Aristotle in the Nicomachean Ethics famously distinguishes three types of friendship according to their objects: virtue friendship, directed at the good (ἀγαθόν); pleasure friendship, directed at the pleasant (ἡδύ); and utility friendship, directed at the advantageous (χρήσιμον).Footnote ¹ All of these qualify as ϕιλία, but the one based on virtue is far superior to the other kinds. As an additional bonus, its practitioners—though motivated by higher considerations—do not lose out on the benefits provided by the other two types since, by definition, virtuous friends who love each other for their moral qualities at the same time find each other pleasant and useful.Footnote ²

By contrast, Cicero in his Laelius de amicitia recognizes only virtue friendship: ‘friendship cannot exist except among good men’, boni, who are inspired by love for each other’s excellence.Footnote ³ Like Aristotle, he stresses that such a high-minded relationship entails usefulness (utilitas); the latter is, as it were, a welcome by-product of amicitia, but is never the motivation for pursuing true friendship, which comes about naturally from affection for the other person.Footnote ⁴ As he pithily puts it (in a description of his speaker Laelius’ friendship with Scipio the Younger), ‘friendship didn’t pursue advantage, but advantage followed on friendship’.Footnote ⁵ Here and elsewhere, Cicero seems to argue against both popular Roman ideas of amicitia as self-serving social networking and the Epicurean view of friendship as a mere instrument for the procurement of pleasure.Footnote ⁶

If Cicero thus grants utilitas a place in the friendship among boni, he never says that this kind of relationship is productive of pleasure. Unlike Aristotle, who freely discusses the ἡδονή inherent in virtue friendship, Cicero uses uolupta(te)s rarely in De amicitia and never to refer to the pleasures specifically of friendship. The word occurs two times in a neutral contextFootnote ⁷ and three times in discussions of putative philosophical goods, typically with outspoken or silent disapproval: ‘some people [propose] pleasures [as the highest good], something that is appropriate for beasts’; ‘people who in the manner of farm animals refer everything to pleasure’; ‘those who have handed themselves over entirely to pleasure’.Footnote ⁸ In these cases, the reference is clearly to the Epicureans, who consider pleasure the summum bonum and who come in for much criticism in De amicitia, even though—like other philosophers and philosophical schools throughout the dialogue—they are never mentioned by name.Footnote ⁹

Does this mean that Cicero does not consider friendship to be enjoyable? On the contrary: Laelius again and again waxes ecstatic about the delights granted by friendship: ‘They seem to take the sun out of heaven, those who remove friendship from life—friendship than which we have no better gift from the gods, none that is more delightful (iucundius).’Footnote ¹⁰ iucundus and iucunditas are repeatedly applied to the experience of friendship, as are delectare and oblectatio.Footnote ¹¹ Perhaps Laelius’ favourite characteristic of amici and amicitia, however, is that they are ‘sweet’, dulcis or suauis.Footnote ¹² He exclaims, ‘What is sweeter than to have someone to whom you dare to talk about all things as if to yourself?’,Footnote ¹³ and reanimates the dead metaphor of taste inherent in the idea of sweetness by calling the suauitas of a friend’s speech and character a ‘not inconsiderable seasoning of friendship’ and claiming that old friendships are the sweetest, just as old wines tend to be.Footnote ¹⁴

In stressing the delights and especially the sweetness of friendship, Cicero is following the general Graeco-Roman consensus. In his 1905 dissertation on the topoi found in ancient discussions of friendship, Gottfried Bohnenblust observes that, despite all disagreements on the origin and purpose of friendship, there is a general stress on ‘das Angenehme (ἡδύ, dulce, suave)’.Footnote ¹⁵ Like Cicero, a number of writers even claim that there exists nothing sweeter or more enjoyable than a friend or friendship itself: ‘while in my right mind, I would compare nothing to a delightful friend’, declares Horace, and Plutarch states point-blank that ‘there is nothing sweeter than friendship and nothing else that offers more joy.’Footnote ¹⁶

Cicero’s strategy in De amicitia is thus clear: he will not allow ‘pleasure’ a place in friendship, leaving the term uoluptas to those ‘beastly’ hedonists, the Epicureans.Footnote ¹⁷ At the same time, he extols the joys of friendship, freely using such terms as delectare, suauis, dulcis and iucundus. Unlike Aristotelian virtue friendship, Ciceronian amicitia does not entail pleasure, but it is definitely sweet.

