The Humean theory of motivation: much ado about nothing?

Abstract

According to the Humean theory of motivation, desire is identified as the primary source of motivation, while cognitive states like beliefs are recognized as necessary but not sufficient conditions. This paper conducts a comprehensive analysis of the established teleological argument supporting the Humean theory of motivation. The analysis finds that recent anti-Humean strategies cannot conclusively challenge the core premises of this argument. While this result may initially imply a strong and convincing defense of the Humean theory against opponents’ criticisms, it will be shown that the argument can withstand objections only through the adoption of an assumption that renders it unfalsifiable.

1 Introduction

Recent decades have witnessed increasing focus in the fields of metaethics, metanormative theory, and action theory on identifying the fundamental requirements of human behavior and motivation. At the core of this inquiry lies a central question: can an individual lacking relevant conative states at a given moment be deemed to possess a motivating reason for a specific action (Hume, 1978, p. 413).¹ This question has ignited a longstanding debate among philosophers and lies at the heart of the dispute between advocates and opponents of the Humean theory of motivation. According to this theory, an individual’s beliefs and other cognitive states serve as only a necessary, but not a sufficient, condition for motivation.² In other words, no cognitive state, whether alone or in combination with other cognitive states, can prompt action without the support of an independent desire. Advocates of this view often challenge other key tenets of Hume’s original account of motivation, including the assertion of the subject’s infallibility in recognizing psychological states (Hume, 1978, p. 189; Qu, 2017), the concept of reason as “the slave of passion” (Hume, 1978, p. 415), and the idea that passions cannot be evaluated by reason (Hume, 1978, p. 416).³

Contemporary philosophical discourse typically retains Hume’s traditional account of motivation either as a negative assertion, positing that beliefs and other purely cognitive states are insufficient to motivate the subject to act (Nuyen, 1984, p. 26; McNaughton, 1988, p. 22; Shaw, 1989, p. 163), or as a positive assertion, contending that the subject’s motivation necessitates the presence of desire or some other conative state (Gregory, 2021, p. 27). All adherents of the Humean theory of motivation also maintain Hume’s proposition that beliefs and desires are distinct entities or conceptually independent psychological states. This implies that, for any given belief-desire pair, it is conceivable that the subject possesses one component without the other (Smith, 1994, p. 7).

The Humean conception of motivation, once broadly embraced as philosophical dogma (Smith, 1987, p. 36), has faced significant criticism in recent years, leading to the exploration of various argumentative strategies. These critiques include attempts to refute (a) the central claim of the Humean view, asserting that beliefs cannot motivate without the assistance of an independent desire, (b) challenges to the thesis positing that beliefs and desires are conceptually independent psychological states, and (c) arguments proposing the existence of psychological states known as “besires”, which purportedly simultaneously exhibit the representative characteristics of beliefs and the motivational characteristics of desires (Zangwill, 2008, p. 51).⁴

In this paper, we will examine the Humean theory of motivation through the lens of one of its highly influential arguments, which has achieved its well-known and developed form in the works of Smith (1987, p. 55; 1989; 1994, p. 116). This argument, widely recognized as the teleological argument (see Wallace, 1990, p. 359; Coleman, 2008, p. 128), can be summarized as follows:

• (1) To have a motivating reason, inter alia, is to have some goal.

• (2) Psychological states, such as beliefs and desires, differ in terms of their direction of fit.

• (3) The direction of fit can be explained in terms of the functional roles of psychological states.

• (4) The belief that p is a state that tends to go out of existence in the presence of a perception with the content that not p.

• (5) Desire that p is a state that tends to endure in the presence of a perception with the content that not p, disposing the subject in that state to bring it about that p.

• (6) There are no psychological states that simultaneously instantiate both directions of fit.

• (7) To have a goal is the same as being in a state that tends to persist in the presence of a perception with the content not p, predisposing the subject in that state to bring about the realization of p.

• (8) Therefore, having a goal is the same as being in a psychological state with a direction of fit characteristic of desires.

• (9) Therefore, to have a motivating reason, inter alia, is to have some desire.⁵

More concisely, the argument asserts that having a motivating reason equates to having a goal, which in turn entails being in a psychological state with a direction of fit characteristic of desires. Beliefs and desires differ based on their direction of fit, explained by their functional roles. Specifically, the belief that p tends to be rejected in the presence of the perception that not p, while the desire that p tends to persist and motivate the subject to make p the case in the presence of the perception that not p. Consequently, a psychological state cannot have both directions of fit simultaneously. Therefore, having a motivating reason is essentially having a desire.⁶

The primary objective of this paper is to conduct a comprehensive analysis of the teleological argument presented above, focusing particularly on its four key premises: (2) asserting that beliefs and desires differ in terms of their direction of fit, (3) explaining this direction of fit in terms of the dispositional role of psychological states, (5) holding that the essential characteristic of the subject’s desire is that it prompts them to act, and (6) claiming that there are no psychological states with both directions of fit.

