Hume's Social Epistemology and the Dialogue Form

Abstract

Hume begins his Dialogues Concerning Natural Religion by discussing what an ideal dialogue ought to look like. Many considerations that Hume raises coincide with similar concerns in contemporary social epistemology. This paper examines three aspects of Hume's social epistemology: epistemic peerhood, inquiry norms and the possibility of rational persuasion. Interestingly, however, I will argue that the conversation between Philo, Cleanthes and Demea falls short of meeting Hume's articulated standard of what an ideal dialogue ought to look like. From this analysis, I defend the less popular view that Demea's decision to leave the conversation (in Part XI) was entirely reasonable and suggest an explanation for why Hume decided to make Cleanthes the ‘hero’ of the Dialogues.

1. Introduction

Historians of philosophy and social epistemologists have emphasised the importance of the dialogue form in our pursuit of our epistemic goals.¹ Suppose you disagree with an epistemic peer over some proposition, and that both you and your epistemic peer are concerned about determining whose belief is true, engaging in dialogue would then be a natural outcome of peer disagreement. But what should a productive philosophical dialogue look like? Are there norms that should govern philosophical conversations, such that participants are treated with respect, and the conversation moves in an epistemically and ethically fruitful direction? Throughout the history of philosophy, many philosophers have employed the dialogue form as a useful inquiry mechanism for pursuing epistemic goals. One such notable figure is David Hume.

Hume's Dialogues Concerning Natural Religion is widely recognised as an influential work in the philosophy of religion. However, while much of the attention is often given to Hume's arguments about the nature and existence of God (found in Parts I to XII of the Dialogues), there is much we may learn from Hume's introduction to the Dialogues about Hume's views on what an ideal dialogue ought to look like. This paper provides an analysis of Hume's discussion of an ideal dialogue by exploring three aspects of his social epistemology, namely, his views on epistemic peerhood, inquiry norms and virtues, and the possibility of rational persuasion.²

Section 2 introduces Hume's discussion of an ideal dialogue. Section 3 discusses three aspects of Hume's social epistemology in relation to his views of an ideal dialogue. Section 4 then considers whether the conversation between Philo, Cleanthes and Demea (in Parts I–VII of the Dialogues) fulfil the standard of Hume's articulation of what an ideal dialogue should look like. Interestingly, I will argue that it does not. This analysis then generates several important implications for understanding Hume's Dialogues and social epistemology, which I will discuss in section 5. For instance, from this analysis, I will defend an unpopular explanation in the literature, concerning Demea's motivation to leave the conversation in Part VI. Additionally, it allows us to consider a response to the contentious question, of why, if Philo was Hume's primary spokesperson, Hume decided to make Cleanthes the ‘hero’ of the Dialogues instead. Section 6 concludes the paper.

2. Hume's Ideal Dialogue

In the Introduction to the Dialogues, Pamphilus distinguishes two ways that a pedagogue (or author) may convey their instruction (or teaching) to a pupil (or reader): “the form of dialogue” or “the methodical and didactic manner” (D 0.1). The former proceeds by “conversation” while the latter by “Accurate and regular argument”. There are, Hume tells us, “some subjects, however, to which dialogue-writing is peculiarly adapted, and where it is still preferable to the direct and simple method of composition” (D 0.2). What are these subjects? Hume explains,

Any question of philosophy, on the other hand, which is so obscure and uncertain, that human reason can reach no fixed determination with regard to it; if it should be treated at all; seems to lead us naturally into the style of dialogue and conversation. Reasonable men may be allowed to differ, where no one can reasonably be positive: Opposite sentiments, even without any decision, afford an agreeable amusement: and if the subject be curious and interesting, the book carries us, in a manner, into company, and unites the two greatest and purest pleasures of human life, study and society. (D 0.4)³

The dialogue form is better suited, or more “peculiarly adapted” (D 0.2), to subjects which are described as being “obscure”, “uncertain”, which “human reason can reach no fixed determination with regard to it”.

There are some subjects, on Hume's view, where reasonable people should not differ about – and those are better suited to the didactic method of writing. This would, it would imply, include the subjects that Hume himself treated in the works which he employed a didactic style, such as those in his Treatise Concerning Human Nature, his Essays and his Enquiries. However, there are some subjects where reasonable people can differ about – and these subjects are better suited to the dialogue method of writing instead.

