Reinventing the Mirror: Reconciling Guyer’s and Allison’s Interpretations of Kant’s Second Analogy

This paper concerns two interpretations of Kant’s second Analogy in the Critique of Pure Reason. On the one hand, Paul Guyer argues that the goal of the second Analogy is to provide grounds for the confirmation of beliefs about causal relations. Henry Allison, on the other hand, argues that the principle of the second Analogy is a condition of the possibility of experiencing succession, whether subjective or objective. Their starkly different views on the aims and coherence of Kant’s overall system clearly influence their interpretive differences in the specific context of the second Analogy. The exegetical investigation required to evaluate each point of divergence between them is beyond the scope of this paper. Instead, as regards the second Analogy in particular, I focus on one point where Guyer’s position has been misrepresented. Allison characterizes Guyer as equating objective validity with empirical truth. I show that Guyer is equating objective validity with justification, or a claim to knowledge, rather than knowledge itself. Therefore, Allison has misinterpreted Guyer’s interpretation of Kant in this regard. This paper concerns two interpretations of Kant’s second Analogy in the Critique of Pure Reason. On the one hand, Paul Guyer argues that the goal of the second Analogy is to provide grounds for the confirmation of beliefs about causal relations. Henry Allison, on the other hand, argues that the principle of the second Analogy is a condition of the possibility of experiencing succession, whether subjective or objective. After summarizing the Analogies generally and the second Analogy in more detail, I present Guyer’s interpretation, followed by Allison’s objections. While Allison’s interpretation may be preferable in general, this paper shows that, as regards the second Analogy, Allison’s primary objection to Guyer is based on an inaccurate reading of Guyer’s position. Among other objections, Allison characterizes Guyer as equating objective validity with empirical truth. I show that Guyer is equating objective validity with justification, or a claim to knowledge, rather than knowledge itself. This interpretation depends upon a particular understanding of Kant’s views on Res Cogitans (2014) 5 Southworth | 163 2155-4838 | commons.pacificu.edu/rescogitans the relationship between experience and belief. The view that Guyer attributes to Kant is counterintuitive, but a thorough examination of his interpretation requires more investigation into Kant’s psychology than is possible here. A note before we begin: there is already a vast literature on Kant’s “Analogies of Experience” due, at least partially, to their later importance in the “Refutation of Idealism” which was added to the second edition of the Critique of Pure Reason. One interpretive issue that comes up in the secondary literature is whether these two sections of the Critique can be taken separately from Kant’s larger project of transcendental idealism or must stand or fall with the larger project. Paul Guyer is a proponent of the former view, while Henry Allison favors a holistic interpretation of the Critique. Their starkly different views on the aims and coherence of Kant’s overall system clearly influence their interpretive differences in the specific context of the second Analogy. Both scholars offer readings faithful to the text, and the exegetical investigation required to evaluate each point of divergence between them is far beyond the scope of this paper. Instead, as regards the second Analogy in particular, I focus on one point where Guyer’s position has been misrepresented. In the second edition formulation of the Analogies of Experience Kant claims “[e]xperience is possible only through the representation of a connection of perceptions” that every experience presupposes a connection to some other experience (Kant 1998, 295). The goal of the Analogies is to demonstrate both the possibility and the necessity of this representation. In the service of this end, Kant offers a proof of each of the categories of relation individually. The a priori presupposition of these categories, substance/accident, cause/effect, and community/reciprocity, grounds experience of the three modes of time, persistence, succession, and simultaneity (Kant 1998: 296). The First Analogy, which focuses on the category of substance, or persistence, offers a proof of the principle that all change of appearances is only alteration grounded in a single substance persisting in time (Kant 1998, 304). This substance is necessary to connect all of our representations. Without this substratum, objects would continually be coming into and going out of existence, without temporal or spatial continuity. The second Analogy, dealing with the category of succession, presents a proof of the principle that all alterations occur according to a priori rules, namely the law of cause and effect (B234, Kant 1998, 304-05). Since I will discuss the second Analogy for the remainder of this paper it does not require a detailed description here. The Third Analogy, of the category of simultaneity, seeks to prove that all substances that are simultaneous are in interaction with one another (Kant 1998, 316). Objects are Paul Guyer, Kant and the Claims of Knowledge (Cambridge [England]; New York: Cambridge University Press, 1987). Henry E Allison, Kant’s Transcendental Idealism (New Haven: Yale University Press, 2004). Res Cogitans (2014) 5 Southworth | 164 2155-4838 | commons.pacificu.edu/rescogitans perceived as simultaneous if the perception of one can follow the perception of the other in no necessary order, as in Kant’s famous example of looking at a house. This is opposed to the principle of succession, in which perceptions must occur in a particular order, as those of a boat traveling downstream. Kant posits that if A and B are simultaneous, they have each determined each other. If this were not the case, if A and B were simultaneous but did not interact, then the fact that the perception of B follows that of A would lead us to believe that the first caused the second, that they were successive. Things that are simultaneous mutually determine each other, or are in constant interaction (Kant 1998; 318). For example, the moon and the earth are two simultaneous objects that interact through the force of gravity. If objects that exist simultaneously were not mutually determining, then we would not be able to distinguish between perceiving them as simultaneous and perceiving them as in sequence. Simultaneity is a symmetric relation, while succession is an asymmetric one. It only goes in one direction. Therefore all objects that exist simultaneously are in mutual interaction with one another. The Analogies of Experience function to deduce our right to employ the relational categories in a way that gives us access to time determination that is objective, a judgement to which you would expect universal assent, rather than purely subjective, a matter of opinion. To return to the second Analogy, it specifically seeks to prove that every event that we experience presupposes a preceding event. It is clear that this is intimately connected to the rationalist doctrine of the Principle of Sufficient Reason. Thus, in the first edition, Kant writes that every event, “presupposes something which it follows in accordance with a