While acknowledging the profound influence of the teleological argument in contemporary philosophical discourse, this paper endeavors to demonstrate that some of the most formidable objections raised by anti-Humean authors can be effectively disarmed by adopting the assumption about the nature of motivation, which suggests that a subject’s motivation to perform an action φ at time t may stem from desires of which the subject is not consciously aware, and that the subject may be mistaken in identifying the relative strength of their own motivation. However, the concluding section of this paper will underscore that, while this assumption plays a pivotal role within the teleological argument, rendering it unfalsifiable, it should not be viewed as an insurmountable limitation of the Humean theory of motivation, for neither the teleological argument nor the aforementioned assumption constitutes integral and indispensable components of this philosophical position.⁷

2 The direction of fit: a challenge to premise [2]

Premise (2) asserts that the distinction between psychological states, such as beliefs and desires, can be explained by their direction of fit. However, the concept of direction of fit faces challenges from several influential objections, casting doubt on its reliability as a fundamental criterion for distinguishing between belief and desire. This section will outline these objections and analyze their implications for Smith’s defense of the Humean theory. However, before we delve into such an analysis, a brief historical clarification is in order.

One of the early foundations of the contemporary notion of direction of fit can be traced back to Hume’s understanding of the nature of human psychology (see Little, 1997, p. 60). Hume argued that there are two main types of psychological states – beliefs and desires (Smith, 1994, p. 7) – to which we can reduce all other intentional states. He further asserted that beliefs and desires maintain a contingent relationship, viewing them as distinct psychological entities. Specifically, Hume viewed the disparity between these states, as well as their mutual independence within an individual’s psychology, in the fact that beliefs are evaluable in terms of truth and falsity, and that only beliefs can be deemed susceptible to criticism as irrational, whereas desires, in line with his anti-rationalistic stance, are considered exempt from truth conditions or rational criticism (Hume, 1978, p. 415). While Hume identified the basic division into two main groups of psychological states, corresponding to the current classification of different directions of fit, the metaphor of direction of fit took its distinctive shape in the second half of the 20th century. Searle and Vanderveken (1985, p. 85) attribute the term “direction of fit” to Austin (1962), although the idea, in its rudimentary form, was initially introduced in Elizabeth Anscombe’s influential monograph Intention (Anscombe 1957/2000, p. 56). This idea was conveyed through an anecdote about a man shopping with a list of items and a detective recording his purchases.

Despite the growing body of literature on the concept of direction of fit, contemporary authors often refer to the following passage from Mark Platts as the quintessential interpretation of this concept:

Miss Anscombe, in her work on intention, has drawn a broad distinction between two kinds of mental state, factual belief being the prime exemplar of one kind and desire a prime exemplar of the other (Anscombe, Intention, §. 2). The distinction is in terms of the direction of fit of mental states with the world. Beliefs aim at the true, and their being true is their fitting the world; falsity is a decisive failing in a belief, and false beliefs should be discarded; beliefs should be changed to fit with the world, not vice versa. Desires aim at realisation, and their realisation is the world fitting with them; the fact that the indicative content of a desire is not realised in the world is not yet a failing in the desire, and not yet any reason to discard the desire; the world, crudely, should be changed to fit with our desires, not vice versa. (Platts 1997, pp. 256–257, emphasis added)

According to the conventional understanding of Platts’ metaphor, the idea that beliefs have a mind-to-world direction of fit suggests that when our beliefs fail to align with the factual state of affairs in the world, the subject must revise or discard those beliefs to bring them in line with reality – in other words, to fit them into the world. Conversely, the idea that desires have a world-to-mind direction of fit implies that when our desires do not correspond with the factual state of affairs, the subject should alter that state of affairs to align it with their desires – that is, to enact changes in the world to fulfill the propositional content of their desires.⁸