Briefly, an ideal dialogue is one where participants are regarded as reasonable, yet they differ on their views about a subject which is complicated (such as natural religion), and after engaging in dialogue with each other, may not come to an agreed upon conclusion. To understand in greater detail what is entailed by each of these aspects (such as, what Hume would have considered ‘reasonable people’), we need to turn to Hume's social epistemology, that can be found in other areas of his works.

3. Three Aspects of Hume's Social Epistemology

This section will clarify Hume's view of an ideal dialogue by focusing on three aspects of his social epistemology: epistemic peerhood, inquiry norms and virtues, and the possibility of rational persuasion.

3.1. Epistemic Peerhood

It is significant that an ideal dialogue is a conversation between “reasonable men”. A helpful way to understand this is to say that disputants in a dialogue ought to be ‘epistemic peers’.⁴ Given the obscurities and uncertainties about the topics suited to the dialogue method, we would reasonably expect there to be disagreement among the participants. Pamphilus explains that within a dialogue, “Reasonable men may be allowed to differ, where no one can reasonably be positive: Opposite sentiments, even without any decision, afford an agreeable amusement …” (D 0.4). This declaration alone appears to distinguish an ideal Humean dialogue from a Socratic dialogue. For while, in a Socratic dialogue, the teacher would seem more reasonable than the pupil (the teacher is an epistemically superior position),⁵ in a Humean dialogue, each disputant is described as being ‘reasonable’. It is important to note that in order for disputants to be epistemic peers, they need not be equally reasonable.⁶

Relatedly, recall Hume's distinction between the didactic and dialogue method noted above. The didactic method is a vehicle for the transfer of information between the teacher and the pupil, presumably, between an epistemic superior and an epistemic inferior. However, the dialogue itself is one that ought to be conducted among epistemic peers.

Another related evidence that Hume thought that dialogue participants should be epistemic peers can be found in a letter he wrote to Gilbert Elliot. Here, he writes to a friend whom he deeply respects: “Had it been my good fortune to live near you, I should have taken on me the character of Philo, in the dialogue, which you'll own I could have supported naturally enough: And you would not have been averse to that of Cleanthes” (HL 154). In fact, in order to ensure that the character of Cleanthes was not viewed in a poor light, Hume invites Elliot to help “to strengthen that side of the argument” (HL 154). His ideal dialogue composition, Hume explains, will be one where “a variety of character and genius being upheld” (HL 154).⁷

What exactly is entailed by ‘epistemic peerhood’ for Hume? What is relevant here is the reasoning ability of participants. ‘Metaphysics and Theology’ seem to be in the category of subjects which exceed the scope of human reason (e.g., EHU 12.5, 12.34). And in another letter to Gilbert Elliot, Hume explains, “But in metaphysics or theology, I cannot see how either of these plain and obvious standards of truth can have place. Nothing there can correct bad reasoning but good reasoning: and sophistry must be opposed by syllogism” (HL 151, emphasis mine). Even though our discussion over obscure subjects may not yield any certain conclusions; nevertheless, reasoning is still possible and important.⁸ In contemporary parlance, therefore, Hume's concern with epistemic peerhood seems more conducive to the notion of reliability peerhood. Participants ought to be able to reason effectively. They may not initially share the same evidence, but they should be able to process the evidence well (once they possess it).⁹

The dialogue, then, is a platform where participants may share their evidence, and together, reason and draw conclusions about their shared evidence. This is exactly what we see in Hume's Dialogues: participants in the Dialogues, at various times, introduce key evidence in favour of their positions, and together, they process the evidence and reason about it. In this way, at the beginning of the dialogue, participants ought to be reliability peers (reliability understood as ‘reasoning ability’ or ‘evidence processing’), and by the end of the dialogue, they are also evidential peers (or ‘evidence possessing’).

3.2. Epistemic Virtues and Inquiry Norms

A related aspect to epistemic peerhood (and arguably, part of what peerhood may entail) is the view that in an ideal dialogue, participants ought to follow certain inquiry norms.

