ROBERT HANNA | Cognition, Content, and the A Priori | Oxford University Press 2015


By Robert Hanna

I am very grateful to David Landy for all the time and effort he has spent reading, analysing and criticising my work on Kantian Non-Conceptualism (KNC) and especially Cognition, Content, and the A Priori (CCAP).

For the purposes of his critical comments on CCAP, Landy has zeroed in on one highly specific issue:

Can we represent complex objects as complex without concepts?

Let’s call this The Representational Complexity Question, a.k.a The RCQ.

In turn, there are two versions of The RCQ that we need to distinguish:

RCQ1: In reality, can we represent complex objects as complex without concepts?

RCQ2: According to Kant, can we represent complex objects as complex without concepts?

I think that the correct answers to RCQ1 and RCQ2, alike, are yes. Landy thinks that the correct answers to RCQ1 and RCQ2, alike, are no. Obviously, however, it is also rationally possible (i.e. conceptually and logically consistent, intelligible, and prima facie plausible) correctly to answer yes to RCQ1 but no to RCQ2, or correctly to answer no to RCQ1 but yes to RCQ2. For me, at least, what matters infinitely more is that the correct answer to RCQ1 be yes, since CCAP is a Kant-inspired systematic study in contemporary philosophy of mind and knowledge, and not an exegetical work on Kant’s philosophy. So it would not worry me very much if, under the weight of a great many Kant-texts showing that the all-things-considered correct answer to RCQ2 is no, I had to fall back from the heroic double yes position to the more cautious RCQ1-yes-but-RCQ2-no position.

In any case, for better or worse, I never was very interested in playing Kant-text-quoting ping pong; indeed, here’s what I said in Kant, Science, and Human Nature about my approach to Kant-interpretation:

Abstracting away now for a moment from the domestic differences between the various approaches to Kant-interpretation … it should be noted that the leading interpretive assumption of Kant, Science, and Human Nature, like that of Kant and the Foundations of Analytic Philosophy, is that the history of philosophy is a genre of philosophy, pure and simple. It is purely and simply the genre in which the fundamental questions of philosophy are addressed through the explication and critical analysis of great (by which I mean the most brilliant, groundbreaking, mind-changing, and trendsetting) old books, and in which the theses, arguments, and theories found in those great old books are directly related to contemporary philosophical debates.

As applied specifically to Kant, however, for me this implies two methodological principles and an overarching dictum. The two methodological principles are as follows. First, charitably attribute to Kant the best philosophical view consistent with all the texts on a given topic. Second, in cases of conflicting texts on a given topic, charitably attribute to Kant the best philosophical view consistent with at least some of his texts, and bracket the texts in which he seems confused or mistaken. Like all philosophers, Kant sometimes errs, or anyhow nods. But we respect him most by critically noting and then setting aside his slips, and by promoting his deepest and most powerful doctrines. So the overarching dictum is this: Kant’s Critical Philosophy is fully worth studying, critically analyzing, charitably explicating, defending, and then independently developing in a contemporary context. This is because, in my opinion, more than any other single-authored body of work in modern philosophy the Critical Philosophy most doggedly pursues and most profoundly captures some non-trivial fragment of the honest-to-goodness truth about rational human animals and the larger natural world that surrounds them. (Hanna 2006:7)

And in the case of defending KNC, I’ve tried my best to follow those principles and that dictum. Fortunately for me, however, there is actually a simple alternative argument for the heroic double yes position that also decisively proves that position, and is also substantively different from the, as it were, ‘classical’ Two Hands Argument for KNC that Landy cites and criticises.

The simple alternative argument depends on a thesis, first explicitly formulated in CCAP, but not discussed by Landy, that I call The Autonomy of Essentially Non-Conceptual Content:

Whether in the intentional states of non-human animals, human infants, or rational human cognizers, some essentially non-conceptual content that is altogether concept-free […]  really exists. (CCAP, p. 72)

Correspondingly, I will call the simple alternative argument that is based on The Autonomy of Essentially Non-Conceptual Content, The Argument from Babes and Beasts, a.k.a The ABB.

