CHAPTER
7
Matthew Roller
This chapter investigates the Roman philosopher Seneca’s habit of presenting multiple exempla in a list, as a way of illustrating a moral value and thereby supporting a particular argument. By “exemplum” I mean a narrative of or reference to an action that has been performed by a social actor before the eyes of members of his or her community, that has been judged by that audience of spectators to be notably “good” or “bad” in terms of one or more of the values or beliefs that the audience and actor share, and that has subsequently been monumentalized as a noteworthy “deed” (or misdeed) carrying normative force. Romans of all eras habitually adduce exempla in hortatory or deliberative contexts, when urging a particular course of action or when measuring an action against a moral standard set by previous performances. Because such contexts occur in texts of every sort, ranging across oratory, historiography, philosophical writing, legal writing, and various kinds of poetry, the use of exempla is pervasive in Roman texts—as it also was, no doubt, in speech. Visual objects or built structures, which often provide a monumentalizing function, may also thereby participate in the construction of exempla.1
As scholars have long observed, exempla have a dual ontology: they are simultaneously unique and typical.2 This duality is baked into exempla from the moment of their creation. On the one hand, any given performance in the public eye is irreducibly unique and contingent: one person, at a particular time and place and under particular circumstances, performs a particular action before the eyes of a particular subset of the community as it is constituted then and there. On the other hand, once that audience judges the performance and assigns it value in one or more socially salient moral categories—determining, for example, that the actor was valorous in battle, or was ungrateful to his benefactors, or (to touch on categories discussed below) that he controlled his anger appropriately, or tolerated insults with moderation, or the opposite of all these—it thereby sorts that performance into pre-constituted bins into which other exemplary performances have also already been sorted. So treated, the particular performance becomes one instance, or one more instance, illustrating or substantiating the overarching moral categories of valorous or ungrateful or suitably self-restrained conduct. This kind of category, in which membership is determined by shared properties constituting necessary or sufficient conditions for membership, and in which all members are “equivalent” insofar as they satisfy these criteria, is sometimes called a “classical” category, due to its theorization by Aristotle.3 The (classical) category thus highlights particular features of the individual exempla it subsumes, namely features that these exempla have in common, while deemphasizing other features that are not shared. Through their relationship to such a category, individual exempla are “typical”—equivalent to, and at least potentially interchangeable with, other exempla assigned to this category by virtue of sharing the relevant feature.
Furthermore, the normativity of exempla depends upon categorization. Exempla become consequential for the community and capable of bearing normative force only upon being processed into categories that are pre-constituted as socially relevant. Categorization is thus the mechanism by which an exemplum in its particularity is made both general and socially relevant.4 The exemplum’s passage from contingent, unique, individual action to socially consequential “deed” of general significance that instantiates a broadly accepted moral category may explain the Senecan (and more broadly Roman) impulse to construct arguments by providing exempla in a list. For a category’s overall social consequence—the validity of its universalist claim—may seem greater, or at least more obvious, if it is shown to contain a number of exempla. The more instances that are listed under a category, then, the stronger an argument invoking that category may appear to be.
Yet the exemplum’s passage from specific to general via categorization does not efface its uniqueness. For as soon as a particular exemplum is cited as an instance of some category, its ineluctable distinctiveness resurfaces: features specific to itself enter into relationships with particular elements of the framing argument in support of which this exemplum and its governing category have been adduced. When multiple exempla are adduced and listed serially, each one in its individuality may generate its own distinctive set of relationships to the framing argument. So notwithstanding their “classical” equivalence as instances of the moral category under which they are marshaled (i.e., their sharing of a single predicate), certain exempla may, nevertheless, appear to be more relevant than others to the argument that frames them. Nor is this all. Exempla in a list also generate relationships among themselves, as the particular features of each exemplum interact with those of its neighbors. Thus any listing of exempla in any given order generates dynamics internal to the list itself, creating themes, hierarchies, and emphases that may appear to support, undercut, or simply exist quite independently of the framing argument that notionally subsumes and organizes them. Such “nonclassical” behavior within categories has caught the attention of post-structuralist philosophers and cognitive linguists in recent decades. Among the former, Derrida contends that serial exempla have a dual allegiance, both to “the law” under which they are expressly marshaled (the classical category) and to an “other law” that emerges from their piecewise interrelationships (accounting for nonclassical behavior). Among the latter, scholars like Lakoff and Rosch have shown that categories often display “prototype effects” or “typicality effects,” where certain members of a category may be deemed “better” examples of the category than others.5
These generalizations are best illustrated via specific examples. In what follows, I investigate the dynamics of uniqueness and typicality associated with three different lists of exempla that appear in the younger Seneca’s De Ira, a philosophical dialogue in three books on controlling and exterminating the vice of anger. I consider how the exempla in these lists relate to the moral categories under which they are marshaled, and to the framing arguments that these moral categories are supposed to illuminate. I also examine the internal dynamics of these lists, particularly how they work to sustain, modify or undermine the argument overtly being made. Finally, I am concerned throughout with the rhetorical mode in which these exempla are adduced. Exempla in a list always at least purport to illustrate the category that subsumes them, and, in some cases, illustration seems to be the full extent of their rhetorical operation. Often, however, befitting a deliberative or hortatory rhetorical context, at least some degree of injunctivity is also present—that is, anything from a subtle hint to an explicit assertion that the exempla provide models of moral action for readers or addressees to imitate or avoid, or at the very least furnish moral standards for readers to adopt as their own when evaluating the actions of others.6 Now, while this study focuses on a single work of Seneca, my broader investigation of Roman exemplarity suggests that many of these observations and conclusions apply equally to lists of exempla found elsewhere in Seneca, and in other Roman authors as well; Seneca’s praxis in this work is no more (or less) illuminating of the phenomenon than that of any other author.7 Accordingly I will point to similar phenomena elsewhere in Seneca and in other authors, and cite parallel arguments by other scholars, whenever possible.