Cicero’s polemical move of eschewing pleasure while still embracing sweetness is made possible by a subtle semantic shift that occurs when the philosophical terminology of pleasure is translated from Greek into Latin. The Greek term for pleasure, used by Aristotle and Epicurus, among others, is ἡδονή, a noun that is obviously related to the adjective ἡδύς ‘sweet’. The adjective and derived adverb are themselves regularly used in philosophical contexts to refer to the pleasant and pleasurable: thus, for example, the acolytes of Aristotelian pleasure friendship pursue exactly τὸ ἡδύ (see above), and the final goal of Epicureans is to live ἡδέως ‘sweetly, pleasantly’.Footnote ¹⁸ In Greek, then, pleasure equals sweetness, and Cicero’s rhetorical ploy of separating the two simply would not work.

In Latin philosophical writing, ἡδονή becomes uoluptas, a choice that appears straightforward, given a large semantic overlap between the words.Footnote ¹⁹ Cicero himself makes this point when discussing Epicurean pleasure in De finibus (2.13):

Cicero goes on to explain that other potential candidates, such as laetitia and gaudium, fail on the grounds that they denote only mental pleasures, while uoluptas, like ἡδονή, is experienced by both the body and the mind.Footnote ²⁰

Even so, something is lost in translation while something else is gained. Frequently found in comedy and apparently a feature of colloquial speech, uoluptas can refer to any ‘agreeable experience or sensation’ (OLD 1a), including the creature comforts that the characters of comedy typically pursue. Thus, for example, one Plautine character declares that he will delight uoluptate uino et amore (Merc. 548), while another lists among uoluptates ‘laughter, jokes, kissing, dancing, sweet talk and good will’.Footnote ²¹ The ubiquitous use in Plautus of mea uoluptas to address one’s beloved further attests to the word’s association with sensual and sexual pleasures.

While there is thus nothing inherently disreputable in the term uoluptas, there is also nothing high-minded; as Cicero puts it, uerbum ipsum uoluptatis non habet dignitatem.Footnote ²² Terminology colours perception: a philosophical school that goes after uoluptas might appear in a more negative light than one that pursues ἡδονή. The same, incidentally, is true for modern languages: the pleasure or plaisir at which hedonists aim in English and French sounds considerably more positive than the Lust that becomes the object once the same doctrine is translated into German.Footnote ²³

Cicero, at any rate, endeavours throughout his philosophical work to make uoluptas into a bad word, among other things by casting personified Pleasure as a rival to Virtus, a rhetorical move made additionally attractive by the alliteration.Footnote ²⁴ In non-philosophical contexts, though, he continues to use uoluptas in its anodyne sense, as when he expresses formulaic ‘pleasure’ at receiving a letter.Footnote ²⁵ The same practice is followed by Seneca, though he jokingly clarifies that he is using the word in the everyday, not the philosophical, sense (Ep. 59.1):

For a philosophically trained Roman, even expressing everyday delight can turn into a semantic minefield.

Another disadvantage of uoluptas is that, unlike ἡδονή, it lacks a proper adjective to describe what is ‘pleasant’ or ‘pleasurable’.Footnote ²⁶ The archaic adverb uolup, from which the noun is derived, is nearly entirely restricted to comedy and to the fairly colourless idiom uolup esse ‘be a source of pleasure’. The derivative adjective uoluptarius, by contrast, meaning ‘characterized by or concerned with sensual pleasure’ (OLD 1a) or ‘devoted to pleasure, pleasure-loving’ (OLD 2), already has the negative connotations that come to the fore when the word is borrowed into English as ‘voluptuary’.