The first norm is the norm of fair representation. A good dialogue, for Hume, is one where the positions of the different participants are presented fairly and are not misrepresented. Hume explains,

I have often thought, that the best way of composing a dialogue, would be for two persons that are of different opinions about any question of importance, to write alternately the different parts of the discourse, and reply to each other. By this means, that vulgar error would be avoided, of putting nothing but nonsense into the mouth of the adversary: And at the same time, a variety of character and genius being upheld, would make the whole look more natural and unaffected. (HL 154)

In fact, his concern with ensuring that Cleanthes’ position is represented well (providing the best possible version of the position) is apparent when he asked Gilbert Elliot to help “to strengthen that side of the argument”.¹⁰ Indeed, Hume's experiences with having his arguments misrepresented surely contributed to the importance he placed on this norm.¹¹

Closely related to the norm of fair representation is the norm that in an ideal dialogue, participants are given a fair hearing by allowing time and space for each speaker to develop their views in conversation. In the letter to Gilbert Elliot discussed above, Hume describes how participants should be fairly given the time and space to respond to each other. And in the Dialogues, Pamphilus explains that an ideal dialogue preserves “a proper balance among the speakers” (D 0.1).

The willingness to give each participant a fair hearing is connected with what Hume takes to be an important virtue. Hume writes,

In conversation, the lively spirit of dialogue is agreeable, even to those who desire not to have any share in the discourse: Hence the teller of long stories, or the pompous declaimer, is very little approved of. But most men desire likewise their turn in the conversation, and regard, with a very evil eye, that loquacity, which deprives them of a right they are naturally so jealous of. (M 8.5)

The third norm in an ideal dialogue is the norm of good company. While the previous norms focus on how participants ought to treat different positions, this norm governs how participants ought to treat each other. Unlike the popular view of debates, which focus on the best argument for different positions, Hume emphasizes the importance of the people and relationships in an ideal dialogue.

Pamphilus explains that the dialogue ought to proceed “in the natural spirit of good company” (D 0.1, emphasis mine). And even if, at the end of the dialogue, participants may not agree, the dialogue nevertheless “carries us, in a manner, into company, and unites the two greatest and purest pleasures of human life, study and society” (D 0.4). That is, even if the dialogue may not be productive in the sense of allowing participants to come to an agreed upon conclusion, the discussion among good company in itself is something that is valuable.

In one sense, this can be seen clearly in Hume's Dialogues – at the beginning and the end.¹² Notice how the Dialogues begin: in a friendly setting (Cleanthes’ library) with Demea paying Cleanthes compliments (D 1.1); and towards the end of the Dialogue, where the disputants describe each other as friends (e.g., D 11.19, 12.1) and where the friendship between Cleanthes and Philo appear quite distinct (e.g., D 12.2) – and the friendship between Cleanthes and Philo is meant, in some way, to be analogous to the friendship between Gilbert Elliot and Hume.¹³ It is not hard to imagine that perhaps much of Hume's view of an ideal dialogue, being conducted in a nice cosy setting among good company, may be modelled after his own experience in salons in France.¹⁴

Many contemporary philosophers, I think, would agree with the importance of these norms: the norm of fair representation, the norm of fair hearing, and the norm of good company. Whenever any of these norms are not honoured in a dialogue, the dialogue would tend to ‘break down’ and prove unfruitful. Indeed, we often teach that ignoring each of these norms, in various contexts, may be viewed as being epistemically vicious. For instance, not honouring the norm of fair representation is often viewed as violating the principle of charity and the intentional adoption of strawmanning. Not honouring the norm of fair hearing, in the context of a dialogue, is often associated with bullying and disrespect. And not treating each other with a certain basic level of respect often dissolves the dialogue into unfruitful debates or ‘ego-fights’.

Unsurprisingly, all three norms are tightly connected to each other. By virtue of participants regarding each other as peers (peerhood), and treating each other cordially (norm of good company), they would invariably understand the importance of respecting the norms of fair representation and fair hearing.

3.3. Rational Persuasion

A third important aspect of an ideal dialogue concerns the possibility of rational persuasion. In contemporary epistemology, some philosophers have raised sceptical concerns about the possibility of rational persuasion. For instance, Fogelin has argued that in cases of Deep Disagreement, rational persuasion is not possible. He writes, “In the end, however, we should tell the truth: there are disagreements, sometimes on important issues, which by their nature, are not subject to rational resolution” (Fogelin 1985: 11).¹⁵ Other philosophers have remained more optimistic, arguing that even in cases of Deep Disagreement, rational persuasion remains possible.¹⁶ Does Hume think that rational persuasion is possible in the context of a dialogue?

