Pensée des sciences
Paperback 22,00 €
Reviewed by: Adam Andreotta (Curtin University)
Kant famously wrote in the Preface to his Prolegomena to Any Future Metaphysics (released two years after the first edition of the Critique of Pure Reason) that: “I freely confess: it was the objection of David Hume which first, many years ago, interrupted my dogmatic slumber” (4:260). In Kant, Hume, and the Interruption of Dogmatic Slumber, Abraham Anderson attempts to understand what Kant meant by this locution. Amongst the central theses that Anderson defends in the book include: [i] the contention that it was Hume’s Enquiry Concerning Human Understanding that awoke Kant from his dogmatic slumber, and not Hume’s Treatise of Human Nature; and [ii] the claim that is was Hume’s challenge to the principle of sufficient reason which awoke Kant from his dogmatic slumber, not a denial of the causal principle governing experience—the thesis that every event has a cause.
In what follows, I will present a summary, and commentary, of Abraham’s defence of these two theses which take place over the course of 5 main chapters, and a lengthy introductory chapter. Before doing so, it is important to clarify some key concepts. First, how does Anderson construe the term ‘principle of sufficient reason’? Anderson tells us that he:
shall use [the term] to refer to the causal principle not restricted to experience, which was supposed to be known by reason, and which Hume led Kant to reject (xii).
Abraham claims (xiv) that Kant was awoken from his dogmatic slumber because he accepted Hume’s criticism of this principle—Hume’s point being that we cannot know causal relations by pure reason. Why is the principle important in the first place? The rationalist principle of sufficient reason is important because without it we cannot know any causal claim that goes beyond experience, such as the claim that something cannot come from nothing. We may be more explicit about this key principle by looking at how it differs from the causal principle—a thesis which Anderson is also concerned with in the book.
The Causal Principle [Hereafter, ‘CP’]: “the principle that every event has a cause” (xi).
We can see how this differs from the former by considering the following:
Principle of Sufficient Reason [Hereafter, ‘PSR’]: “the causal principle extending beyond experience” (xi).
What is the key difference between these two theses? PSR is concerned with what we are justified in believing—that is, it limits our knowledge of causes to experience. Whereas CP is making a definitive claim (albeit one that is negative). PSR claims that we are entitled to hold causal beliefs only insofar as they cohere to experience. If a claim about a matter of fact goes beyond experience, then we are not justified in believing it, even if doing so is natural or useful. CP, on the other hand, is making a negative metaphysical claim—namely, that it is false that every event has a cause. Anderson claims that PSR more accurately allows us to see Hume’s attack as one about metaphysics—the term ‘metaphysics’ in this context referring to the science of objects “beyond experience” (xi). Hume, according to Anderson, is not attacking the causal principle: what he is doing is presenting the limits of our grounds of justification—which is of course limited to experience for Hume. This dispute is an important one to solve, Anderson claims, because it gives us a “clue to the meaning of the Critique” (xi). Anderson point outs (xv) that since Hume is not explicit about his rejection of PSR in either the Treatise or Enquiry, his own proposal is controversial.
In the introductory chapter, titled the ‘The State of the Question’, Anderson provides a survey of the secondary literature which focuses on the issue of how to interpret Kant’s claim in the Prolegomena that, “I freely confess: it was the objection of David Hume that first, many years ago, interrupted my dogmatic slumber” (4:260). The introduction is the longest chapter of the book (42 pages). In it, Anderson considers several different answers to the question of what Kant meant by “the objection of David Hume”, and how such an objection awoke Kant from his dogmatic slumber. Given there is no scholarly consensus about how to understand what Kant meant by the “objection of David Hume”, or how the objection interrupted Kant from his “dogmatic slumber”, Anderson summarises the different perspectives that have been taken on the issue. Anderson also points out that the issue is not only about how to interpret Kant’s famous locution, but also whether Kant should be taken at his word. Anderson states that some think Kant’s claim about being awoken by Hume is a “confusion and misremembering” (1) and should not be taken literally. Anderson thinks that Kant should be understood literally, but he also considers reasons for thinking he should not be. These include Kant’s 1798 letter to Christian Garve, where Kant states that it was the Antinomy that awoke him from his dogmatic slumber—thus apparently contradicting what Kant himself says in the Preface to the Prolegomena. Given that Hume is not mentioned by name in the first edition of the Critique until the very last part, Anderson considers views which propose that this letter lends support to the claim that Kant was not awoken by Hume.
