story here, too, a story about our senses and their relation to the world that would have to be very different from the story told in the first part of the movie "the Matrix." But why believe that story? Or maybe there isn't a story here at all. Maybe we just believe the deliverances of our senses, period. There is no reason. We just can't help it.

 

Leibniz thought that the Principle of Sufficient Reason (PSR), much like the principle of non-contradiction, was a self-evident principle (self-evident to our intellect) which could reliably guide our reasoning. As such, he thought he could reason from some pretty uncontroversial propositions to the conclusion that a necessary being must exist. We might say that, based on his understanding of "reason," the existence of contingent things is evidence for the existence of a necessary thing. Why does Leibniz accept PSR? Why does Hume reject it?

 

In each of these cases we are led to think about foundations. Our reasoning has to have starting points from which we reason--things that are either immediately taken as having evidential value, or are taken to have evidential value on the basis of some broader system of beliefs--perhaps organized into a narrative framework. And our reasoning doesn't merely need evidential starting points. We also need to have standards of reasoning--principles which guide our inferences, leading us to take some body of evidence to be evidence for a conclusion.

 

In short, to base our beliefs on reason and evidence, we need rational and evidential starting points. Foundationalist epistemologies treat our starting points as that upon which our whole network of beliefs is ultimately built.  Following the metaphor of building construction, we need a solid foundation or the whole edifice becomes unsound.

 

In my last post, I suggested that Hegel could be viewed as a foundationalist, but of a special kind--namely, a fallibilistic foundationalist. In this post, I want to distinguish some different species of foundationalism, locate Hegel's dialectic within this taxonomy, and offer a case for the advantages of Hegel's approach.

 

I'll begin with a broad distinction between what I’ll call absolute foundationalism and fallibilistic foundationalism. Absolute foundationalism treats one’s foundations as beyond challenge, and might be further divided into two sub-species: what I’ll call Cartesian-style foundationalism, and what I’ll call dogmatic foundationalism. Cartesian-style foundationalism seeks to discover which foundations are, in fact, beyond dispute—through something like the methodological doubt pursued by Descartes in his Meditations—and then rely only on such foundations. Dogmatic foundationalism, by contrast, treats its starting points as beyond dispute but makes no attempt to offer a philosophical case for doing so.

 

In fact, dogmatic foundationalism might be viewed as arising out of the perceived failure of Cartesian-style foundationalism, based on a pair of prima facie compelling critiques. The first critique runs basically as follows. To rely on nothing other than those starting points which are indubitable is to rely on starting points that can’t get you anywhere. If you insist on believing only what can be grounded on such starting points, you are driven to a kind of solipsism: the only thing you can trust is the existence of your own thoughts, and you are unwarranted in believing that any of those thoughts correspond to any reality outside your own mind. This is a kind of extreme skeptical outcome that is intolerable—it makes living a human life impossible—and can only be overcome by allowing in foundations that don’t have the Cartesian stamp of indubitability.

 

The second critique should be familiar to followers of this blog: No attempt to prove that one’s foundations are beyond dispute can be successful, because no such attempt can be ungrounded. How, then, can the attempt itself avoid relying on foundational starting points? And these will be either the very ones one is attempting to establish, or different foundations which, therefore, have not themselves been proven to be beyond dispute.

 

In effect, dogmatic foundationists draw three lessons from these critiques of Cartesian-style foundationalism: (a) everyone has foundations, (b) everyone has more foundations than the class of indubitable ones (assuming this class has members at all), and (c) no set of foundations can be justified or shown to be the right ones, since all such efforts are either circular (and hence dogmatic) or appeal to deeper, unconsidered foundations that are held to dogmatically.

 

This third lesson is what leads them straight to dogmatic foundationalism: we have no choice but to be dogmatic foundationalists, but we do have a choice about which dogma to embrace. On this view, the only difference between the empiricist and the biblical fundamentalist is which foundations are being held to dogmatically—"who" one has chosen to treat as one’s ultimate epistemic authority.

 

The problem with this argument, of course, is that it is a false dilemma. There are more options than the two versions of absolute foundationalism (Cartesian-style and dogmatic). There is fallibilistic foundationalism.

 

The fallibilistic foundationalist accepts (a) and (b), but isn’t prepared to embrace an unqualified form of (c). Instead of insisting that “no set of foundations can be justified or shown to be the right ones,” the fallibilistic foundationalist wants to say instead that “no set of foundations can be justified or shown to be the right ones prior to their adoption.” But once one has adopted a set of foundations, a critical stance becomes possible with respect to specific foundational beliefs.