Now to Seneca. Towards the middle of De Ira book one he addresses the question whether anger is ever useful.8 In particular, he evaluates the a fortiori claim (Ira 1.7.1) that anger is useful in war, if nowhere else, since it rouses the spirit and incites brave deeds by spurring men into danger. As a first step towards refuting this claim, Seneca argues generally that vices cannot be controlled and therefore should never be admitted intentionally (§§7–10). He then brings on an interlocutor to endorse the utility of anger in war, while expressly rejecting this contention in his own authorial voice (§11.1): “’But against the enemy,’ someone says, ‘anger is necessary.’ Nowhere less so, where attacks ought not to be disorderly but controlled and managed.” He then supports his counterassertion by adducing a series of exempla. First, he lists barbarian peoples whose native courage has been nullified when fighting the Romans because, lacking discipline, they yielded to anger (§11.2–4). Then he turns to a series of three Roman generals who, he declares, succeeded precisely because they controlled their anger:
(1.11.5) How else did Fabius restore the state’s debilitated forces than by knowing how to hesitate, drag things out, and delay (quod cunctari et trahere et morari sciit)—all things that angry men are unacquainted with? The state, which was then standing on the precipice, was done for had Fabius been as rash as anger was urging. But he kept the state’s circumstances (fortuna publica) under advisement and, having assessed its resources, from which nothing could now be lost without losing everything, he set aside grief and revenge (dolorem ultionemque seposuit), focusing exclusively on advantage and opportunity: he conquered his anger before he conquered Hannibal (iram ante vicit quam Hannibalem). (6) What of Scipio? Did he not abandon Hannibal, the Punic army, and everything at which one should get angry (relicto … omnibus … quibus irascendum erat), and transfer the war to Africa—but so sluggishly that he gave spiteful people the impression that he was slothful and self-indulgent (tam lentus ut opinionem … segnitiae … daret)? (7) What of the other Scipio? Did he not sit a good long time before Numantia, calmly bearing the chagrin that was his own and the people’s (hunc suum publicumque dolorem aequo animo tulit), that it was taking longer to conquer Numantia than Carthage? While he invested and blockaded the enemy, he brought them to the extremity that they fell by their own swords. (8) Consequently, not even in battle or war is anger useful: the fact is, it’s prone to rashness, and while it longs to create danger, it does not guard against danger. Martial valor is surest which looks around itself a good long time, governs itself, and advances slowly and deliberately.
Let us first distinguish the moral category under which these exempla are marshaled from the framing argument in support of which that moral category is invoked. The framing argument is “anger is neither beneficial nor useful in war”—initially articulated at §11.1, and then reprised at greater length in the summing-up at §11.8. The moral implications of this argument are clear: Seneca is examining the interaction of the vice ira (anger) with the virtue virtus (military valor). The moral category he constructs, then, consists of “warriors who were valorous and either (1) yielded to anger or (2) suppressed it.” The exempla of courageous barbarians defeated due to uncontrolled anger substantiate the first subset of this moral category, while the exempla of Roman generals victorious due to restraining angry impulses substantiate the second.9 Yet Seneca has gone to some length to make these generals fit into this moral category. Consider first his treatment of Fabius Cunctator, the famous “delayer” of the Hannibalic war. Three themes that commonly surface when Fabius’ “delaying” episode is invoked—that Fabius is “slow,” that he is waiting for fickle fortune to turn to Roman advantage, and that he prioritizes the safety of the commonwealth over the pursuit of military glory—all appear in Seneca’s handling of the exemplum: in the cluster of thematic “delaying” words (cunctari, trahere, morari); in the statement that Fabius kept the state’s overall circumstances (fortuna publica) in mind, and in the claim that he marshaled its depleted resources prudently. Yet Seneca here organizes these themes in a novel way by suggesting that they all manifest Fabius’ control over anger, a motif not prominent in the Fabian tradition.10 Thus, Seneca introduces a small but significant modification to a hoary exemplum, making it fit the moral category and support the framing argument.
