Grounding empirical in transcendental reality

Anja Jauernig's book provides an exceptionally clear, rigorous, and illuminating account of Kant's critical idealism as an ontological doctrine that posits two layers of being or reality. Jauernig seeks to reconstruct “Kant's account of what there is in the world, understood as the sum total of everything that has reality, including, in particular, his account of appearances and things in themselves and their relation to one another” (xii). Central to Kant's view is the claim that our mind cannot actively produce the intuitive data that it needs for representing sensible appearances: it must receive such data from without via affection by things in themselves. In Jauernig's analysis, this appeal to things in themselves derives from a “basic”, “rock-bottom” commitment (307): Kant holds that our faculty of intuition is passive and requires affection by things in themselves because he regards the human mind as finite. This finitude, our “inability to be ontologically creative” stands “in contrast to the ontologically creative nature of a divine infinite mind, which is purely active and does not stand in need of being given any material to bring forth representations or genuine existents” (314). Jauernig stresses that “From Kant's point of view, it is obviously true that the human mind is finite” and “this finitude entails that the human mind is ontologically uncreative” (316). Since our finite theoretical intellect lacks the power to produce sensible objects, we cannot represent such objects unless we receive sensations from existing things in themselves. The sensible world of human experience is mind-dependent. Objects like cats, including their spatiotemporal and causal properties, ontologically depend upon our shared cognitive capacities: “empirical objects have all of their spatial and temporal determinations in virtue of being represented by finite minds to have them” (155). But although spatiotemporal objects are “fully mind-dependent”, they have genuine empirical reality: “a mind-dependent fact” or “a mind-dependent existence is still a kind of genuine fact” or “a kind of genuine existence” (156). Jauernig posits two conditions under which the objects of our representations genuinely exist: “empirical objects exist because” (this is the first criterion) “they not only conform to Kant's formal conditions of proper objecthood but” (this is the second criterion) “also are grounded in things in themselves in a properly direct way. This is in contrast to other intentional objects such the dragon from my dream, which do not satisfy all of Kant's formal conditions of proper objecthood and are not grounded in things in themselves in the right way, and thus” have a mere pseudo-existence (156). The first criterion for empirical reality requires that our representations satisfy the formal conditions of proper empirical objecthood. Our minds must actively combine sensible data according to valid rules so that we may cognize public objects of an intersubjective, coherent, and unified experience. This act of synthesis must conform both to the spatiotemporal forms of human sensibility and to intellectual-conceptual forms such as ‘substance’ or ‘causality,’ including the corresponding laws of understanding, such as the general causal principle. These forms of representation impose a general lawfulness, unity, and coherence upon empirical reality. It is only if our representations satisfy the formal constraints deriving from our mental faculties that they allow us grasp real spatiotemporal relations and lawful interactions among empirical objects. But for Jauernig, this is not yet sufficient for empirical reality: “The empirical reality of empirical objects consists also in the grounding of empirical objects in transcendentally real things.” Mere formal “coherence and lawfulness is not enough for experience; experience also depends on ‘friction’ supplied by entities that exist at the transcendental level of reality” (247). On Jauernig's reading, this second condition allows Kant to avoid problems that Berkeley's idealism faces in distinguishing empirically real objects of human experience from empirically ideal objects of dreams or illusions. “In contrast to other fully mind-dependent objects such as the intentional objects of fictions, dreams, illusions, and hallucinations, empirical objects are properly directly grounded in things in themselves, which gives them enough reality to count as existents” (116). For Jauernig, the “intentional objects of experience are the only intentional objects that both conform to Kant's formal conditions of proper objecthood…and are grounded in things in themselves in a properly direct way so that they can be said to exist” (122). Whereas, “empirical objects on Berkeley's account are not grounded in the right kind of transcendentally real things.” Thus, “empirical objects” in Berkeley “end up on a par with pseudo-existing entities such as” a hallucinated oasis (126-127). The cited passages suggest that Jauernig's two conditions are individually necessary but only jointly sufficient for empirical reality. Our representations might exhibit the coherence and lawfulness that is required for objective experience without yielding objective experience of empirical realities, if these representations are not properly grounded in things in themselves. Moreover, our representations might be properly grounded in things in themselves without yielding objective experience of empirical realities – for instance, when cognizing subjects do not properly synthesize given data. Jauernig provides an intriguing, sophisticated