rule.” In the second edition he instead writes that all events, “occur in accordance with the law of cause and effect”. The principle espoused asserts that to experience any event requires the assumption of a previous event containing the conditions of the possibility of the event that we experience. Thus, the first key component of this Analogy is the claim that we must assume every event is caused as a necessary precondition for experiencing events. Just as the First Analogy proves that experiences of events just are experiences of alteration, so the Second proves that to experience events we must assume that this alteration is caused. Without the assumption that every event has a cause we would not perceive distinct events, objects, or appearances at all. Instead we would be confronted with the undifferentiated manifold of experience, James’ “blooming, buzzing confusion.” The This is a position which can be found in both Guyer and Allison, as well as Arthur Melnick, Kant’s Analogies of Experience. (Chicago: University of Chicago Press, 1973) and Eric Watkins, Kant and the Metaphysics of Causality (Cambridge University Press, 2005). 3 There is some interpretive controversy over whether the Second Analogy refers to sequences of events, or to sequences of representations constituting an event. However, Allison and Guyer both agree on the latter interpretation so I will not delve into the issue here. 4 William James. “The Principles of Psychology” (1890), 462. Res Cogitans (2014) 5 Southworth | 165 2155-4838 | commons.pacificu.edu/rescogitans second Analogy specifies the conditions in which we can subsume our experience under the law of cause and effect. It provides the rule by which we determine that our constant perception of successive apprehensions is representative of an actual succession of appearances. Kant claims that when we perceive appearances as changing over time (a state exists at one time, the opposite of which exists at another time), we are really linking two distinct appearances (Kant 1998: 304). Although we represent these appearances as connected, we do not directly experience cause and effect. Rather than perceiving some state of affairs, A, as causing another state of affairs, B, we merely perceive A as preceding B. Subsuming this experience under the law of causality requires an inferential move which the mere succession of our representations does not sufficiently substantiate. The rule that justifies this inference is necessity: that if A causes B, we cannot experience B prior to A. That the fluctuation in representations constituting an event occurred in a particular, determined order is what makes the difference between a mere succession of apprehensions, which are subjective, and a succession of states of affairs, which are objective (Kant 1998: 308). Here Kant acknowledges what may appear to be a contradiction. In previous remarks on the understanding, Kant has argued that only through the perception and comparison of many appearances does the understanding come to discover rules, whereas here he claims that causality is an a priori principle of the understanding (A196/B241). Kant’s response is that this is only the case with particular empirical rules. The category of causality is pure a priori and, like other pure a priori representations, “we can extract [them] as clear concepts from experience only because we have put them into experience, and experience is hence first brought about through them” (308-09). Which is to say that although beliefs about particular causal laws (i.e. that if you apply heat to a gas it will increase in volume, and vice versa), are formed through repeated experience, the broader concept that events have causes is a necessary precondition of the experience of events in the first place. It is not derived from experience and cannot be directly experienced, but necessarily grounds the kinds of experiences that we, as humans who receive appearances through the twelve categories, have. The question of how a particular state A gives rise to another state B is an empirical question of which we can have no knowledge a priori. This does not prevent us from having knowledge of the form of alteration, of the conditions under which it is possible that one state gives rise to another. The relationship between causality and justification is the focus of Guyer’s interpretation of the second Analogy, discussed more closely below. According to Guyer, the Analogies provide an “epistemological theory of the necessary conditions of time determination.” He argues that Kant’s theory of time determination must be understood, not as a psychological model of belief generation 5 Guyer, Kant and the Claims of Knowledge, 207. Res Cogitans (2014) 5 Southworth | 166 2155-4838 | commons.pacificu.edu/rescogitans but rather as an epistemological model of the confirmation of beliefs. For Guyer then, Kant’s argument in the second Analogy is as follows: given that we must perceive states of affairs as following from preceding states of affairs, and that a succession of perceptions does not guarantee a successive state of affairs, some rule is required to justify our belief that a succession has occurred that is not merely subjective, but objective. In Kant’s famous examples, this rule justifies us in distinguishing between the succession of representations constituting the apprehension of a house, in which we can look at either the top or the bottom first, and those constituting the apprehension of a ship moving downstream, which we experience in a necessary or determinate order. On Guyer’s interpretation, particular causal laws are a requirement, not for successive representations generally, but for being justified in the belief that such a succession actually constitutes an event. As noted above, if Kant is understood as appealing to causal laws as evidence for the validity of causal laws, then he makes empirical laws necessary conditions of experience when, on his own view, they must be based on experience. Guyer’s reading of the second Analogy avoids characterizing Kant’s position as circular or question begging. So long as Kant is making epistemic claims about the confirmation of beliefs, rather than psychological claims about the generation of beliefs he is not appealing to causal laws as evidence for the validity of causal laws. Instead Kant is appealing to causal laws for confirmation of beliefs about the causal relationships governing particular events. Guyer holds that the Analogies provide conditions for making knowledge claims, for confirming beliefs, not conditions for the possibility of experience, conditions for forming beliefs in the first place. Thus, the second Analogy is a proof of the conditions that must obtain for us to be justified in claiming that we have perceived an objective succession, as opposed to a subjective one. Allison criticizes Guyer’s position on several points. Setting aside his apparent dislike of Guyer’s insistence on a tight connection between the second Analogy and the Refutation of Idealism, he also regards Guyer as equating objective validity with empirical truth. There are two other main reasons that Allison considers Guyer’s reading to be inaccurate. First, Allison interprets Guyer as arguing that we require knowledge of particular causal laws to recognize instances of objective succession. Therefore, Allison claims that Guyer’s position is “implausible on the face of it, since