It goes like this.


1.  In reality, all normal infant human animals (‘babes’) and most non-human animals (‘beasts’), e.g. dogs and cats, lack a conceptual cogitive capacity and a capacity for self-consciousness (a.k.a understanding or Verstand, and apperception), although they do indeed possess an essentially non-conceptual cognitive capacity and a capacity for first-order consciousness (a.k.a sensibility or Sinnlichkeit).

In other words, babes and beasts are sentient, sensible animals, but not sapient, discursive/reflective animals.

2.  In reality, most adult human animals, e.g. all actual and possible human readers of these words, possess not only an essentially non-conceptual cognitive capacity and a capacity for first-order consciousness, or sensibility, but also a conceptual cognitive capacity, or understanding, and a capacity for self-consciousness, or apperception.

In other words, we are not only sentient, sensible animals, but also sapient, discursive/reflective animals.

3. In reality, since normal human infants, non-human animals, and adult humans, alike, are animals, they share the very same essentially non-conceptual cognitive capacity and the very same capacity for first-order consciousness: in particular, we ourselves were once human infants, so obviously we share the same essentially non-conceptual capacity and the same capacity for consciousness as our earlier selves.

In other words, babes, beasts, and most adult human animals, alike, are all sentient, sensible animals in that they share the very same essentially non-conceptual capacity and capacity for first-order consciousness.

4. In reality, all normal human infants, and many non-human animals, e.g. cats and dogs, can and do tell the difference between the left hand side of their bodies (e.g. their left hands or left forepaws) and the right-hand side of their bodies (e.g. their right hands or right forepaws).

5. So, in reality, all normal human infants and many non-human animals, e.g. cats and dogs, can and do represent complex objects as complex without concepts.

6. So, in reality, since we share with all normal human infants (e.g. our earlier selves) and most non-human animals the very same essentially non-conceptual capacity and the very same capacity for first-order consciousness—the very same sensibility—then we can and do represent complex objects as complex without concepts too.

7. Kant knew all this, as is clearly and distinctly shown by his remarks about infant human minds and non-human animal minds (see e.g. McLear 2011).

8. Therefore, the correct answers to RCQ1 and RCQ2, respectively, are yes and yes.

One extremely important further implication of ‘The ABB’ is that it answers Landy’s very good question about the cognitive subject of essentially non-conceptual content:

So who or what is it that is the single subject of Hanna’s non-conceptual representations? The natural Kantian answer to this question—the transcendental unity of apperception—is ex hypothesi unavailable to him. As having a non-conceptual representation is, for Hanna, being in what he calls The Grip of the Given—our bodies’ tracking of objects in the causal structure of the world—and, as Hanna repeatedly emphasises, the necessity of our being embodied, my guess here is that he wants to say that the single (non self-conscious) subject of a non-conceptual representation is the human body itself.

Wrong guess! In fact, what I want to say is that “the single (non self-conscious) subject of a non-conceptual representation” is the minded (non-human or human) animal (see CCAP, Section 1.3, et passim; see also Hanna and Maiese [2009], chs 1–2, esp. pp. 19–20; and Hanna [2011]). In Kant’s lingo, the minded animal is also the non-human or human subject of the power of choice or Willkür (see e.g. MS, 6:213–14). The minded human animal is also the same as the real human person, or in Kant’s lingo, the subject of “human choice” (MS, 6:213); and correspondingly, I’ve written another Kant-inspired book about the metaphysics of free will and real human persons called Deep Freedom and Real Persons (Hanna MS), that is all about minded human animals, their free agency and their identity.