account of empirical reality that is worthy of careful examination. I shall consider some epistemological implication of Jauernig's account. Specifically, I worry that it renders empirical reality epistemically inaccessible to cognizing subjects. Jauernig acknowledges that we cannot rely upon the formal conditions of experience if we seek to ascertain whether our representations relate to empirically real things. One might explain this epistemic shortcoming of the formal conditions in two different ways. First, one might hold that only genuine experience, as opposed to dreams or illusions, truly satisfies the formal conditions. Jauernig suggests this view when she says: “Kant is right to claim that experience is distinguished from illusions, hallucinations, and dreams through its conformity to his formal conditions of experience” (189). Here the reason why we cannot employ the formal conditions as an epistemic criterion for determining empirical reality is that we cannot ascertain whether our representations conform to these conditions and thus yield genuine experiences. Alternatively, one might hold that even if we can ascertain that a representation satisfies the formal conditions of experience, this is neither here nor there as far as genuine experience is concerned because dreams or illusions also satisfy these conditions. Jauernig suggests this explanation when she says: “Experience and pseudo-experience are both constructed in such a way as to conform to Kant's formal conditions of experience” (189). Note that only this second explanation ensures that the formal conditions are individually insufficient for genuine experience. The first explanation would entail that whenever representations do satisfy the formal conditions, this suffices for the genuine experience of empirically real objects. But the two explanations converge on the same epistemic problem. Whether (via the second explanation) illusory representations exhibit the same (epistemically accessible) formal unity and lawfulness that also characterizes genuine experience, or whether (via the first explanation) illusory representations fail to satisfy this formal constraint for reasons that are not epistemically transparent to us: we cannot use the formal conditions to determine that the intentional objects of our representations have empirical reality. Jauernig's material condition of empirical reality fares no better here: it is, as Jauernig admits, not “usable as” an “epistemic criteri(on)” (189). Since we lack cognitive access to things in themselves, we cannot discern whether our representations are properly grounded in things in themselves. Thus, we cannot use Jauernig's two ontological conditions for empirical reality as epistemic criteria for determining whether we have genuine experience of empirical reality. This seems to imply, among other things, that we can never know whether we are eating real bread or playing with genuine cats. If Kant's view had this implication, it would invite Berkeley's complaint (directed against Descartes and Locke) that we “are plunged into the…most deplorable skepticism.”1 Jauernig suggests that Kant's Refutation of Idealism enables him to avoid such skepticism. In this argument, Kant begins from the premise that we can cognize an objective temporal order among our inner mental states. He claims that such cognition requires that we also cognize empirically real outer objects: spatial things that persist through time. The argument seeks to show that our beliefs in the existence of outer spatial objects are as certain as our beliefs in the existence of our empirical self and its (objectively ordered) inner mental states. As Jauernig notes, one striking question here is why “we can represent something persistent in time only on the basis of genuine perceptions, in outer experience but not on the basis of mere hallucinations or dreams” (190). She suggests that Kant relies on an inference to the best explanation: “the desired representation of something persistent requires that the perceptions on which it is based exhibit strong, persistent regularities as far as their content is concerned, regularities that would be nothing short of miraculous if these perceptions in truth were mere dreams, hallucinations, illusions. By contrast, if they are genuine perceptions” that “are due to affections of sensibility by things in themselves, the persistent, strong regularities in their contents can be seen as reflecting certain (unknown) regularities in the realm of things in themselves” (191). This reconstruction of the Refutation argument raises various problems. For one, the inference Jauernig envisages does not seem to yield the intended conclusion. The conclusion that we (or Kant) want to reach is this: we can be certain that we cognize persisting outer empirical objects which stand in regular causal connections and which ground a corresponding regular order among our perceptions. Jauernig's inference, if successful, instead leads to the conclusion that there are regularities among things in themselves. Since that conclusion pertains to transcendentally real things which are outside us in a non-spatial sense, it seems consistent with a skeptical stance regarding empirically real spatial appearances. It is logically possible that regularities in the domain of things in themselves might affect us with representations that yield regular, persisting dreams or illusions. This also looks like an epistemic possibility from our finite cognitive vantage point. Thus, even if Jauernig's argument allowed us to infer the existence of transcendental regularities, this would not provide