Constructing Reason

Abstract

In The Architectonic of Reason, Lea Ypi provides an illuminating and innovative interpretation of the Architectonic in the first Critique. Ypi argues that Kant’s project of uniting practical and theoretical uses of reason in a critical metaphysics ultimately fails because practical reason does not have its own domain in which to legislate. This article challenges Ypi’s objection to practical reason’s lack of a domain in the first Critique. Its main contention is that reason’s need for unity in legislation may be satisfied by a belief in God as a necessary practical presupposition rather than a dogmatic metaphysical reality.

1. Introduction

In The Architectonic of Reason, Lea Ypi (Ypi 2021; cited throughout by page number) provides an illuminating and innovative interpretation of the often neglected Architectonic in the first Critique. Ypi argues that Kant’s project of uniting practical and theoretical uses of reason in a critical metaphysics ultimately fails in the first Critique but that the section nevertheless warrants serious study and reflection. The main reason is that the Architectonic shows the systematic importance of the unity of reason in its theoretical and practical uses. Even though its solution appears to fall within dogmatic rather than critical metaphysics, the Architectonic sets a task for critical philosophy that is later solved in the second and third Critiques. While Kant moves on to provide a domain for reason’s practical legislation in the second Critique and an account of purposiveness as normativity in the third, the Architectonic shows why the unity of reason is crucial and places systematicity at the centre of the critical project. Ypi shows how the Architectonic teaches us that a systematic endeavour must begin with an idea of the whole to which parts are added organically as limbs to a body and for this reason, anyone interested in the constructive unity of Kant’s critical project ought to begin with the Architectonic where the idea of reason’s unity is first presented. In this article, I focus on reason’s unity as arising from its legislation and discuss three points: (1) the idea that reason’s legislation constructs the objects of experience and is itself constructed; (2) the notion of a domain for legislation and its relation to legislation as both producing the content of laws and imposing these as binding; (3) cosmic philosophy as the legislation of reason.

Ypi makes a strong case for the importance of the Architectonic, which Kant defines as the art of systems (A832/B860). ¹ After providing the elements of pure reason in the first part of the first Critique, Kant turns to the question of how they fit together in the Architectonic. This is crucial since systematic unity is what turns cognition into science. Ypi’s aim is to show that the project of uniting the different uses of reason in a systematic unity must necessarily fail in the first Critique because this work does not include a distinct domain for reason’s practical legislation. Instead, the Architectonic presents purposiveness as design as the systematic idea uniting the theoretical and practical use of reason. Unlike the third Critique’s notion of purposiveness as normativity, this uniting idea presupposes a purposeful creator as a necessary idea of reason. While Ypi focuses on the development of this notion in the first Critique, she aims to show why the problems confronted by the Architectonic receive a more satisfactory solution in the second and third Critiques. For Ypi, the Architectonic shows that the first Critique cannot unite the two uses of reason systematically and this unsolved conundrum then calls for the solutions presented in the second Critique, which provides the practical use of reason with its own domain, and the third Critique, which reconsiders purposiveness as normativity. ² On her reading, even if the Architectonic fails to provide a critical solution to the problem of reason’s systematic unity, it succeeds in setting unity as a task for further inquiry.

The perceived failure of the Architectonic might account for the lack of interest in the section in current Kant scholarship. Why should we be concerned with the first Critique’s failed attempt at systematic unity when Kant’s solution to the problem is available to us in the subsequent Critiques? Apart from an historical interest in the development of Kant’s thought, one might wonder whether Kant’s lapse into metaphysics in search of systematic unity should interest us today when we can read his development of the problem directly in his later works. Despite these objections, Ypi argues that the Architectonic warrants interest because it is

a significant and unique text for understanding why the unity of reason is necessary, what role the principle of purposiveness plays in integrating theoretical and practical uses of reason, and how engaging with Kant’s challenge may be relevant to reflect on the problem of systematic unity in philosophy more generally. (p. 2)