The exempla of the Scipios, as presented here, neatly replicate the main features of the Fabius exemplum. The elder Scipio, like Fabius, is said to have separated himself from anger-inducing stimuli (relicto omnibus … quibus irascendum erat, cf. [Fabius]dolorem ultionemque seposuit), and was strikingly “slow” in his military movements—to the extent that he incurred blame, as if his failure to rush into battle indicated a lack of courage or valor (tam lentus ut opinionem… segnitiae … daret). While Fabius himself is not described here as cowardly, this accusation appears elsewhere in the Fabian tradition, hence is part of the broader Fabian discourse in which this presentation of Scipio participates.11 Yet Scipio is more usually presented as an aggressive figure, whose style of generalship contrasts sharply with the cautious Fabian approach.12 To present Scipio in Fabian guise seems to be Seneca’s own innovation, as he seeks additional examples to substantiate this moral category and to support his framing argument. The younger Scipio is likewise presented here as Fabian: immune to the stimuli that are elsewhere said to arouse anger, conspicuously slow and deliberate in his military movements, and—like his elder kinsman, and like Fabius elsewhere in the tradition—allegedly subject to moral condemnation (hunc suum publicumque dolorem) for not moving more quickly. Taken at face value, then, these three exempla substantiate their governing moral category (valorous, anger-suppressing warriors) and support the framing argument (that anger is useless in war)—but only once the received Fabian exemplum has been modified to include anger control, and the received Scipionic exempla been entirely reengineered to resemble Fabius. If the moral category provides “the law” (as Derrida puts it) for these three exempla, then Fabius, the first exemplum of the three, provides “the other law” that equally governs the remaining two, and causes the Scipios, unexpectedly, to appear in Fabian guise. In the terms of cognitive linguistics, Fabius appears to stand (in Seneca’s mind, at least) as the “best example” of this category—the standard by which the quality of other members of the category is judged.13 Fabius’ centrality is shown by the attraction he exerts on the representation of the Scipios.
But are “illustration” and “substantiation” all that these exempla achieve rhetorically? Is Seneca’s argument here sufficiently relevant to the situation of his addressee that these exempla may carry injunctive force as well? The dialogue De Ira is expressly addressed to Seneca’s brother L. Annaeus Novatus. However, as is usual in Seneca’s philosophical prose, the second-person interlocutor (“you”), with whom Seneca speaks and often argues, slides readily from the specificity of the named addressee to the generality of the intended reader—it is really “us” he’s talking to, us Roman aristocrats who, Seneca imagines, constitute this work’s broader readership. Nothing in the text suggests that Novatus commands, has commanded, or imminently will command an army; hence these exemplary generals do not seem to provide immediately resonant, compelling models for how Novatus himself should act (i.e., suppressing anger to achieve victory).14 In due course, however, Novatus or any other Roman aristocrat of the Julio-Claudian era (“we,” the implied addressees) may indeed find himself commanding troops as a governor, prefect, or imperial legate. Thus, while these exempla represent a scenario that is not out of the question for a contemporary aristocrat, their rhetorical mode here seems primarily illustrative, and not particularly injunctive.
These exempla do, however, contribute to the larger argument in De Ira that anger is never useful, an argument of which other parts relate more directly to Seneca’s addressees. To examine these parts, we jump to the third book. This book deals with the anger aroused by our interactions with other people, especially those distinctly higher or lower in status than ourselves (i.e., our rulers and our slaves), and with how we may control or avoid such anger.15 The book’s central portion contains numerous exempla of rulers who grew angry at provocations by their underlings, to the cost of both parties; a smaller number of exempla involve rulers who beneficially repressed or controlled their anger in such situations. Let us consider the latter group first. The latter group is introduced as follows (Ira 3.22.1): “These [preceding] exempla should be regarded as examples to avoid; while the following, conversely, are to be followed (haec … exempla quae vites, et illa … quae sequaris): they are moderate and gentle, where neither a reason to grow angry, nor the power to punish, was lacking.” Here we may again distinguish the moral category that subsumes the exempla from the framing argument in support of which the moral category is adduced. The framing argument is that anger indulged by powerful people is destructive to their underlings and themselves. The moral category containing the exempla is, as earlier, constructed from the interplay of a canonical virtue and vice: it consists of rulers in whom the virtue moderatio (moderation) prevails over the vice ira. Here, I summarize the exempla, as they are narrated at length. First Seneca offers three separate anecdotes involving King Antigonus the One-Eyed, two describing his restraint in the face of insults from his soldiers, and one describing his lenient treatment of defeated enemies who had verbally abused him (§22.2–5). He then relates a tale in which King Philip II of Macedon responded moderately to insults received from Athenian ambassadors (§23.2–3). Finally, he describes at greater length how the Roman emperor Augustus handled an historian called Timagenes who wrote insultingly about the imperial family (§23.4–8). After several warnings, Augustus banned Timagenes from his house, whereupon he took up residence with the prominent orator and ex-consul Asinius Pollio and attained great notoriety. Augustus did nothing further than to needle Pollio by saying he kept wild animals at his house. Here the exempla conclude, and Seneca returns to the frame:
Therefore let each person say to himself, whenever he is provoked: “Am I more powerful than Philip? Yet he was spoken ill of without retaliation. Have I greater power over my own household than the deified Augustus had over the whole world? Yet he was content to withdraw from his abuser. (2) Why should I avenge with whips and shackles my slave’s cheeky reply, or insolent look, or a mutter I can’t quite make out? Who am I, that it’s a sacrilege for my ears to be offended? Many have pardoned foreign enemies; am I not to pardon the lazy, the careless, the talkative? (3) Let a boy be excused by his age; a woman by her sex; someone from outside the household by his being a free man; a member of the household by his intimacy (etc.) (3.24.1)
Let us consider the powerfully injunctive rhetoric of the framing argument. Seneca introduces these exempla, as noted earlier, by explicitly declaring, “you should avoid those, and follow these” (quae vites, quae sequaris, §3.22.1), and concludes with the passage just quoted, where he puts into our mouth (for the addressees are still Novatus and, implicitly, “us”) words by which we may exhort ourselves to proper conduct. Besides this, the concluding exemplum involves Augustus, a Roman like “us” who lived recently: in contrast the prior exempla, besides being narrated more briefly, are temporally and culturally more distant, as they involve foreign kings from earlier times.16 Furthermore, the Augustus exemplum concerns the emperor’s treatment of an intimate member of his own household. This exemplum is most relevant to “us,” as Seneca argues in the passage just quoted, precisely insofar as “we” Roman aristocrats analogously dominate other members of our own households. The argument proceeds a fortiori: if those kings (especially Augustus), with their greater power over their underlings, could restrain their justified anger, all the more should I, whose power over my underlings is less.17 Thus, as the list of exempla is extended, there is a crescendo in their relevance to “us.” All the exemplary rulers demonstrate the requisite moderatio when provoked, but the concluding and longest exemplum of Augustus is most similar to “our” situation and, consequently, most highly injunctive. Augustus, in short, shows us how we should behave.18 In this list, then, the (Derridean) “other law,” or the (cognitive) “best example,” is embodied not in the first exemplum of the list, as with Fabius and the Scipios, but in the last. This structure supports a rhetoric that, in addition to being illustrative, is also strongly injunctive.
