Grand Central Station, New York City (circa 1930)
The essay below will be published in two installments; this is the second.
You can also download and read or share a .pdf of the complete text of this essay, including the list of REFERENCES, by scrolling down to the bottom of this post and clicking on the Download tab.
What is Human Knowledge? Categorical Normativity and Categorical Epistemology, #2.
… [T]he categorical epistemology conception of a philosophical explanation of the normatively best and highest kind of knowledge—that it adequately establishes an inherent or intrinsic connection between the sufficient conscious-evidence-based reason for a believer’s assertion of her belief-claim, via her properly-functioning cognitive capacities or mechanisms, and the truth of her belief—is (perhaps surprisingly) largely compatible with Timothy Williamson’s highly plausible “knowledge first” approach to epistemology in Knowledge and its Limits (Williamson, 2000: p. v).
There are three reasons for large measure of compatibility, all of which flow directly from my Kant-inflected and Essential Embodiment-oriented conception of the theory of knowledge, categorical epistemology.
First, High-Bar knowledge—i.e., intrinsically compelling, cognitively virtuous, essentially reliable justified true belief—which is the normatively highest kind of knowledge, is the primitive, non-analyzable, irreducible, immanently structured, and categorically normative highest good and ideal standard of rational human cognition with which epistemology is fundamentally concerned. Second, High-Bar justification contains the three basic elements of (i) intrinsically compelling, cognitively virtuous, essentially reliable justification, (ii) truth, and (iii) belief, and these are the metaphysically non-detachable, essentially-related elements of High-Bar Knowledge. And third, a priori knowledge via basic authoritative objectively necessarily true rational intuition is the perfection of our capacities for rational human cognition, and therefore counts as the normative paradigm of High-Bar Knowledge.
Or in other words, categorical epistemology is a perfectionist Kantian morality of essentially embodied rational human cognition. No doubt, Williamson would sharply disagree with me about the robust rational normativity of authentic a priori knowledge. But at the same time, we do both hold that knowledge as such is a primitive, non-analyzable, irreducible cognitive phenomenon with which all serious explanatory epistemology must begin, even though I would contend, contra Williamson, that the non-analyzability of the proper parts of the cognitive phenomenon of knowledge is explained by their being connected synthetically a priori. Furthemore, we do agree that knowledge is inherently mentalistic and factive. So there is significant philosophical common ground shared between us, alongside some important differences.
More generally, categorical epistemology is both non-trivially similar to and also non-trivially dissimilar to other contemporary approaches to epistemology. On the one hand, categorical epistemology shares with virtue epistemology (Brady and Pritchard, 2003; Fairweather and Zagzebski, 2001; Sosa, 2007) and other recent or contemporary practically-oriented approaches to epistemology (Stanley, 2005) the basic idea that both the ascription and also the actual occurrence of human knowledge have the following characteristics: they are inherently sensitive to our properly-functioning cognitive capacities or mechanisms; inherently motivated by rational human interests; inherently governed by rational human ideals, values, and reasons (i.e., norms); and ultimately grounded on the real fact of (or in at least the non-eliminable conception of ourselves as having) free agency. But on the other hand, categorical epistemology sharply differs from other practically-oriented approaches to human knowledge in the following respect. According to categorical epistemology, the principles of rational human animal knowledge are grounded on categorically normative principles, which in turn are all ultimately subsumable under the Categorical Imperative. Hence the governing norms of knowledge are explicitly and irreducibly categorical—i.e., unconditional, strictly universal, non-instrumental, and a priori—and also ultimately constrained by the Categorical Imperative.