Throughout the work, Ypi makes references to Kant’s metaphors from construction and their role in presenting Kant’s notion of the architectonic. By relating architectonic with the metaphors of construction, Ypi differs from Paula Manchester who has argued that we should understand Kant’s ‘architectonic’ as a technical term whose relation to construction is merely etymological (Manchester 2003: 188). Ypi instead argues that the metaphors of construction and architecture give us insight into Kant’s understanding of architectonic as the art of building systems. For Ypi, architectonic as construction is a continuous unifying process:

The speculative and practical unity of the system is guaranteed only if human beings take seriously their nature as moral self-legislating agents and the rational order they are able to create on the basis of the free development of their moral ends. Far from being guaranteed by nature, and settled once and for all, systematic unity is a work of continuous construction. (p. 33)

This last sentence shows how Ypi’s understanding of systematic unity points beyond the first Critique: without the notion of purposiveness as normativity, the construction lacks an inherent guiding thread which only appears in the third Critique’s account of reflective judgement. Ypi uses the third Critique as a measure for the first; she shows that the first Critique lacks the tools for accounting for this type of constructive process. Although she takes the third Critique as a lens through which to understand the systematic endeavour of the first, she does not pretend to find a complete plan for the construction of reason ready in the first Critique. Instead, she argues that the first Critique sets the task of the unity of reason but does not yet provide a solution that lives up to the demands of critical metaphysics. In particular, the first Critique relies on the notion of purposiveness as design that implies postulating a purposive creator and thus backsliding into dogmatic metaphysics. Ypi thus opposes the standard constructivist account of Kant, which sees the unity of reason as a single project through Kant’s work (O’Neill 1989).

In the following, I discuss three points in dialogue with Ypi’s reading of the Architectonic. My engagement with Ypi’s study circles around the conundrum of Kant’s three central metaphors of systematicity: construction, law and geography. The first concerns the relationship between Architectonic and construction; I discuss whether Ypi’s study should be read as a constructive approach to reason, and to clarify the differences and similarities between Ypi’s reading and Kantian constructivism, I confront her reading with that of Onora O’Neill. The second is the relationship between reason’s legislation and its domain. Ypi argues that the unification of reason turns dogmatic because the practical use of reason does not have its own domain in the first Critique. I discuss how Ypi understands reason’s legislation and why this implies a need for a domain. Against her reading, I argue that reason’s need for unity in legislation may be satisfied by a belief in God as a necessary practical presupposition rather than a dogmatic metaphysical reality. The final question is that of systematicity. Here I discuss how the notion of cosmic philosophy as legislative relates to the practical and theoretical legislation of pure reason.

2. Kant’s plan to construct a unitary system

Like the critique of pure reason, also the construction of reason is fundamentally ambiguous: we can understand reason as both the constructing subject and the constructed object. As constructing, reason constructs experience through the forms of intuition and the categories of the understanding. ³ This constructive view of experience allows us to map a critical metaphysics in correspondence with the necessary conditions of experience. The constructive approach builds on the idea that if reason constructs experience then critical metaphysics can construct reason and the result will match possible experience, which means that the objective aspect of construction depends on the subjective aspect. In the critical endeavour, reason becomes the object of construction and here the unity of reason becomes a problematic task for the critical philosopher: even if the critical philosopher manages to correctly list the elements of reason’s constructive activity, these elements need to be put together in a unitary account. Once the elements of reason are in place, as they are after the first Critique’s Doctrine of Elements, the challenge is putting the different elements together according to a plan rather than as a random aggregate. Without a unifying principle, the elements can only be united into an unsystematic aggregate that does not qualify as a science: ‘A “science” deserving of that name is therefore for Kant not reducible to the availability of a series of cognitions, be they rational or sensible, a priori or empirical, but requires a principle for their organic relation and functional systematization’ (p. 22). The critique of reason aims to provide the science of reason that successfully reconstructs its systematic unity. On the constructivist view, this leads to a double construction since the task of the philosopher is to map reason’s unity and thereby reconstruct its construction. This is why Kant turns to the unity of reason in the Architectonic, in which he argues that the elements of reason are articulated like the limbs of an organic body and not like randomly aggregated parts. Because Kant conceives of reason’s unity as organic unity, the search for a unifying schema leads him to the question of purposiveness. To put together reason’s various elements organically we must conceive of them as having a common purpose (A833/B861). The biggest difficulty for this common purpose is uniting theoretical and practical uses of reason under a single principle. As Ypi remarks, ‘The interest in systematic unity is therefore theoretical and practical at the same time’ (p. 20). Since systematic unity is supposed to unite both uses, the first Critique’s account is clearly disadvantaged because it lacks the notion of practical autonomy that Kant later introduces in the Groundwork for the Metaphysics of Morals. ⁴ Instead, Kant relies on a stipulation of purposiveness as design, meaning that we need to assume that reason’s uniting purpose is part of its design rather than a normative feature of its use. Ypi argues that the first Critique’s stipulation of this purpose as founded on design lets the critical endeavour slide into dogmatic metaphysics. Her point is that this task is better fulfilled by the third Critique’s account of purposiveness as normativity, which places the purpose with the cognizing subject rather than an external creator.

