CHRISTOPHER INSOLE | Kant and the Creation of Freedom. A Theological Problem | Oxford University Press, 2013 


By Christopher Insole

I would like to thank David Sussman for his patient, insightful, and full account of the sweep of my book’s argument, and the important questions that he raises at key points. I am deeply grateful for the time and attention taken, and the care shown. Sussman pushes me helpfully in some areas where further reflection and discussion is needed.

I shall focus on four points where Sussman challenges me, or asks for further clarification: (1) the “theological argument” for transcendental idealism; (2) noumena regarded as “ontologically distinct” from phenomena; (3) the defence of concurrence; and (4) the “incredible” nature of Kant’s world-view (on my interpretation). The first three areas involve more of a clarification of my position, than a refutation. The fourth area needs more extensive reflection.[1]

1. The ‘Theological Argument’ for Transcendental Idealism

Sussman comments on the ‘oddness’ of Kant’s ‘theological case’ for transcendental idealism. This consists of Kant’s claim that if God were to be directly the creator of space and time, God would be “spatialised and temporalised”. This is reflected in Kant’s expressed worry that if space is a condition “of all existence in general”, then it would also have to be a condition “of the existence of God” (B72). Sussman comments that this is a weak argument, reflecting that

God would seem to escape space and time whether these are direct or indirect creations by the simple fact that he is prior to his own creative acts and therefore to whatever follows from them.

I entirely agree with Sussman’s criticism of Kant’s argument. Indeed, I say something along these lines in the book itself, writing that

even if space is conceived of as a condition of created existence, it is not clear why such spatiality would be contagious for the divine nature. The existence of space would still be dependent directly upon the divine will to create, with space having no existence independent of God’s decision to create. (p. 176)

I do go on to offer some brief reflections, which I expand upon elsewhere,[2] as to why Kant might have thought that this was an important problem. But this should not be mistaken for an endorsement of Kant’s argument.

2. Noumena Regarded as ‘Ontologically Distinct’ from Phenomena[3]

Sussman paraphrases my interpretation of Kant along the following lines, saying that “[i]f Insole is right,”

Kant, for good or ill, did indeed take noumena to be ontologically distinct from phenomena, and believed both that we could know that such noumena exist and that they really do affect the phenomenal world. (boldface added)

It all depends, of course, on what precisely one means by noumena being “ontologically distinct”. On some plausible interpretations, this might be a correct characterisation of my position: if one means, for example, that the properties and nature of noumena are not exhausted by the properties that appear (the ‘phenomenal properties’). But, typically, something far less plausible is meant, when the notion of the ‘ontological distinctness’ of noumena arises. Usually, such a description of the ‘noumenal affection’ account is intended to align it with so-called ‘two world’ interpretations of Kant, where one world is non-spatial and non-temporal (the ‘noumenal realm’), and the other world (the ‘phenomenal realm’) is spatial and temporal. Noumenal affection accounts, it is claimed, are committed to the view that ‘somehow’ non-spatial and non-temporal things in themselves interact with the world of spatial and temporal things. All manner of problems cluster around the nature and coherence of the causal relationship supposedly envisaged by Kant, on this ‘two-world’ interpretation, between spatio-temporal and non-spatio-temporal realities.

Such a picture is not at all what Kant has in mind, on my interpretation, and it is not what a plausible noumenal affection account should claim. Part of the problem here is a now standard taxonomy for classifying interpretations of transcendental idealism (‘one or two worlds’), which can be found in learned journals, monographs, and textbooks. This taxonomy is in fact deeply unhelpful, not least because it solidifies the assumption that ‘one world’ interpretations must be ‘epistemological rather than metaphysical’, with no trace of traditional metaphysics, and that more metaphysically committed interpretations are freighted with all the difficulties of the ‘two world’ interpretation, as set out above. Kant’s position meets these two rival models rather like a three-dimensional object passing through two-dimensional planes. Rather than starting with the problematic distinction, and fitting Kant into this, it is better just to state Kant’s position, gesturing to aspects of this position that both camps have partially understood.

Theologically sensitive noumenal affection accounts understand that we are dealing with one created world (so far, so good, for one-realm accounts), which therefore does not include the uncreated God, who is not part of any world. This is not typically addressed at all by one-realm accounts, which tend to occlude any sort of traditional metaphysics. God, who is not part of the ‘world’ (or any other world), creates the world, which is the realm of things in themselves. Amongst these things in themselves are ourselves. Our reception of things in themselves, both of our own noumenal selves, and of other things, is always and everywhere mediated through the forms of intuition, space and time, and the categories of thought (such as substance and causation), insofar as these are received through space and time. There are degrees of appearance and disclosure to rational creatures who are part of the created (noumenal) world, where “epistemic humility” consists in understanding the principled boundaries of what we can and cannot know within this framework, and what we must rationally believe, given these boundaries, alongside our non-negotiable commitment to irreducibly important projects such as morality.