Correspondingly, it should also be fully noted that the fundamental distinction in categorical epistemology between High-Bar justification and knowledge, and Low-Bar justification and knowledge, is itself only a specification of a more general and necessary structure of human rationality, which I call Two-Dimensional rational normativity. Two-Dimensional rational normativity is the fact that the conditions on normative evaluations of rationality fall into two importantly different kinds: (i) Low-Bar rational normativity: the necessary and sufficient conditions for minimal or nonideal rationality, which include the possession of online, uncompromised versions of all the cognitive and practical capacities constitutive of intentional agency, and (ii) High-Bar rational normativity: the necessary and sufficient conditions for maximal or ideal rationality, which include all the necessary and sufficient conditions for Low-Bar rational normativity as individually necessary but not jointly sufficient conditions, and also include the perfection, or correct and full self-realization, of all the cognitive and practical capacities constitutive of intentional agency, as individually necessary and jointly sufficient conditions. Non-satisfaction of the conditions for Low-Bar rational normativity entails non-rationality and non-agency. As we’ll see below, in a certain special range of cases of the non-satisfaction of the conditions for Low-Bar knowledge, Low-Bar rational normativity further allows for the possibility of what, following Frank Hofmann, I will call Non-Conceptual Knowledge, in non-human animals such as cats or horses, and also in non-rational human animals such as infants or unfortunate adult victims of various pathological cognitive conditions. Hofmann compellingly argues that non-conceptual perception not only is regularly called “knowledge” by cognitive scientists, and furthermore satisfies four basic conditions on any cognitive activity that plays the “knowledge role,” but also grounds conceptual/doxastic perceptual knowledge and justification by putting the cognitive subject in a position to have them (Hofmann, 2014).
Nevertheless, by sharp contrast, it is not the case that non-satisfaction of the conditions of High-Bar rational normativity entails either non-rationality or non-agency. This point, in turn, makes it possible to see very clearly the fundamental flaw in One-Dimensional theories of rational normativity, no matter how plausible and sophisticated these theories might otherwise be (see, e.g., Korsgaard, 2009). According to a One-Dimensional theory, any failure to meet the ideal standards of rational normativity entails non-rationality, non-agency, and non-responsibility. To be sure, on a sophisticated One-Dimensional theory, there can be a continuum of degrees of rationality with a variety of significant thresholds along the way. But the basic fact remains that in a One-Dimensional framework, any degree of rationality short of the ideal standards is to that extent non-rational. Or in other words, if you are not ideally or perfectly rational, then you are a rationally defective or irrational animal, and off the hook. For example, if you fail to know in the highest sense (i.e., if you fail to have High-Bar justified true belief), then you are not in any sense a rational or responsible cognitive agent, although you may approach that epistemically blessed state to a greater or lesser degree. Or if you fail to act in the practically or morally highest way (i.e., if you fail to have a good will in Kant’s sense), then you are not in any sense a rational or responsible practical or moral agent, although you may approach that morally blessed state to a greater or lesser degree.
Disastrously, these results of One-Dimensionalism play directly into the hands of radical cognitive, practical, and moral skeptics, since as a matter of fact no actual rational human animal ever manages to meet all or even most of the High-Bar standards of rational normativity, but instead is doing extremely well indeed if she ever manages to meet some of them—for example, successfully performing some basic authoritative a priori objectively necessarily true rational intuitions in mathematics, logic, or philosophy. How convenient for the radical skeptic, then, that most or all of us, most or all of the time, turn out to be irrational animals. Perhaps even more disastrously, these results also play directly into the hands of “human, all too human” intentional agents looking for a fast track out of their everyday cognitive and practical difficulties in a thoroughly nonideal actual natural world. How convenient for them that falling short of rational perfection should entail the suspension of responsibility: If rationality—like God—is dead, then everything is permitted, and they can take the nihilist’s way out, like the pathetically wicked character Smerdyakov in The Brothers Karamazov:
“Take that money away with you, sir,” Smerdyakov said with a sigh.
“Of course, I’ll take it! But why are you giving it to me if you committed a murder to get it?” Ivan asked, looking at him with intense surprise.
“I don’t want it at all,” Smerdyakov said in a shaking voice, with a wave of the hand. “I did have an idea of starting a new life in Moscow, but that was just a dream, sir, and mostly because ‘everything is permitted’. This you did teach me, sir, for you talked to me a lot about such things: for if there’s no everlasting God, there’s no such thing as virtue, and there’s no need of it at all.
Yes, sir, you were right about that. That’s the way I reasoned.” (Dostoevsky, 1958: vol. 2, p. 743)
For these reasons, then, it is clear that One-Dimensional theories of rational normativity are false.