To work around this problem, Roman philosophical writers discussing pleasure, especially in the context of Epicureanism, take recourse (whether consciously or not) to the original meaning of ἡδονή and describe uoluptas, too, as sweet, suauis or dulcis. A good example is the (in)famous opening of Lucretius’ second book, where, in addition to ‘sweet’, pleasure is also said to be ‘delightful’ (iucundus):Footnote ²⁷

Torquatus, the Epicurean spokesman in the first book of De finibus likewise uses suauis, iucundus and derivatives to describe Epicurean pleasure, as does Cicero himself in his rebuttal in Book 2, adding dulcis to the mix.Footnote ²⁸ His very definition of uoluptas is a showcase of the Latin philosophical vocabulary of pleasure as it had established itself in Cicero’s time: laetitiam in animo, commotionem suauem iucunditatis in corpore.Footnote ²⁹

In light of this linguistic usage, then, Cicero’s strict separation of sweetness and pleasure in De amicitia is striking. Subhuman uoluptas is reserved for the Epicureans, while friendship is extolled for being sweet and delightful—that is, affording exactly what both popular sentiment and such philosophers as Aristotle would describe as pleasure, but what Cicero seems to be at pains never to refer to as such. This seems especially jarring in light of the fact that Epicurus famously promoted friendship, declaring—as attests Cicero’s own character Torquatus in De finibus—that ‘of all things that wisdom has provided for the happy life, nothing is greater than friendship, nothing more productive, nothing more delightful (iucundius).’Footnote ³⁰ For the Epicureans, friendship provides a significant avenue to the pleasure that is the summum bonum, but you would not be able to tell this from Cicero’s De amicitia.

Cicero’s animosity towards Epicureanism is well known,Footnote ³¹ and we might chalk up his pointed use of vocabulary simply to a zealous and unfair strategy of denying the disreputable hedonists a place in the delightful world of virtue friendship that Cicero is expounding. However, while the anti-Epicurean slant of De amicitia is clear, especially in the dialogue’s strong stance against the idea that friendship is pursued for the sake of utilitas (see above), I believe that there is more to Cicero’s avoidance of the term uoluptas for the delights of amicitia than a knee-jerk dislike of the Garden. In Cicero’s mind, or at least in his linguistic use in the dialogue, uoluptas is reserved specifically for Epicurean pleasure. However, I suggest that Cicero believes—and may well be correct in believing—that the pleasures of friendship as he envisages them would not actually qualify as pleasures according to Epicurus.

In De amicitia, Cicero does not advance an overt argument to this end, but we can draw on his earlier discussion of Epicurean ethics, including friendship, in the first two books of De finibus for a clearer sense of his opinions.Footnote ³² Cicero’s treatment there points to a number of putative problems or inconsistencies within Epicurean thought and terminology, with many of which scholars of Epicureanism are still grappling today. Owing to the facts that, apart from three letters and a couple of collections of maxims, Epicurus’ own works survive only in fragments and that we have to rely on later reports (including notably Cicero’s) in reconstructing his doctrine, many aspects of what at first glance appears to be a straightforward philosophical system turn out to be quite unclear and remain hotly debated.

As for the Epicurean view of friendship itself, this is today, and was clearly already in antiquity, one of the greatest bones of interpretative contention.Footnote ³³ Briefly, the problem is that for Epicureans pleasure is the summum bonum, the final goal that is pursued for its own sake, whereas everything else is pursued for the sake of pleasure. As we have seen, friendship is especially privileged among the things that are instrumental for pleasure—an attractive proposition, but one that leads to serious problems. Since in the eyes of many readers, ancient and modern, friendship implies a genuine altruistic valuing of friends for their own sake, the Epicurean promotion of friendship seems to undermine fatally the basic Epicurean tenet that pleasure is the only thing that is a good in and of itself. As it happens, Epicurus himself on occasion seems to come dangerously close to endorsing the pursuit of friendship as a good in its own right.Footnote ³⁴ This ambivalent attitude to friendship is comparable to Epicurus’ treatment of the virtues, including justice, which are also famously viewed as not choiceworthy in themselves, but only in so far as they are necessary instruments for securing a pleasurable life.Footnote ³⁵ Even so, readers from antiquity to the present have occasionally suspected that Epicurus must somehow value the virtues for themselves, something that would be a fatal blow to his minimalist, single-good system.Footnote ³⁶

While these issues touch on the subject of this article and will continue to come up, my main concern in what follows is another: which of the manifold experiences and feelings that people might describe as pleasurable, sweet, delightful, etc. actually qualify as pleasures in the Epicurean sense? While there is ample literature on many controversial aspects of Epicurean hedonism, this question receives very little airtime in contemporary scholarship, even though it was, I believe, of great concern to Cicero, as well as other ancient critics of Epicureanism, including, as we will see, Plutarch in his polemical A Pleasant Life Is Impossible according to Epicurus.