With these clarifications in mind, let us now embark on a more nuanced examination of the complexities inherent in this issue. The metaphor of direction of fit encounters several problems. Firstly, regarding the direction of fit of beliefs, it remains unclear whether the subject should align their beliefs with the world (as proposed by Platts), or if beliefs inherently tend to align with the world, or if there is another mechanism at play (see Gregory, 2012, p. 604). The main difficulty here stems from the fact that it is questionable, to say the least, whether the subject can have voluntary control over the adoption, rejection, and modification of their beliefs. According to the traditional view of doxastic voluntarism – espoused by Augustine, Aquinas, Pascal, Descartes, James, and many others – the subject has the same kind of voluntary control over their beliefs as they do over their actions. In other words, the subject can decide whether to accept or reject a certain belief in the light of evidence. In contrast, Thomas Hobbes argued that the subject’s beliefs are entirely determined by valid arguments and evidence, with the will exerting no influence whatsoever on the adoption or rejection of these states (see Hobbes 1641/2006, obj. 13). This viewpoint, known as doxastic involuntarism, is now widely accepted by the majority of philosophers (e.g. Alston, 1988; Bennett, 1990; Buckareff, 2004, 2006; Pojman, 1999; Williams 1973). Note, however, that embracing the stance of doxastic involuntarism presents a significant challenge for the concept of direction of fit, as it denies the subject’s ability to revise or reject their beliefs when they conflict with the factual state of affairs. Consequently, the metaphor of direction of fit would need to be reevaluated, as the direction of fit for beliefs would not be achievable under this framework.

Another challenge to the concept of direction of fit is addressed in Frost’s extensive study (2014), which argues that there are discernible differences in how beliefs ‘tend towards truth’ and desires ‘tend towards realization’, undermining the notion of direction of fit as a symmetrical relation. This difficulty is particularly evident in Platts’ metaphor. Namely, Frost contends that, while Platts attempts to present the two directions of fit in a completely symmetrical fashion, it is difficult to avoid the impression that this symmetry is unnatural, as it can only be said in a figurative and non-literal manner that the world should be changed to fit our desires (Frost, 2014, p. 433). However, within an acceptable and successful explanation of direction of fit, figurative phrases should not find their place. Frost refers to such explanations as committed theories, asserting that they must satisfy the following three conditions: (a) there exists a unified type of “fitting” relationship inherent to the theory of direction of fit, (b) this relationship exhibits a strict symmetry, justifying the assertion that there are two and only two directions of fit, and (c) the combination of these conditions jointly clarifies the structure and nature of psychological states, thus justifying the use of the technical term “direction of fit” (2014, p. 441). Frost claims that no explanation of direction of fit proposed so far has managed to satisfy all three conditions because they either rely on metaphorical language, fail to preserve the symmetry of the direction of fit relation, or seek to reduce the concept of direction of fit to other concepts. Unlike earlier explanations – e.g., Platts (1997) and Searle (1983, pp. 7–8) – which typically retain metaphorical components, the most common issue with recent analyses of direction of fit is their failure to satisfy condition (c). For example, Smith’s (1994, p. 118) analysis essentially reduces this concept to the dispositional role of psychological states, Gregory (2012) explains it by reducing it to objective and subjective reasons,⁹ and Archer (2015) tries to explain it in terms of the inferential properties of psychological attitudes. While Frost’s objections are undoubtedly serious, they still do not demonstrate why a completely reductive and uncommitted analysis of the concept of direction of fit could not function as a component of the teleological argument and, consequently, as a sufficiently good conceptual support for the Humean theory of motivation. The absence of sufficient grounds within these analyses to retain the technical term “direction of fit” does not necessarily undermine its explanatory value.

Finally, Frost challenges the supposed opposition between the mind and the world implied by the concept of direction of fit. In the case of theoretical attitudes – such as beliefs and other cognitive attitudes – it makes sense to speak of the mind as an internal factor representing something and the world as an external factor being represented. However, Frost argues that in the case of practical attitudes – such as desires and other conative attitudes – the representation of the world as “external” stems from the attempt to achieve symmetry in the two directions of fit, but is not plausible because the “world” here is not some “non-mental piece” or something “other” or “external” to the subject (Frost, 2014, pp. 460–461). Proponents of the concept of direction of fit could counter Frost’s objection by appealing to the traditional view that the object of desire cannot be something that the subject already possesses (Lauria & Deona 2017, p. 4). For example, if subject S already has phone x, it is impossible for S to desire to obtain x; if S has quit smoking, it is impossible for S to desire to stop smoking, and so forth. These examples demonstrate that the object of desire (and other conations) is inaccessible to the subject, or something they do not currently possess. Thus, Frost’s criticism can be circumvented, as there is a clear sense in which the objects of our desires are “external” to the subject.