The relevant distinction (as one-realm interpretations typically emphasise) is indeed between things as they are in themselves, and things as they appear: and the distinction applies from top-to-bottom and everywhere. The whole event of our being affected by a spatio-temporal object, is the appearance of a more fundamental noumenal interaction (although we know not what). There is no mysterious interaction between spatio-temporal and non-spatio-temporal objects, although there is a spatio-temporal appearing of fundamentally non-spatio-temporal objects. Standard one-realm interpretations tend to occlude (whereas two-realm interpretations emphasise) Kant’s undoubted fundamental commitment to things in themselves being non-spatial and non-temporal. This does not lead to extravagant metaphysical speculation, as we could never in principle know how spatial and temporal objects ‘map’ onto things in themselves. Does an empirical tree in some sense ‘correspond’, in a one-to-one way, with a thing in itself? We know that our experience of a tree is somehow answerable to and determined by a supersensible reality, but the mechanism of the grounding is inaccessible to us. We simply cannot know what it is (singular or plural) that grounds the appearance that is a tree. That, we might say, ‘is the whole point’, as the claim that there must be a one-to-one correspondence between the objects of experience and things in themselves risks violating the very epistemic discipline that one-realm interpretations value so highly. We can, though, on the basis of the rational and disciplined needs of practical reason come to warranted beliefs about the realm of things in themselves, as long as these beliefs do not contradict what theoretical reason is able to know.

Given all this, the less misleading thing to say about the ‘noumenal realm’ is, I think, this: that the noumenal realm is ontologically identical with the world that we experience (the ‘phenomenal realm’), but that, epistemologically speaking, we only enjoy partial degrees of disclosure about the fundamental nature of this world. The world discloses itself to us as spatial and temporal, but, for a variety of theoretical and practical reasons, we can be assured that this is a feature of our reception of the world, rather than being in the world independent of, and prior to, this reception.

3. The Defence of Concurrence

Sussman comments that I think “that some form of concurrence is the correct model of action”, and objects that “it is not entirely clear”, from my presentation, “why Kant’s view [which involves the denial of concurrence, C.I.] should be rejected”.

This is, I think, a bit too quick, and also, perhaps, a little unfair. First of all, I do not think the main impact of my treatment of concurrence is a ‘defence’ of it. It is true, as I also comment in my response to Ertl, that I reveal my allegiance to a concurrentist account in just one sentence of the book (on p. 214). But the overall intention of the treatment of concurrence is to bring to light, in as fair a way as possible, what is fundamentally at stake in rejecting or accepting concurrence accounts. So, in explaining the attraction of ‘mere conservationist’ accounts I concede that they are clearer, more obviously coherent, more elegant, and more able to give an articulate response to the problem of theodicy. On the other side, I spend quite a lot of time showing why theologians have wanted to affirm concurrence accounts. I do this both in a more formal technical key, but also in a more scriptural, as well as a more ‘intuitive’ way. Put briskly, all of these different approaches revolve around one simple commitment: a desire to say the ‘biggest thing possible’ about the total dependence of all of creation, all of which is created ex nihilo, upon God, and God alone, with a corresponding sense that the relationship between the creator ex nihilo and creation should and ought to be unique, sui generis, and, to a degree, mysterious.

On one related point, in connection with the question of grace, I am inclined to agree with Sussman, at least to the extent that I ought to refine and rephrase the way I express myself. Sussan objects to me along the following lines:

Insole argues that Kant just has very little interesting to say about grace and vicarious atonement, a high price to pay for anyone operating in a Christian context.

As evidence for this, Sussman quotes me where I say that “a recurring problem for interpreters of Religion is that the conceptual space permitted for divine action is restricted at best, and incoherent at worst” (p. 240). In response, Sussman writes that “the charge is puzzling, insofar as Kant does in fact seem to have much to say about grace and atonement in Religion within the Limits of Reason Alone“. The shape of the work done by grace, finds Sussman, is as follows. We are obliged “to strive toward a kind of virtue that we can never reach in any finite span of time”, such that we “attain true holiness only in such an infinite progress in virtue taken as totality”:

Grace, for Kant, is God’s crediting this totality to us at some point in time as if we had already attained it in full. The view is complicated and it may not ultimately be coherent, but it is not clear how it suffers from an excessively restricted conceptual space.