sufficient certainty that lawfully connected outer objects exist and persist in empirical reality. This worry is aggravated by the fact that there is not just an epistemic but an ontological gap between whatever ‘regularities’ might obtain among things in themselves and the regularities we believe to exist among phenomenal objects. Phenomenal regularity essentially involves the persistence of empirical objects and the laws governing their interactions. But since persistence or permanence is a temporal, hence mind-dependent feature, it cannot characterize mind-independent things in themselves. It is unclear why a non-temporal form of ‘regularity’ – whatever that might be – should translate into the entirely different, temporally persistent regularity that we take to characterize the natural world of our experience. Finally, Jauernig's argument invites an objection that Kant raises against realists who identify spatial objects with things in themselves: they are committed to deplorable skepticism about spatial objects because on their view, the existence of such objects could not “be cognized in any way other than by an inference from effect to cause, in which case it must always remain doubtful whether the cause is in us or outside us” (A372). Jauernig's argument also seems to rest on an inference from an effect, namely our inner mental states, to a transcendentally real outside cause. She might respond that this inference is innocuous: since for Kant our faculty of sensible intuition is inherently passive and uncreative, he must endorse an inference from actually given intuitions to their transcendentally real source. But this inference faces two crucial limitations (even waiving traditional worries about the very concept of transcendent causation). First, it does not allow us to infer anything specific about the transcendentally real sources of our intuitions: for instance, it does not show that the seeming physical regularities which our outer intuitions purport to display are grounded in transcendental regularities. We cannot rule out the epistemic possibility that our perceptions of seeming physical regularities derive from transcendental irregularities, from a noumenal chaos or from an erratic fickle genius. Second, even if the presence of representations in a finite mind indicates that their material-sensational components must derive from something outside the finite mind, it seems epistemically possible that on any particular occasion the composition of our representations is due to our own subjective constructions – as when we fabricate seemingly regular, coherent dream scenarios. Here our finite mind rather than some real outer object produces a seemingly regular sequence of representations. Thus, the inference from a given effect, namely some representational content, to a specific cause, namely objects which exist, persist, and lawfully interact outside our mind, is never reliable: “it must always remain doubtful whether the cause is in us or outside us.” In sum, Jauernig's appeal to the Refutation of Idealism does not seem to mitigate the skeptical worries which arise from her view that we lack sufficient empirical access to what she posits as the ontological conditions for empirical reality. When Jauernig appeals to transcendental affection as a necessary condition for empirical reality, she invokes a specific grounding relation. Our representations relate to empirically real objects only if they are grounded in things in themselves in a properly direct fashion: “empirical objects are properly directly grounded in things in themselves, which gives them enough reality to count as existents” (116). For Jauernig, this also distinguishes genuine sensations from mere pseudo-sensations: “A pseudo-sensation is a representation that is exactly like a sensation except that it is not the direct result of an affection of sensibility by things in themselves. Sensations are produced by sensibility in response to transcendental affections; pseudo-sensations are produced by the imagination through the reproduction and recombination of previous sensations” (313). Jauernig's approach here is vaguely reminiscent of contemporary views like the causal theory of perception or epistemic reliabilism. On such views, our representations have good epistemic credentials if they stand in the right causal connections to the real world. Whether the appeal to causal connections can carry the normative weight that is required by epistemic concepts such as justification, knowledge, or objective experience is a matter of debate. Leaving this issue aside, I am not sure that one can easily adopt something (vaguely) analogous to the reliabilist view in the context of Kantian idealism. For contemporary reliabilists, our perceptions qualify as objective since they are caused in a stable, non-deviant way by precisely those mind-independent outer objects which our perceptions purport to grasp. By contrast, in Kant's view perceptions which yield objective experience do not grasp the mind-independent things that ultimately ground them. Even when we have genuine experience, we do not perceive the transcendentally real things in themselves which do the ontological grounding work. Rather, our outer experience represents empirically real outer appearances which, for Jauernig, are ontologically distinct from things in themselves. But if there is no representational connection between a mental content R and its ontological basis, then I am not sure what it means to say that this ontological basis grounds R in a properly direct as opposed to an improperly indirect fashion. If R entirely fails to represent