The discussion above raises doubts about the credibility of premise (2) and, consequently, the validity of Smith’s argument. However, these concerns are not insuperable barriers for either Smith’s argument or the Humean theory of motivation. This is because premise (2) asserts that the distinction between belief and desire can be articulated via their direction of fit, without specifying any additional qualifications of this concept. As previously discussed, proponents of the Humean theory, aiming to reinforce their position with the concept of direction of fit, may present it through positions classified as uncommitted by Frost, which offer a replacement of this concept rather than its full explication (Frost, 2014, p. 441). Such uncommitted positions include those that view direction of fit as the dispositional role of psychological states (as proposed by Smith), or, with the necessary modification of premises (3), (4), and (5), positions that explain this concept in terms of subjective and objective normative reasons (Gregory, 2012), or, alternatively, positions that interpret it in terms of the inferential properties of psychological attitudes (Archer, 2015). Considering this, even though the current understanding of direction of fit is more critical compared to the views of philosophers three decades ago when Smith first articulated his teleological argument, there are no conclusive reasons to reject this argument by challenging premise (2).¹⁰

3 The functional roles of psychological states: a challenge to premise [3]

The preceding section has outlined significant critiques of the notion of direction of fit, a pivotal component of Smith’s argument in support of the Humean theory of motivation. While some readers may be inclined to dismiss this argument due to the aforementioned criticisms, gaining a more comprehensive understanding of its persuasiveness requires temporarily setting aside such objections and proceeding to examine its premise (3). This premise presents Smith’s proposal for the dispositional characterization of psychological states, wherein the concept of counterfactual dependence occupies a central position. This proposal seeks to eliminate the controversial and metaphorical nature of the widely acknowledged philosophical distinction between belief and desire, which is traditionally based on direction of fit. Smith states that

… the difference between belief and desire in terms of direction of fit can be seen to amount to a difference in the functional roles of belief and desire. Very roughly, and simplifying somewhat, it amounts, inter alia, to a difference in the counterfactual dependence of a belief that p and a desire that p on a perception with the content that not p: a belief that p tends to go out of existence in the presence of a perception with the content that not p, whereas a desire that p tends to endure, disposing the subject in that state to bring it about that p. (Smith, 1994, p. 115; see also 1987, p. 54)

Even if we disregard Frost’s criticism that Smith’s analysis of the direction of fit reduces this concept to something else, two significant issues persist with this approach. Firstly, there is the challenge of providing a satisfactory explanation of the concept of “perception”, which is central to Smith’s analysis. If we consider perception to be a type of belief, Smith’s explanation of direction of fit becomes circular, as the concept of belief is used to analyze and ultimately reduce both directions of fit.¹¹ Alternatively, if we instead consider perceptions as appearances or seemings, Smith’s explanation is obviously incorrect. This is because it is not always the case that we reject our beliefs based on what appears or seems false to us. For example, despite objects around us appearing colorless or black when the light is off, we do not typically believe this to be the case. Similarly, we are not inclined to believe that a stick submerged in water becomes bent, despite the illusionary appearance (see Gregory, 2012, p. 604). In summary, the inclusion of the concept of perception in Smith’s analysis of the direction of fit results in either circularity or a fundamentally flawed conclusion.

The second issue with Smith’s analysis arises from the fact that there are numerous instances of belief that p wherein the subject is not inclined to reject it in light of a perception with the content that not p, as well as cases of desires which the subject rejects in light of a perception with the content that not p. In their critique of Smith’s position, Sobel and Copp cite several such examples (Sobel & Copp, 2001). For instance, subject A persists in their belief in the existence of a God who is omniscient, omnipotent, and perfectly good, despite being aware that the existence of such an entity is incompatible with the presence of evil in the world. Subject B believes that their lottery number will be drawn tonight, despite never having won any games of chance before. Subject C believes that it will not snow tonight, despite the weather forecast stating otherwise and being considered a reliable source (2001, pp. 47–48). Should we conclude that the subjects in these examples do not actually believe that p, but rather desire that p, given that they retain their beliefs despite their perceptions with the content that not p, aligning with the direction of fit characteristic of desire in Smith’s analysis? Sobel and Copp argue against this. Namely, it would be highly unusual to claim that the psychological states of the subjects in the aforementioned examples are actually their desires rather than their (irrational) beliefs. And the fact that such psychological states do not align with Smith’s account of the direction of fit of beliefs calls into question the persuasiveness of his account rather than leading us to no longer consider them a type of belief. Moreover, there are instances that challenge Smith’s explanation of the direction of fit of desires. Consider the case of a basketball fan who passionately desires their team, Team X, to win. However, this fan abruptly stops cheering for Team X and vehemently rejects their desire for the team to win as soon as the team starts losing (2001, p. 48). This scenario presents a psychological state that mirrors the functional role typically associated with beliefs according to Smith’s framework. Given this example, one might question whether we should conclude that this fan actually believes that Team X will win while cheering, rather than desiring their victory (2001, p. 48). It seems that, according to Smith’s analysis of direction of fit, we would need to embrace such a conclusion. Once again, this conclusion speaks far more convincingly against Smith’s analysis than it does to classify the psychological state of the fan in the example as belief.