I think that Sussman is onto something here, albeit with a few caveats. First of all, the caveats:[4]

(i) The significance of what Kant says about ‘grace’ here, depends in part upon how one reads the significance of Religion within the Limits of Reason. By Kant’s own account, it is an attempt at a ‘translation’ from the categories of historical and revealed religion, into the ‘pure religion of reason’. So, he has to say something about ‘grace’, because this is a core part of historical revealed Christianity. This is different from needing to say something about ‘grace’, because he considers it to be a core feature of the demands of his philosophical system.

(ii) The translation that Kant gives involves conceiving of all our moral action from a perspective outside of time, as if from a single eternal moment. But how significant is this, really, given that Kant believes that moral effort is, in any case, something that occurs in the space of noumenal moral freedom, which is to say, outside of time? Regarding our moral actions as ‘outside of time’ is, in any case, just what we are required to do. It is not clear that very much is being added to Kant’s system here.

(iii) The translation that Kant performs conspicuously takes us away from what the Christian tradition, as Kant would have received it, unanimously affirms about grace, which is that grace precedes moral conversion, rather than following it.

Nonetheless, with these caveats in place, I think I would rephrase myself. Rather than saying that Kant ‘does not say very much’ about grace, I would now say that when Kant does talk about grace, what he says about it is different, in important ways, from the tradition that he receives.

4. The ‘Incredible’ Nature of Kant’s World-View (On My Interpretation)

An anxiety that arises, in relation to Kant’s notion of noumenal freedom (as I interpret it), is that human beings can seem to become somehow responsible for “everything that is determined by causal law”. Sussman does not find my reply, where I draw upon Allen Wood, to be persuasive. This reply consists in reflecting that “epistemic humility” about the noumenal realm “blocks us from any extensive speculation about how far” facts about our own moral character “are implicated in other features of the natural world”, such that perhaps we are “only […] responsible for the psychological facts of our own character”. Sussman does not find this convincing, commenting that because of the interconnectedness of empirical facts

there seems to be no way my choice of character can be treated as anything less than a choice of my entire phenomenal world, insofar as my character is a necessary element of that world as it is.

I am inclined not to defend myself here: in part, because I think Sussman is correct about the difficulty of stopping the “spread” of noumenal affection to one’s character alone; but also, in part, because the main focus of this response is supposed to be on the “epistemic humility”, rather than any particular suggestion about how the fundamental noumenal realm might be expressed in the phenomenal realm. It is vital, for Kant, that we do not say more about noumenal freedom, and about God, than (he thinks) we need to, in order to secure the needs of practical reason. All we need to say is that noumenal freedom is required by practical reason, and that, because it is not theoretically impossible, we are permitted and required to believe in it, such that believing in noumenal freedom secures the foundations of morality, and, by extension of religion (within the limits of reason alone). I think that, occasionally, Kant is tempted to say a bit more than this. One intriguing passage is to be found in Reflexion 5612. Here Kant reflects that “actions here in the world are mere Schemata of the intelligible [actions]”, where “these appearances […] are still interconnected in accordance with empirical laws”. In the empirical realm we

cognize [our] own character only from the phaenomenis, [and] impute it to [ourselves], although it is, to be sure, itself determined by external causes.

There now follows, from Kant, an intriguing suggestion:

If one knew it in itself, then all good and evil would not be ascribed to external causes but only to the subject alone, together with the good and the disadvantageous consequences. (Refl. 5612, AA 18:253–4)

If Kant means this, the idea seems to be that ‘all good and evil’, including perhaps earthquakes, would not ultimately be ascribable to external causes, but would arise, just as Sussman suggests (albeit that Sussman intends it as a sort of reductio) from our own noumenal freedom. Although this passage might offer some support for the ‘noumenal freedom’ account that I offer, it exacerbates another aspect of a criticism pushed on me by Sussman.

Sussman reflects that even if the exegesis of Kant is correct, it all seems, as he pithily puts it: “pretty incredible”. He goes on to comment that

such choice seems to be less and less something we really do in any familiar sense, and instead becomes a kind of inscrutable fact that lies behind all our choices, somehow both ‘always already’ and ‘never yet’ the particular way it is.

I feel the force of this point, and need to reflect upon it further. At this point, I have a few initial, and rather brisk, reflections.