We now turn back to the earlier and larger set of exempla in De Ira 3—those involving rulers who grew angry when irritated by underlings, to the cost of both parties. To control such anger, Seneca exhorts us to erect barriers (impedimenta) to our vices, and to order our minds such that, even if we suffer an unexpected affront, we either do not feel anger or we suppress it (Ira 3.13.6). He continues, “It will be apparent that this can be done, if I offer up a few exempla out of a huge array, from which one can learn two things: how much evil anger has when it exploits the full power of extremely powerful men; and how much it can command itself, when it is constrained by greater fear” (§13.7). The framing argument, then, is that we can obstruct anger and thereby avoid the damage caused by yielding to it. The moral category supporting this argument is twofold: rulers who yield to anger, to their own cost and that of others; and people in whom anger is suppressed by fear. Two exempla follow, each adapted (with major alterations) from Herodotus, in which an angry ruler commits an outrage against a courtier, while the courtier, though angry about this mistreatment, responds spinelessly.
The first exemplum is as follows.19 At a Persian royal feast, a courtier called Praexaspes makes so bold as to suggest to the king, Cambyses, that he should drink more moderately. The king, affronted, fuels his pique by drinking all the more. He then demonstrates his lack of impairment by shooting an arrow through the heart of Praexaspes’ son, even cutting open his chest to show the arrowhead lodged in the boy’s heart. He then asks Praexaspes how his aim is. The courtier replies that not even Apollo could have shot truer (§14.1–2). Seneca’s commentary on this tale follows. He starts by excoriating the courtier: “May the gods curse that man, a slave in spirit more than in station! He praised a deed that it was too much even to witness. The splitting of his son’s chest into two parts and the quivering of the heart from its wound he deemed an opportunity for flattery.”20 Then he excoriates the king: “O bloody king, worthy that the bows of all his nearest and dearest be turned against him!…”—though, he adds, on balance the courtier is worse: “… But though we may execrate the man who ends a banquet with punishments and corpses, still it was more criminal to praise the shot than to make it.” Now, Seneca’s description of the king’s actions duly substantiates the moral category of rulers who indulge their anger to the cost of others (§13.7). His invective stresses the evil done, and constructs the king’s performance as a misdeed, an example to be avoided. This invective may also implicitly support the framing argument that barriers to anger can (and should) be erected, since Seneca criticizes the king for doing precisely the opposite—namely, removing all inhibitions by drinking more heavily. The example of the father/courtier Praexaspes, meanwhile, seems to demonstrate that the anger he felt (iratus, §14.5) has indeed been suppressed—whether by fear, as stated in the moral category, or by the opportunity to offer flattery, as Seneca subsequently says. Yet Seneca’s denunciation of Praexaspes for his “slavish” mind and “criminal” words does not seem to mesh closely with the framing argument, which states negatively that people should suppress or avoid feeling anger when provoked, but does not indicate positively what an acceptable response might look like. Praexaspes does seem to have suppressed or avoided feeling anger: so why does Seneca criticize him, rather than praise his self-restraint?21 In fact, as though perceiving that his invective against Praexaspes has exceeded his own remit, Seneca now makes a show of calling himself back to the precise terms of his framing argument: “We shall see how the father should have conducted himself … for now, the point under discussion [i.e., as asserted at §13.7] is clear: anger can be suppressed” (§14.4).
To clarify the operation of this exemplum, let us now examine the second exemplum involving an angry ruler and his courtier—another Herodotean anecdote, heavily modified to render it a virtual doublet of the first. A Persian king, angered at unspecified advice given by his courtier Harpagus, cooks up Harpagus’ children and serves them to him at a banquet. He then reveals the crime by presenting their heads to Harpagus, and inquires how he liked the food. The courtier replies, “At the king’s table, every dish is delightful” (§15.1).22 Seneca’s commentary on this anecdote is as follows:
What did he gain by this flattery? That he not be treated to what remained. I don’t forbid the father to condemn his king’s deed, or to seek a worthy punishment for such a savage monster. But for now I conclude the following (sed hoc interim colligo): that anger arising even from monstrous evils can be buried away and compelled to speak words antithetical to itself. (3) Such reining-in of one’s anguish is necessary, especially for those allotted this kind of life and invited to a king’s table. (3.15.2)
Here, in contrast to the previous exemplum, Seneca refrains from excoriating the courtier, and even finds a way to excuse him (“he might have had to finish the dish …”). Still, Seneca makes a point of asserting that Harpagus would be justified in seeking to punish the king for the injury suffered, and only then expressly calls himself back to his argument (“but for now I conclude this: …”), which is the narrower contention that anger can be suppressed.
