1 Introduction

There was a heated debate surrounding intellectualism during the first half of the twentieth century. The recent scholarship on the history of intellectualism is very much indebted to Michael Kremer (2017), who explored the intellectual landscape against which Gilbert Ryle (1945 [2009]; 1949 [2009]) proposed his epistemological and action-theoretic criticism of intellectualism. Intellectualism was a thesis about the typical source of motivation in human action. Intellectualists held that this typical source is the intellect, which may operate in the format of thinking, judging, deliberating, reasoning, or other similar acts of mind. This intellectualist notion was sometimes called “intellectualist psychology” or “the intellectualist theory of action” (see, e.g., McDougall (1908 [1919], 406); Russell (1927, 2–3)). Opponents of this notion argued that the typical sources of motivation in human action are non-intellectual elements, such as instincts, habits, and impulses. This opposing view was sometimes referred to as “anti-intellectualist psychology” (see, e.g., Parsons 1935, 423, 435).¹

Although the controversy over intellectualist psychology was centered around the empirical facts about human nature, it was sometimes carried out toward a discussion around the conceptual issue of what renders a behavior intelligent (see Kremer 2017). The intellectualists claimed that intelligent behavior requires to be guided by the intellect, among other things. The anti-intellectualists, however, denied this necessity condition, holding that instincts or habits may operate in an intelligent way in the absence of any input from the intellect. The political psychologist Graham Wallas and the social psychologist William McDougall were among the anti-intellectualists who argued along this line (Kremer 2017). In this debate, intelligence was identified by the behavioral mark of adaptability to situational features². Note that this is a pre-rational conception of intelligent behavior. That is, the attribution of intelligence to a piece of behavior along this conception has no implication about the rational status of that behavior (see Lanz 2000). According to Kremer (2017), intellectualist and anti-intellectualist psychologists agreed over the claim that for a behavior to be reasonable or rational it requires to be preceded by the conscious consultation of the intellect.

I present the evidence showing that Frank Ramsey was exposed and attracted to a wide range of anti-intellectualist psychological works. It is a well-known fact that Ramsey was a passionate reader of classical pragmatism and that he went on to develop his own version of pragmatist philosophy in the second half of the 1920s. I show that classical pragmatists and anti-intellectualist psychologists adopted similar accounts of human nature, the fact that explains Ramsey’s joint interest in the anti-intellectualist tradition in psychology and the pragmatist tradition in philosophy.

Ramsey made three substantial contributions to the anti-intellectualist tradition via his pragmatist notion of habit. First, he contributed to anti-intellectualist psychology by arguing that judgments (as instances of the intellectual acts of mind) take shape by habits. Second, he blurred the distinction between the domain of reason, on one hand, and the domain of instincts and habits, on the other. He thus departed from the orthodox view of the active figures of the intellectualism debate by arguing that reasonability could be predicated on instincts and habits. That said, Ramsey adhered to a view similar to that of Gilbert Ryle’s that appeared in the mid-twentieth century, according to which there is a class of rational or reasonable actions that are not informed by conscious consultation of the intellect. Third, Ramsey’s notion of habit provided the ground to explore the nature of knowledge employed in acting from habit. It was Margaret MacDonald who extended the epistemological dimension of Ramsey’s notion and eventually came up with the distinction between knowing that something is the case and knowing how to apply a rule (or habit). As Michael Kremer (2022) shows, MacDonald’s account of knowledge found its way to Ryle’s philosophy. Hence, I shall suggest that Ramsey influenced Ryle’s account of knowledge through the channel of MacDonald.

2 Anti-Intellectualism and Classical Pragmatism

The close connection between anti-intellectualism and classical pragmatism was noted by those exploring the philosophical insights of anti-intellectualism in the early twentieth century. John Dewey, for instance, found it apt to take anti-intellectualism as a pragmatist philosophy (Dewey 1910, 478–79), and Ralph Perry discussed pragmatism as a philosophical position residing within the anti-intellectualist tradition (Perry 1918, 287). The affinity between anti-intellectualism and classical pragmatism should not come as a surprise if we take this into account that the classical pragmatists adopted an account of human nature similar to what was advocated by anti-intellectualist psychologists.

The classical pragmatists are famous for their Humean-like account of human nature, an account that is set against intellectualist psychology. For the classical pragmatists, human actions are ultimately cases of instincts and habits. For William James, “living creatures” are “bundles of habit” (James 1890, 104). Habit is “the enormous fly-wheel of society”, and “it alone is what keeps us all within the bounds of ordinance” (James 1890, 121). In other words, “all our life, so far as it has definite form, is but a mass of habits” (James 1899 [1983], 47). For C. S. Peirce, “our logically controlled thoughts compose a small part of the mind, the mere blossom of a vast complexus, which we may call the instinctive mind” (Peirce 1978a, 130), and human inferences are guided by “habit[s] of mind” (Peirce 1978a, 227–28). In John Dewey’s words, “habits… constitute the self… [T]hey are will. They form our effective desires… They rule our thoughts” (Dewey 1922, 25).³

The pragmatist account of human nature could be interpreted in two opposing ways, depending on how one conceives the intelligent status of instincts and habits. First, if one holds that instincts and habits involve no intelligence whatsoever, then it implies that the pragmatist suggests that humans are non-intelligent beings. Second, if one holds that at least certain cases of instincts and habits involve intelligence, then it implies that the pragmatist conceives humans as intelligent beings. Something similar to the first interpretation of the pragmatist account of human nature could be found in John Watson’s behaviorism, a popular theory of psychology that emerged in the 1910s. But we shall see that the classical pragmatists defended the claim that instincts and habits might involve intelligence, and thus adopted the second interpretation mentioned above.