One of the most important views considered in the introduction is Norman Kemp Smith’s, which comes from his 1923 Commentary to Kant’s “Critique of Pure Reason”. Kemp Smith claims that Kant was awoken by Hume’s attack on the ‘causal axiom’ (referred to as ‘CP’ above)—the thesis that every event has a cause. This is a view Anderson returns to throughout the book. It represents an important rival to Anderson’s own view. The view is considered by Hume in the Treatise as follows.
’Tis a general maxim in philosophy, that whatever begins to exist, must have a cause of existence (T 188.8.131.52; SBN 78-9).
Kemp Smith’s claim that Kant was awoken by the Treatise, and his claim that Hume denied CP, are controversial because, as Anderson points out, the Treatise was not translated into German until the Critique of Pure Reason was already published. This is a problem because it is commonly understood that Kant could not read English. Anderson gives several other reasons for doubting Kemp Smith’s proposal. One of the examples he cites comes from Hume’s 1754 letter to John Stewart, which says the following:
I never asserted so absurd a Proposition as that any thing might arise without a Cause: I only maintain’d, that our Certainty of the Falshood of that Proposition proceeded neither from Intuition nor Demonstration; but from another Source (Hume, 1754).
This passage seems to support Anderson’s reading, as Hume is quite upfront here about the nature of his scepticism about causation. Anderson, further, quotes Kant who says in the Prolegomena that Hume’s question “was not, whether the concept of cause is correct, usable, and indispensable for the whole knowledge of nature, for this Hume never doubted” (4:258). Hume’s passage, and Kant’s own admission, seem to go against Kemp Smith’s view, as Anderson suggests.
The rest of the introduction is concerned with several other controversial topics and summaries of scholarly views. For example, Anderson considers the remarks of Manfred Kuehn, Günter Gawlick and Lothar Kreimendah, who argue that it was Treatise 1.4.7. (The conclusion to Book 1) that awakened Kant from his slumber. He considers Lorne Falkenstein’s view, which says that the seeming contradiction between the letter to Garve and Kant’s Preface can be reconciled by accepting that Kant had a gradual awakening. And also, Eric Watkins’s view, which says that Kant is trying to refute Hume’s sceptical challenge to the idea of having any causal knowledge.
In chapter one, Anderson begins to address the book’s central question about what Kant meant by the ‘objection of David Hume’ in the Preface to his Prolegomena. Further, Anderson seeks to understand what Kant meant by being awoken from a ‘dogmatic slumber’. Anderson’s contention, which is further developed in subsequent chapters, is that the objection of David Hume equates to Hume’s attack on Metaphysics (Anderson call this “another name for the objection of David Hume” p. 44). This attack, Anderson tells us, is seen by Kant as a contribution to the Enlightenment because of its implications for the liberation of the human mind—one of which includes a challenge to theological authority.
The chapter touches upon many important issues. One is the directness of Kant’s writing style. Anderson notes that it was dangerous at the time to make attacks on metaphysics too openly (50), since the battle over metaphysics had significant implications for certain religious and political matters. Another issue has to do with Kant’s actual references to Hume. If it was really the objection of David Hume which awoke Kant, then why isn’t Hume mentioned by name in the first edition of the Critique of Pure Reason until late in the work (until the Discipline of Pure Reason)? Might this be a reason to doubt the veracity of Kant’s claim? Anderson thinks not, given the way Hume’s work had been received at the time. He notes of the hostile reception that Hume’s Dialogues of Natural Religion received upon its release. Anderson (53) references a 1779 review of Dialogues featured in the Göttingische Anzeigen von gelehrten Sachen, which was quite critical of the work. The review charged the text of corrupting the youth. This is an interesting reason for why Kant may have been relucent to refer to Hume explicitly initially, and Anderson does a good job of exploring it. It is interesting to note that Hume himself was also conscious of the reception of his own work, which affected the way it was written. In Hume’s December 1737 letter to Henry Home, he says: “I am at present castrating my work, that is, cutting off its nobler parts: that is, endeavouring it shall give as little offence as possible” (Hume, 1737).
To make sense of Kant’s defence of Hume, Anderson also discusses what Kant said about Hume’s critics. These include Thomas Reid, James Oswald, James Beattie, who appealed to common sense to overcome Hume’s concerns about the causal principle. Kant rejects these kinds of appeals to common sense, and Anderson shows why Kant takes Hume’s objections seriously, and how they were misconstrued by others. On page 62, for example, he looks at Priestley’s claim that Hume actually doubted the concept of cause and that the concept was useful. But as Anderson points out, Hume did not think the notion of causation was useless; and neither did he cease to believe in it. Such discussions help to show why Kant found Hume so troubling and help to understand the nature of Hume’s scepticism.