These are the relevant tenets of Epicurean hedonism:

• (i) There are bodily and mental pleasures.Footnote ³⁷

• (ii) Bodily pleasures are primary; based on sense perception, they occur only in the moment. Mental pleasures are derivative of bodily pleasures, but are superior, having the advantage that—through memory and anticipation—they extend to the past and future.

• (iii) Some pleasures involve a process by which a desire is fulfilled and a pain removed. For example, the pleasure of drinking fulfils the desire for hydration and removes the pain that is thirst.

• (iv) The greatest pleasure, however, is a steady state characterized by the absence of all bodily and mental pain: ἀπονία (freedom from bodily pain) and ἀταραξία (freedom from mental disturbance).Footnote ³⁸

So what would be examples of pleasures that fit this Epicurean scheme? Unfortunately, the surviving Epicurean literature offers only few concrete illustrations. As for bodily pleasures, apart from the highest pleasure of painlessness, the typical examples are eating, drinking and sex. More generally, Epicurus included all agreeable sensations furnished by the senses, as Cicero reports in the Tusculans:Footnote ³⁹

As for the mental pleasures, these consist in the enjoyment and the recollection of bodily pleasures, as well as in their anticipation, including ideally the confident belief that one will be able securely to extend one’s present state of bodily painlessness into the future.Footnote ⁴⁰ Thus, according to Plutarch, Epicurus considered ‘the stable condition of the flesh and the confident expectation of this’ to be ‘the highest and most secure joy’.Footnote ⁴¹

Even though there is ample evidence that, as Cicero’s Torquatus puts it, the Epicureans ‘hold that the pleasures and pains of the mind arise from the pleasures and pains of the body’,Footnote ⁴² this tenet has met with doubt. Thus, for example, A.A. Long and D.N. Sedley write in The Hellenistic Philosophers:Footnote ⁴³

It is not clear, however, that Epicureans really were unable to explain the kinds of mental pleasures enumerated by Long and Sedley as ultimately based on bodily pleasure. Clearly, Epicurus does not reckon only with such straightforward mental delights as the memory of a good meal had today and the anticipation of another one tomorrow, but includes more complicated chains of hedonic cause and effect. Thus, the freedom from fear of death and of the gods—a major element of ἀταραξία—is the pleasure of the firm knowledge that, lacking existence and hence sensation after death, one will suffer no physical harm (for example through decay of the body or tortures in the underworld), and that in one’s lifetime the gods will never punish or capriciously hurt one. What Epicureanism teaches is exactly the kind of mindset and behaviour that gives people both the ability to avoid bodily pain in the present and the certainty that they will be able to do so in the future. This might explain also Epicurus’ contention that, in studying philosophy, pleasure accompanies the very acquisition of knowledge (Sent. Vat. 27): this does not mean that the intellectual activity of philosophizing is pleasurable in itself, but that the budding Epicurean’s realization that one need not fear physical pain, and will be able to live ‘undisturbed’, is an immense mental pleasure ultimately derived from a physical one.Footnote ⁴⁴

A similar argument can be made about Long and Sedley’s ‘joyous remembrance of philosophical conversations’, a reference to Epicurus’ famous death-bed letter to Idomeneus, in which the philosopher describes his ability to counteract excruciating physical pain with τῇ τῶν ἡμῖν γϵγονότων διαλογισμῶν μνήμῃ.Footnote ⁴⁵ If, as seems likely, the remembered discussion was indeed what Epicurus elsewhere calls ‘διαλογισμοί conducive to a blessed life’, the mental pleasure both at the time and in recollection would have consisted in the joyful realization that pain need not be feared and that our ἀταραξία is entirely up to us.Footnote ⁴⁶ As for Long and Sedley’s final point, the pleasure of friendship is of course what is at issue in this article, but, on a purely instrumental understanding of Epicurean friendship, its pleasure would be the experience of confidence not only that our friends are useful to us right now but especially that they will remain so in the future.Footnote ⁴⁷ We can rely on our friends for assistance in achieving a life free from bodily and mental pain, whether such help takes the form of material support or—especially if both we and our friends are Epicureans—the joint philosophizing that leads to our desired ἀταραξία.