On The Two-Dimensional theory, however, things are sharply different. Satisfaction of the conditions for Low-Bar rational normativity is a necessary and sufficient condition of the cognitive, practical, and moral responsibility of intentional agents, but it does not guarantee that any of the further conditions of High-Bar rational normativity are actually satisfied. In other words, it is fully possible for an intentional agent to be minimally and nonideally rational, but in a bad or wrong way, to any degree of badness or wrongness, all the way down to the lowest limiting case of cognitive or practical monstrosity within its kind. For example, at any point short of the limiting case of an utter disregard for, and a complete inability to heed, any and all canons of reasonable belief, truth, and validity/consistency in logical reasoning—at any point short of sheer madness—the intentional agent remains cognitively responsible to some degree. So too at any point short of the limiting case of an utter disregard for, and a complete inability to heed, any and all moral principles grounded on the dignity of persons, and any and all canons of validity/consistency in practical reasoning—at any point short of sheer sociopathy or the complete disintegration of agentive coherence—the intentional agent remains morally and practically responsible to some degree.
As a consequence of that fact, it’s also fully possible for an intentional agent to be minimally and nonideally rational in a good or right way, to any degree of goodness or rightness, all the way up to the highest limiting case of cognitive or practical perfection within its kind—for example, successfully performing some basic authoritative a priori objectively necessarily true rational intuitions in mathematics, logic, or philosophy—for all of which, again, the intentional agent is also fully cognitively and practically responsible.
As my discussion so far implies, explicitly situating categorical epistemology within the framework of Two-Dimensional rational normativity yields a fourfold classification of different, basic, normatively-graded kinds of cognition. This fourfold classification comes clearly into view when we recognize the notion of context-sensitive causal reliability, together with the fact that certain kinds of cognitive acts or states in non-human animals, and in non-rational human animals, fall short of Low Bar knowledge, yet still include what I call essentially non-conceptual content and direct sense perception (Hanna, 2015: chs. 2-3) and also a context-sensitive causally reliable cognitive mechanism for evidentially connecting sense perception with its worldly objects. So non-human animals, non-rational human animals, and rational human animals share the minimally basic epistemic capacities, and by exercising those capacities well, they thereby can all achieve Non-Conceptual Knowledge.
In a nutshell, my rationale for this claim is grounded on the following three points. First, direct sense perception based on essentially non-conceptual content is perceptual knowledge-by-acquaintance. Second, perceptual knowledge-by-acquaintance is genuine knowledge in at least three important senses, namely (i) that it guarantees an essentially reliable, non-accidental connection between cognition and the world, (ii) that it involves the successful exercise of the minimally basic epistemic capacities, and (iii) that its cognitive phenomenology is maximally evidential in that context. And third, therefore direct sense perception based on essentially non-conceptual content is also genuine knowledge in at least three important senses, even though it fails the belief condition and the truth-condition on Low Bar knowledge and High Bar Knowledge. More explicitly, then, the larger Two-Dimensional framework that comprehends categorical epistemology provides for a non-conceptual, non-doxastic, non-alethic, and distinctively different fourth kind of minimally basic epistemic activity, namely Non-Conceptual Knowledge, to go along with mere Low-Bar knowledge, with context-sensitive causally reliable Low-Bar knowledge, and with High-Bar knowledge.
Non-Conceptual Knowledge is similar in several important ways to what Ernest Sosa calls “animal knowledge” (Sosa, 2001), but with two crucial additions: first, Non-Conceptual Knowledge is cognitively driven by essentially non-conceptual content, and second, it both occurs and also makes sense only within the larger four-levelled, Two-Dimensional explanatory framework of categorical epistemology, whereas Sosa’s explanatory framework utilizes a more compact binary contrast between animal knowledge and reflective knowledge (Sosa, 2007, 2009). As such, some classes of cases of Sosa’s animal knowledge fall under Non-Conceptual Knowledge, and some of them fall under one or another of the kinds of Low-Bar knowledge. Correspondingly, some classes of cases of Sosa’s reflective knowledge fall under the more Internalistically-sophisticated kinds of Low-Bar knowledge, and some of them fall under High-Bar knowledge. All things considered, I do think that Sosa’s “virtue reliabilist” account is in many ways fundamentally correct, but also that the more complex structure of categorical epistemology, embedded within a cognitive-semantic theory of essentially non-conceptual content and conceptual content, ultimately does more explanatory work, and also characterizes the highest kind of knowledge more completely.