The objections raised in this section may suggest that Smith’s analysis of the direction of fit in terms of the functional roles of beliefs and desires undermines his argument in favor of the Humean theory of motivation. However, adopting an anti-Humean position based on these objections would be premature and unjustified. It remains possible to modify premise (3) to rely on a more plausible analysis of direction of fit that addresses the previously mentioned criticisms. As discussed earlier, compelling explanations of direction of fit include situating it within the framework of objective and subjective reasons (Gregory, 2012) or characterizing it in terms of the inferential properties of psychological attitudes (Archer, 2015). While these accounts may not strictly align with Frost’s concept of a committed theory, they offer promising avenues for fulfilling the explanatory role in a satisfactory and compelling manner. Thus, while anti-Humeans can effectively challenge Smith’s analysis of direction of fit, they may not conclusively prevail in the ongoing debate over the proper determination of the necessary and sufficient conditions for motivating reasons. In short, the objections presented thus far have underscored weaknesses in Smith’s argument, but have not demonstrated the insurmountable shortcomings of the Humean theory of motivation. A conclusive refutation would necessitate an argument showing the inherent futility of all attempts to explicate the concept of direction of fit. However, it remains uncertain whether such an argument will ever be formulated.

4 Desires and motivation: a challenge to premise [5]

Recall that premise (5) characterizes the desire that p as a psychological state that persists in the presence of perception with the content that not p and motivates or prompts the subject to bring about the case that p. This motivational conception of desire is directly derived from Hume’s understanding of passions, which holds that desire is a psychological state that is fundamentally characterized by its ability to motivate the subject (Sinhababu, 2009, p. 468). Philosophers adhering to this view use the term “desire“ generically to refer to any motivating state (e.g. Nagel, 1970; Schueler, 1995; Armstrong, 1968; Stampe, 1986; Stalnaker, 1984; Smith, 1994; Dretske, 1988; Dancy, 2000; Millikan, 2005). This view considers desire equivalent to the term “pro-attitude” introduced by Davidson (1963), encompassing a wide range of urges, compulsions, emotions, and other psychological states distinguished by their property of motivating the subject (Shaw, 2021a, b).

This section does not address the issue of whether the concept of desire should be clearly distinguished from other pro-attitudes, as such discussions extend beyond its scope. Instead, it critically analyzes two philosophical strategies aimed at challenging premise (5). The first strategy, termed the empirical strategy, involves presenting actual cases of individuals who lack the motivation to act in accordance with their desires due to specific psychological influences. The second strategy, referred to as the conceptual strategy, is based on an “affective conception” of desire as expounded by Smithies and Weiss (2021), suggesting a possible dissociation between desire and motivation. The objective of this section is to demonstrate that these two strategies are insufficient to offer conclusive evidence for disputing premise (5). Each strategy will be examined in sequence.

4.1 Empirical strategy

In contemporary philosophical literature, a significant number of authors have endeavored to substantiate the thesis of the dissociation of the relationship between desire and motivation by drawing on cases from clinical practice, with depressed individuals’ reports being the most commonly utilized (see e.g. Roberts, 2001; Law, 2009; Ratcliffe, 2010; Smith, 2013).¹² In this section, our focus is on Swartzer’s (2015) analysis of the Humean theory of motivation, which incorporates the testimonies of two patients with clinical depression. These patients report their inability to act in accordance with their desires, indicating a complete lack of motivation to carry out the actions they desire (Swartzer, 2015, pp. 8–9). These and other similar examples lead to a clear anti-Humean conclusion as they provide empirical evidence that challenges premise (5), and consequently, the Humean theory of motivation. The theory, in this respect, is believed to capture the relationship between desire and motivation too narrowly by either asserting that desire is identical to motivation or that the degree of motivation is on par to the degree of desire. Swartzer concludes that this understanding of cases of clinical depression