Among the central messages of Watson’s behaviorism was that “each of us is the sum total of all the habits we have acquired since our birth” (Mills 1998, 52). For Watson, habitual behavior was reducible to a set of reflexes, which, for him, were mechanistic and devoid of intelligence (see Watson 1924, 184–85). Then, Watson’s account of the nature of reflex arc and its extension to all human behaviors, in McDougall’s words, resulted in “the mechanistic behaviorism”, which ignored “the simple obvious and commonplace truth that man is a purposive being” and instead argued that “men are… Robots” (McDougall 1926, 73).⁴ But, for the pragmatists, human beings were not Robots in the sense in which McDougall used the term.

The classical pragmatists argued that instinctive and habitual behaviors have the capacity to be intelligent. Peirce argued that habit, unlike “mechanical law”, manifests itself in a flexible way. This form of “modality” is “the central principle of habit” (Peirce 1931, 207). Without this feature, for Peirce, the “possibility of habit developing into intelligence would be cut off at the outset” (Peirce 1931, 208). In his famous chapter on habit in The Principles of Psychology (1890), James argued that reflexive, instinctive, and habitual behaviors involve intelligence.⁵ In a similar spirit to Peirce and James, Dewey (1922) argued at length that habits have the capacity to be more and more sensitive to situational features and to be developed into full-blown intelligent ones. Indeed, as Bermúdez and Felletti (2021, 588) put it, the pragmatist tradition provides rich conceptual resources for those contemporary scholars who want to argue that there is at least a subset of instinctive or habitual behaviors that involve intelligence.⁶ We shall see that Ramsey’s version of pragmatism remained faithful to the pragmatist tradition in holding that instincts and habits may involve intelligence. But Ramsey showed awareness and attraction to anti-intellectualist psychology before he started to develop his pragmatism in 1926.

3 Ramsey and Anti-Intellectualist Psychology

Ramsey had been an avid reader of left-wing politics since he was a school-boy in the late 1910s (see Misak 2020, 44–46, 79, 86, 250; Marouzi 2022). He wrote a couple of economic and political essays with a heavy leftist tone in the early months of 1920. In one of them, quoting Wallas’s definition of socialism as mentioned in Human Nature in Politics (1908, 92), Ramsey appears to express his sympathy with anti-intellectualist psychology by asserting that “social instincts” are “necessary antecedents of harmonious cooperation” and the basis of “the moral and intellectual consciousness of man” (ASP.1983.01: 007-02-02). In the same essay, he comes close to endorsing the claim that instincts involve intelligence by discussing the nature of intelligence along Darwinian lines. He argues that all organisms that move and interact with their environments are intelligent to a degree—a thesis implied by the anti-intellectualism of Wallas (1908) and McDougall (1908 [1928]).⁷ In another essay written in February 1920 (ASP.1983.01: 007-02-01), we can find Ramsey following up the implications of his preferred evolutionary account of intelligence for pedagogical theory by discussing Wallas’s “A Criticism of Froebelian Pedagogy” (1901) with some details in an approving way without naming him. In fact, the extent of Ramsey’s acquaintance with Wallas’s anti-intellectualism might go deeper than what his essays suggest, for Ramsey must have had the chance to talk to Wallas in person as Wallas was an occasional visitor of the Ramseys (see Misak 2020, 10). Although Ramsey never referred to Wallas (or McDougall) in his later mature writings, it appears that it was through his reading of Wallas that the seeds of anti-intellectualist psychology were planted in his mind.

Max Lerner, a political scientist in the mid-twentieth century, once wrote that the anti-intellectualists such as Sigmund Freud share a “loose and ramshackle instinct psychology” (quoted in Kremer 2017, 21). An advocate of Freud, of course, would drop the terms “loose and ramshackle”. Nonetheless, the attribution of anti-intellectualism to Freud is apt. A lesson to be learned from Freud’s psychology is that it is our instincts and drives, residing at the bottom of our unconsciousness, which are the springs of our actions. Hence, intellectualist psychology gives us a misleading account of the motives in human action (see Deigh 2001, 1254–55). In addition, on Freud’s theory, our instinctive behaviors involve intelligence: if, for instance, certain means are not available for us to arrive at the end of sexual desire, we would be disposed to employ some other means to achieve that end.⁸ Hence, Freud’s theory could be categorized as an instance of anti-intellectualist psychology.

For Ramsey, although the utilitarian psychology had some useful applications, it was much less advanced than Freud’s psychology (Ramsey 1924 [1991], 308).⁹ In 1924, Ramsey spent six months in Vienna to be psychoanalyzed by Theodor Reik, Freud’s student and colleague, to cure his anxieties about sex (Misak 2020, 150–77). Ramsey’s letters to his mother during this period reveal his optimism about the high explanatory power of psychoanalysis in telling us about human dispositions, both at the individual and aggregate level. He believed that our views and attitudes are driven by unconscious desires and emotions: “Psychoanalysis is very important even, I think, to one’s work. You see obscure unconscious things may decide your attitude about certain things, especially personal factors in a controversial subject. Lots of work on the Foundations of Mathematics is emotionally determined…” (TFL MS/COLL/735, 3/5).