In what follows, by a contingently reliable cognitive capacity or mechanism I mean a cognitive capacity or mechanism that tracks truth in the actual world. The notion of a contingently reliable cognitive capacity or mechanism can then be put alongside the two notions of a context-sensitive causally reliable cognitive capacity or mechanism and an essentially reliable cognitive capacity or mechanism, that I previously formulated.
Granting all that, then here are contextual-definition-style formulations of the four basic kinds of knowledge recognized by categorical epistemology: (i) Non-Conceptual Knowledge: Perception P in an animal subject S is Non-Conceptual Knowledge if and only if (ia) P is based on essentially non-conceptual content, and (ib) S possesses a properly-functioning and context-sensitive causally reliable cognitive capacity or mechanism that yields S’s conscious evidence E for P. (ii) Low-Bar Knowledge: Belief B in an animal subject S is Low-Bar Knowledge if and only if (iia) B is true, (iib) S possesses a properly-functioning and at least contingently reliable cognitive capacity or mechanism that yields S’s conscious evidence E for B, and (iic) S has a reason for asserting B based on E, i.e., S has a Low-Bar justification for B. (iii) Context-Sensitive Causally Reliable Low-Bar Knowledge: Belief B in an animal subject S is context-sensitive causally reliable Low-Bar Knowledge if and only if (iiia) B is true, (iiib) S possesses a properly-functioning and context-sensitive causally reliable cognitive capacity or mechanism that yields S’s conscious evidence E for B, and (iiic) S has a reason for asserting B based on E, i.e., S has a Low-Bar justification for B. (iv) High-Bar Knowledge: Belief B in an animal subject S is High-Bar Knowledge if and only if (iva) B is true, (ivb) S possesses a properly-functioning and essentially reliable cognitive capacity or mechanism that yields S’s intrinsically compelling conscious evidence E for B, and (ivc) S has a sufficient reason for asserting B based on E, i.e., S has a High-Bar justification for B.
This fourfold classification of kinds of cognition combines elements of epistemic internalism, epistemic externalism, virtue epistemology, and contextualism (Steup and Neta, 2024) within the progressively larger frameworks of categorical epistemology and Two-Dimensional rational normativity, while also sustaining the classical thesis that (conceptual, doxastic, rational) knowledge is justified true belief. In this connection, it should be specifically noted that although Non-Conceptual Knowledge is not in any way subject to Gettier considerations—that is, not subject to the possibility of a merely accidental or contingent connection between conscious evidence and the world—nevertheless Non-Conceptual Knowledge is not conceptual and not doxastic, and therefore not “in the logical space of reasons” (Sellars, 1963: p. 169), or directly subject to the constraints of of even Low-Bar rational normativity. So Non-Conceptual Knowledge flows from the successful exercise of minimally basic epistemic capacities, and is knowledge in a genuine sense—namely, the sense in which “knowledge by acquaintance” is genuine knowledge. Moroever, Non-Conceptual Knowledge constitutes a kind of essentially non-conceptual and context-sensitively causally reliable animal cognition that grounds all the other kinds of knowledge. Furthermore, Non-Conceptual Knowledge anticipates some of the necessary features of rational human knowledge in the normatively highest sense. Nevertheless, Non-Conceptual Knowledge is at most pre-rational and proto-rational. Therefore, strictly speaking, it is neither Low-Bar knowledge nor High-Bar knowledge.