Before concluding this response to Landy’s critical comments, I want especially to emphasise, and even re-emphasise, something else. The something else is that CCAP is itself a big, complex book about many fundamental topics in the philosophy of mind and knowledge, and not merely about whether there are essentially non-conceptual representations of complex objects as complex, or not. My discussion of that issue takes up exactly one footnote and one-third of one page in a book that is 441 pages long.

In fact, then, I am now going to tell you briefly what CCAP is actually all about. In CCAP, I work out a unified contemporary Kantian theory of rational human cognition and knowledge. Along the way, I provide detailed accounts of:

(i) intentionality and its contents, including non-conceptual content and conceptual content,

(ii) sense perception and perceptual knowledge, including perceptual self-knowledge,

(iii) the analytic-synthetic distinction,

(iv) the nature of logic,


(v) a priori truth and knowledge in mathematics, logic and philosophy.[1]

Moreover, CCAP is specifically intended to reach out to two very different audiences: contemporary analytic philosophers of mind and knowledge, on the one hand, and contemporary Kantian philosophers or Kant-scholars, on the other. At the same time, CCAP is also riding the crest of a wave of revolutionary new trends and new work in the philosophy of mind and epistemology, with a special concentration on the philosophy of perception. What is revolutionary in this new wave are its strong emphases on action, on cognitive phenomenology, on disjunctivist direct realism, on embodiment, and on sense perception as a primitive proto-rational capacity for cognising the world. Therefore, CCAP aims to make a fundamental contribution to this philosophical revolution by giving it a specifically contemporary Kantian twist, and by pushing these new lines of investigation radically further.

In Chapter 1, “Introduction: Cognition, Content, and Knowledge Revisited”I provide an introductory account of a Kant-inflected approach to the philosophy of mind and knowledge, categorical epistemology. I also develop several new arguments against narrowly naturalistic approaches to intentionality and intentional content, and contrastively promote the idea of a liberal naturalist approach, based on the thesis that the conscious rational human mind is essentially embodied.

In Chapter 2, “The Grip of the Given: A Kantian Theory of Non-Conceptual Content”, I work out out a general theory of non-conceptual content and conceptual content. Highlights of the account include, first, a defence of non-conceptualism by means of an extended, step-by-step argument for the existence, concept-autonomy, and concept-independence of essentially non-conceptual content; second, a corresponding refutation of conceptualism; and third, an application of the doctrine that necessarily whatever is mental is saliently, even if only pre-reflectively, conscious—The Deep Consciousness Thesis—to essentially non-conceptual content.

In Chapter 3, “Radically Naïve Realism”, I use the theory of non-conceptual content developed in Chapter 2 as the basis of a new theory of direct or naïve realism about sense perception, including a new solution to Molyneux’s problem, and a theory of perceptual self-knowledge. A central feature of the account is its radical version of metaphysical disjunctivism, the doctrine that all sense perception is veridical and categorically different from hallucination and illusion.


Reading Chapters 2 and 3 in tandem reveals that I have a direct reply available to Landy’s passing, but philosophically important, objection about the normativity of essentially non-conceptual representations:

I do not hold that my having such mental states represents those features as features of the martini. Here is one reason for that: if such representations represent the martini as a martini, they do so in a way that does not allow, even in principle, for misrepresentation. Non-conceptual representations represent whatsoever features of the objects that de facto cause them to have the content that they do. Thus, they will necessarily (by definition) represent those objects as having the features that they in fact have. Of course, many philosophers have thought that the notion of a representation that cannot possibly misrepresent is a confused one, and I think this should give us pause about the precise role of non-conceptual representations in Hanna’s system.

My reply, worked out in those chapters, is that the very idea of a ‘veridical representation’ contains two distinct conceptions of veridicality, first, a directly referential (a.k.a knowledge-by-acquaintance) conception, and second, an accuracy or correctness (a.k.a knowledge-by-description) conception, which I shall call, respectively, veridicality1 and veridicality2:

(i) veridical1: a representation R of an object O is veridical1 if and only if R cannot fail to refer directly to O, and O actually exists,


(ii) veridical2: a representation R of an object O is veridical2 if and only if R accurately/correctly picks out the basic features of O.