It thus seems to me that, pace Long and Sedley, Epicureans are perfectly capable of explaining what they consider mental pleasures by ultimately predicating them on bodily ones, whether past, present or future.Footnote ⁴⁸ For an Epicurean argument contemporary to Cicero that posits such a chain reaction in order to explain hedonic experience and motivation, we may take another look at the opening to De rerum natura Book 2, partly quoted above. Lucretius makes clear that the pleasure one feels at watching storm-tossed sailors from safe land is not sadism or Schadenfreude (2.3 non quia uexari quemquamst iucunda uoluptas) but the knowledge of being free from the evils from which others are suffering (2.4 sed quibus ipse malis careas quia cernere suaue est)—both the bodily pains and dangers of the seafarers in distress and, metaphorically, the wrong attitudes of non-Epicureans, which cause them both to experience bodily pains in the present and, worse, to anticipate them for the future. When Lucretius concludes that ‘nothing is sweeter than to inhabit the well-fortified serene temples of the wise erected by learning’,Footnote ⁴⁹ he does not mean that practising philosophy is a pleasure in and of itself, but that it is ‘sweet’ exactly because it affords the certainty that one is immune to any unbearable bodily or related mental pain, both present and future.

On this understanding of Epicurean hedonism, then, it is questionable whether certain objects or activities that ordinary people might consider pleasurable would actually qualify as pleasures in the Epicurean sense. This includes even some activities that modern scholars in discussions of Epicureanism have casually used as examples of Epicurean pleasures. Panos Dimas, for instance, mentions ‘reading a book’, but, unless we are talking about a work of Epicurean philosophy, it is hard to see how this would cause bodily pleasure or a mental pleasure based on a bodily one.Footnote ⁵⁰ Raphael Woolf’s ‘taking a walk’ is also at least open to doubt:Footnote ⁵¹ one could argue, of course, that the exercise involved is providing bodily pleasure and/or the mental pleasure of knowing that one’s fitness regime helps preserve one’s bodily ἀπονία for the future; if the walk leads through a beautiful landscape or city, one might also undergo some ‘pleasant motions’ of visual sense perception. Even so, I suspect that for many people in our post-romantic era the experience of taking a walk goes beyond these pedestrian elements and involves a more holistic pleasure, one that may no longer be the pleasure of Epicurus. Finally, ‘doing crossword puzzles’ and ‘throwing darts in contests at the pub’, suggested by John Cooper as part of a whole series of pastimes an Epicurean might enjoy,Footnote ⁵² appear to be activities that—however pleasurable we ourselves may find them—are well-nigh impossible to interpret as bodily pleasures or mental pleasures derived from bodily ones.Footnote ⁵³

Modern readers who enjoy the activities adduced by Dimas, Woolf and Cooper would, of course, most likely maintain that they find them pleasurable as such. However, as Cicero in De finibus delights in pointing out, the problem is that orthodox Epicureans cannot claim haec enim ipsa mihi sunt uoluptati (Fin. 1.25)—that is, that anything provides pleasure in and of itself, unless it consists in, or in some way derives from, bodily pleasure. Disqualified from being such putative ‘pleasures in their own right’ are not only the intellectual pursuits of Cicero’s interlocutors Torquatus and Triarius (who, like Dimas, enjoy reading books)Footnote ⁵⁴ but also the virtuous actions of Torquatus’ ancestor M. Torquatus Imperiosus, who notoriously had his own son executed for disobeying military orders, something that according to Cicero cannot be explained as hedonistically motivated.Footnote ⁵⁵ Epicureans cannot hold that virtue is intrinsically pleasurable, only that it is conducive to pleasure—which in the case of a father who orders the death of his son is open to serious doubt.Footnote ⁵⁶

Cicero returns to this argument at the end of his speech against Epicureanism in De finibus Book 2, bombarding Torquatus with a series of questions (107):

Intellectual activity, beautiful man-made or natural objects and leisurely pastimes are things in which Torquatus and no doubt many other people take what in ordinary language would be called pleasure—but this is not pleasure in the Epicurean sense.Footnote ⁵⁷

Cicero’s point is that Epicureanism’s single-minded refusal to ascribe intrinsic value to anything other than bodily pleasure and its mental anticipation, appreciation and recollection puts the system under enormous pressure. Not only can there be no goods other than pleasure (which means, among other things, that virtue or friendship cannot be considered good and choiceworthy in and of itself), but there cannot even be any pleasures beyond enjoyable sense-stimulation, its mental reflection, and ἀπονία and ἀταραξία. As soon as Epicurus or one of his followers can be shown to value as intrinsically good and/or pleasurable something —whether it be hunting, philosophizing, friendship or virtue—in addition to their restrictive summum bonum, then the entire doctrine, predicated as it is on a single narrowly defined good, is in danger of collapsing (Fin. 1.25):

The existence of too many pleasures would be ruinous to the minimal brand of hedonism promoted by Epicurus.