In relating these two anecdotes, then, Seneca presents his exempla as becoming unruly, as burgeoning with meaning that exceeds the strictures of the framing argument and demands extra moralizing commentary. Specifically at issue is a moral conflict between the framing argument’s contention that controlling anger is both possible and beneficial, and the awkward fact, revealed by these courtiers’ craven replies, that controlling one’s anger does not necessarily result in virtuous action. And why would it? For to control anger through fear or the desire to flatter is merely to subject one vicious affect to the tyranny of another. This moral complexity, boiling up from within these two exempla in their particularity, seriously complicates the contention of the framing argument that controlling this vicious passion is ipso facto good, or at least avoids harm.23 As with the lists of exempla discussed earlier, then, these anecdotes involving kings and courtiers generate particular moral dynamics that seem to undermine—and draw Seneca away from—the framing argument. In Derridean terms, an “other law” emerging from the exempla themselves—revealed in the first exemplum, and reiterated in the second—threatens to overturn altogether “the law” ensconced in the framing argument.
Yet Seneca needn’t have presented his argument in this way. He could simply have included in the framing argument (“the law”) the assertion that controlling anger by means of other vicious passions neither avoids harm nor leads to virtuous action. Had he done so, the moral complexities generated by the particulars of these two exempla, and the moral commentary he consequently provides, would have had their place supporting the framing argument. Instead, however, Seneca has chosen to present a narrower framing argument, to adduce exempla whose implications (as he makes a point of showing) exceed and even resist that argument’s strictures, and to depict himself as being initially swept away by outrage, before recalling himself to the narrower bounds of the framing argument.24 Why present this material in this way? I offer two conjectures. First, Seneca as author performs precisely the feat that he here argues is possible, and that he has adduced these exempla to illustrate. For having represented himself as growing angry at Praexaspes’ craven conduct, and as harboring reservations about that of Harpagus, he makes a show of suppressing and controlling his anger or concern, thus preventing such affects from carrying him away from his purpose—which is, precisely, to persuade us that anger can be controlled.25 Second, Seneca’s stated moral reservations about the courtiers’ conduct, especially his dramatic denunciation of Praexaspes, introduces a strongly injunctive dimension to exempla otherwise presented as “merely” illustrative. For while he insists he is demonstrating only that anger can be controlled even in the face of grave provocation, and while he presents these exempla as illustrating this possibility, his moralizing interventions nevertheless make clear that “we” should do it differently from the courtiers in these anecdotes (i.e., find a way to conjoin anger suppression with virtuous action). He thereby foreshadows his next major contention: that an aristocrat faced with unspeakable, intolerable evils at the hands of his ruler can escape by committing suicide (§15.3–5). Indeed, the relevance of these Persian exempla to Seneca’s Roman aristocratic addressees, whether Novatus or “us,” is patent. As he declares in the passage quoted above, the plight of aristocrats interacting with rulers is always and everywhere the same: “[so it is] for those who are allotted this kind of life, and invited to a king’s table.”26 We (Roman aristocrats) also sometimes dine and have other social interactions with the emperor. We are thereby exposed to the damage a ruler who indulges his anger can inflict, which we, in turn, may aggravate if we show our own anger. For we are in his power as (if) slaves to masters, as Seneca’s language in this and subsequent sections stresses.27 The paradox of the aristocratic courtier’s position within a monarchic structure is that, in his relationship to his ruler, he is literally a subject but may be figured metaphorically as a slave; whereas in his relationships to his numerous slaves, freedmen, clients, and other dependents, he is literally a master and paterfamilias but may be figured metaphorically as a king. Hence, in this section of the work, Seneca can present us to ourselves in “slavish” guise as inferiors who suffer injury, while in De Ira 3.24, as discussed above, he presents us to ourselves in “kingly” guise as superiors who potentially inflict injury.28
The kind of analysis I have provided in this chapter could be extended indefinitely. There are many dozens of passages in Seneca, and hundreds—perhaps thousands—of further passages elsewhere in Roman literature, in which a list of exempla is gathered under a moral category that is adduced to support some aspect of a framing argument. In most and perhaps all such passages, understanding how the list works in context involves questions of the sort I have raised in this chapter: how similar the individual exempla are to the situation by which they are framed, and to what extent they substantiate their governing moral category; how the exempla individually interact with the framing argument, and with one another; what kinds of “other laws” or “prototype effects” emerge from these interactions; what consequences such dynamics have for the framing argument (i.e., supporting it, undermining it, or doing something else); and to what extent the exempla are illustrative or injunctive. That is to say, I am claiming that the three passages analyzed in this chapter are “typical” of lists of exempla: that the kinds of phenomenon observed here may be found in other lists as well, and that the kind of analysis carried out here may be useful in understanding those other lists. For if this is not the case, why bother writing this chapter at all? Nevertheless, as these analyses also demonstrate, each particular list of exempla works differently in terms of how the individual exempla do or do not substantiate the moral category subsuming them, what kinds of dynamics the exempla generate among themselves, and what kinds of resonance they create individually and collectively with the framing argument. That is to say, I am also claiming that each of the three lists analyzed in this chapter is unique, and therefore that none of my analyses here will be precisely applicable to any other list of exempla, in Seneca or elsewhere. It therefore seems fitting, in conclusion, to emphasize that my own analysis—based as it is on creating a list of examples of exempla in a list—cannot escape, and indeed necessarily (re)performs, precisely the problematic it has set out to analyze.29
Endnotes
1For this model of Roman exemplarity, Roller 2004: 1–7. Scholars use the word “exemplum” variously to refer to the actor/performer him- or herself, to the action s/he performs, or to a reference to or narrative thereof (as here). When narrated, an exemplum typically displays the “paratextual” or “paranarrative” feature of being exterior to the main narrative, with different characters and temporality, even as it interacts with the main narrative in various ways: e.g., Lyons 1989: 30–1; Alden 2000: 1, 23–38.