Psychoanalysis could be also used to explain certain societal tendencies, Ramsey thought. He wrote to his mother: “I’ve read a great deal of psychoanalytic literature… I’m becoming rather, an enthusiast for psychoanalysis. I’ve been reading a great book by Reik on the psychology of religion which is most awfully good” (TFL MS/COLL/735, 3/3). The book of Reik that Ramsey is referring to must be his Der Eigene und der Fremde Gott? (1923), one of the early applied psychoanalytic works that has yet to be translated into English. In it, Reik offers a psychoanalytic explanation of people’s interests in religion. Shortly after his return to Cambridge from Vienna, Ramsey became a formative member of the 1925 Psych And Society group that held weekly meetings to discuss the recent works of the Freudian tradition (Forrester and Cameron 2017, chap. 6).¹⁰ Around this time, he delivered an Apostle talk titled “Civilization and Happiness” (1925 [1991]), which was “akin to Freud’s Civilization and its Discontents, which had yet to be written” (Forrester and Cameron 2017, 404). In it he offered a psychoanalytic explanation of some of the societal tendencies of his time in Britain. In fact, Ramsey took Freud’s instinct theory so seriously so as to use it as a basis of his welfare policy advocacy in his 1923–1925 Apostle talks (see Marouzi 2022). Ramsey thought highly of Freud’s anti-intellectualism.

4 Ramsey’s Notion of Habit

Ramsey’s diary notes in January 1924, a few months before his trip to Vienna to meet Reik, reveal the early signs of his deep interest and engagement with Peirce’s pragmatism (see Misak 2020, 144). Ramsey would soon go on to develop his own version of pragmatism. Here I shall confine my discussion of Ramsey’s pragmatism to his notion of habit, a notion central to the anti-intellectualism of classical pragmatists.

What brought Ramsey to the topic of habit was his analysis of belief. In this Ramsey followed the footsteps of his pragmatist predecessors who, roughly speaking, gave us an account of belief in terms of its effects on action. It was Alexander Bain who inspired this pragmatist insight: belief involves “acting, or being prepared to act, when the occasion arises” (Bain 1872, 372, emphasis added). Peirce adopted this view and found it apt to use the concept of habit to refer to this preparedness to act that constitutes belief: belief “involves the establishment in our nature of a rule of action, or, say for short, a habit” (Peirce 1978a, 255, emphasis added).

The initial presentation of Ramsey’s pragmatism appeared in “Truth and Probability” (1926 [1990]) and “Facts and Propositions” (1927 [1990]). In both papers, we find Ramsey pointing to the tight connection between beliefs and actions: “any set of actions for whose utility p is a necessary and sufficient condition might be called a belief that p” (1927 [1990], 40). By saying that “beliefs” are “bases of possible actions”, what Ramsey asserts is that holding a belief comes with a set of dispositions to act (1926 [1990], 68). In a similar spirit to Peirce, Ramsey refers to these dispositions as habits (1929 [1990], 150). Ramsey uses “habit in the most general possible sense to mean simply rule or law of behavior, including instinct”, and he mentions his debt to Peirce as the primary source of inspiration of his discussions around habit, belief, and the nature of human inferences (1926 [1990], 90fn2). In short, Ramsey followed Peirce in arguing that “belief involves a habit or disposition to behave” (Misak 2016, 168).¹¹

Some of Ramsey’s remarks suggest that he adopted the pragmatist Humean-like account of human nature. He writes, “the human mind works essentially according to general rules or habits” (1926 [1990], 90), and “whenever I make an inference, I do so according to some rule or habit” (1926 [1990], 91). We do not only have the habits of inference, but also the habits of “observation”, “memory”, “induction”, and so on (1926 [1990], 92–93). That is, Ramsey appears to hold that habits lie at the bottom of all human actions—the thesis that puts him against intellectualist psychology.

Ramsey endorses the claim that habits involve intelligence, too, for three reasons. First, he gestures toward the idea that habitual dispositions are sensitive to situational features when claiming that holding a belief-habit does not lead one to act in a “uniform” way (1927 [1990], 44). In addition, he suggests that the heterogenous manifestations of a belief-habit might, or might not, involve an intellectual act of mind, implying that there are cases of intelligent habits that are not guided by the intellect. On Ramsey’s account, a belief-habit may dispose one to make an assertion or to judge an instance in a certain way. It may also dispose one’s body to move in a certain direction in the absence of any intellectual operations. He writes:

[M]any of our dispositional beliefs are manifested far more in our actions than in our thoughts. For instance, I have a dispositional belief (or perhaps I should rather say knowledge) that the Cambridge Union is in Bridge Street; but this belief is very rarely manifested in an act of thought; I do not often have occasion to judge that that is where the union is: I only do this when I have to inform a stranger, or just now when I thought of it for an example. On the other hand, this belief of mine is frequently manifested by my turning my steps that way when I want a book from the Union Library, which I do without any process of thought which could properly called thinking that the Union is in Bridge Street. In Oxford, I should have to think where the Union was, but in Cambridge, where I am at home, I go there habitually without having to think (Ramsey 1991b, 44–45).

The passage above suggests that there are habitual actions of the kind that serve our genuine purposes in the absence of any thoughts or judgments. In these cases, in Ramsey’s words, “habit or instinct has made the intermediate stage of judgment disappear; thought has been ‘telescoped’ away and the stimulus leads straight to action. [As in so many cases of habitual response, the intermediate stage of judgment has disappeared; in a phrase sometimes used it has been ‘telescoped’]” (Ramsey 1991b, 51, the insertion is original).¹² Ramsey illustrates his point with the following example: “‘It’s a fly!’ is a judgment; brushing it off, not” (1991b, 50). Brushing off the fly is an instance of intelligent habitual behavior that operates in the absence of any intellectual acts of mind.