Another interesting puzzle has to do with why Kant is so explicit about Hume’s influence in the Prolegomena. If Kant wanted to avoid the controversies associated with the Dialogues, as Anderson proposes, then why is Kant so open about his debt to Hume in the Prolegomena—two years later after the release of the first edition of the Critique, where he is not so explicit? Anderson’s claim is that Kant’s avowal of his debt to Hume in the Prolegomena is a response to the Göttingische Anzeigen review of the Critique, which came out 2 years after it was released. It may have been that Kant wanted to make his point more explicit since, as Anderson notes (55), Kant regarded the review as a radical misunderstanding of the text. By that point Kant may have felt he had nothing to lose. Anderson offers a second reason for why Kant is more ready to acknowledge his debt to Hume in the Prolegomena. Also published in 1781 was the edition of the works of Sulzer, published by Blanckenburg. Anderson notes that in the Preface to the work, Blanckenburg evoked Sulzer’s Preface to Hume’s Enquiry, which was also included in the work (originally published 26 years earlier), that Hume’s writings would “pull German philosophers by the sleeve and rouse them from their peaceful rest” (cited in Anderson’s text on p. 64). It is hard to know for certain that Kant is responding directly to this passage, but it certainly looks very similar to what Kant writes in his Preface as Anderson points out (p.65).
In the second chapter, Anderson attempts to define the “Objection of David Hume.” After claiming that the objection of David Hume is really attack on metaphysics, Anderson attempts to be more specific about what this attack amounts to. According to Anderson, this attack on metaphysics has three steps, which are divided up further in the chapter. These include:
 “no one can know from pure concepts a priori that because one thing is, another must necessarily exist also.” (72)
 leads, in Anderson’s view, to two implications. The first is:
[C1] “cause is not a legitimate child of reason but a bastard of the imagination, and that all the other purportedly a-priori-subsisting cognitions of reason are mere falsely reminted common experiences.” (72)
And the second is,
[C2] “That there is no metaphysics and cannot be any” (72). (Here Anderson takes metaphysics to be reasoning beyond experience.)
Anderson suggests that Kant located this attack on metaphysics (what Anderson calls ‘Hume’s Objection) in the Enquiry and not in the Treatise, as some commentators such as Kemp Smith have suggested. This attack, Anderson tell us, is substantial because it undermines Leibniz’s principle of sufficient reason, and the causal principles of Descartes and Locke. This is consequential, as such arguments were employed to prove God’s existence (76). So  is clearly a significant result.
Anderson is right, in my view, to characterise Hume’s attack on the “rational origin of the concept of cause” (77–78). This seems to cohere more succinctly with Hume’s radical empiricism, rather than a denial of the causal principle, as Kemp Smith maintains. Further it also seems to cohere with what Kant himself says in his Preface to the Prolegomena. Kant claims Hume’s question:
was not whether the concept of cause is correct, useful, and indispensable for the whole knowledge of nature, for this Hume had never doubted; but whether it is thought by reason a priori (4:258–59).
Next, Anderson engages with the question of whether it was the Treatise or Enquiry that was the key source which awoke Kant from his slumber. Anderson describes the view of Kemp Smith, who follows Vaihinger and Erdmann, in thinking that it was Treatise 1.3.3 that awoke Kant from his dogmatic slumber. Anderson rehearses points made in previous chapters, noting that the Enquiry was published in German in 1755; but the Treatise was not published until 1790, putting it after the 1781 edition of the First Critique. And given that Kant did not know English, this timeline is problematic. This is not a knock down argument, of course, as there were parts of the Treatise translated and Kant knew people who could have read it. For example, Treatise 1.4.7—where Hume advanced a series of sceptical claims—was translated. Yet Anderson claims (89) that the Treatise was less well known in relevant circles.
So where in the Enquiry, then, does Anderson claim Kant located Hume’s attack? There are various places he cites—not all of them are discussed in this chapter. One claim Anderson makes is that [C1] is stated in parts 1 and 2 of Enquiry 7, “Of the Idea of Necessary Connexion.” (90). This is where Hume argues that we have no idea of necessary connection beyond constant conjunction. Another is Anderson’s discussion of [C2], which says that there cannot be any metaphysics, by looking at section 8, 9, 10 and 11 of the Enquiry. For example, he focuses on Hume’s claim at 11.30 which says that, since the idea of necessary connection is grounded in constant conjunction, there is a problem of determining a unique cause. This has implications for the concept of a divine cause. Anderson suggests that the first step, , is to be found in Section 4, part 1 and at 12.29 note (d) of the Enquiry. Such a position is defended in later chapters.