… poses a problem for those who would insist on too tight a connection between desire and motivation. In particular, this understanding gives us reason to doubt that being motivated to φ is identical to desiring to φ, or that strength of motivation is, of necessity, identical to degree of desire. For it seems quite coherent to say that [the depressed individuals] strongly desire or strongly care about doing things that, because of their depression, they lack motivation to do. Whatever motivation is, it seems to be at least conceptually separable from desire. (Swartzer, 2015, pp. 9, emphasis added)

However, it is noteworthy that proponents of the Humean position possess argumentative strategies that challenge the conclusion regarding the conceptual separability of desire and motivation. Swartzer, in particular, commits a conceptual error by confusing the subject’s incapacity to act with the subject’s lack of motivation. Swartzer’s reference to depressed individuals articulating their inability to fulfill desires does not provide sufficient evidence to establish that these individuals lack motivation to act in accordance with their desires. Thus, Swartzer’s conclusion, asserting that motivation is conceptually independent of desire, lacks adequate support. The desires of these individuals may be accompanied by motivation, but the effect of depression on them is so strong that their motivational influence is practically imperceptible.¹³ It is crucial to note that the undetectability of a psychological factor does not equate to its complete absence. Therefore, the motivational impact of the desires of depressed individuals cited by Swartzer may be so weak that it leads them to mistakenly believe they possess a desire lacking any accompanying motivation.¹⁴ In fact, Swartzer himself admits that the anti-Humean conclusion that could be drawn based on these examples is not decisive:

[A] Humean might insist that when one’s motivation is so slight that there is little chance of effectively leading to action, it might be easy to mistakenly think of oneself as having no motivation whatsoever. For these reasons, the Humean might be skeptical that these cases provide us decisive examples of motivationless desires. (2015, p. 11, emphasis added)

The counterexample of premise (5) analyzed above is based on the direct influence of depression on the relationship between desires and motivation. Yet, the same model can also be applied to other counterexamples of this premise that rely on various psychological factors, such as apathy, psychological or physical fatigue, and others (see e.g. Milevski, 2017). In summary, the first strategy involves citing clinical cases that demonstrate a dissociation between the subject’s desire and corresponding motivation. It turns out, however, that this approach is insufficient to successfully challenge premise (5) and, consequently, cannot be used to undermine the Humean conception of motivation.

4.2 Conceptual strategy

We will now examine the second strategy for challenging premise (5), which is based on the relatively new concept of desire. This concept has been developed in a recently published study by Declan Smithies and Jeremy Weiss (2021), building upon Ruth Chang’s (2004) thesis on affective desires. Smithies and Weiss advance a thesis regarding “basic” desires, which are desires that are not attributed to a subject on the basis of any other belief or desire (Smithies & Weiss 2021, p. 45). They suggest that basic desires are dispositions to cause the phenomenal character of the affective experience of desire (2021, p. 49), To distinguish and identify basic desires, the authors list two categories of affective experience, namely, feelings of attraction and aversion (2021, p. 27). They further contend that these experiences are intentional and conscious episodes. It is important to point out that Smithies and Weiss differentiate attraction and aversion from hedonic phenomena, such as pleasantness and unpleasantness, because a subject may be attracted to the prospect of an action or outcome without necessarily finding it pleasant (2021, p. 44). In accordance with their conception, when an individual desires to perform an action, they have a disposition or inclination to experience a particular type of affective experience. This may involve feeling attracted to the prospect of performing a particular action or feeling a sense of aversion or repulsion towards the idea of not performing that action.¹⁵

Now, the fundamental feature of this conception, which presents a challenge to the Humean theory, is the supposition that desires, as construed, furnish us with normative reasons for acting due to their dispositional nature towards affective experiences. Namely, desires that provide normative reasons for action possess a characteristic that permits a dissociation between desire and motivation. This results from the indisputable fact that subjects do not invariably act in harmony with their subjective normative reasons. Hence, these authors reject the commonly held belief that motivation is the defining aspect of desire, known as the motivational theory of desire (2021, pp. 49–50). Instead, they assert that the essential feature of desire is the presence of a subjective normative reason or affective experience, rather than just motivation. Smithies and Weiss point out this as follows:

[W]e sometimes have good reasons for acting that we’re not motivated to act upon. Indeed, we’re not always capable of acting on the good reasons we have – say, to remain calm in the face of adversity or to stop obsessively thinking about things that make us upset. If we were fully rational, then of course we would always be motivated to act on the balance of reasons. Sadly, however, we’re not fully rational. Perhaps there are principled limits on how much irrationality is compatible with having beliefs and desires at all, but that goes beyond anything we’ve argued for here. In any case, it’s not just a remote possibility, but an undeniable fact about the actual world that we’re not always motivated to act rationally on the basis of our reasons. (2021, p. 52, emphasis added)