Second, holding that an adequate philosophical position would take the Jamesian metaphor of “the stream of experience” seriously, Ramsey writes: “any system such as behaviorism which does not include experience is evidently wrong or at least incomplete” (Ramsey 1991a, 52). Ramsey’s worry about behaviorism seems to be two-fold. On the one hand, the strong behaviorism ignores our phenomenal experience and mental life—Ramsey, for example, holds that we have feelings of belief (1926 [1990], 65). On the other hand, pace the Watsonian behaviorist, Ramsey does not believe that “people are automata”. He says the idea that a human being is “an automaton” implies the false view that “the consciousness is really useless” (Ramsey 1991a, 68). This means that Ramsey would have sided with James in his critical debate with Huxley. For James in part rejected Huxley’s automaton theory on the ground that it implies the absurd idea that “consciousness would be useless” from the standpoint of evolution (James 1879a, 3; see Klein 2019). Thus, Ramsey’s anti-intellectualist psychology does not imply the Watsonian picture that humans are automata. From this it follows that Ramsey, contra Watson, did not conceive habitual dispositions in mechanistic terms.

Third, we can turn to the main sources of inspiration for Ramsey’s pragmatism and see what kind of views on the intelligence of habit he might have inherited from them. We have already discussed Peirce’s view. But Ramsey’s pragmatism was also influenced by Russell. He writes, “my pragmatism is derived from Mr Russell”, where by pragmatism he means the idea that “the meaning of a sentence is to be defined by reference to the actions to which asserting it would lead, or, more vaguely still, by its possible causes and effects” (1927 [1990], 51). The account of meaning that Ramsey attributes to Russell seems to capture the gist of Peirce’s pragmatic maxim, the application of which to mental concepts (such as belief) gives us an analysis of them in dispositional terms. In addition, Ramsey thinks that the kind of pragmatism that he sees in Russell’s philosophy is behaviorist-friendly: “In the Theory of Knowledge Mr. Russell’s earlier Rationalism has been considerably modified in a pragmatist or behaviourist direction” (Ramsey 1991a, 137, emphasis added). It is true that the dispositional analysis of mental content brings pragmatists close to the behaviorist treatment of mind, but we have seen that the pragmatists and Watsonian behaviorists diverge in a radical way when it comes to their views on the nature of dispositions. The joint attribution of pragmatism and behaviorism to Russell demands further explanation.

There was a branch of behaviorism in Ramsey’s time known as “purposive behaviorism” which, unlike the behaviorism of Watson, advanced the thesis that human dispositions are intelligent (Mills 1998, 33–36). Purposive behaviorism originated in the New Realism of Ralph Perry and Edwin Holt, among others, and flourished in the writings of Edward Tolman in the 1920s. The similarity between the New Realist and the pragmatist treatments of dispositions can be explained by the fact that the leading figures of New Realism “were heavily influenced by William James” and “like their mentor”, they “were pragmatists” (Mills 1998, 32). When Russell visited Harvard University in 1914 to teach two courses and to deliver the Lowell Lectures, he discussed the core ideas of New Realism and pragmatism (of both James and Peirce) with Holt, Perry, and others (Misak, forthcoming; Neuber 2023). From that point, until the late 1920s, one can find some pragmatist or (purposive) behaviorist themes in Russell’s works - such as his “On Propositions” (1919) and The Analysis of Mind (1921). Kitchener (2004) argues that this period of Russell’s career marks his “psychological turn” and his foray into behaviorism as the result of his interest in naturalized epistemology (see also Misak 2018).

Russell was not on board with Watson’s mechanistic treatment of dispositions, and, instead, advocated a view similar to that of McDougall and the New Realists on which human dispositions have the capacity to be intelligent (Kitchener 2004, 285fn11). In some of his works Russell was explicit on this point, in some others not.¹³ Thus, Ramsey’s reference to Russell’s pragmatism or behaviorism implies that he saw the conceptual similarities between Russell’s philosophy and the classical pragmatists, and that he took himself to be inspired by both. One of those conceptual similarities could be their conceptions of human dispositions which suggested that those dispositions may involve intelligence.¹⁴

5 Ramsey’s Contributions to Anti-Intellectualism

Ramsey’s notion of habit enriched the anti-intellectualist tradition. He offered an anti-intellectualist account of judgment-formation, he rejected the separation between the domain of reason on the one hand and the domain of habit on the other, and he provided the ground to explore the nature of knowledge employed in acting from habit.

5.1 Judgment-Formation

We have seen that the debates surrounding intellectualist psychology concerned the nature of typical source of motivation in human action. We shall understand “human action” in its broad sense so it includes the mental act of judgement, for the active figures in the debates took the intellectualist psychologist to make a claim about the formation-process of human judgments, among other things. McDougall, for instance, discusses two opposing views on how moral judgments are formed. First, the view held by “the intellectualists” (1908 [1928], 186) or “the traditional doctrine”, according to which “the intellectual process… on which we pass moral judgment is the primary and essential step in exerting moral judgment, and that any emotion involved in the process is the consequent on this intellectual process” (1908 [1928], 184). Second, the view that reverses the order of process by holding that “moral judgments are expressions of moral emotions”, not their consequents (1908 [1928], 184). Siding with the second view, the anti-intellectualist McDougall argues that “moral judgments are ultimately based on the emotions” (1908 [1928], 185).¹⁵