 

Hegel's dialectic can be seen as falling into this class. I will concede that there are those who would resist using the term "foundationalism" to describe Hegel's method, but this is a linguistic quibble in which I'm not much invested. I recall that once, at a party in graduate school, I announced that Hegel "has a coherence theory of knowledge but a correspondence theory of truth." In so doing I classed him as a coherentist rather than as a foundationalist. But these are terms of art in the discipline--and there is something about Hegel's epistemology that strikes me as warranting the foundationalist label (as I noted in my last post).

In describing Hegel's approach as it relates specifically to philosophy, Michael Allen Fox notes in his surprisingly accessible book on Hegel (called, interestingly enough, The Accessible Hegel) that Hegel was largely indifferent to one's starting point in the philosophical journey. Wherever we start, "our philosophical journey will inevitably be a prolonged process of self-examination in which thought interrogates itself and remedies its deficiencies as it progresses." The end result of this process will be what "'justifies' the starting point we have chosen by proving its fruitfulness in yielding knowledge."

We have to start somewhere--to think at all requires us to have rational and evidential starting points that characterize what that thinking will look like. But as we engage in whatever style of thinking these starting points entail, we will inevitably encounter and transcend deficiencies. Whatever we think about, we will eventually be forced inward to thought itself, and called upon by its failures to revise our starting points.

In fact, Hegel thought this was true not just of philosophy. The same dialectical process plays out not just in the history of philosophy, but in human history. The efforts of human societies to engage with the world drive humanity towards "self-interrogation" in which deficiencies are discovered and remedied.

As a way of clarifying and defending this Hegelian approach, I want to consider a version of fallibilistic foundationalism out of which Hegel's species might be thought to arise. I’m inclined to give it a big mouthful of a name: defeasible-but-trustworthy doxastic practice foundationalism. After all, it’s important for academics to come up with pretentious-sounding names, or they wouldn’t be academics. But to save space, I’ll call it DTDP foundationalism for short.

 

To understand this species of foundationalism, we need to know what a “doxastic practice” is. The term is one I get from William Alston, and it basically refers to any belief-forming practice one might engage in. Here are some examples: 

Assign a set of contrary beliefs to heads and tails on a coin, flip the coin, and believe in accord with the outcome of the coin flip.

Use your senses to investigate your environment, and believe whatever propositions you find yourself immediately inclined to believe based on such sensory investigation.

In attempting to form beliefs about what happened to you in your past, consult your memory.

In forming beliefs about regularities in nature, make predictions about what one ought to observe under given conditions if the supposed regularity holds, then create or put oneself into the given conditions and determine if one observes what was predicted. Keep doing this. If the supposed regularity remains unfalsified after extensive testing, accept it as true.

In forming beliefs about what happened in Europe in the 19th Century, consult your old “History of Western Europe” textbook.

In forming beliefs about what happens after bodily death, consult the Upanishads and the Vedas and believe whatever is clearly stated on this matter in these texts.

 DTDP foundationalism attaches credibility to some but not all of the range of possible doxastic practices, in effect treating as foundational any belief that comes from this select list of “trustworthy” practices. But a belief coming from a “generally trusted” practice is not immune to criticism. Rather, any such belief, call it B, is believed unless and until it is confronted with “defeaters,” that is, other beliefs that arise out of trusted doxastic practices and which call B into question.

 

There are different sorts of defeaters, but I won’t get into that now. What this rough characterization makes clear is that this species of foundationalism acknowledges the fallibility of our foundations, and builds a kind of checks-and-balances system into our belief-forming practices. The checks and balances cannot be implemented apart from embracing the presumptive trustworthiness of a set of doxastic practices, but insofar as the trustworthiness of their deliverances is only presumptive rather than absolute, it allows for specific foundational beliefs produced by the agreed set of doxastic practices to be rejected based on the total deliverances of the entire set.

 

But how do we arrive at which doxastic practices to trust in the first place? Here, there seem to be a range of alternatives. You could take an essentially social approach: identify those doxastic practices that enjoy both wide social use (they are widely relied on in the formation of beliefs) and a high degree of intersubjective corroboration (they typically lead most people to the same beliefs when used in the same circumstances).

 

Then there’s the introspective approach. You look to the “phenomenological features” of the doxastic practice and its deliverances—that is, those qualities or characteristics available to introspection. Perhaps the beliefs formed by the practice consistently just seem right to you, so much so that you essentially “cannot help but believe them.” Or maybe the practice itself has an inner “veridical feel”—a sense of putting you in touch with truth as you’re engaging in it.

Or maybe you identify those doxastic practices you pretty much have to trust if you want to live anything remotely like an ordinary human life, and you trust them and only them.