2Among many scholars who discuss the dual ontology of exempla are Barchiesi 2009: 45–7, Belknap 2004: 15–16 (speaking generally of items in a list), Gelley 1995: 1–3, Harvey 1992: 208, 214–15.
3For Aristotelian categories, see e.g., Arist. Cat. 1b9–2a10, Top. 103b20–38, cf. Metaph. 1068a8–10. In Aristotelian theory, membership in a category (e.g., “people who controlled their anger”) is affirmed or denied via a predicate attached to the subject (“So and so controlled/did not control his anger”). Aristotle’s term for “predicate” is κατηγορία—hence the term “category” that moderns use to describe such groupings. The term “classical category” (in the sense of Aristotelian predicate) is current both in philosophical discussion of categorization (e.g., Harvey 1992: 198) and in cognitive-linguistic discussion (e.g., Lakoff 1987: 6–11, Murphy 2002: 11–28).
4The question of how we abstract and generalize at all, since we live in a phenomenal world that confronts us only with discrete, particular objects and events, has engaged philosophers from Plato (who articulated the theory of forms) to Derrida (who labels the whole question “exemplarism”: e.g., Hollander 2008: 1–2, 51–4). Cognitive linguists (next note) have pursued this question from a different perspective. On the “passage” of the exemplum from individuality to generality, Hollander loc. cit.; Harvey 1992: 199, 213–15; also Waldenfels and Goodrich in this volume.
5On the internal dynamics of exempla in lists, Sammons 2010: 17–20 (on Homeric catalogues), Langlands 2008: 162–3, Lucarelli 2007: 287 (both on Valerius Maximus); Hinds 1999 (Ovidian and Hellenistic catalogues of virtuous women); below on Seneca. For such dynamics in lists more generally, Belknap 2004: 6–8, 16–35; Mainberger 2003: 42–61; Barney 1982: 192–6. On Derrida’s “law” and “other law,” Harvey 1992: 196–206. On “prototype” or “typicality” effects in cognitive linguistics, Lakoff 1987: 12–154, with Rosch 1978: 35–41 and Murphy 2002: 22–64.
6The terms “illustrative” and “injunctive” were suggested to me by Noel Carroll. On this contrast, e.g. Barchiesi 2009: 46, Roller 2004: 52–3, Chaplin 2000: 137–40 (with different terminology); Harvey 1992: 195, 208.
7Seneca is, however, the only Roman author who engages in explicit philosophical reflection on exemplarity, discussing how individual actions pass into moral concepts of general relevance: see Ep. 120, with Inwood 2005: 283–99, Langlands (this volume), and Roller (forthcoming). Reflection on the rhetorical function of exempla, conversely, is widespread among Roman authors: e.g., Rhet. Her. 4.62; Cic. Inv. 1.49, Or. 120, Off. 1.115–21; Quint. Inst. 5.11–13.
8For this book’s structure, Fillion-Lahille 1984: 284–6.
9The angry and therefore militarily vulnerable barbarian is an ancient stereotype: Harris 2001: 222–3; Giacchero 1980: 182 (particularly on the exempla in De Ira 3 of Persian kings indulging anger: see below).
10On the Fabian tradition, Roller 2011: 185–200 (delay), 186–8 (fortune), 191–3 (safety). Only in Silius, to my knowledge, is the idea of Fabian anger broached: in a speech, Fabius urges his son to set aside anger (7.555–65), perhaps implying that Fabius also controls his own.
11On the Fabian strategy as cowardly and dishonorable, Roller 2011: 189–91.
12For sources, Roller 2011: 197 n. 52, 199 n. 65.
13Lakoff 1987: 39–46 for these terms (summarizing Rosch); the “best example” is a kind of “prototype effect.” Similarly, Langlands 2008: 175–8 discusses a list of exempla in Val. Max. 2.7, showing how later figures in this list explicitly or implicitly emulate earlier ones.