At the same time, however, although Low-Bar Knowledge is indeed “in the logical space of reasons,” and thereby subject to the constraints of rational normativity, it is open both to Gettier considerations, and also to global skeptical worries. More specifically, in some introspectively indistinguishable conceivably possible worlds the very same conscious-evidence-based reason for S’s belief is connected to a falsity-maker, not a truth-maker (Cohen, 1984). Thus Low-Bar Knowledge falls well short of knowledge in the normatively highest sense. By sharp contrast to both Non-Conceptual Knowledge and Low-Bar Knowledge, however, High-Bar Knowledge is not only “in the logical space of reasons,” and thereby subject to the constraints of rational normativity, and both contingently and causally reliable. It is also essentially reliable, as well as sufficiently justified by a conscious-evidence-based reason, via a properly-functioning cognitive capacity or mechanism. High Bar Knowledge is thereby impervious to Gettier worries and to global or radical skepticism alike. Hence, again, High-Bar Knowledge is the highest good or summum bonum of epistemology.
Now, what about context-sensitive causally reliable Low-Bar knowledge? If S possesses knowledge in this sense, then S possesses context-sensitive causally reliable Low-Bar a posteriori knowledge, which is a pretty good kind of knowledge to have—say, via trustworthy testimony—but at the same time context-sensitive causally reliable Low-Bar knowledge is without complete conviction, intrinsic compellingness, or self-evidence, and also without essential reliability. For one thing, just as with Low-Bar Knowledge, so too with context-sensitive causally reliable Low-Bar knowledge, in some introspectively indistinguishable conceivably possible worlds the very same conscious-evidence-based reason for S’s belief is connected to a falsity-maker, not a truth-maker. This possibility leaves context-sensitive causally reliable Low-Bar knowledge wide open to radical or global skepticism. And for another thing, as I pointed out earlier in this section, because context-sensitive causally reliable Low-Bar knowledge does not necessarily include rational insight into the underlying structure of what connects S’s conscious-evidence-based reason for believing to the truthmaker of her belief, her conscious-evidence-based reason for believing could be epistemically flawed in various ways, including greater or lesser irrelevance to the situation at hand, greater or lesser superficiality, greater or less triviality, or more or less obvious formal inconsistency with other beliefs she holds. However, by sharp contrast, when I look carefully at this sequence of strokes, i.e.,
| | | | | | |
and thereby come to believe that there are seven strokes on the page, then I possess High-Bar a posteriori knowledge. This is because my evidence-based reason for believing that there are seven strokes on the page is inherently or intrinsically connected to the truth-maker for that belief, via veridical, direct sense perception. This in turn constitutes an epistemically appropriate, properly-functioning cognitive capacity or mechanism. And the cognitive phenomenology—i.e., the subjectively-experiential specific characters (Bayne and Montague, 2011; Smithies, 2013a, 2013b)—of my perceptual belief is also intrinsically compelling or self-evident. By another important contrast, when a normal, healthy, minimally linguistically competent 3-year old child comes to believe that 3+4=7 by counting aloud on her fingers, which for her is at best a semi-reliable cognitive process and clearly not mathematical rational intuition, then she possesses Low-Bar a priori knowledge. And by a final important contrast, in the now-familiar case in which I know that
3+4=7, i.e., | | | + | | | | = | | | | | | |
via mathematical authoritative rational intuition, then I possess High-Bar a priori knowledge, which is the very best and highest of all kinds of knowledge, even better than High-Bar a posteriori knowledge. In so doing, I have thereby achieved membership in the indefinitely large class of cases of knowing that collectively constitute the jewel in the crown of the summum bonum of epistemology.[i]
NOTE
[i] For a full-dress presentation, elaboration, and defense of the material discussed in this essay, see (Hanna, 2015).
REFERENCES
(Bayne and Montague, 2011). Bayne, T. and Montague, M. (eds.), Cognitive Phenomenology. Oxford: Oxford Univ. Press.
(Brady and Pritchard, 2003). Brady, M. and Pritchard, D. (eds.), Moral and Epistemic Virtues. Oxford: Blackwell.
(Chalmers, 1996). Chalmers, D., The Conscious Mind: In Search of a Fundamental Theory. New York: Oxford Univ. Press.
(Cohen, 1984) Cohen. S., “Justification and Truth.” Philosophical Studies 46 (1984): 279–295.