Ramsey’s account of the nature of judgement has been read against the background of the lively philosophical debates within the Cambridge of the 1920s.¹⁶ On this reading, Ramsey’s account embodied his reaction to the controversies occurring then around Russell’s writings on judgement: Russell’s multiple relation theory of judgment had it that a judgment consists of a relation between a mind and the constituents of a proposition, and Ramsey developed an improved version of the multiple relation theory in “Facts and Propositions” (1927 [1990]).¹⁷ There is no evidence suggesting that Ramsey changed his mind about the nature of judgment until the end of his short life in January 1930. Nonetheless, Ramsey’s theory of judgement was not only about the nature of judgment but also about how judgments are formed. Ramsey’s account of judgment-formation was in conversation with the debates surrounding intellectualist psychology, which went beyond the Cambridge controversies around Russell’s theories.

Ramsey holds that judgments are those “acts of thought… which have propositional reference and an affirmative character” (1991b, 45–46). He does not put any restriction on how judgements are formed. This important aspect of Ramsey’s account of judgement was a critical response to John Cook Wilson’s narrow conception of what judgements are. Cook Wilson, an Oxford philosopher, had presented his view in the posthumously published Statement and Inference (1926). In it, in Ramsey’s words, Cook Wilson holds that “judgement is a decision reached from doubt, and presupposes a preliminary process of inquiry and indecision” (1991b, 46). That is, for Cook Wilson, judgements are limited to those “cases in which we come to a conclusion after a process of reflexion” (1991b, 47). Cook Wilson was an apt representative of the early twentieth century intellectualists (Bengson and Moffett 2011, 10fn14). His account of judgement fits nicely into the intellectualist view of judgment-formation discussed by McDougall (1908 [1928], 214–16); that is, the view that judgments are formed by the intellectual processes. In response to Cook Wilson, Ramsey sets to broaden the range of instances covered by the term “judgement”:

[W]e shall use the word much more widely… Judgement in our usage presupposes no process of reflexion or weighing of evidence; we may reflect and weight the evidence before we judge but only too often we jump to a conclusion without any such process (Ramsey 1991b, 46).

Ramsey argues that judgements are formed by “rules for judging,” and that these rules are the expressions of habits of mind (1929 [1990], 149-50; 1926 [1990], 91; for further details, see Section 5.3). That is, while McDougall held that emotions play the central role in judgment-formation, the pragmatist Ramsey replaced emotions with habits.

5.2 Reasonable habits

In the last section of “Truth and Probability” (1926 [1990]), we find Ramsey exploring in what senses the word “reasonable” is used. Sometimes “to be reasonable means to think like a scientist, or to be guided only by ratiocination and induction or something of the sort (i.e., reasonable means reflective)”. We use this sense of the word “when we contrast reason and superstition or instinct” (1926 [1990], 90fn2). Nonetheless, Ramsey finds another sense of reasonableness more attractive:

We may go to the root of why we admire the scientist and criticize not primarily an individual opinion but a mental habit as being conducive or otherwise to the discovery of truth or to entertaining such degrees of belief as will be most useful… Then we can criticize an opinion according to the habit which produced it. This is clearly right because it all depends on this habit (Ramsey 1926 [1990], 90fn2).

There is no contrast between reason and habit (or instinct) in Ramsey’s anti-intellectualism. In one occasion Ramsey tells us that his account of reasonableness comes from Peirce: “Following Peirce, we predicate it [reasonableness] of a habit not of an individual judgment” (1928 [1990], 97). Peirce, in fact, once argued that even “a decapitated frog”—which is able to act in a purposive way (see Klein 2018)—“almost reasons”. Had the decapitated frog had “the power of preparatory meditation”, we could even drop “almost” and claim that the frog reasons (Peirce 1978b, 189). For Peirce, “a habit, or disposition to respond to a given kind of stimulus in a given kind of way” can be taken to involve “reasoning” (1978a, 294). The idea that reasoning can be predicated on non-intellectual elements such as habit seems to be shared by the other classical pragmatists, as well. Dewey argues that “the real opposition is not between reason and habit but between routine, unintelligent habit, and intelligent habit or art” (Dewey 1922, 55), and he blurs the distinction between reason and emotion in moral inquiries (Dewey and Tufts 1932, chap. XIV; see Henne 2020). James, too, blurs this distinction in “The Sentiment of Rationality” (1879b)—see Crippen (2018); Klein (2017). Thus, the pragmatists expand the domain of reason beyond human intellect. They predicate reasonability on non-intellectual elements, such as habits, instincts, and emotions.¹⁸

This pragmatist insight appears to be an anomaly among anti-intellectualist psychologists, for the sample of anti-intellectualist works studied by Kremer has led him to suggest that anti-intellectualist psychologists (together with intellectualist psychologists) adhered to the following classification of human behavior: “human action is motivated in one of three ways: either it stems from some non-rational factor such as instinct or emotion; or it is the result of a non-rational automatism produced by habit; or it is guided by explicit intellectual thought” (Kremer 2017, 22). That is, for Kremer, both intellectualists and anti-intellectualists shared the assumption that rationality is to be predicated only on actions guided by intellectual operations and that there is nothing rational or irrational about human instincts or habits, regardless of whether they exhibit the mark of intelligence. They all ignored “manifestations of human rationality—reasonableness—that do not depend on reasoning” (Kremer 2017, 23). But I hope I have shown that the anti-intellectualism of Ramsey (or, in general, the classical pragmatists) did not ignore such manifestations of human rationality. For Ramsey, there are instances of habitual behaviors that are reasonable which do not come with reasoning by means of intellectual operations.