14For L. Annaeus Novatus/L. Iunius Gallio Annaeanus (as he had become, by adoption, before 52 CE), RE 1 cols. 2236–7, s.v. Annaeus (12) (Rossbach, 1899); PIR2 I 757. His attested magistracies are a proconsulship in Achaea in ca. 51–2 CE, and a two-month suffect consulship in 55 or 56. In neither office would he have commanded soldiers or faced an enemy, and it is unlikely he ever did so, as he is not known to have gone on to a proconsulship or governorship of a province with legions. In any case, De Ira was almost certainly completed no later than 52 (Griffin 1976: 396–8), well before he might have assumed such a post.
15On the structure of book three, Fillion-Lahille 1984: 287–90.
16Val. Max. 1.6. ext. 1 declares that “external” exempla are less authoritative than Roman ones (Skidmore 1996: 89)—a bias not limited to Valerius.
17Malchow 1986: 516 discusses the argument. A fortiori argumentation is commonly used when exempla are adduced to support an exhortation. Quint. Inst. 5.11.9–12 discusses two varieties: “from lesser to greater” and “from greater to lesser” (as here). For Homeric examples, Willcock 1964: 141–3; Alden 2000: 25, 37 (describing such argumentation without naming it). Examples in Seneca abound: e.g., Ira 3.13.4 (if Socrates allowed his friends to criticize him, all the more should you); also 3.25.1–2. Exempla adduced a fortiori are especially common in consolation: e.g., Polyb. 14.2–16.4, and passim in the consolatory dialogues, with Mayer 1992: 158–62; also Willcock 1964: 145–6 (Homeric examples).
18Mayer 1991: 155–57 discusses the “crescendo” structure in Senecan lists of exempla generally; see Malchow 1986: 513 on this passage.
19Herodotean “original:” Hdt. 3.34–5. On Seneca’s version, Setaioli 1981: 385–8, Lavery 1987: 281, Giacchero 1980: 183–6.
20In Hdt. 3.35, the courtier’s reply is said to be motivated by fear for himself. Had Seneca provided this motive, the exemplum would more neatly substantiate the moral category of “cases of anger suppressed by fear”—though would still likely earn Seneca’s disapproval: cf. Ira 2.33.6 (n. 23 below), where fear for oneself is derogated.
21Indeed, in §14.5–6 Seneca somewhat grudgingly rehearses arguments that extenuate Praexaspes’ words (potest dici merito devorasse verba, etc.: see Malchow 1986: 474). Also, in the closely parallel cases of the courtiers Harpagus (Ira 3.15.1–2; below) and Pastor (Ira 2.33.3–6; n. 23 below) he refrains from invective, recognizing that they could suffer even greater harm should they object.
22In Hdt. 1.108–19, Harpagus is a courtier to Astyages, king of the Medes. Astyages is angered not at Harpagus’ advice, but at his failure to do away with the infant Cyrus (§117–18). Seneca has changed these details in order to make this exemplum align more closely with that of Praexaspes, as Malchow 1986: 475–6 correctly sees. Contra Setaioli 1981: 387–8, who attributes the divergences from the Herodotean source to Seneca’s faulty memory. Homeric paradeigmata are also likely constructed or altered to bring them into significant concurrence with the framing situation: Willcock 1964; Alden 2000: 26–7, 39. On the Harpagus episode cf. Lavery 1987: 281–2, Giacchero 1980: 182–3.
23In the closely parallel case of the courtier Pastor whose son Caligula murders, Pastor restrains his anger in order to protect a second son. Here Seneca can commend Pastor’s response, since it was not fear for himself (timor, a vice) but a feeling of obligation to a family member (pietas, a virtue) that suppressed his anger: contempsissem … patrem, si sibi timuisset: nunc iram compescuitpietas (Ira 2.33.6).
24Many scholars have observed that examples, in their particularity, inevitably produce “excess” signification beyond what any governing generalization can cover (esp. Lyons 1989: 34; also Harvey 1992: 195–6, Goldhill 1994: 58–9, 70). In general I agree (pp. 82–3 above): here I contend only that Seneca could have included this particular detail in his framing argument, and that in not doing so he gives himself scope to develop the theme of anger control in surprising ways.
25I thank Susanne Lüdemann for pointing out the self-referential quality of Seneca’s own display of anger control. Note, however, that the courtiers’ anger is reined in by the power of another vicious affect (fear, ambition), while Seneca’s anger is reined in by the power of his own rational discourse. Thus he “tops” his exemplary courtiers. Brev. 13.8 displays a similarly studied calling-back to the argument, following a moralizing “digression” that actually supports the point being made.
26Ira 3.15.3; similarly at §14.6. In contrast, in the list of exempla at §3.22–3, analyzed earlier, the foreign exempla are presented as more distant and less authoritative than the concluding Roman example of Augustus. Regarding the current passage, Giacchero 1980: 181–9 connects the high relevance of these Persian courtiers with Seneca’s condemnation of Caligula in this work (e.g., Ira 3.18.3–19.5, and elsewhere), suggesting that Seneca aims to present the latter as an oriental despot; similarly Ramondetti 1996: 239–53.
27Aristocrats presented as slaves: those who dine with kings potentially constitute a slave workshop (ergastulum, §16.3), and their condition is slavery (servitus/servientes, §15.3, 4, 16.1); they must not wait for someone else to release them from slavery (vindicet, §15.4); suicide is in their power and provides a direct path to freedom (libertas, 4x in §16.4).
28On the master–slave relationship as a metaphor for the emperor–aristocrat relationship (and vice versa) in the early Imperial period, Roller 2001: 213–87.