Alternatively, you might take the approach of someone like Alvin Plantinga, saying that those doxastic practices are trustworthy which involve the use of cognitive faculties under those conditions in which the faculties are generally reliable. The problem here, of course, is that there may be a difference between a doxastic practice being reliable and our being able to tell that it is reliable. Plantinga has famously proposed that we have a cognitive faculty dubbed "the sensus divinitatus" which is responsible for theistic belief and which produces that belief reliably. His critics instantly cry foul--but part of Plantinga's point is that we have no inner guarantee or non-circular proof that any of the things we treat as reliable cognitive faculties are reliable cognitive faculties. That a congitive faculty is trustworthy, in the sense of consistently delivering true beliefs, does not clearly entail that we have immediate access to this fact.  And that, of course, is part of the problem. How do you decide which doxastic practices to presumptively (not absolutely) trust?

 

However you decide, it appears at least initially as if you confront variants of the same troubles that plagued absolute foundationalism. The problems have simply shifted from the level of foundational beliefs to the level of foundational doxastic practices.

 

Consider: Suppose I adopt a social version of DTDP foundationalism—trusting my senses, the pronouncements of certain socially agreed-upon authorities, etc. Either I have grounds for believing that doxastic practices embodying the relevant social features are more likely to produce true beliefs than those that lack these features, or I don’t. If I do, what would those grounds be like? Wouldn’t they either be circular (relying for their support on beliefs that are ultimately derived from these very same doxastic practices) or based on different doxastic practices that then fall outside the scope of justification in terms of their social features? Aren’t we then driven to being dogmatic in our embrace of the doxastic practices we embrace, simply shifting the problem from unjustified trust in certain foundational beliefs to unjustified trust in certain foundational belief-forming practices?

 

Here is where we are, I think, pushed towards something more Hegelian. The reason why DTDP foundationalism offers a basis for critiquing foundational beliefs, even when they come from trusted doxastic practices, is because the whole set of trusted practices can generate defeaters for individual beliefs. But by the very same mechanism, DTDP foundationalism also provides a basis for critiquing the doxastic practices themselves.

 

Here’s what I have in mind. If a doxastic practice that was initially put into the “trustworthy” class keeps encountering defeaters for the beliefs it generates, it might eventually be bumped out of that class. Or if the doxastic practice encounters defeaters in a certain context, it might be judged untrustworthy in that context. We might find that it encounters fewer defeaters when certain refinements are made to the practice, and so the practice might evolve to embody these refinements. It is very easy to see the scientific method as emerging out of precisely this sort of refinement.

 

But keep in mind that all of this evolution is influenced by one’s starting points—that is, by which doxastic practices one has initially included in the generally trustworthy set. What set of doxastic practices one treats as trustworthy will shape what defeaters, if any, drive the evolution of the whole set. Start with too narrow a set, or get “unlucky” in your starting points, and you may be left with only a few refined doxastic practices that you trust—and can you really trust even them? Maybe their having escaped defeat is mostly a function of the shrinking set of possible sources of defeat.

 

The worry here is this: The smaller the set of trusted doxastic practices, the less likely any of them are to encounter defeaters—simply because of how few they are, rather than because of their intrinsic merits. And as the evolutionary process described above continues, you won’t be pumping out new doxastic practices. Defeaters destroy. They don't create. As the evolution continues and the number of trustworthy doxastic practices shrinks, we’ll thus be put into a position in which we should be increasingly less confident that continued-failure-to-be-defeated is a good gauge of trustworthiness.

Perhaps this worry isn't decisive, but a different worry looms nearby. Even if one's remaining doxastic practices are deemed trustworthy based on their success in producing undefeated beliefs, we might worry that the subset of beliefs generated by these practices is far too narrow to be representative of reality, or that it is too narrow to help us live our lives in a way that meaningfully embodies the truth about ourselves and our world. We may have evolved certain practices to a point of high refinement but left ourselves with a set of practices that is nevertheless deficient, because the boundaries of the knowable that they establish are too narrow.

 

But once one treats doxastic practices as mutually evolving in the light of each others’ deliverances, one can begin to ask whether this fixation on nothing but doxastic practices might not itself speak to a set of unconsidered presuppositions—some hidden starting points that shape our ideas about what should be our starting points. Why should we make doxastic practices the sole underpinnings of our belief system? Why not, instead, begin with an interpretation of the whole of experience, embrace a set of doxastic practices based on that interpretation, and then critically live out this whole?

In other words, starting with doxastic practices as the foundation for all our beliefs is not the only way to start, nor is it the only thing that can lead to an evolutionary process. So, why not allow as a legitimate starting point a perspective which treats a narrative picture of reality as foundational, and doxastic practices as derivative?