Here's one reason for suspecting that Hume thinks that rational persuasion is highly improbable, or impossible: as argued above, Hume thinks that the dialogue form is most appropriate for topics where reasonable men may differ, even if none of them can be reasonably positive. The topic selected in an ideal dialogue would be one “which is so obscure and uncertain, that human reason can reach no fixed determination with regard to it” (D 0.4). Due to this, it would appear that appealing to human reason to resolve dialogues, or to rationally persuade another, would be highly unlikely since our views on these issues inevitably extend beyond the scope of human reason.

In this regard, rational persuasion, for Hume, appears highly improbable. This makes sense of remarks such as:

Reasonable men may be allowed to differ, where no one can reasonably be positive: Opposite sentiments, even without any decision, afford an agreeable amusement … (D 0.4, emphasis mine)

It is true; if men attempt the discussion of questions, which lie entirely beyond the reach of human capacity, such as those concerning the origin of worlds, or the œconomy of the intellectual system or region of spirits, they may long beat the air in their fruitless contests, and never arrive at any determinate conclusion. (E 8.1, emphasis mine)

[T]he facts are here so complicated and dispersed, that a certain conclusion can never be formed from them, and that no single convert will ever be made by any disputes upon this subject; but each disputant will still go off the field with a stronger confirmation of those opinions and prejudices, which he brought to it. (FE 111, emphasis mine)

In the passages above, Hume explains that in disputes upon certain subjects, it is highly unlikely that participants would be rationally persuaded to change their mind. There is a sense in which this is illustrated by Hume's own Dialogues, where participants, such as Demea, clearly did not conciliate.¹⁷ These passages thus suggest that rational persuasion over topics which exceed the scope of our reason is highly unlikely. To be clear, because these subjects exceed the scope of human reason, we cannot therefore appeal to reason to provide a definite and clear conclusion which we would expect all rational people to accept.

However, there is another sense of rational persuasion which Hume appears more optimistic about. Namely, we may employ a certain form of reasoning to persuade someone else to change their mind. Let me explain. Hume thinks that disputants share something similar, and this is important to prevent the dialogue from being fruitless. He writes,

For as the faculties of the mind are supposed to be naturally alike in every individual; otherwise nothing could be more fruitless than to reason or dispute together … (E 8.1)

In this case, given that disputants share “the faculties of the mind”, reasoning or disputing would not be altogether fruitless. As argued above, if participants are viewed as epistemic peers, they would thus have similar reasoning skills. Even if all participants possess similar reasoning skills (reliability peers) and similar evidence (evidential peers), they may still disagree since they may not share the same presuppositions (Ooi 2022). But Hume nevertheless provides a way for participants to rationally persuade each other: they are to employ what Lynch (2016) calls ‘irenic reasons’. According to Lynch, “A gives an irenic reason R to B for some P, only if were B aware of her principles, and reasoned consistently with them, B would recognize that R is a reason for P” (Lynch 2016: 252). That is, participants may still rationally persuade each other through reasoning with each other on their own terms, given their own presuppositions. This may be seen in the Dialogues where participants often step into the shoes of each other in order to try and persuade them. For instance, Philo's use of the expression “in your sense of these attributes” (D 10.35, emphasis mine) illustrates Philo's use of irenic reasoning – that is, ‘according to your understanding, definitions and presuppositions, Cleanthes …’

The distinction drawn above between two forms of rational persuasion maps onto a helpful distinction by Ranalli (2018). Ranalli distinguishes between what he calls rational resolvability and rational persuasion:

Rational Resolvability: A and B's disagreement over p is rationally resolvable if and only if there is some doxastic attitude D that A and B can jointly take to p which is the (uniquely) rational attitude for A and B to have towards p. (Ranalli 2018: 4977)

Rational Persuasion: A rationally persuades B to adopt A's doxastic attitude D to p if and only if there is a set of premises accepted by A that A can appeal to in an argument that rationally ought to persuade B into adopting D towards p (and vice versa). (Ranalli 2018: 4978)

Here, we may argue that in the context of the dialogue, Hume is a sceptic with regards to Rational Resolvability but is more optimistic about the possibility of Rational Persuasion. If this distinction is correct, it clarifies the nature of the arguments and expectations participants should adopt in dialogues.