In chapter 3, Anderson attempts to locate where in the Enquiry Hume’s first step in Hume’s attack on Metaphysics is (recall this is the thesis that we cannot have knowledge of causation independent of experience). Anderson also attempts to defend the thesis that the Enquiry supports his own proposal that Hume’s first step is really an attack on the principle of sufficient reason.
Anderson begins by focusing on section 4.11 of the Enquiry, and its debt to 4.2, where Hume talks about our knowledge of matters of fact—namely that, when it comes to matter of fact it is always possible to imagine things being different to the way they, are or what we are used to. While it would be odd, we can easily imagine that a rolling white billiard ball will float up when it hits the black or stop completely. This is because no contradiction materialises: as long as we reason a priori, anything can cause anything. It follows from this, Anderson claims (102), that we cannot know a causal necessity a priori. And further, Anderson states: this “implies a denial of the principle of sufficient reason” (102). This is because we could not say of a cause that it was a sufficient reason of its effect. To put things more precisely, Anderson claims this means that we cannot know a priori, of anything at all, that it must have cause (102). This is drawn from what Hume says at 4.13 of the Enquiry:
When we reason a priori, and consider merely any object or cause, as it appears to the mind, independent of all observation, it never could suggest to us the notion of any distinct object, such as its effect; much less, show us the inseparable and inviolable connection between them (EHU 4.13; SBN 31-32).
Does Anderson’s suggestion do a better job of explaining such a passage compared to the one put forward by Kemp Smith—namely, that Hume denies that every event has a cause? I think so. Anderson’s account—that Hume is rejecting the principle of sufficient reason—seems to capture the spirit of this passage in a more adequate way than Kemp Smith’s.
The chapter also features an interesting discussion about Hume’s disavowal of a thesis that Lucretius called ‘Ex nihilo, nihil’ (119)—nothing comes from nothing. This idea is important because it was taken by some to prove the existence God (as Locke and Clarke tried to do.) Anderson claims that Kant would have seen Hume’s rejection of this the principle as a rejection of the principle of sufficient reason. Anderson claims that
In rejecting Ex nihilo, nihil fit, then, Hume is not rejecting the principle that every event has a cause, which he emphatically accepts. Rather, he is rejecting the principle that Descartes, Locke, and Clarke had used to prove the existence of a divine Cause (109).
It is important, Anderson points out, that this principle does not disprove god; only that it cannot be used to prove god. Again, I think this does a good job of capturing the spirit of Hume’s sceptical empiricism, which is to draw the limits of what we can be justified in believing—namely, to experience.
In chapter 4, Anderson supports his reading of Kant’s interpretation of Hume by examining the Treatise. His main contention is that Treatise 1.3.3 is not, as Kemp Smith supposed, an attack on the causal principle governing experience. He investigates Treatise 1.3.3 in order to undermine Kemp Smith’s claim.
It is important for Anderson to consider Treatise 1.3.3. because, as he states, Hume does not say explicitly in the Enquiry that he is attacking the principle of sufficient reason “in so many words” (123). In addition to arguing against Kemp Smith’s interpretation of 1.3.3, Anderson also draws upon Hume’s letter to Henry Home: the ‘Letter from a Gentleman.’ This letter is important for several reason. First, because it features a candid remark by Hume about the construction of his text—namely, that he went about “castrating” the Treatise, meaning that he cut “off its noblest parts.” Anderson notes that this is most likely because of its implications for theology. What this means is that some interpretive work is needed to determine what Hume is claiming. And second, and more importantly for the content of his argument, Anderson notes of Hume’s reply to critics of the Treatise. Hume claims:
The Author is charged with Opinions leading to downright atheism, chiefly by denying this principle, that whatever begins to exist must have a cause of existence (cited on p. 135 of Anderson’s text).
This is the causal principle listed above—the one which Kemp Smith claims Hume is denying. Hume’s response to this charge is interesting, however. He claims that he is not denying the principle, but rather disputing that the principle was “founded on demonstrative or intuitive Certainty”. This passage supports Anderson’s reading because it shows Hume’s focus is on justification, not on whether the causal principle is false.