The italicized segments of the paragraph suggest that, similar to Swartzer, Smithies and Weiss confuse behavior with motivation. Even when a subject is unable to align their actions with subjective normative reasons, as illustrated in their example where the subject struggles to stop obsessively recalling past traumatic experiences, it does not necessarily imply a lack of motivation to cease such behavior. This is exemplified by a scenario where the subject earnestly explains to their psychiatrist that they are genuinely trying to stop dwelling on past traumas, employing various techniques, but are unsuccessful. This portrayal corresponds with the idea that an individual’s normative reasons are typically accompanied by motivation, which may be overridden but not entirely nullified.

Upon closely examining all the evidence presented thus far, it becomes apparent that the critical analysis of both empirical and conceptual strategies falls short of offering conclusive reasons to dismiss premise (5). It is crucial to note that this finding should not be construed as a definitive victory for Humean motivation theory; rather, it indicates that this form of argumentation is inadequate in conclusively undermining the theory.

5 Psychological states with two directions of fit: a challenge to premise [6]

As is well known, within the traditional paradigm concerning the nature of human psychology, originating from Hume’s views, there is no necessary and indissoluble connection between cognitive and conative states. For this reason, the majority of authors who still adhere to this view decisively reject the possibility of the existence of unitary psychological states – known as “besires” – which would simultaneously instantiate the representational features of beliefs and the motivational characteristics of desires, and consequently, which would have both directions of fit (e.g., Ridge, 2008, p. 53). Smith contests the possibility of besires, relying on his account according to which the direction of fit is explained in terms of the dispositional roles of beliefs and desires, as we have discussed in detail in Sect. 2, asserting that:

[T]hough it might sound like a coherent possibility that there be such a state, it isn’t really, at least not if we take the suggestion quite literally. For, as we have understood the concept of direction of fit, the direction of fit of a state with the content that p is determined, inter alia, by its counterfactual dependence on a perception with the content that not-p. A state with both directions of fit would therefore have to be such that, both, in the presence of such a perception it tends to go out of existence, and, in the presence of such a perception, it tends to endure, leading the subject who has it to bring it about that p. Taken quite literally, then, the idea that there may be a state having both directions of fit is just plain incoherent. (Smith, 1994, p. 118)

At first glance, Smith’s conclusion seems well-founded. A state with a “two-way direction of fit” would simultaneously instantiate fundamentally opposing dispositional roles. The assumption of a dual direction of fit was originally introduced to support the thesis that at least some of our beliefs (e.g., moral or normative ones) can be motivationally effective without the assistance of independent conative states. However, as Smith observes, if such a state – for example, a besire that p – simultaneously tends to go out of existence and tends to endure when the subject is confronted with evidence that not p, it is inevitable to conclude that it not only fails to motivate the subject to act but also raises questions about its very coherence as a psychological state.

Despite the initial persuasiveness of Smith’s conclusion, many contemporary philosophers do not consider such states problematic. Some philosophers propose that our moral and normative beliefs are essentially besires (Little, 1997; Bedke, 2009). There are authors who extend this argument and suggest other psychological states as plausible candidates for besires. For instance, Swartzer contends that appetitive desires such as hunger, thirst, and urges are forms of besires (Swartzer, 2013). Bromwich goes even further by asserting that all beliefs are besires (Bromwich, 2010), and Fileva maintains that the state of hope constitutes a clear instance of besires (Fileva, 2021). To assess the plausibility of premise (6), which challenges the coherence of psychological states with two directions of fit, this section will exclusively focus on the arguments presented by Danielle Bromwich, although the conclusion reached here will apply to all current and future positions within which the possibility of the concept of besires is asserted.