Thus, in an ideal dialogue, participants cannot simplistically appeal to reason to settle the debate (there is no one rational view that everyone ought to adopt); instead, they may attempt to persuade each other by clarifying each participant's reasoning in light of the evidence each participant brings to the table.

However, Hume also discusses another form of persuasion – this time, non-rational persuasion; what we may call Rhetorical Persuasion. That is, Hume acknowledges that it is possible to use rhetoric to persuade another participant (non-rationally) to change their mind. This is, of course, non-philosophical (even anti-philosophical) and ought not be practiced in an ideal dialogue. Consider the following passages for instance:

Should I enumerate all the evils, incident to human life, and display them, with eloquence, in their proper colours, I should certainly gain the cause with most readers … (FE 111)

Does any man pretend to have more good sense than Julius Cæsar? Yet that haughty conqueror, we know, was so subdued by the charms of Cicero's eloquence, that he was, in a manner, constrained to change his settled purpose and resolution, and to absolve a criminal, whom, before that orator pleaded, he was determined to condemn. (EL 14)

I am indeed persuaded, said Philo, that the best and indeed the only method of bringing every one to a due sense of religion, is by just representations of the misery and wickedness of men. And for that purpose a talent of eloquence and strong imagery is more requisite than that of reasoning and argument. (D 10.2)

Given Hume's views on the psychology of belief, one may employ eloquence or rhetoric instead of reasoning to persuade someone. In fact, eloquence is more effective than reasoning in persuasion.¹⁸ However, while more effective, this strategy violates the inquiry norm of good company – instead of rationally reasoning with others, we attempt to use non-rational techniques to change their view (to manipulate them, as it were).

4. Re-Reading the Dialogues as a Dialogue

In the previous section, I have argued that for Hume, an ideal dialogue is one where participants enter the dialogue viewing each other as reliability peers (they possess roughly similar reasoning skills) – not necessarily evidential peers (they may not share the same evidence base), or presuppositional peers (they may not share the same presuppositions). In an ideal dialogue, participants adhere to three inquiry norms: the norms of fair representation, fair hearing and good company. In so doing, they do not expect that, in order to be rational, the other participants ought to adopt the same view as them. Instead, they readily expect that participants may remain rational even if they disagree. They thus attempt to reason, by introducing new evidence and by providing irenic reasons for each participants’ view.

In this section, I want to apply the norms and rules of an ideal Humean dialogue (as sketched out above) to Hume's Dialogues Concerning Natural Religion. Put differently, how does Hume's Dialogues fare when compared with his own articulation of what an ideal dialogue ought to look like?

To begin, notice that in general, many of the features of an ideal Humean dialogue can be observed throughout the Dialogues. In the beginning of the Dialogues, it appears that the norm of good company is established quite quickly. In a friendly manner, the participants sit together in Cleanthes’ library, and they pay compliments to each other (e.g., D 1.1, D 1.3). At various points, they explicitly refer to each other as ‘friends’ (D 3.9, D 7.9, D 11.19, D 12.1). Further, Philo tells us that he lives with Cleanthes “in unreserved intimacy” (D 12.2). And Cleanthes’ familiarity with Philo can be observed from Pamphilius’ descriptions of Cleanthes’ response to Philo (e.g., D 1.4, D 11.9).

Additionally, the structure of the Dialogues suggests that participants adhered to the norms of fair representation and fair hearing. Throughout the Dialogues, the different participants take turns to respond to each other. Hume clearly attempted to compose the Dialogues by writing in a way that allowed participants, as it were, to write “alternately the different parts of the discourse, and reply to each other” (HL 154). Hume (through Pamphilus) makes this feature clear – participants are allowed to develop their responses, and constantly respond to each other (for instance, the phrase “replied Philo” appears 15 times, “replied Cleanthes” 17 times, and “replied Demea” 8 times).