 

A process that treats only doxastic processes as foundational, and then evolves by successively refining and discarding doxastic practices in the light of experience, is in danger of ultimately leaving one with too small a set of foundational beliefs on which to build. As noted above, while experience with the failures of doxastic practices can refine and rule out ones that were already part of your epistemic arsenal, it doesn’t generally introduce into your arsenal wholly new doxastic practices that you’ve never thought to try.

 

Consider the analogy of natural selection: to take advantage of all the "niches" available in the environment, it’s not enough to have a mechanism that kills off unsuccessful species. You also need some creative wellspring from which new species arise. Storytelling, speculative philosophy and theology, mythological narrative—these things are the creative side of the human endeavor, and it is in part from them that new ways of engaging with the world, of wrestling with what to believe and what not to believe, come about. Perhaps, then, we shouldn't just allow holistic interpretations and narratives as potential starting points, but should allow a place in our epistemological lives for the creative process by which such interpretations and narratives arise. In other words, we should treat storytelling and speculative thinking as not just an entertaining diversion but as helping us in the quest for truth.

This is not to say, of course, that every story we tell should be immediately embraced as true or anything like that. But some stories resonate in such a way that people are moved to live them out as if they were true. And in some cases these initial experiments don't immediately come crashing down by virtue of a volley of defeaters. Instead, the experiments knock down some parts of the story but not others. And some elements in the narrative motivate new doxastic practices which themselves prove extremely fruitful, in the sense of pumping out beliefs that resist defeat. Many see the birth of the scientific method in these terms: the Christian conviction that the universe is essentially orderly, that there must be a regularity and structure beneath even the most chaotic-seeming dimensions of our experience, set the stage for science.

 

And of course there’s more to life than telling stories and engaging in doxastic practices. We have artistic and athletic activities, cultural and religious rituals, spiritual practices, ways of life that embody value systems and moral codes. Engaging in the business of living impacts, in all sorts of ways, our beliefs and our ways of forming beliefs.

 

The point is that doxastic practices are not isolated from human life in all its richness and diversity. To allow them to evolve in isolation from the rest is to presuppose that there isn’t anything in the rest that has the power to promote the quest for coming into alignment with the Truth. And I would say that the fallibilism implicit in DTDP foundationalism calls one to be suspicious of any such presupposition. So, not only does DTDP foundationalism move beyond criticism of the deliverances of one’s adopted doxastic practices and towards lived-out criticism of the practices themselves; it also moves towards allowing for more than just doxastic practices into the class of legitimate starting points one might legitimately live out critically.

And so we've made our way towards something that looks more like the Hegelian approach. But how does it work, you may ask? How does one framework give rise to another? What sorts of deficiencies motivate a change in one's system?

 

Here I must confess to an attempt to be clever. In case no one’s noticed yet, this post has not merely been an attempt to make a case for the Hegelian dialectic. It’s also been laid out, however imperfectly, to serve as an example of the Hegelian dialectic at work.

What is Naturalism? Part V: Alternative Ontologies and Why We Should Care

So far, my primary aim in my series of posts on naturalism has been to offer an account of metaphysical naturalism—that is, naturalism construed as an ontology (an account of the nature of reality). I have proposed to define metaphysical naturalism as follows: it is the view that the picture of the world that emerges through the exercise of the scientific method (roughly as described in previous posts) is an exhaustive picture of the fundamental reality behind the world of appearances. In other words, it is the view that the scientific “picture” that emerges through the scientific method describes not merely the world that is constructed by the subject as it receives and organizes data coming in from things-in-themselves; it actually describes for us what the underlying noumenal reality is like. Or, more precisely, this is what would be achieved by the picture of the world that science would ultimately produce if it were able to achieve and adequately test an ultimate model of everything. In any event, science is bringing us ever closer to an exhaustive grasp of noumenal reality.
If this is what metaphysical naturalism maintains, then metaphysical supernaturalism becomes the view that even if science were to achieve that elusive theory of everything, and test it rigorously and successfully, the picture it gave us would not be an exhaustive depiction of noumenal reality.

But if we define metaphysical supernaturalism in those terms, then a further division becomes possible, between two broad species. The first species shares the naturalist premise that the scientific method is generating an account of reality as it is in itself, and not merely as it appears to us—but it denies that science gives us all of reality. This is the view that there are two kinds of reality—what might be called material or physical reality, and then some other kind (spiritual or mental, perhaps). Descartes can be seen as epitomizing metaphysical supernaturalism in this sense, insofar as he endorses a dualistic ontology that sharply divided the physical world and the mental, and insofar as he takes science to be capable of studying only the former.