Ryle (1949 [2009], 30–31) followed a path similar to that of Watson’s behaviorism in holding that habits are mechanistic acts and they do not involve intelligence. That is, he would not have agreed with the pragmatist insight that habits (which may not involve the act of intellect) can be the cases of intelligent, reasonable, or rational behaviors. Nonetheless, he introduced a distinct category of behaviors that count as intelligent ones, though they may not involve conscious consultation of the intellect: skillful behaviors.¹⁹ Kremer’s characterization of the intellectualists’ and anti-intellectualists’ accounts of reasonableness has led him to argue that a significant aspect of Ryle’s philosophy was that he showed us a third way between the two sides of the debate as he found it apt to predicate reasonableness on skillful behaviors. That is, Ryle was neither an intellectualist nor an anti-intellectualist (Kremer 2017). But if my exposition of the pragmatist notion of reasonableness is right, then it means that Ryle did not show us a third way between intellectualism and anti-intellectualism; his position must be rather understood in continuation of the kind of anti-intellectualism that one can find in the pragmatist tradition. Misak (forthcoming) presents evidence suggesting that Ryle had extensive knowledge of Ramsey’s and Peirce’s pragmatism. If so, it is plausible that Ryle took his notion of reasonableness from them.²⁰

5.3 Non-Reductive rules

We have seen that the intellectualist advances the idea that intelligent action requires to be guided by the intellectual operations. If we take those intellectual operations to consist of contemplation or consideration of a set of propositions, we may understand the intellectualist thesis as follows: what renders an action intelligent is putting the knowledge of some propositions to work. This is roughly how Ryle (1945 [2009]; 1949 [2009], chapter two) formulated the intellectualist view.²¹ Pace the intellectualist, Ryle argued that the body of knowledge employed in intelligent action resists reduction to the knowledge of some propositions.²² Ryle thus held that knowing how to act in an intelligent way cannot be reduced to knowing that such and such propositions are the case.²³ I shall argue that Ramsey’s notion of habit provided the ground for Ryle’s account of knowledge.

As mentioned earlier, Ramsey makes a distinction between what he calls “genuine judgements” and “rules for judging”. A genuine judgement is “a map of neighboring space by which we steer” (1929 [1990], 146). This map analogy illustrates two features of genuine judgments: that they have some representational content, and that they dispose us to act in certain ways. Rules for judging, on the other hand, are what Ramsey calls “variable hypotheticals”. A variable hypothetical is a generalization with infinite domain such as “all men are mortal” (1929 [1990], 145). Variable hypotheticals are the logical expressions of one’s habits and they “form the system with which the speaker meets the future” (1929 [1990], 149). It is by means of this system of habits that we make judgements. If I have the habit (x) ϕx ⊃ ψx, it means that “if I meet a ϕ, I shall regard it as a ψ” (1929 [1990], 149). These habits are the typical motives in human actions as Ramsey’s anti-intellectualist psychology suggests. Ramsey sometimes uses the term “dispositional belief function” for habits (1991b, 49–51), implying that, unlike genuine judgments (which are maps by which we steer), they do not have representational content. The content of habits consists in how they dispose us to act in the world, not in what they tell us about what the world looks like.

But in what sense do (genuine) judgements and rules (for judging) differ from each other? It seems plausible to argue that the difference between them is only the matter of complexity: if a rule is an empirical generalization with infinite domain, then it can be taken to be equivalent to infinite conjunctions of singular propositions. Then, each singular proposition would stand for a judgement. Nonetheless, Ramsey insists on a difference in kind: rules cannot be reduced to judgements. According to him, if one takes a rule such as “all men are mortal” as equivalent to infinite conjunctions of singular propositions (i.e., Socrates is mortal, and Keynes is mortal, and Wittgenstein is mortal, and so on), then the map-analogy for judgements cannot be invoked to cash out the role that rules play in actions because “if we professedly extend it [the map] to infinity, it is no longer a map; we cannot take it in or steer by it. Our journey is over before we need its remote parts” (1929 [1990], 146). Ramsey explicitly warns us that rules are not “propositions,” but are “cognitive attitudes” (1929 [1990], 147). His non-reductive treatment of rules puts him against what he had himself held in “Facts and Propositions” (1927 [1990], 48–49), where he had followed Wittgenstein’s reductionist treatment of generalizations in Tractatus (1921)—see Methven (2014).

I shall suggest that Ramsey’s non-reductive treatment of rules ought to be understood along with Lewis Carroll’s (1895) treatment of rules of inference. Given that Ryle (1945 [2009]) uses Carroll’s piece to motivate his distinction between knowing that and knowing how, this reading of Ramsey may shed some light on how he contributed to the pre-history of Ryle’s account of knowledge.

Assume that propositions A and B together entail proposition Z. Carroll (1895) considers the case when someone (say, a student of an elementary logic course) accepts A and B, but does not accept Z. That is, the student fails to see how A and B entail Z. In an attempt to teach the student the relevant lesson of logic, the teacher might ask the student to write down an additional proposition C (which states that “if A and B are true, Z must be true”) and to put C next to A and B as a new premise. Nonetheless, the student may still not see how to arrive at Z by accepting A, B, and C. Then, the student needs to be told to add another premise like D, which states that “if A and B and C are true, Z must be true”. The challenge may go on ad infinitum, and the student who failed to see how to arrive at Z by accepting A and B, may never realize how to do so by being told additional propositions such as C and D which are supposed to capture the normative force of the relevant logical rule of inference. In short, Carroll (1895) shows that a propositional explanation of the normative content of a logical rule results in an infinite regress problem for the epistemic agent. In other words, Carroll can be read to have shown that rules are not propositions; they rather operate upon propositions, and the normative aspect of that operation cannot be fully explained by a set of propositions.