Further, that Philo, for instance, focused on rational persuasion through the use of irenic reasons rather than rational resolvability can be evidenced from the five times he used the phrase, “according to your …” – that is, assuming the principles or presuppositions of another dialogue participant, what kinds of reasonings may follow, or conclusions may be drawn? Consider:¹⁹

You are honouring with the appellation of Atheist all the sound, orthodox divines almost, who have treated of this subject; and you will, at last be, yourself, found, according to your reckoning, the only sound Theist in the world. (D 4.4, underline mine)

How therefore shall we satisfy ourselves concerning the cause of that Being, whom you suppose the Author of Nature, or, according to your system of Anthropomorphism, the ideal world, into which you trace the material? (D 4.9, underline mine)

but according to your hypothesis of experimental Theism, they become so many objections, by removing the effect still farther from all resemblance to the effects of human art and contrivance. (D 5.2, underline mine)

But according to your method of reasoning, these difficulties become all real; and perhaps will be insisted on, as new instances of likeness to human art and contrivance. (D 5.6, underline mine)

Your conclusion, even according to your own principles, is therefore … (D 7.1, underline mine)

Here, it seems that Philo first has to correctly understand the positions of the other participants; and when reasoning, argue that, according to their own presuppositions, their conclusions are more (or less) reasonable. In this way, from the outset, it is worth acknowledging that many aspects of Hume's Dialogues appear to meet the standard of an ideal Humean dialogue.

However, in analysing the details and conversations in the Dialogues, it would appear that in many ways, the Dialogues violate many expectations of an ideal Humean dialogue. Perhaps the most obvious concern has to do with the norm of good company. More specifically, was Philo a good dialogue partner? Did he treat his fellow dialogue partners with respect, or as peers? It is uncontroversial, I think, to argue that he was a much better dialogue partner to Cleanthes than he was to Demea. Even in the beginning of the Dialogues, Pamphilus remarks that Cleanthes (who was, presumably, an intimate friend of Philo) seemed to have “perceived some raillery or artificial malice in the reasonings of Philo” (D 1.4). And in Part 11, Cleanthes remarks to Demea,

And are you so late in perceiving it? replied Cleanthes. Believe me, Demea; your friend Philo, from the beginning, has been amusing himself at both our expence; and it must be confessed, that the injudicious reasoning of our vulgar theology has given him but too just a handle of ridicule … (D 11.19)

Following this, Demea decides to leave the chat. And the short exchange which follows Demea's leaving, between Cleanthes and Philo, is of extreme importance (something often overlooked by commentators). We may wonder if, by this point, Hume has lost sight of the norms of an ideal dialogue, and, as some commentators suggest, the dialogue just gets away from Hume – he's lost control of the characters.²⁰ Against this reading, I suggest that Hume knew exactly what he was doing – in fact, in the short exchange which follows, he makes it clear that Hume (the author) is perfectly aware that Philo was violating the norm of good company. First, Hume's awareness that Philo is violating this norm can be seen in Cleanthes’ immediate response.

After Demea's departure, Cleanthes and Philo continued the conversation in the following manner. Our friend, I am afraid, said Cleanthes, will have little inclination to revive this topic of discourse, while you are in company; and to tell truth, Philo, I should rather wish to reason with either of you apart on a subject, so sublime and interesting. Your spirit of controversy, joined to your abhorrence of vulgar superstition, carries you strange lengths, when engaged in an argument; and there is nothing so sacred and venerable, even in your own eyes, which you spare on that occasion. (D 12.1)

Here, Cleanthes clearly accuses Philo of violating this norm. In response, we read,

I must confess, replied Philo, that I am less cautious on the subject of Natural Religion than on any other … (D 12.2)

Philo admits that he has taken less care to honour the norm toward Demea – though reiterating his friendship with Cleanthes. So Hume here appears to be very aware of Philo's attitude. In addition to violating the norm of good company, at various points, the norm of fair hearing does not appear to be honoured well. One example of this are the interruptions that the characters engage in (e.g., D 2.10), where all three participants were guilty of interrupting another, and being interrupted by another: Philo interrupting Cleanthes (D 2.25–2.26), Demea interrupting Philo (D 7.6–7.7) and Cleanthes interrupting Demea (D 9.1–9.2). It would appear that, if we were to honour the norms of fair hearing, we would allow each participant to proceed without interruption – but clearly, this norm was not honoured in the Dialogues.²¹

One additional point: in Part 11, it appears that Philo began to employ rhetorical strategies to convince Demea of Philo's own view: in effect, he stopped attempting to rationally persuade Demea, and instead attempted to non-rationally persuade him. Part 11 of the Dialogues is made up of two main arguments: the inferential argument from evil (D 11.2–12) and the evidential argument from evil (D 11.13–16). However, we have reason to doubt that in presenting both of these arguments, Philo was engaging in rational persuasion; in fact, we have good reason to suspect that he was, in fact, employing rhetorical persuasion instead.