This in turn gives rise to two distinct conceptions of misrepresentation:

(i) misrepresentation1: a representation R of an object O misrepresents1 O if and only if O fails to refer directly to O, and O fails to actually exist,


(ii) misrepresentation2: a representation R of an object O misrepresents2 O if and only if R fails to accurately/correctly pick out all or some of the basic features of O, even if R refers directly to O, and O actually exists, that is, even if R is still veridical1.

For essentially non-conceptual representations, veridicality1 and misrepresentation1 are disjunctive or on/off—either representations do or do not refer directly to actually existing objects—whereas veridicality2 and misrepresentation2 are matters of degree: here, representations are more or less accurate/correct or inaccurate/incorrect.

(It is a different story for conceptual representations and judgements/propositions—all of veridicality1/misrepresentation1 and also veridicality2/misrepresentation2 are disjunctive or on/off, and none of them are matters of degree, i.e. there are no ‘degrees of truth’ in a semantic sense, whenever the background logic is classical or at least non-deviant—but that’s a different story for a different day.)

My view in CCAP, then, is that necessarily, all essentially non-conceptual representations are veridical1, but not necessarily all of them are veridical2: therefore, essentially non-conceptual representations never misrepresent1 their objects, but nevetheless they can and often actually do misrepresent2 their objects.

The paradigmatic sort of veridicality1 + misrepresentation2 that I have in mind, is when we essentially non-conceptually veridically1 perceive, remember, projectively imagine, etc., actually existing objects, yet also fail to track their location, velocity and/or causal-dynamic profile accurately/correctly.

One good example of this from sports, which actually happened to me many times, to my great chagrin, is my essentially non-conceptual veridical1 representation of a soccer ball flying towards my head, which I misrepresent2 by failing to track its spatiotemporal location, velocity and/or causal-dynamic profile accurately/correctly, so that I then stupidly head the ball out of bounds instead of deftly into the corner of the net like my all-time soccer-hero Pele. Bummer.

Another good, although less morally upright and soberly Kantian example is reaching out for a lovely, thirst-quenching martini sitting on the table and clumsily knocking over the glass instead of skillfully snatching it up. Double bummer.


Now back to our regularly-scheduled programming.

In Chapter 4, “Truth in Virtue of Intentionality, Or, The Return of the Analytic-Synthetic Distinction”, I deploy the earlier accounts of conceptual content and non-conceptual content in order to provide a full explanation and vindication of the analytic-synthetic distinction, including a theory of synthetic a priori truth. This vindication includes an extended critique of Quine’s critique of the analytic-synthetic distinction, and also an explicit argument against the Kripke-Putnam conceptions of the necessary a posteriori and the contingent a priori.

In Chapter 5, “The Morality of Logic”, I work out an argument for the categorical normativity of logic and its intrinsic role in rational mental representation or intentionality per se, including all modes of rational human cognition and intentional action. A central pay-off of this argument is that it can be used to solve the problems of the explanatory and justificatory status of logic (the logocentric predicament), the epistemic status of logic, and Quine’s Predicament (that every truth is revisable, but ‘sheer logic’ is unrevisable).

In Chapter 6, “Rationalism Regained 1: The Benacerraf Dilemmas”, I carefully re-present and re-deploy the famous Benacerraf Dilemma, by extending from its original version in the context of mathematical truth and knowledge, to logical truth and knowledge, and then generalising it to all kinds of a priori knowledge. I also show that Kant worked out his own sophisticated version of the Dilemma in 1772, two centuries before Benacerraf, and presciently postulated transcendental idealism as the right solution.

In Chapter 7, “Rationalism Regained 2: A Priori Knowledge and the Nature of Intuitions”, in the context of working out solutions to the three versions of The Benacerraf Dilemma, I discuss the nature and epistemic status of intuitions, with special reference to rational intuitions, and also offer a corresponding critique of the contemporary critique of intuitions by Experimental Philosophy.