Ramsey gives us a hint to conceive his non-reductive account of rules along this line: in the mixture of proposing his reasons for why rules resist reduction to judgments, he writes “cf. Mill on ‘All men are mortal’ and ‘the Duke of Wellington is mortal’” (1929 [1990], 146). The reference is to the second chapter of Book II of John Stuart Mill’s A System of Logic, where Mill argues that a syllogism begs the question. In section 3 of that chapter, Mill writes, “the proposition that the Duke of Wellington is mortal, is evidently an inference… but do we, in reality, conclude it from the proposition, All men are mortal? I answer no” (Mill 1843 [1872], 212). Mill argues that if we take the proposition “all men are mortal” among the premises of our deductive inference and if we take that generalization to be reducible to infinite conjunctions of singular propositions, then it implies that the conclusion (i.e., “the Duke of Wellington is mortal”) is already present among our premises, and so we have committed a logical fallacy in our inference, a petitio principii.

According to Mill, this whole problematic picture of inference is the consequence of “overlooking the distinction between two parts of the process of philosophizing, the inferring part, and the registering part, and ascribing to the latter the functions of the former” (1843 [1872], 213). In the next section of that very chapter, Mill writes: “All inference is from particulars to particulars: General propositions are merely registers of such inferences already made, and short formulae for making more… the conclusion is not an inference drawn from the formula, but an inference drawn according to the formula” (1843 [1872], 221). As John Woods puts it, Mill’s remarks in this passage anticipate Carroll’s regress problem and attempt to escape this problem by holding that “general propositions are ‘registers’ of inference rules and are not, as such, eligible to be premises of real inferences” (Woods 1999, 320).²⁴

According to Ryle, Carroll’s regress is generated because “knowing how to reason was assumed to be analyzable into the knowledge or supposal of some propositions” (1945 [2009], 227). That is, one who takes knowing how to be reducible to knowing that faces the puzzle introduced by Carroll (1895), for which, according to Ryle, no successful solution has been proposed (1945 [2009], 227). The way out of the puzzle is to acknowledge that “the principle of an inference cannot be one of its premisses or part of its premiss. Conclusions are drawn from premisses in accordance with principles, not from premisses that embody those principles” (Ryle 1950 [2009], 248–49).

Ramsey died shortly after he wrote down “General Propositions and Causality” (1929 [1990]) and never saw it in its published form. Nonetheless, one wonders, had Ramsey lived longer, whether he would have found the next natural move of his project to follow up the epistemological implications of his non-reductive account of rules: if one holds that the normative content of rules cannot be captured by a set of propositions, then one may argue with Ryle that the knowledge associated with applying rules (or, as Ramsey might put it: acting from habits) differs in kind from the knowledge associated with holding propositions. The germ of Ryle’s account of knowledge was present in Ramsey’s notion of rules (or habits).

6 From Ramsey to Ryle

As Kremer’s (2022) fascinating narrative suggests, it was Margaret MacDonald who passed on the distinction between knowing that and knowing how to Ryle. MacDonald was an Oxford philosopher who, with Alice Ambrose, was among the circle of Wittgenstein’s students in the 1930s who had first-hand access to Wittgenstein’s emerging philosophical ideas that were drifting away from Tractatus. Among these emerging ideas was Wittgenstein’s account of rules, which, as Kremer argues, inspired MacDonald’s account of knowledge, which in turn influenced Ryle. I shall supplement Kremer’s narrative by suggesting that, in addition to Wittgenstein and MacDonald, we should take the influence of Ramsey on Ryle’s account of knowledge seriously. Misak (forthcoming) and Kremer and Misak (forthcoming) show that MacDonald was an astute reader of pragmatism and that she was in particular an authority on the philosophy of Ramsey and Peirce. Below I briefly discuss a few of her works to show how she drew on Ramsey’s pragmatism to motivate her account of knowledge.²⁵

MacDonald defended her PhD dissertation under Susan Stebbing on the relationship between language and the world in 1934. It was titled “The Logical Characteristics of Expression” (UCL 368-2-B), with lengthy discussions around Peirce’s pragmatism and his theory of signs. At one point in her thesis, MacDonald discusses Ramsey’s reductive treatment of rules in “Facts and Propositions” (1927 [1990]), but then carefully writes that “Ramsey later changed his view of general propositions and called them ‘variable hypotheticals’, rules for constructing propositions rather than propositions themselves”, citing Ramsey’s “General Propositions and Causality” (1929 [1990]) (UCL 368-2-B: 37).²⁶ One year after defending her thesis, MacDonald wrote “C. S. Peirce on Language” (1935). In it she gives us a clear picture of Peirce’s account of rules of inference, which resembled Ramsey’s:

The Kantians said that the conclusion of a deductive argument was ‘thought with’ the premises, not explicitly but implicitly or confusedly… But… [w]hat does happen Peirce thought was much more like a game of chess in which symbols are moved according to principles of inference much as chess men are moved according to rules of chess… If it is a valid move it will be made with these pieces and in accordance with these rules (MacDonald 1935, 113–14).