In her study (2010), Bromwich articulates an ambitious attempt to challenge the Humean theory of motivation, demonstrating that the concept of besire is not incoherent. More specifically, she accepts Smith’s account, which explains the direction of fit in terms of the dispositional roles of beliefs and desires, but argues that the conclusion Smith draws from it does not actually follow (2010, p. 345). While acknowledging that Smith is correct in asserting that a state with both directions of fit regarding the same content would be incoherent, Bromwich suggests that a belief that p could still be a state that tends to dissipate in the presence of a perception with the content that not p and yet prompts the subject to bring about q, where q represents a different propositional content. Bromwich thus explains the coherence of a psychological state with both directions of fit as follows:

A belief that p could be a state that tends to go out of existence on the perception that not-p and yet be a state that tends to dispose the subject to bring it about that q, where q is not the same content as p. If I believe that ‘Adultery is wrong’, I would give this belief up if I came across a decisive reason to think that it is false. I would give this belief up on the perception that not p. But notice, in the absence of such a reason, my belief does not dispose me to bring it about that p: it does not dispose me to bring it about that adultery is wrong […] my belief disposes me to q, when I believe that p: it disposes me to be faithful, endorse faithful relationships, condemn adulterers, and so on. (Bromwich, 2010, pp. 345–346)

The point raised by Bromwich is valid. However, this point does not provide conclusive reasons for drawing a much more ambitious conclusion that challenges the Humean theory of motivation, which Bromwich attempts to derive in her study. The main difficulty in her anti-Humean argument arises from the fact that her understanding of the concept of direction of fit remains within the framework of Smith’s account, which, as we have already seen, faces serious problems. However, if we were to adopt a different account of direction of fit, such as the one advanced by Archer (2015), according to which it is understood as a set of inferential properties of psychological attitudes, Bromwich’s anti-Humean conclusion would simply not follow. Archer’s account explicitly rules out the possibility of psychological attitudes having both directions of fit, as such attitudes logically cannot simultaneously instantiate inferential characteristics that make them suitable for inclusion in classically valid inferences, and inferential characteristics that would exclude them from such arguments (Archer, 2015, pp. 176–177). Bromwich’s conclusion thus faces a paradoxical outcome: although she correctly concludes that the concept of a psychological state with both directions of fit is not incoherent within Smith’s account, her anti-Humean position can only be sustained if Smith’s problematic account of direction of fit is retained. Yet, if we replace Smith’s account with Archer’s, we once again arrive at the conclusion that the idea of two-directional psychological states is inherently incoherent.¹⁶

In summary, Bromwich’s thesis, advocating the conceptual coherence of psychological states with two directions of fit, is excellently presented and defended. However, this thesis seems accurate only within the framework of a fundamentally flawed account of direction of fit, such as Smith’s. However, in the context of more plausible theories, her thesis is evidently false. If we reject Smith’s account of direction of fit – a decision supported by the reasons outlined in Sects. 2 and 3 – and opt, for instance, for Archer’s account, we would find it necessary to dismiss not only Bromwich’s anti-Humean solution but also all endeavors to defend the anti-Humean position by invoking the notion that psychological states can simultaneously exhibit both directions of fit.

6 Concluding remarks

The examination of Smith’s teleological argument has revealed that four of its most contentious premises resist conclusive refutation. This outcome could readily suggest that the argument offers a compelling, straightforward, and potent means of defending the Humean theory of motivation against criticisms raised by its detractors, as explicitly asserted by Smith (1994, p. 116). Despite Smith’s optimistic stance toward his argument, the analysis presented in this study strongly suggests a conclusion that goes in the opposite direction. Specifically, we have observed that the primary shortcoming of premises (2), (3), and (6) stems from the inadequacies inherent in Smith’s treatment of the concept of direction of fit. Precisely because these inadequacies impact Smith’s account, it has been proposed that these premises could be defended against anti-Humean objections through a relatively straightforward maneuver. This entails relying on an alternative explanation of this concept that is not afflicted by the same deficiencies as Smith’s.

However, upon examining premise (5), it has been revealed that the teleological argument can be salvaged only if one adopts the assumption that the subject’s motivation to φ at time t may be derived from desires that the subject is not consciously aware of at that time, and that the subject may be mistaken in identifying the relative strength of their own motivation.¹⁷ While this assumption is compelling, as demonstrated in the course of this study, it constitutes the primary source of difficulties encountered by the teleological argument. This is because the assumption effectively invalidates any counterarguments presented by critics of the central thesis of the Humean theory of motivation, irrespective of their initial persuasiveness, ultimately ascribing an unfalsifiable character to that thesis. It is crucial to emphasize, however, that this point should not be interpreted as indicative of an unavoidable and inherent flaw in the Humean theory of motivation. Instead, it highlights a significant shortcoming of the teleological argument, though not necessarily of other arguments employed in defense of this theory.¹⁸ Neither the teleological argument nor the mentioned assumption are indispensable components of the Humean theory. Consequently, I propose concluding this study with a negative inference that is, nonetheless, restricted to the teleological argument. This approach allows for the possibility that other arguments might prove significantly more successful in defending the Humean theory.