The second kind of supernaturalism, by contrast, denies the naturalist premise altogether. That is, it denies that science and its methods have taken us past the domain of phenomena. This would be Kant’s contention. He would argue that at every stage in the process of science, we are dealing with the objective pole of consciousness, which is constructed by the engagement of the self with the world. What science gives us, then, is an increasingly useful phenomenal picture—that is, an understanding of the interface between self and external reality that facilitates at least one species of successful engagement, namely prediction and control. But this useful picture is still just a picture—the world as it appears to us, mediated through our faculties, our concepts, and our categories—and so is not to be identified with reality as it is in itself.

With respect to this second kind of supernaturalism, a further subdivision becomes possible, based on whether noumenal reality is thought to be utterly unknowable or simply unknowable through scientific means. The former is Kant’s position. The latter position admits of even further variations, in terms of both how much of the noumenal realm is thought to be accessible to human cognition and what means are thought to be of use in helping us understand the noumenal. These further subdivisions will not interest me here. It should be clear from what I have said in previous posts, however, that both Hegel and Schleiermacher fall within this broad category, as do many others—including, most significantly, a philosopher who preceded Kant by many, many centuries, and who might be regarded as the founder of Western philosophy. I mean, of course, Plato.

So, let us lay out these alternatives in more formal terms:

Metaphysical Naturalism: The theory that noumenal reality is to be exhaustively identified with the “idealized scientific picture of the world”—that is, the picture that would be generated by the scientific process of careful observation, modeling, and empirical testing of models, if only that process were allowed to continue to completion.

Metaphysical Supernaturalism A (Cartesian Supernaturalism): The theory that the idealized scientific picture would give us an accurate portrait of one kind or dimension of noumenal reality (what we might call material or physical reality) but that there is another kind or dimension (mental or spiritual) that lies outside the scope of science to adequately represent.

Metaphysical Supernaturalism B (Kantian Supernaturalism): The theory that the idealized scientific picture does not give us noumenal reality at all, but offers at best only an ever-more-useful phenomenal picture. Noumenal reality is something entirely other than what emerges in the scientific picture, and is furthermore inaccessible through any other means of inquiry. Noumenal reality, in short, is unknowable.

Metaphysical Supernaturalism C (Platonic Supernaturalism): The theory that the idealized scientific picture does not give us noumenal reality at all, but offers at best only an ever-more-useful phenomenal picture. Noumenal reality is something entirely other than what emerges in the scientific picture, but can be at least partly grasped through other means.

Now none of these categories offers a specific account of the nature of reality. What I have offered here are categories into which more precisely characterized ontologies might fall—but I have done so primarily in epistemological terms (that is, in terms of what and how we know). I have done so in large measure because my starting point was with naturalism, and because at least as I understand naturalism, the naturalist’s ultimate allegiance is epistemological.

Let me explain. If you are a metaphysical naturalist as I have defined it, that means you identify ultimate reality with the picture that would emerge through an idealized science. But insofar as contemporary scientists disagree about numerous things at a very basic level, being a naturalist in this sense doesn’t commit you to a specific picture of what ultimate reality is like. A metaphysical naturalist may have such a picture (and most usually do), but an allegiance to metaphysical naturalism entails a fallibilism with respect to any such picture, insofar as science might advance in ways that force the picture’s abandonment. There are surely metaphysical naturalists who believe string theory offers a (barely comprehensible) portrait of ultimate reality. But to be a metaphysical naturalist in the sense I have been articulating here is to care more about the scientific method than about any specific scientific means of modeling the world. If we advanced to a point where string theory could be empirically tested, and if it were then falsified, anyone who clung to string theory at that point would not be a metaphysical naturalist in my sense.

But this also means that, given my taxonomy, those who thinks noumenal reality is utterly inaccessible through any means, including scientific ones, are metaphysical supernaturalists (of the Kantian sort)—even if they also believe that (a) the inaccessibility of noumenal reality entails its irrelevance for human life, (b) all we should care about is understanding the phenomenal world in which we live our lives, and (c) we ought to rely entirely on scientific methods (and, perhaps, scientifically endorsed methods) in shaping our beliefs.

But persons who embrace (a)-(c) are not apt to call themselves supernaturalists, even though in terms of their views about ultimate reality they deny that an idealized science would put us in touch with that reality (and hence, in effect, believe that there is more to reality than meets the scientific eye). They are, instead, apt to call themselves naturalists. And of course, there is something very significant they have in common with metaphysical naturalists as defined above: they share a deep allegiance to the scientific method and its outcomes.

What this shows is that metaphysical naturalism is only one species of naturalism, and that for a full understanding of naturalism we need to make some further distinctions. In effect, the Kantian “supernaturalists” described above dismiss the entire category of “noumenal reality” as unimportant for human existence, and identify “reality” in the only significant sense as that which we must contend with in experience. The world of experience is the world we live in, and that world can be more or less accurately depicted. Accuracy is measured in terms of the depiction’s usefulness for prediction and control—and the method that's proven itself to consistently produce the best depictions for these purposes is the scientific method.