For MacDonald, Peirce rejects “the descriptionist” account of laws, which takes laws to be reducible to singular propositions as if laws are “memory-saving devices”. Peirce thought this nominalist notion fails to make sense of the success of scientific predictions and so it must be replaced with an account of laws that take them as “a way of thinking or behaving, or establishment of habit” (MacDonald 1935, 125–26). In MacDonald’s words, “it is this which is the meaning of generality and not any metaphysical question about the substantial existence of universals” (1935, 128).

As Kremer (2022) shows, it was in the 1937 symposium on “Induction and Hypothesis” that MacDonald introduced her account of knowledge. MacDonald gave the lead paper in which she surveyed the rationalist and empiricist treatments of the problem of induction. Halfway through her discussion of the empiricist approach, she argued for two senses of “know”: one whose content is a proposition, and the other which applies to rules and how they are to be used in practice. What makes this MacDonald piece related to our story about the Ramsey-Ryle connection is that the empiricist treatment of induction out of which MacDonald’s account of knowledge was born was Ramsey’s (and A. J. Ayer’s, whose view was an extension of Ramsey’s treatment of empirical generalizations to all empirical propositions). MacDonald first put forward an exegesis of Ramsey’s view as follows:

Ramsey, in fact, denied that an empirical generalization or law is a proposition at all. It is a rule of conduct or guide for behavior of the form ‘If you meet an X treat it as having ϕ…’. Sometimes, like C. S. Peirce, he speaks of laws both as rules and as habits (MacDonald 1937, 25).

She then introduced her bifurcated account of knowledge:

In the sense in which believing an empirical proposition is opposed to knowing it, to believe and act on P is a way of discovering the truth or falsity of P and so of coming to know either P or not-P. But there is another sense of “know”, I think, which applies to rules but is not opposed to “believe”… “Knowing the rules” here means understanding and being able to apply them (MacDonald 1937, 26–27).

MacDonald saw the epistemological gap in Ramsey’s non-reductive account of rules and filled it in with what became the basis of Ryle’s philosophy: it is one thing to know that something is the case, and quite another to know how to apply a rule. Later on, Ryle defended MacDonald’s account of knowledge in his famous “Knowing How and Knowing That” (1945 [2009]), though with no mention of Ramsey or MacDonald.²⁷ In his subsequent writings, Ryle (1949 [2009]; 1950 [2009]; 1951 [2009]) used the analogy of inference-ticket for rules, which implied that they are not propositions (and, a fortiori, not premises of arguments), but they are rather means, similar to tickets, by which we move from the station of premises to the station of conclusion.²⁸

Kremer (2022) argues that MacDonald’s account of knowledge was inspired by Wittgenstein’s emerging account of rules in the 1930s. I do not intend to reject this claim, but I hope I have shown that we should take the influence of Ramsey on MacDonald and Ryle seriously for three reasons. First, Ramsey’s account of rules was itself an important source of inspiration for Wittgenstein’s account of rules (Misak 2023). Second, Ramsey’s anti-intellectualism had contributed to the distinction between judgments and rules, the epistemological aspect of which could be extended to motivate the distinction between knowing that and knowing how. Third, as the quoted passages from MacDonald (1937) suggests, MacDonald motivated her account of knowledge by discussing Ramsey’s and Peirce’s views. Ramsey’s anti-intellectualism influenced Ryle’s account of knowledge.

7 Concluding Remarks

The pragmatist tradition in philosophy and the anti-intellectualist tradition in psychology offer us similar accounts of human nature: human actions are ultimately guided by non-intellectual elements. Ramsey’s account of human nature was shaped by his joint interest in these two traditions. He argued that human life is generally guided by habits. This was not to suggest that human beings are automata as John Watson would say, but rather to suggest that the intelligence of human beings is largely rooted in the nature of habits. Ramsey enriched the pragmatist notion of habit and thus made three substantial contributions to the anti-intellectualist tradition. First, he sided with anti-intellectualist psychologists in arguing that human judgments are ultimately shaped by habits of mind, not intellectual processes. Second, he expanded the domain of reason beyond human intellect, arguing that reasonability could be predicated on habits. Ryle has been read as to suggest that there exists a class of rational or reasonable actions that do not involve any conscious consultation of the intellect. A similar view could be also found in Ramsey’s writings on habit, which are in the spirit of the pragmatist tradition. Third, Ramsey laid the ground to explore the nature of knowledge involved in acting from habits. The epistemological aspect of Ramsey’s notion of habit was enriched by Margaret MacDonald, who in turn came up with the distinction between knowing that something is the case and knowing how to apply a rule (or habit), the distinction that found its way to Ryle’s philosophy in the mid-twentieth century. Ramsey thus influenced Ryle’s account of knowledge through the channel of MacDonald.

Acknowledgements

I would like to express my deep appreciation to Cheryl Misak, Joseph Berkovitz, Kevin D. Hoover, Michael Kremer, and William E. Seager for their extensive and helpful comments on the earlier draft of this paper. I am also immensely grateful to Alexander Klein, Ata Heshmati, Filippo M. Sposini, Céline Henne, Allan Olley, Kobi Finestone, Mehdi Ebrahimpour, Christoph Limbeck-Lilienau, and two anonymous referees for their invaluable comments and guidance throughout this project. I also thank the audiences at Young Scholars Pragmatism Conference (University of Cambridge) on October 2021, HAPSAT Graduate Workshop (University of Toronto) on December 2021, and the annual conference of the British Society for the History of Philosophy (the Royal Society of Edinburgh) on April 2022 for their constructive feedback.