Although Ypi emphasizes the differences between the role of the cognizing subject in the first and third Critiques, she also shows that the constructive project of metaphysics is a collective and public endeavour:

the authority of reason is asserted by placing it at the heart of a constructive project which is recursive and public in character. The coherence of its principles is established by going back and forth in the process of assessing human knowledge and confronting its cumulative results with an ongoing conversation based on rules that everyone can in principle endorse. (p. 6)

Ypi recognizes that her account of the Architectonic is intended to be read as a continuation of constructivist readings of the first Critique. Since the Architectonic provides a plan for the construction of reason, Ypi aims to ‘sketch the contours of that plan, to explore how Kant brings it to completion, and to illustrate where its weaknesses lie’ (p. 173). Ypi builds on recent constructivist readings of the critical project but admonishes against reading the constructive project as fully developed in the Architectonic. Although the overall project of uniting the theoretical and practical use of reason remains as it is set out in the first Critique, the interpretations of reason’s practical legislation and the understanding of purposiveness change dramatically as Kant develops his philosophy.

To understand the contribution of Ypi’s study, we might confront it with Onora O’Neill’s account of the Canon and Architectonic in The Construction of Reason. Both Ypi and O’Neill are interested in the way reason’s unity is constructed in the Doctrine of Method. Yet their conclusions are dramatically different: while Ypi concludes that Kant fails to provide a critical construction of unity in the Architectonic, O’Neill argues that the Canon provides an account of how this unity can be constructed in a community. Both Ypi and O’Neill draw on a passage in the Discipline of Pure Reason where Kant asserts the importance of community and debate for the construction of reason: ‘The very existence of reason depends upon this freedom [of critique], which has no dictatorial authority, but whose claim is never anything more than the agreement of free citizens, each of whom must be able to express his reservations, indeed even his veto, without holding back’ (A738/B766). While Ypi argues that the first Critique’s attempt at systematic unity in the Architectonic fails to live up to this ideal of free critique, O’Neill sees an early formulation of the categorical imperative in this passage:

The only route by which we can vindicate certain ways of thinking and acting, and claim that those ways have authority, is by considering how we must discipline our thinking if we are to think or act at all. This disciplining leads us not to algorithms of reason, but to certain constraints on all thinking, communication and interaction among any plurality. In particular, we are led to the principle of rejecting thought, action or communication that is guided by principles that others cannot adopt, and thus to the Categorical Imperative. (O’Neill 1989: 27)