Is Cicero’s depiction of Epicurean doctrine correct on this point? It is often claimed that Cicero’s understanding of Epicureanism was faulty and/or that he unfairly distorted Epicurean views for his polemical ends. The first is unlikely. As Carlos Lévy points out, Cicero studied with two Epicurean scholarchs, Phaedrus and Zeno of Sidon, and was personally acquainted with many Epicureans, Greek and Roman alike.Footnote ⁵⁸ He also had access to a considerable body of Epicurean writing, including the works of Epicurus himself, nearly all of which is inaccessible to us. This alone should make us wary of assuming that modern scholars would be in a position to have a better grasp on Epicurean doctrine than such a well-trained individual much closer in time and with a much better library.

As for the second point, Cicero makes no bones about his disapproval of Epicureanism and is often highly polemical in discussing the school. Even so, I do not think that he would deliberately misrepresent Epicurean tenets, something that it would be hard to get away with, given the widespread interest in Epicureanism among the upper classes in mid-first-century Rome.Footnote ⁵⁹ Cicero himself in De finibus introduces Epicureanism as a philosophical system that is ‘very well known to most people’Footnote ⁶⁰ ; there would thus be little scope for telling downright falsehoods without being called out by his readers.

As it happens, most—but not all (see below)—modern scholars who consider the issue seem to agree with Cicero’s restrictive interpretation of Epicurean pleasure, as does, notably, Plutarch in his dialogue A Pleasant Life Is Impossible according to Epicurus.Footnote ⁶¹ In this work, the main speaker Theon sets out to prove the apparently paradoxical proposition that ‘it is not even possible to live pleasantly according to them’, them being the hedonistic Epicureans.Footnote ⁶² At great length, Plutarch makes essentially the same point as Cicero in De finibus: if, according to Epicurus, pleasure is restricted to bodily pleasures and those mental pleasures derivative of them,Footnote ⁶³ then there is very little left for Epicureans to enjoy. Notably, neither the contemplative nor the active life can hold any charm for the deprived denizen of the Garden, who will never know the many pleasures other people derive from intellectual investigation or virtuous action—since, of course, for Epicureans, these are not pleasures at all.Footnote ⁶⁴

Both in antiquity and today, however, there have been people with a different understanding of Epicureanism, believing that Epicurus must have held one or both of the following:

• (i) There are things that are goods in their own right beyond bodily pleasure and the mental pleasure derived from it; and

• (ii) There are things that are pleasurable in their own right beyond bodily pleasure and the mental pleasure derived from it.Footnote ⁶⁵

As for the first, virtue and friendship are the obvious candidates. In De finibus, Torquatus reports that a subsection of ‘more timid’ (timidiores) Epicureans fear that, ‘if we think that friendship should be sought for the sake of pleasure, then friendship seems, as it were, completely lame.’Footnote ⁶⁶ They therefore believe that we first select friends in a purely utilitarian fashion, but then come to love them ‘for themselves’ (propter se ipsos). Among modern scholars, for example, Julia Annas maintains that, on the Epicurean view, ‘living virtuously … is something we seek for its own sake’, and so is friendship.Footnote ⁶⁷ In her opinion, ‘we achieve it [pleasure] precisely by having non-instrumental concern for virtuous action and the interests of others.’Footnote ⁶⁸ Similarly, Dorothea Frede takes a view of friendship reminiscent of that of Cicero’s ‘timid’ Epicureans, claiming that, according to Epicurus, ‘friendship consists in the love of the others for their own sake.’Footnote ⁶⁹