Later in the chapter Anderson considers why readers have failed to see the Enquiry as the source of Kant’s awakening. He considers the claim that the causal principle is attacked in Treatise 1.3.3. Anderson disputes this on two grounds because he thinks that:
a) “The causal principle is attacked in the Enquiry too” (139)
b) “The causal principle [Hume] attacks is not the [CP] but the [PSR]” (139)
Anderson considers why Erdmann, Vaihinger, and Kemp Smith failed to see this. One reason he suggests is that while Hume in Enquiry 12.29 note (d) is direct in his rejection of the causal principle Ex nihilo nihil fit, he is indirect in his rejection of the PSR. Another reason he offers is that, while the Treatise is long and detailed in its steps, the Enquiry is “brief and elliptical” (140).
In the final chapter, titled ‘Hume’s Attack on the “Impious Maxim” as the Hidden Spine of the Critique’, Anderson attempts to locate several places in Kant’s Critique which support his contention about the PSR. He does so by examining four places in the Critique that recall Hume’s rejection of the impious maxim (Ex nihilo, nihil fit) at 12.29 note (d). Recall this is the claim that Anderson says is the most direct attack on the PSR. The four places include: the Transcendental Ideal, the Postulates, the Analogies and the Antinomy.
One example that Anderson cites is from the ‘Postulates of Empirical Thought’, in the ‘General Note on the System of the Principles.’ There Kant says that by beginning with mere categories, “We can easily think the non-existence of matter. From this the ancients did not, however, infer its contingency” (B290n). Anderson notes that Kant discusses this matter not to argue that matter is necessary, or contingent, but to suggest that we cannot prove that it is contingent or necessary. Anderson notes that this resembles a discussion Hume makes at 12.28-29 note (d), where Hume rejects the Ex nihilo, nihil fit maxim. The two sections are as Anderson suggests, quite similar. It is one example of the interesting connections Anderson makes between the two works.
In closing, we can ask: is the central claim that Anderson defends in the book plausible? Recall that this is:
Hume interrupted Kant’s dogmatic slumber…by attacking the rationalist principle of sufficient reason, and showing that we are not entitled to it, since we cannot conceive effects as logically necessary given causes, or vice versa, and since we cannot know, either intuitively or demonstratively, that there can be nothing without a reason why it is thus and not otherwise (159).
To my mind, Anderson’s contention does better than some of his rivals—which are, it should be noted, charitably considered in the book. Anderson is, further, careful in his analysis and does not draw any hasty conclusions when advancing his own views. Where there is speculation, it is supported with passages from Kant’s and Hume’s texts, historical documents, and possible counter interpretations. This careful nature of proceeding is one of the virtues of the book.
The book will obviously be of interest to Hume and Kant scholars who seek to understand how Hume’s ideas influenced Kant’s. But it will also be of interest to those seeking to understand the nature of Hume’s scepticism. Given this, I did wonder why Anderson did not discuss how Hume’s radical scepticism affected Kant. As Kevin Meeker (2013, 2) points out, many early readers of Hume—he includes Kant here—interpreted Hume as a radical sceptic. (An interpretation that goes against the scholarly consensus today.) Thinkers like Thomas Reid, for example, thought that if we accept Hume’s system, then we would have to say that we lack rational grounds for holding our everyday common sensical beliefs. It would have been interesting to see whether Anderson thought this radical scepticism played an integral part in Kant’s awakening.
I have only been able to touch upon a few of the issues of the book in this review. It is my hope that I conveyed the great interest of it. I found the book to offer a thorough and convincing account of the influence Hume had on Kant’s thought.
Hume, David.  1932/2011. “Hume to Henry Home, December 2, 1737, Letter 6.” In The Letters of David Hume, ed. J. Y. T. Greig, 2 vols, 1:23– 25. Oxford: Oxford University Press.
Hume, David. 1739-40 [2000 ]. A Treatise of Human Nature. Edited by David Fate Norton and Mary J. Norton. Oxford: Oxford University Press.
Hume, David. 1748 . An Enquiry Concerning Human Understanding. Edited by Tom L. Beauchamp. Oxford: Clarendon Press.
Hume, David.  1932/2011. “David Hume to John Stewart, February 1754.” In The Letters of David Hume, ed. J. Y. T. Greig, 2 vols, 1:187. Oxford: Oxford University Press.
Meeker, Kevin. 2013. Hume’s Radical Scepticism and the Fate of Naturalized Epistemology. London: Palgrave Macmillan.
Kant, Immanuel.  2003. Critique of Pure Reason. Translated by Norman Kemp Smith. With a new introduction by Howard Caygill. 2nd ed. Houndmills, UK: Palgrave Macmillan.
Kant, Immanuel.  2004. Prolegomena to Any Future Metaphysics. Ed. Gary Hatfield. Cambridge: Cambridge University Press.