Consider first Philo's presentation of the inferential argument. In Hume's Fragment on evil, Hume explains that he is unable to prove, through reason and experience, that pain exceeds pleasure (i.e., rational persuasion). Instead, he admits, that in this regard, rhetorical persuasion may be more effective, but Hume decides not to engage in rhetorical persuasion. He writes,

Should I enumerate all the evils, incident to human life, and display them, with eloquence, in their proper colours, I should certainly gain the cause with most readers, who would be apt to despise, as frivolous, all the pleasures, which could be placed in opposition to them… But I take no advantage of this circumstance, and shall not employ any rhetoric in a philosophical argument, where reason alone ought to be hearkened to. (FE 111)

In attempting to demonstrate that pain exceeds pleasure in the Dialogues, in D 10.32–34, Philo runs into the same problems that Hume did (namely, that he is unable to prove that conclusion through the use of reason and experience).²² Unlike the Fragment, however, being unable to rationally persuade his interlocutors of his views, Philo decides to employ rhetorical persuasion. He literally began to “enumerate all the evils, incident to human life, and display them, with eloquence, in their proper colours”. He begins the inferential argument by doing exactly this, Philo enumerates the evils – “There seem to be four circumstances, on which depend all, or the greatest part of the ills, that molest sensible creatures” (D 11.5) – and then goes on to display them clearly, with great eloquence (D 11.5–11).

Next, notice that after enumerating the evils (inferential argument from evil), he continues by presenting the evidential argument from evil. The evidential argument, as Holden (2010) notes is clearly an argument that Philo cannot endorse – it clearly contradicts his own epistemological position (Holden 2010: 173–6). Instead, unlike his presentation of the logical argument from evil (in D 10), Philo's presentation of the evidential argument may be viewed as a clear act of rhetorical persuasion, using strong and colourful language:

Look round this universe. What an immense profusion of beings, animated and organized, sensible and active! You admire this prodigious variety and fecundity. But inspect a little more narrowly these living existences, the only beings worth regarding. How hostile and destructive to each other! How insufficient all of them for their own happiness! How contemptible or odious to the spectator! The whole presents nothing but the idea of a blind Nature, impregnated by a great vivifying principle, and pouring forth from her lap, without discernment or parental care, her maimed and abortive children. (D 11.13)

Notice that Philo knew exactly what he was doing. At the start of the discussion on pain and suffering (on the moral attributes of God), Philo declares that in order to demonstrate the misery and wickedness of men, “a talent of eloquence and strong imagery is more requisite than that of reasoning and argument” (D 10.2) – that is, rhetorical persuasion is more effective than rational persuasion. That the fellow participants caught on to this can also be seen by the responses of Demea and Cleanthes, highlighted above (D 11.18–19).

If my analysis above is correct, then notice that while the Dialogues may be viewed, on a superficial level, as an attempt by Hume to sketch out an ideal dialogue, it is, in fact, in many ways, less than ideal. In the next section, I articulate several important implications of my analysis.

5. Implications for Interpreting Hume's Dialogues

One immediate implication of the analysis above is that it appears to view Demea's decision to leave the conversation as a sensible one. Often, commentators argue that Demea's decision to leave the chat demonstrates his weakness as a dialogue partner. For instance, O'Connor argues that Demea's decision to leave reveals his “naïveté” (O'Connor 2001: 191). Fogelin thinks that “Hume obviously took pleasure in composing this satirical account of Demea's acceptance of hyperbolic mockery as sincere support” (Fogelin 2017: 77). Instead, the interpretation provided here paints a more optimistic view of Demea's decision to leave: he decided to leave because, by this point, the dialogue “was no longer a genuine inquiry” (Newlands 2016: 638).²³