(i) We must always distinguish between what is “incredible” to us, and what might have been incredible to Kant. One of the joys and challenges of engaging in the history of philosophy, is to expand our sense of what extraordinary things might have been credible to thinkers who, in other ways, we feel so close to. I would submit that the notion of noumenal freedom would not be so “incredible” in a context where many thinkers agreed with Leibniz, that the experienced world is a ‘well-founded phenomenon’ of a fundamentally non-spatial and non-temporal monadic universe.

(ii) The way in which, I think, the conception of noumenal freedom features in Kant’s system is as a guarantor and consolation: of the real possibility of transcendental freedom, and of the validity of religious hope. I do not think, though, that Kant intends our moral life to consist in futile attempts to somehow, impossibly, ‘peer into’ this realm, or to speculate about how it works. Rather, our moral task is to engage with the complex textures of the way in which the world, and ourselves in the world, do appear spatially and temporally. In a parallel way, in relation to natural science, Kant’s commitment to the fundamental non-spatiality and non-temporality of the world in itself, in no way reduces his curiosity about, or study of, the textures of the spatial and temporal appearance of this world. This relates to my response to Sussman’s second objection (noumena regarded as ‘ontologically distinct’ from phenomena): it is not that the noumenal realm is ‘somewhere else’, bizarrely interacting with the spatial and temporal world. Rather the noumenal realm is this world, prior to and independent of our spatial and temporal reception of it.

(iii) From various parts of Kant’s writings, we can identify the following experiences and thoughts, which we experience (in space and time): rational thought; moral conflict, where we overcome a temptation to do other than what we ought; the experience of beauty; and a sense of the sublime. It is not that any of these experiences are directly encounters with the ‘noumenal realm’. They could not be that. But we are able to ask, say, when we have an experience as of being free, or of encountering beauty, or sublimity, ‘is this experience well-founded?’, ‘does it go right down, as it were, into the nature of things?’. And here, Kant is able to give a fundamentally reassuring and consoling answer, by the lights of what he thinks practical reason is able to deliver.

In relation to the notion of noumenal freedom, Sussman raises another excellent objection, which is that we seem to have traded “complete subjection to God” with “an equally mysterious subjection to ourselves”:

Kant seems not to have explained how we can be truly free, but instead merely shown how we might experience our own choice as an alien force to which we are subordinated, despite somehow bearing ultimate responsibility for it.

Correctly, Sussman writes that what would help here is to think more about the so-called “noumenal community”, where “members are coordinated but not subordinated to each other”. I agree, although I do not think it is quite the lacuna in the book that Sussman suggests. It seems reasonable to write, first of all, a book about the conception of human freedom, in relation to divine action, with a plan of writing a further book about the implications of what is discovered therein for the notion of human ‘autonomy’, in relation to God. This, indeed, is my plan.[5]

Received: 22 June 2017.


[1] Sussman also considers, reasonably enough, that Kant’s distaste for, or neglect of, theodicy warrants further discussion. I agree that this is an important and interesting topic. But because Sussman does not dispute my claim that Kant does indeed show such a neglect, or distaste, I shall not tackle it here, as it does not directly impinge upon the thesis of Kant and the Creation of Freedom.

[2] See Insole (2011).

[3] For some of my response to the ‘noumena as ontologically distinct’ issue, I draw upon material from my later book, Insole (2015), ch. 6.

[4] For more extended accounts of these caveats, see my comments, in this journal, to Lawrence Pasternack’s Kant on Religion within the Boundaries of Mere Reason (Oxford & New York: Routledge, 2014). See also Insole (2015), chs 7–8, and Insole (2016).

[5] Some of this material is sketched out in Insole (2015), ch. 8. I am currently working on a full treatment.


Insole, C. (2011), ‘Kant’s Transcendental Idealism and Newton’s Divine Sensorium‘, Journal of the History of Ideas 72(3): 413–36.

——— (2015), The Intolerable God: Kant’s Theological Journey (Grand Rapids, MI: William B. Eerdmans).

——— (2016), ‘Kant on Christianity, Religion, and Politics: Three Hopes, Three Limits’, Studies in Christian Ethics 29(1): 14–33.

© Christopher J. Insole, 2017.

Christopher Insole is Professor in the Department of Theology and Religion at Durham University, UK. Insole’s work is concerned with the relationship between fundamental metaphysical and doctrinal commitments, and patterns of thought in meta-ethics and practical reasoning. His latest publication is The Intolerable God: Kant’s Theological Journey