And finally, in Chapter 8, “Rationalism Regained 3: Kantian Structuralism and Kantian Intuitionism”, I present solutions to all three versions of The Benacerraf Dilemma, and develop a new structuralist account of the nature of mathematics and logic (Kantian Structuralism) alongside a new rational-intuition-based account of the nature of mathematical and logical a priori knowledge (Kantian Intuitionism). These solutions and accounts, in turn, presuppose and indirectly vindicate a Kant-inspired rational-intuition-based approach to meta-philosophy.

So now you can see the much larger philosophical context in which KNC is embedded. And in fact, The Two Hands Argument in Chapter 2, Section 2.5, important as it is for my account, is not even the only argument I offer for the characteristic thesis of KNC, that is:

Essentially non-conceptual content really can, actually does, and necessarily does, exist.

First, in Sections 2.1 and 2.2, I analyse and criticise all the basic theories and doctrines in the debate about non-conceptual content and conclude that, on dialectical grounds, KNC is uniquely placed to resolve the underlying problems in the debate.

Second, in Section 2.7, I also argue that KNC is vindicated by inference-to-the-best-explanation, because KNC, and it alone, solves a generalised version of the well-known Causal Pairing Problem against Cartesian dualist interactionism, first spelled out by Jaegwon Kim (see Kim [2005], ch. 3, esp. pp. 78–80).

Nothing Landy criticises in his comments, affects those arguments at all.

OK. I now conclude triumphantly (yet also in a very nice, polite way) that, as regards this particular issue—in reality or according to Kant, can we represent complex objects as complex without concepts?—I am right and Landy is wrong.

Nevertheless I also urge you, the reader—or someone, anyone—to read all of CCAP, and to engage philosophically with all or even most of it, charitably, critically, and constructively, before it slides Titanic-like into the inky void of institutional amnesia that is the watery graveyard of all big, ‘ambitious’ (classic reviewer-speak for damning with faint praise), complicated, original, unorthodox philosophical books that (as Hume said of the Treatise) “f[a]ll dead-born from the press”, precisely because their authors quite intentionally, but also very imprudently and ‘unprofessionally’, fail to conform to the all-too-familiar rules of the game of contemporary mainstream professional academic philosophy.

Received: 6 January 2017.


[1] Hanna’s accounts of (iii) and (iv) will be discussed in a forthcoming commentary on CCAP, on this site, by Dennis Schulting.


Hanna, R. (2006), Kant, Science, and Human Nature (Oxford: Oxford University Press).

——— (2011), ‘Minding the Body’, Philosophical Topics 39: 15–40.

——— (MS), Deep Freedom and Real Persons: A Study in Metaphysics (Fall/Winter 2016/27 version).

Hanna R. and M. Maiese (2009), Embodied Minds in Action (Oxford: Oxford University Press).

Kim, J. (2005), Physicalism, Or Something Near Enough (Princeton, NJ: Princeton University Press).

McLear, C. (2011), ‘Kant on Animal Consciousness’, Philosophers’ Imprint 11: 1–16.

© Robert Hanna, 2017.

Robert Hanna is the Director of the CSKP & CKP projects, which form part of the Critique & Contemporary Kantian Philosophy online platform. He received his Ph.D. from Yale University in 1989, and has held research or teaching positions at the University of Cambridge, the University of Colorado at Boulder, USA, the University of Luxembourg, PUC-PR Brazil, Yale, and York University, Canada. His work has a broadly Kantian orientation, and he also has strong interests in the history of modern philosophy from Bacon/Hobbes/Descartes to contemporary philosophy, in the philosophy of nature and natural science, and in critical meta-philosophy. He has authored or co-authored six books and is currently working on a four-book series on the nature of human rationality, entitled The Rational Human Condition.