This perspective is what might be called “pragmatic naturalism.” For the purposes of living our lives successfully, noumenal reality is irrelevant. The only thing that matters is what the phenomenal world is like, in terms of such things as the rules it follows and the state it is in—in short, in terms of the features that we must contend with whether we like it or not. This is “reality” in the pragmatic sense, and the best tool for putting us in touch with that reality is science. Theological speculation, by contrast, is utterly useless in telling us what we need to contend with—and so is useless in helping us to better predict what will happen next and how we can better control outcomes. Praying over grandpa doesn’t restore his heart to health and so keep him alive longer. If we want to do that, we need to learn through scientific methods how the heart works and what interferes with its ability to do its job—and then use that knowledge to devise life-saving surgeries.

Now from this broadly pragmatic perspective, there is no doubt at all that science is extremely important and that we ignore it at our peril. For the sake of predicting and controlling outcomes in the world we live in (however merely phenomenal that world might be), science is really the only game in town. But once we’ve stepped away from the question of what is ultimately real and into the question of what is most useful for living our lives in the world of appearances we inhabit, we need to acknowledge that there can be different ways in which a picture of things is useful. Prediction and control is only one species of usefulness. Conduciveness to a sense of wholeness or meaning should surely be construed as useful as well—and why should usefulness in terms of prediction and control be given priority over other forms of usefulness?

Furthermore, if the whole reason for attaching a special priority to the lessons of science is that they enable us to more effectively engage with our world, science is being assessed within a context of agency. That is, we are assuming that we have the power to make decisions based on our understanding of things. But decision-making depends not only on an understanding of what effects we are most likely to produce by this course of action as opposed to that course of action. Decision-making also depends on an assessment of which effects are most worthy of pursuing, and on judgments about the legitimacy of the available means for pursuing those results. A Nazi researcher conducting experiments on unwilling Jewish subjects might have been as committed to the scientific method as anyone else, and as conversant with its lessons as anyone of his era…and yet it doesn’t follow that this scientist is operating with the most useful belief system for successfully living a human life.

Put more simply, that science offers the most successful beliefs for the sake of prediction and control doesn't mean it offers the most successful beliefs in all areas. It doesn't mean it can ground ethics. In fact, some are convinced that the picture of the world generated by science is entirely value-neutral—reducing values to mere epiphenomenal facts about brains. But agency cannot operate without values. In fact, the entire pragmatic approach to deciding how and what we should believe is a value-laden one: “usefulness” is a value concept. And if values are wholly subjective, then the usefulness of any enterprise—including science—is radically relativized. And so it no longer becomes possible to say, as the pragmatic naturalist does, that the scientific method offers the “best” way of forming beliefs about our world. It is only the best relative to certain subjective value systems, that is, relative to certain brain impulses.

A pragmatic naturalist might attempt to avoid these problems by maintaining, as Sam Harris does, that science can be relied upon to give us the ethical norms we need for decision-making, and that these scientifically-grounded norms are in some sense “better” than any others. I have already expressed my skepticism about that, but for now it is enough to note that in looking to science for an objective grounding for ethics and values, Harris is at best expressing a pious hope. It certainly isn’t the case that the scientific method has demonstrated that science can ground objective morality.

There's another issue rooted in the fact that pragmatic naturalism operates in the context of agency. Specifically, what needs to be the case for “agency” to be possible at all? To be an agent is to make decisions based on reasons—but is that even possible under a purely scientific view of the self? To act based on reasons is a very different thing than to exhibit certain behaviors based on the causal influence of pre-existing conditions in accord with natural laws. The latter accounts for behavior in terms of past events producing an outcome. The former would have it that we are in some way capable of bringing about a course of action because of a judgment to the effect that something is worthy of being done.

One might attempt to take the “judgment that this act is worthy of being done” to be just folk-psychology short-hand for a complex predecessor brain state that produces further brain stimuli in accord with natural laws. But in that case, the conscious judgment as such is simply an epiphenomenal by-product of the brain state that causes the behavior.

This is important for understanding why Kant thought we had to postulate something distinctive about the noumenal self in order for morality to be possible. For Kant, moral principles are truths of reason. They are true by virtue of the nature of rational consistency itself. But unlike other truths of reason, such as mathematical truths (2+2=4, say) these are expressly about what behavior it is rational to engage in (whether or not anyone actually engages in it or has any desire to do so).