O’Neill’s point is that the universality principle expressed in the categorical imperative is already present in Kant’s description of the freedom of critique in the Discipline of Pure Reason. She thus understands the systematic unity of reason as a continuous construction between Kant’s different philosophical works. Although Kant does not introduce the categorical imperative explicitly in the first Critique, O’Neill reads it as the ultimate principle of reason, which is already guiding this early work. This is where Ypi’s critique of Kantian constructivism sets in: her point is that Kant’s understanding of the unity of reason undergoes fundamental changes in the subsequent Critiques. Unlike O’Neill, she argues that the first Critique fails because it lacks the resources introduced in the later works, in particular moral autonomy and purposiveness as normativity. Ypi’s unity of reason is not a unity of Kant’s critical works; instead, she aims to show why Kant had to rethink practical and teleological reason to unify theoretical and practical reason systematically. While Ypi agrees with O’Neill that Kant’s intentions for the critical project and the unity of reason are set up in the first Critique, she convincingly argues that Kant still fails to live up to his own ideals for critique. The reason for this is his understanding of reason’s legislation, which does not yet include moral autonomy.

3. Reason’s legislation

In recent years, there has been an increased interest in Kant’s accounts of reason’s laws and legislation (Massimi and Breitenbach 2017; Watkins 2019) and Ypi continues this discussion by investigating the role of legislation for the unity of reason. In the Architectonic, Kant defines legislation as the uniting feature of reason: ‘Now the legislation of human reason (philosophy) has two objects, nature and freedom, and thus contains the natural law as well as the moral law, initially in two separate systems but ultimately in a single philosophical system’ (A840/B868). Despite formulating this ideal for philosophy, Ypi argues that Kant fails to provide a unitary philosophical system in the first Critique because of its deficient account of practical legislation. Combining Kant’s metaphors of construction with those of geography and law, Ypi argues that reason’s practical legislation does not fit into the account in the first Critique because it lacks a designated domain: ‘in the first Critique, reason does not have a territory in which its concepts are legislative, despite Kant’s attempts to define practical obligation as an act through which reason gives laws’ (p. 147). This problem is resolved by the introduction of autonomy in the Groundwork and the second Critique, in which ‘transcendental freedom is no longer a problematic concept in relation to practical freedom (A 803, B 831; 676). Ideas do not have merely practical use, rather reason has its own practical legislation which authorizes it to have a proper “domain” with its own principles and objects’ (p. 148). In her account of legislation in relation to a domain, Ypi refers to the third Critique’s sophisticated account of a domain as a space where the concepts of reason legislate, and argues that without the fact of reason and a notion of transcendental freedom, there is no domain in which the practical concepts of reason are legislative since ‘there is no domain in which the will legislates autonomously’ (p. 147). While ideas have a practical use in the first Critique, only in the second Critique does reason have a practical legislation in the domain of transcendental freedom.

Although Ypi focuses on reason’s legislative function, she does not mention that the term ‘legislation’ is ambiguous in the first Critique, since it can indicate both the production of a law and its imposition as binding (Timmermann 2018: 103). Legislation in both these senses is necessary for moral autonomy. While reason contains or produces the principles of both theoretical and practical laws in the first sense in the first Critique, it lacks the ability to make the moral law binding, which is the innovation introduced with the notion of autonomy. As Dieter Henrich writes:

Were reason only to know the right, without at the same time being the ground of its obligatory power, then it would indeed be the faculty through which we came to know the right, but not its origin. There would be a autognosy of reason with regard to rational action, but in no sense an autonomy of reason. Autonomy requires that reason have the power to effect actions which take place solely because they are rational. (Henrich 1994: 94–5)

While reason produces laws for objects of experience and makes them binding by its cognitive activity, the question of a domain concerns the type of entities for which reason could make the moral law binding. The account of moral legislation in the Canon might be helpful to give a more nuanced picture of Kant’s account of moral legislation in the first Critique. Here Kant presupposes that ‘there actually are pure moral laws that determine behaviour, i.e. the use of freedom of a rational being as such, completely a priori (without regard to empirical motives, i.e. happiness), and that these laws command absolutely’ (A807/B835). However, these commands are not made obligatory by reason, instead ‘God and a future life are two presuppositions that are not to be separated from the obligation that pure reason imposes on us in accordance with principles of that very same reason’ (A811/B839, my italics). Reason’s two types of legislation cannot be united in a single system since one of them depends on the postulate of God, who imposes the moral law as binding even if its content is articulated by reason. Without the ability to make the moral law binding, reason is not yet practically legislative. This lack of moral autonomy is what leads Ypi to argue that reason cannot attain unity with the tools available in the Critique of Pure Reason.