Indeed, Philo's violation of the norms of good company, and use of rhetorical persuasion instead of rational persuasion in Part 11 of the Dialogues, turned the conversation, at the point, into a less than ideal (instead, a ‘toxic’) dialogue. Demea did not leave simply because Philo and Cleanthes argued against him – they have been arguing against him throughout the Dialogues.²⁴ But what makes the disagreement in Part 11 unique is that unlike previous arguments, Philo is no longer appealing to rational persuasion but now proceeds to employ rhetorical persuasion instead – and that, for Demea, was sufficient reason to think that the conversation was no longer a genuine inquiry. One upside, I think, to this reading is that Demea isn't cast as a mere satirical character, as some commentators make him out to be. On their presentation of Demea, he is defending unreasonable views, and has a petty character. But this would conflict with Pamphilus' explanation that one of the upshots of cashing out the Dialogues through the dialogue method is that it allows reasonable people to disagree (D 0.4).

Another upshot of the above interpretation is that it provides an interpretation of Parts 10–11 of the Dialogues which takes into account the literary nuances that Hume makes. The distinction between philosophical reasoning and rhetoric is an important one in Hume's philosophy – and here, if we are sensitive to the dialectical moves that Philo makes, particularly by paying attention to his use of rhetoric, we then have an interesting interpretation of Parts 10–11 of the Dialogues. In Part 10, Philo wages the logical argument from evil. However, unable to decisively prove, from reason and experience that pain predominates (i.e., when rational persuasion fails), he then decides to employ rhetoric (i.e., rhetorical persuasion) in order to convince his interlocutors of his views – and this interpretation further supports the movements in the Dialogue: it explains why, at this point, both Cleanthes and Demea criticise Philo for being disingenuous – amusing himself at their expense instead of engaging in philosophical reasoning.

Finally, consider a question that naturally arises from our analysis: why did Hume present the Dialogues as one which did not meet his standard of an ideal dialogue? In answering this question, I propose that we also begin to resolve a related question which has troubled commentators. Namely, if Philo was Hume's primary spokesperson,²⁵ why did Hume write the Dialogues in such a way that Cleanthes was the hero of the dialogue? And why did Pamphilus, Cleanthes and Demea constantly perceive Philo as being insincere? Hume clearly intended Cleanthes to be the hero. This is seen both in his letter to Elliot where he writes, “You would perceive by the sample I have given you, that I make Cleanthes the hero of the dialogue” (HL 153) and Pamphilius’ conclusion to the Dialogue, that the principles “of Cleanthes approach still nearer to the truth” (D 12.34). Let me offer a suggestion here. In some sense, it seems to me that there is a strong autobiographical element to this. Given the above analysis, one's objection to Philo (as evidenced by the responses of Pamphilus, Cleanthes and Demea) concerns his attitude and not his views. Put differently, they object to Philo's manner rather than his matter. One may wonder whether Hume himself often felt that his own views were often obscured by other's perception of his attitude. First, notice that in his recasting of the Treatise into the Enquiry, he focused more on revising the manner than the matter (MOL 8) – the idea here being that Hume thought that his want of success was not due to his views being false, but that his views were obscured by how people viewed the manner of his writings. Similarly, notice that towards the end of My Own Life, after providing account of his works (his ‘matter’, as it were), he ended the account with a defence of his character (his ‘manner’).

I suggest, therefore, that when Hume wrote, “I should have taken on me the character of Philo” (HL 153), he meant more than that he endorsed Philo's views (as many commentators have pointed out, interpreting this phrase simply as Philo being the exclusive spokesperson of Hume's views is problematic). In effect, Hume wanted to convey that like Philo, his own views are often obscured by others’ views of his attitude. Demea, for instance, did not leave the conversation merely because he disagreed with Philo's views, but as I have argued, he disliked Philo's attitude.

6. Concluding Remarks

In this paper, I have sketched out aspects of what an ideal Humean dialogue ought to look like, focusing on concerns in contemporary social epistemology. I hope to have shown ways in which Hume's philosophy and pedagogy has important implications and contributions for contemporary social epistemology. Unsurprisingly, one upshot of getting clear on what an ideal Humean dialogue ought to look like is that it contributes to our interpretation of the Dialogues. On this interpretation, Demea's decision to leave the dialogue was a sensible one; that an important movement in Parts 10–11 is Philo's change from depending on reason to depending on rhetoric; and provides an explanation for why Hume decided not to make Philo the hero of the Dialogues.²⁶