its ontological ground regardless of whether R counts as veridical or illusory, then the way in which R is grounded seems no better (no more proper or direct) in the veridical case. Let me restate this worry. The notion of a ‘properly direct’ grounding relation seems intended to have a normative, justificatory upshot. This upshot hinges on the special epistemic features that a representation supposedly acquires from the way in which it is grounded. In contemporary realist views, this means (roughly speaking): a mental content R qualifies as objective because there is a stable, reliable connection between the outer sources which produce R in the mind and what R presents as being the case in the outer world. By contrast, in Kant's idealism, our sensations never present us with their transcendentally real outer sources. Hence, there is no epistemic, representational connection between the mental contents that make up our objective experience and the things in themselves which give rise to these mental contents. In light of this, I am not sure what to make of the normative idea that in some but not all cases where we combine representations, these representations are supplied by things in themselves in an epistemically appropriate (properly direct) manner. To illustrate this concern, consider a representation that fails to present an empirically real outer object, e.g., my hallucination of an oasis in the desert. For Jauernig, this representation is not properly directly grounded in things in themselves; rather, it is composed by my imagination. But our finite, uncreative imagination must ultimately draw the simpler representational units required for its fictitious compositions from sensible data supplied by things in themselves. Thus, my hallucination is the result of how my imagination interprets transcendentally given data, just like my veridical representation of my cat Achilles is the result of my imaginative processing. Of course, the imagination processes material content very differently in these two cases. But this difference has nothing to do with transcendental grounding relations; rather, it concerns the mental activities that interpret given mental content either in accordance with or in violation of the formal condition of experience. Thus, the formal criterion for genuine experience carries all the weight here. And, to repeat, it is difficult to see how appealing to a special transcendental grounding relation could add any further weight: even if our representations qualify as genuine experiences that disclose empirical realities, they do not epistemically connect with or disclose the transcendental grounds of their material content any more than pseudo-experiences do. Jauernig's appeal to the envisaged grounding relation might not be intended to have any epistemological upshot: perhaps it is meant to provide only the ontological ‘oomph’ required for empirical reality. But if all our sensible data ultimately derive from transcendental grounding, it is unclear why this ‘oomph’ should be limited to some specific set of sensible data, especially if the non-sensible source of the ‘oomph’ is completely detached from the content of any sensible data. Of course (as I mentioned earlier), the appeal to some special transcendental grounding relation cannot solve the epistemic worries I discussed in the previous section. Since we cannot cognize things in themselves, we (a fortiori) cannot ascertain whether our representations present empirical realities due to being properly grounded in transcendental reality. One might suggest that even if we cannot reflectively grasp the special transcendental ‘oomph,’ the presence of this ‘oomph’ nevertheless does objectify some of our representations so that they can contribute to human knowledge. This suggestion has an externalist flair that strikes me as foreign to Kant's epistemology: a necessary condition for Kantian knowledge qua “certainty for everyone” is that the grounds for one's assent be universally communicable (A821/B849; A829/B857; JL, 9:70) and (thus) reflectively accessible. The attempt to secure our empirical knowledge of real objects by invoking a transcendent, cognitively elusive proper grounding relation seems like a leap of faith, which Kant views as a “scandal” for philosophy’’ (Bxxxix). Jauernig is right to stress that receiving sensations from things in themselves is a necessary condition for our experience of empirical realities. These sensations yield the material input that our finite mind requires for its active of experience. But I am not that the appeal to transcendental affection carries any weight either as an ontological constraint upon empirical reality or as an epistemological constraint upon objective experience, and the fact that such affection our finite mind with mental content that might both in our objective experience of empirical realities and in the pseudo-experience of dreams or illusions. I that we are with the formal conditions of experience as the of genuine experience and empirical reality. For “in order to according to the is connected with a perception according to empirical is we can skeptical “in by the connections of appearances in both and time according to laws of experience, and if the representation of outer things we cannot that they should be These suggest that the accordance of a representation with the formal conditions of experience is individually sufficient for experience and empirical reality. The question whether Kant is to claim that the of our representations with these formal conditions is an epistemically reflectively transparent criterion for empirical reality.