As for the second, we have already seen the doubts of Long and Sedley as to whether all mental pleasures can really be derivative of bodily ones. Cicero himself tells us that it is a common misunderstanding among contemporary Epicureans to believe that ‘that which is morally right itself causes joy, that is, pleasure.’Footnote ⁷⁰ This mistaken belief makes Epicureanism attractive to the masses, but it is unorthodox, ‘for Epicurus or Metrodorus never made this kind of argument, nor any other Epicurean who had any sense or had studied these matters’.Footnote ⁷¹ Torquatus concurs (Fin. 1.55):

In this context, it is interesting that, as David Armstrong has shown, Cicero’s contemporary Philodemus apparently promoted a view of friendship as not purely instrumental but explicitly a pleasure in itself.Footnote ⁷²

While certainty on Epicurus’ own views is impossible to ascertain (and it is, of course, perfectly possible that the founder himself was not entirely consistent on these matters), it seems to me most likely that the minimalist reading of Epicurean hedonism of Cicero, Plutarch and many modern scholars is in fact correct. That both in antiquity and today more expansive views (involving a larger number of goods and/or pleasures) have been promoted attests both to the radicalism of Epicurus’ vision and to the discontent this radicalism has caused.Footnote ⁷³ However, even if Cicero’s understanding of Epicurean hedonism is wrong on this point, it still explains his use of language in De amicitia. Which, finally, brings us back to the pleasures of friendship.

What exactly is it about friendship that Cicero considers so ‘sweet’ and ‘enjoyable’ (and that a less scrupulous language-user would probably call ‘pleasurable’)? First, it is the character of our friends, their virtuous nature that initially induced us to love them, the love they in turn feel for us and, more generally, their delightful talk and manners.Footnote ⁷⁴ Second, there is an inherent delight in socializing with other people and simply having someone with whom one can share experiences.Footnote ⁷⁵ With a close friend, this leads to a pleasurable exchange not only of the most intimate thoughtsFootnote ⁷⁶ but also of mutual good will and kind services.Footnote ⁷⁷ Ideally, friends will spend as much time together as they can, something Laelius fondly remembers he did with Scipio: together, the two friends enjoyed all aspects of life from politics, warfare and travel to study and leisurely pastimes (Amic. 103–4).Footnote ⁷⁸

From a minimalist Epicurean point of view, this list of putative pleasures does not look promising: as Epicureans, we cannot consider our friends and their characters pleasurable in their own right, and it is not clear how the joys of a good heart-to-heart conversation could be made to fit the narrow compass of Epicurean pleasure.Footnote ⁷⁹ How pleasurable in the Epicurean sense it is to hang out with our friends will depend on our activities: good food, drink or sex would of course provide bodily pleasures, but, unless our friends are instrumental in obtaining them, it is hard to see what friendship would add to the experience; by contrast, politics, warfare and study, whether with friends or alone, would never clear the Epicurean hedonistic bar. This leaves only the mutual good will and exchange of services, which does indeed have the potential to contribute to our ἀταραξία in giving us the knowledge that out friends are there to help ensure our freedom from bodily pain now and in the future. Even there, though, what Cicero finds especially delightful is not the advantages one gains from a friend (which throughout De amicitia he considers simply a by-product of friendship; see above) but the friend’s affection that becomes apparent in bestowing them (Amic. 51):

Epicureans, one could argue, might also take pleasure in recognizing their friends’ genuine affection since it adds even more to their confidence in future support—but they could not very well take pleasure in such affection in its own right (amor ipse).

To conclude, even though Cicero never says so, his virtue friendship—just like that of Aristotle—also by definition provides not only utility but also pleasure. Unlike with the first, Cicero never makes an explicit argument as to the second and in fact avoids the term uoluptas, reserving it for references to Epicurean doctrine and instead (availing himself of a quirk of the Latin vocabulary of pleasure) describing the joys of friendship as sweet and delightful. As a comparison with the discussion of Epicurean hedonism in De finibus has shown, Cicero’s avoidance of the Epicurean term uoluptas goes hand in hand with his critique of the minimalism of Epicurean ethics, which restricts the good to pleasure, and pleasure itself to bodily pleasures and the mental pleasures derived from them. On Cicero’s reading (which is also found in Plutarch), many things in which people ordinarily take pleasure do therefore not qualify as pleasures in the Epicurean sense. This includes the pleasures of friendship celebrated in De amicitia: they are not Epicurean uoluptates, but nevertheless remain the sweetest things the gods have bestowed on us.