For Kant, being moral is about being directly motivated by one of these distinctive truths of reason rather than being motivated by what he calls “inclination” (by which he means any appetite, desire, instinct, emotional impulse, etc.). So, in order for it to be possible to be moral, we have to be capable of doing something, not because there is some prior inclination that causes us to do it, but rather because we are directly motivated by a truth of reason. But truths of reason are not brain states. 2+2=4 is not identical with any brain state, even if a brain state might represent mathematical truths. What makes 2+2=4 true is not some neurological event in the brain but the fact that the concept of “2” is related to the concept of “4” in the indicated way. It’s an abstract truth, not a physical fact.

But here’s the key thing: being moral means, for Kant, being motivated to action by an abstract truth rather than by a physical fact. If I do what I do necessarily and exclusively because of complex brain states that bring it about, then I am necessarily removed from the domain of moral agency.

And if I am removed from the domain of moral agency, then I am removed from the domain in which there might be any objective values that can guide my behavior (as opposed to the subjective preferences that are really just physiological impulses which feel a certain way). My behavior then falls into the same category as the behavior of the weather or the tides, except that the causal mechanism is far more complex and so much harder to predict (and is wedded to epiphenomena that are causally inefficacious). But if my behavior is in the same category as that of the weather or tides, then offering reasons to value science or speaking of the pragmatic value of science becomes utterly pointless. Giving reasons and arguments, making the case for valuing this or that—all of this presupposes a genuine capacity to be responsive to abstract ideas and judgments and principles, as opposed to being motivated by nothing but causal forces.

Or consider artistic activity. If all human actions are caused by prior physical brain states rather than being motivated by abstract principles and ideas, then a poem produced by a great artist is not in a different category than (to borrow B.F. Skinner’s crude metaphor) a turd squeezed out by toddler. The causal mechanism responsible for the former is more complex—but without the capacity for human beings to be guided behaviorally by abstracted ideas themselves (even if we concede that these ideas are always represented by brain states), none of the things we would like to say about the poem and the poet (things it makes no sense to say about the toddler and the turd) can meaningfully be said.

That’s how both Kant and I see it, in any event—and I cannot help but think that those who think materialism can easily make sense of human moral and artist agency just haven’t thought about it long enough to understand the problem. I’m not saying that the problem is necessarily insurmountable (although I can’t see a way to surmount it). What I’m saying is that the problem is real and it’s HUGE. Unless the problem can be surmounted, a worldview that admits in nothing but what science offers us about human nature and the mind is a worldview that systematically strips human life of everything that, well, makes us human.

If we are operating purely at a pragmatic level, rather than at the level of ultimate truth, then I cannot see how anyone could view such a worldview as pragmatically the best. While I fully accept that any pragmatically useful worldview must value the scientific method and treat its conclusions concerning the phenomenal world (the world we must contend with) as invaluable for the sake of prediction and control, the fact is that there isn’t just a phenomenal world I have to contend with. There is also a “me” who has to contend with the world. And there are some presumptions about this “me” that I cannot help but adopt as I engage with the world. One of these is that I can be and really am responsive to ideas and reasons and arguments—abstract entities that, like 2+2=4, cannot coherently be identified with any physical state of affairs (even if, as may very well be the case, these abstract entities are represented by a physical state of affairs, that is, by a brain state).

For Kant, this is precisely the point at which an unknowable noumenal reality becomes relevant despite its unknowability. Because Kant was convinced that the capacity for agency was not part of our phenomenal experience of the world. This may sound strange, since most of us have an instinctive understanding of ourselves as capable of making choices based on reasons. But this capacity for agency isn’t what we experience when we make ourselves the object of experience. When we look at ourselves in that way—when we put ourselves at the objective pole of consciousness and study ourselves the way that we study the stars and the inner working of biological organisms—the result is an understanding of ourselves as causally determined to behave as we do. And so as soon as we make of the self an object of experience--as soon as it becomes part of the phenomenal world--we are presented with an object about which we cannot attribute free agency.

Our sense of freedom and agency is present only when we don’t turn our attention to ourselves, only when we don’t make of ourselves an object of experience. It is immediately felt, we might say—part of our subjective experience of ourselves rather than part of our objective experience of ourselves. That this feeling isn’t corroborated when we investigate ourselves empirically would be devastating for the possibility of moral agency if the outcomes of such empirical investigation represented reality as it is in itself. But because it doesn’t, the determinism found in the empirical world doesn’t preclude us from assuming what we need to assume in order for human life to have the meaningfulness and dignity that (for Kant as well as myself) only comes from moral agency.

In other words, for Kant the chief value of an unknowable noumenal reality is that it makes it possible for us to coherently make postulates that transcend what science can teach us about the empirical world, postulates that we must make in order for our overall worldview to really work, not merely for the sake of prediction and control but for the sake of acknowledging ourselves to be what we intuitively take ourselves to be, and what we must be if our lives and projects are to be more meaningful than a turd.