(Dick, 1968). Dick, P.K. Do Androids Dream of Electric Sheep? New York: Doubleday.
(Dostoevsky, 1958). Dostoyevsky, F., The Brothers Karamazov. Trans. D. Magarshack. 2 vols. Harmondsworth, Middlesex: Penguin Books.
(Fairweather and Zagzebski, 2001). Fairweather, A. and Zagzebski, L. (eds.), Virtue Epistemology. Oxford: Oxford Univ. Press.
(Gettier, 1963). Gettier, E., “Is Justified True Belief Knowledge?” Analysis 23: 121-123.
(Hanna, 2006). Hanna, R., Rationality and Logic. Cambridge MA: MIT Press. Also available online in preview HERE.
(Hanna, 2015). Hanna, R., Cognition, Content, and the A Priori: A Study in the Philosophy of Mind and Knowledge. THE RATIONAL HUMAN CONDITION, Vol. 5). Oxford: Oxford Univ. Press. 2015. Also available online in preview HERE.
(Hofmann, 2014). Hofmann, F., “Non-Conceptual Knowledge.” Philosophical Issues 24:184-208.
(Horgan, 1993). Horgan, T., “From Supervenience to Superdupervenience: Meeting the Demands of a Material World.” Mind 102: 555-586.
(Ichikawa and Steup, 2024). Ichikawa, J. and Steup, M., “The Analysis of Knowledge.” In E.N. Zalta and U. Nodelman (eds.), The Stanford Encyclopedia of Philosophy. Fall Edition. Availanble online at URL = <https://plato.stanford.edu/archives/fall2024/entries/knowledge-analysis/>.
(Kim, 1993). Kim, J. Supervenience and Mind. Cambridge: Cambridge Univ. Press.
(Korsgaard, 2009). Korsgaard, C., Self-Constitution: Agency, Identity, and Integrity. Oxford: Oxford Univ. Press.
(Lehrer, 1990). Lehrer, K., Theory of Knowledge. London: Routledge.
(Pritchard, 2012). Pritchard, D., “Anti-Luck Virtue Epistemology.” Journal of Philosophy 109: 247-279.
(Putnam, 1983). Putnam, H. “There is at Least One A Priori Truth.” In Realism and Reason: Philosophical Papers, Vol. 3. Cambridge: Cambridge Univ. Press. Pp. 98-114.
(Scott, 1982). Scott, R. (dir.) Blade Runner. USA: Warner Bros.
(Sellars, 1963). Sellars, W., “Empiricism and the Philosophy of Mind.” In W. Sellars, Science, Perception, and Reality. London: Routledge & Kegan Paul. Pp. 127-196.
(Shope, 1983). Shope, R., The Analysis of Knowing. A Decade of Research. Princeton: Princeton Univ. Press.
(Smithies, 2013a). Smithies, D., “The Significance of Cognitive Phenomenology,” Philosophy Compass 8: 731-743.
(Smithies, 2013b). Smithies, D., “The Nature of Cognitive Phenomenology,” Philosophy Compass 8: 744-754.
(Sosa, 2001) Sosa, E., “Human Knowledge, Animal and Reflective.” Philosophical Studies 106: 193-196.
(Sosa, 2007). Sosa, E., A Virtue Epistemology. Oxford: Oxford Univ. Press.
(Sosa, 2009). Sosa, E., Reflective Knowledge. Oxford: Oxford Univ. Press.
(Stanley, 2005). Stanley, J., Knowledge and Practical Interests. Oxford: Clarendon/Oxford Univ. Press.
(Steup and Neta, 2024). Steup, M. and Neta, R. “Epistemology.” In E.N. Zalta and U. Nodelman (eds.), The Stanford Encyclopedia of Philosophy. Winter Edition. Available online at URL = <https://plato.stanford.edu/archives/win2024/entries/epistemology/>.
Williamson, T., Knowledge and its Limits. Oxford: Oxford Univ. Press.
Against Professional Philosophy is a sub-project of the online mega-project Philosophy Without Borders, which is home-based on Patreon here.
Please consider becoming a patron!