Yet the lack of a domain for practical reason in the first Critique is only part of the story: while reason in the first Critique is capable of formulating moral laws, it lacks the power to give these laws binding force, which is why Kant needs to introduce God as the authority behind the obligatory nature of the moral law. However, Kant does not postulate God’s existence as a metaphysical given or an apodictic truth; only belief in God’s existence is necessary for the imposition of practical laws as binding. Here, the imposition of moral laws as binding requires belief (Glaube) in God and an immortal soul but not the existence of these things. Without the notion of autonomy, the bindingness of moral laws does not yet appear to need a separate domain in which the moral laws have jurisdiction. Instead, what is required is merely the belief in such a domain, as a kingdom of grace, whose existence remains metaphysically inaccessible to us. We might read these suggestions in either a theistic or an agnostic way. A theistic reading, such as the one provided by Ypi, understands God as legislator of the moral law in the first Critique in the sense that God is necessary to make the moral law binding. This reading would indeed imply that the first Critique slides into dogmatic metaphysics in its account of why moral laws are binding. But we might also take an agnostic approach to these suggestions, which would be more in line with Kant’s critical philosophy. An agnostic reading would suggest that the mere belief in God and an immortal soul is sufficient to provide the moral law with a postulated domain. The domain does not depend on the existence of God but only in the belief therein. Reason’s ability to obtain unity in the Critique of Pure Reason does indeed not depend on providing a separate domain for practical reason: ‘So far as practical reason has the right to lead us, we will not hold actions to be obligatory because they are God’s commands, but will rather regard them as divine commands because we are internally obligated to them’ (A819/B847). Here Kant emphasizes the internal obligation already imposed by reason, which is far from a dogmatic introduction of God as a metaphysical truth. Instead, a belief in God satisfies an internal need of reason and relies on a mere presupposition for practical purposes rather than a dogmatic theoretical claim. Reason’s unity is here obtained by a belief in a legislator that makes the moral laws binding since these are already internally imposed by reason. In the discussion on believing that follows, Kant argues that reason has a practical need for a subjectively sufficient belief in God as the creator of the world as a unitary whole and as the legislator of the moral law. Again, all that appears to be needed is a subjectively sufficient belief in a legislator that makes the laws imposed by reason obligatory. This claim would be compatible with an agnostic stance regarding God’s existence and merely a presupposition for practical purposes rather than a slide into dogmatic metaphysics.

4. Purposiveness and systematic unity

To understand reason’s legislation in the first Critique, we need to untangle Kant’s three overlapping metaphors for his proposed system: law, geography and construction. ⁵ Ypi’s reference to the lacking domain for practical reason combines law and geography: in analogy to the way that legislators only have legislative authority over a particular territory, reason can only be legislative if it possesses a domain for its principles. Once this domain is secured, reason can construct its system on it (pp. 132–4). As I argued in the previous section, these complex metaphors point to a fundamental ambiguity in Kant’s conception of legislation: if reason is autonomous then it is both the source of laws and the authority that makes them binding. Ypi correctly points out that reason’s practical legislation lacks a domain in the first Critique but she does not discuss the relationship between this domain and reason’s authority for making its moral laws binding. While this conundrum is addressed by moral autonomy, Ypi points to an additional desideratum raised by the Architectonic, which is satisfied by the conception of purposiveness as normativity in the third Critique. Kant expresses this problem in his distinction between scholastic and cosmic philosophy: ‘philosophy in its cosmic definition aims to resolve the problem of systematic unity in general by integrating the sum total of theoretical cognition with the practical interest of reason’ (pp. 29–30) and ‘[t]he interest in systematic unity is therefore theoretical and practical at the same time’ (p. 20). Ypi cites Kant’s definition of the philosopher who pursues cosmic philosophy as a legal scholar of reason’s legislation, unlike the sophists who pursue scholastic philosophy and behave like pettifogging lawyers (pp. 28–9, citing the Vienna Logic, 24: 800). In the Vienna Logic, Kant thus distinguishes between cosmic and scholastic philosophy by referring to their different relationships to the laws of reason. While the cosmic philosopher strives to study the laws of reason, the scholastic philosopher twists them for his own purposes. In the Architectonic, Kant describes the cosmic philosopher as the legislator of human reason while the scholastic is an artist of reason (A839/B867). He stipulates in the Architectonic that cosmic philosophy unites practical and theoretical uses of reason under a single purpose. Ypi argues that his argument fails for the following reasons:

Kant’s argument is here based on three crucial, and yet undefended assumptions: 1) that identifying the final end (Endabsicht) of nature on the basis of the supreme end that human beings display in their practical actions is possible; 2) that a similar ‘final end’ consists in the moral destination of humankind; and 3) that in a similar constitution of reason nature has proved itself to be both ‘provident’ (versorgenden) and ‘wise’ (weislich). (p. 32)

Ypi’s main contention is that Kant needs to presuppose a wise and provident creator in order to guarantee that reason’s legislation applies equally to both of its objects: nature and freedom. Without the notion of purposiveness as design, this task is left to a dogmatically imposed creator rather than seen as a normative ideal of reason (p. 62). Again, Ypi’s point is that Kant correctly identifies the desiderata for philosophy, namely to unite the practical and theoretical uses of reason under a single purpose, but that his solution to this problem remains within dogmatic metaphysics. Still, Ypi recognizes that ‘the schematism of reason in the first Critique essentially plays an equivalent function to that which the principle of purposiveness plays in the third’ (p. 136). The notion of philosophy as legislative brings us back to the construction of reason: while the Vienna Logic suggests that the cosmic philosopher is a student of laws that have already been imposed, the description of the cosmic philosopher as the legislator of reason in the Architectonic suggests a more active role of imposing the laws of reason on cognitive activity.

5. Conclusion

Lea Ypi shows that engaging with the first Critique’s Architectonic is worthwhile for understanding Kant’s project of unifying theoretical and practical uses of reason. To show this, she engages with Kant’s intricate metaphors of law, geography and construction. Although Ypi grants Kant that philosophy of pure reason is intended to be ‘unchangeable and legislative’, her study confirms that ‘there are several points here which could arouse reservations and weaken the conviction of its lawfulness’ (A847/B875). The Architectonic is thus a section that points beyond the Critique of Pure Reason towards the reformulations of reason’s legislation that Kant brings in his following works. Ypi correctly points to the lack of a domain for practical legislation in the first Critique, and to this contribution I have added the notion of obligation as an additional problem for the concept of moral legislation in this work. While Kant sketches the idea that ‘morality is the only lawfulness of actions which can be derived entirely a priori from principles’ (A841/B869) in the Architectonic, he does not provide an account of how reason can make these binding. Instead, as Ypi shows, his account risks lapsing into dogmatic metaphysics since it depends on the presupposition of God as the authority behind the moral law. Yet I have argued that Kant’s account of a practically necessary belief in God’s existence is neither dogmatic nor metaphysical. The tools of the first Critique may provide unity for reason through a subjectively necessary yet objectively insufficient belief in God, whether as legislator of the moral law or as creator of a unitary world. Belief in God is indeed a necessary presupposition of the first Critique’s account of reason’s unity and the bindingness of the moral law but this should not be conflated into a dogmatic metaphysical claim about God’s existence. Instead, an agnostic reading would suggest that this belief is provided by practical reason itself and that a presupposition of a creator satisfies reason’s need of practical unity.