All of this is a way of saying that Kant is not merely a metaphysical supernaturalist of type B, but also a pragmatic supernaturalist in something analogous to Metaphysical Supernaturalism A. If a pragmatic naturalist finds the most useful holistic picture of the world to be a purely scientific one that makes no room for positing entities or powers that defy scientific modeling, then Kant is a pragmatic supernaturalist in the sense that he finds a holistic picture that includes “transcendent” elements to be more useful overall. These transcendent elements are not objects of knowledge—because they are postulates about what falls outside the phenomenal world and hence outside the realm in which (for Kant) knowledge is possible. That there is a noumenal reality beyond the phenomenal world, even if we can't know anything about it,  make it possible to include in our overall worldview postulates that don't describe the empirical world and aren’t knowable—and this in turn makes it possible for us to operate in the phenomenal world with a worldview that is pragmatically more useful than any worldview limited to affirming what empirical investigation can support.

In effect, then, I am interpreting Kant as embracing Metaphysical Supernaturalism A and pairing it with pragmatic supernaturalism—in contrast with those who embrace Metaphysical Supernaturalism A and pair it with pragmatic naturalism.

Both of these positions need to be further contrasted with someone who embraces Metaphysical Naturalism but then pairs it with pragmatic supernaturalism. This would be someone who thinks that supernatural beliefs are false but useful and so should be encouraged despite being out of touch with reality. Daniel Dennett seems to think that many modern defenders of religion fall into this category, when he talks about “belief in belief” in Breaking the Spell. Richard Dawkins seems to treat Karen Armstrong in these terms as well.

I suspect they’re wrong. I think the theologians and philosophers they pigeonhole in this way are, like Kant, pragmatically-minded adherents to Metaphysical Supernaturalism B. That is, they think that ultimate reality is beyond our reach—including the reach of science—but that we nevertheless need to operate in terms of a (fallible) picture of reality, and that the best procedure for doing so is pragmatic. A picture that denies the models produced by the best science is less useful than any picture that affirms them—but one that affirms nothing but them is less useful than some pictures that affirm something more.

As anyone who’s read my book should know, I lean towards a kind of pragmatic supernaturalism—but I don’t identify reality with what is pragmatically useful. That is, I think reality is what it is, so to speak. But I also think that what we must contend with in experience, and which ways of thinking are most helpful in contending successfully, offer important guidance into bringing our worldview into alignment with a deeper reality (in this respect I am closer to Hegel than to Kant).

I’ve also been convinced by certain philosophical arguments (especially the Leibniz/Clarke Cosmological Argument which I discuss in my book) that the kind of reality pictured in scientific modeling cannot offer ultimate explanations. And so, if there is to be a reason for why there is something rather than nothing (a principle I find myself unable to set aside), then there must be a noumenal reality crucially unlike what science depicts. As such, I cannot accept Metaphysical Naturalism. But when it comes to which species of Metaphysical Supernaturalism I adopt, I find myself waffling quite a bit--but I think I ultimately lean towards the Platonic variety.

But I'm not content to simply go with my leanings here, because I think Schleiermacher and Hegel are right about something important: Whatever the noumenal reality is, I’m a part of it. Not the "me" who is an object of experience—that’s the phenomenal me. And the naïve “phenomenal me” that comes from immediate introspection is no less phenomenal than what scientists look at when they study my brain. What bearing it has on the “noumenal me” remains an open question.

But still, I am what I am—and so in being me (as opposed to putting myself at the objective pole of conscious observation and then studying me) I am being part of noumenal reality. And there may be a way to leverage that fact into some kind of understanding of noumenal reality. That’s what Hegel tries to do in The Phenomenology of Spirit--and whether a project like that is going to work is, I think, a question we should not attempt to answer in advance of pursuing it. I think the way to decide whether noumenal reality can be accessed in some fashion or other is to really try out different strategies and methodologies. The one Hegel recommends is, in effect, a pragmatic one—but one that fills in some crucial holes in more conventional forms of pragmatism.

And so there’s my set-up for talking about Hegel in future posts. But it may take awhile to get those posts put up, in part because I'm getting ready to leave for vacation, but more significantly because talking about Hegel is hard. There are two reasons for this. First, his writing is so danged obscure, so loaded with his own invented jargon. Second, Hegel has an authorial voice I can’t stand, one that is weighted down by the man’s obvious arrogance and sense of self-importance. But unlike certain others whose authorial voices turn me off, I’ve become convinced that Hegel’s is one I need to contend with—because, even if he is a pompous ass, he’s also brilliant.