29I thank Michèle Lowrie and Susanne Lüdemann for stimulating conversation and critiques, along with the other participants in the splendid conference at the University of Chicago where this argument was first aired. I further thank Andrew Riggsby and Gareth Williams for horizon-broadening comments on an earlier version.
References
Alden, M. (2000) Homer beside Himself: Para-narratives in the Iliad, Oxford: Oxford University Press.
Barchiesi, A. (2009) “Exemplarity: between Practice and Text,” in Y. Maes, J. Papy, and W. Verbaal (eds), Latinitas Perennis Volume II: Appropriation and Latin Literature, Leiden: Brill, 41–62.
Barney, S. (1982) “Chaucer’s Lists,” in L. Benson and S. Wenzel (eds), The Wisdom of Poetry, Kalamazoo: Western Michigan University Press, 189–223.
Belknap, R. E. (2004) The List: the Uses and Pleasures of Cataloguing, New Haven: Yale University Press.
Chaplin, J. (2000) Livy’s Exemplary History, Oxford: Oxford University Press.
Fillion-Lahille, J. (1984) Le De Ira de Sénèque et la philosophie stoïcienne des passions, Paris: Klincksieck.
Gelley, A. (1995) “Introduction,” in A. Gelley (ed.), Unruly Examples: On the Rhetoric of Exemplarity, Stanford, CA: Stanford University Press, 1–24.
Giacchero, M. (1980) “Le reminiscenze erodotee in Seneca e la condanna di Caligola,” Sandalion, 3: 175–89.
Goldhill, S. (1994) “The Failure of Exemplarity,” in I.J.F. De Jong and J.P. Sullivan (eds), Modern Critical Theory and Classical Literature, Leiden: Brill, 51–73.
Griffin, M. (1976) Seneca: a Philosopher in Politics, Oxford: Clarendon Press.
Harris, W. V. (2001) Restraining Rage: the Ideology of Anger Control in Classical Antiquity, Cambridge, MA: Harvard University Press.
Harvey, I. (1992) “Derrida and the Issues of Exemplarity,” in D. Wood (ed.), Derrida: a Critical Reader, Oxford: Blackwell, 193–217.
Hinds, S. (1999) “First Among Women: Ovid, Tristia 1.6 and the Traditions of ‘Exemplary’ Catalogue,” in S. Braund and R. Mayer (eds), Amor: Roma: Love & Latin Literature, Cambridge: Cambridge University Press, 123–142.
Hollander, D. (2008) Exemplarity and Chosenness: Rosenzweig and Derrida on the Nation of Philosophy, Stanford, CA: Stanford University Press.
Inwood, B. (2005) Reading Seneca: Stoic Philosophy at Rome, Oxford: Oxford University Press.
Lakoff, G. (1987) Women, Fire, and Dangerous Things: What Categories reveal about the Mind, Chicago: University of Chicago Press.
Langlands, R. (2008) “‘Reading for the Moral’ in Valerius Maximus: the Case of Severitas,” CCJ, 54: 160–85.
Lavery, G. B. (1987) “Sons and Rulers: Paradox in Seneca’s De Ira,” L’Antiquité classique, 56: 279–83.
Lucarelli, U. (2007) Exemplarische Vergangenheit: Valerius Maximus und die Konstruktion des sozialen Raumes in der frühen Kaiserzeit, Göttingen: Vandenhoeck & Rupprecht.
Lyons, J. D. (1989) Exemplum. The Rhetoric of Example in Early Modern France and Italy, Princeton: Princeton University Press.
Mainberger, S. (2003) Die Kunst des Aufzählens. Elemente zu einer Poetik des Enumerativen, Berlin: De Gruyter.
Malchow, R. (1986) Kommentar zum zweiten und dritten Buch von Senecas Schrift “de ira” (= dial. 4 und 5), Erlangen-Nürnberg: Diss.
Mayer, R. (1991) “Roman Historical Exempla in Seneca,” in P. Grimal (ed.), Sénèque et la Prose Latine, Geneva: Vandoeuvres, 141–69.
Murphy, G. L. (2002) The Big Book of Concepts, Cambridge, MA: MIT Press.
Ramondetti, P. (1996) “Il tema della cena nel ‘De Ira’ di Seneca,” Atti della Accademia delle Scienze di Torino, t. 2: Classe di Scienze Morali, Storiche e Filologiche, 130: 213–53.
Roller, M. (2001) Constructing Autocracy: Aristocrats and Emperors in Julio-Claudian Rome, Princeton: Princeton University Press.
Roller, M. (2004) “Exemplarity in Roman Culture: the Cases of Horatius Cocles and Cloelia,” CP, 99: 1–56.
Roller, M. (2011) “The Consul(ar) as Exemplum: Fabius Cunctator’s Paradoxical Glory,” in H. Beck et al. (eds), Consuls and Res Publica: Holding High Office in the Roman Republic, Cambridge: Cambridge University Press, 182–210.
Roller, M. (forthcoming) “Precept(or) and Example in Seneca,” in K. Volk and G. Williams (eds), Latin Philosophy, Oxford: Oxford University Press.
Rosch, E. (1978) “Principles of Categorization,” in E. Rosch and B. Lloyd (eds), Cognition and Categorization, Hillsdale, NJ: L. Erlbaum Associates, 27–48.
Sammons, B. (2010) The Art and Rhetoric of the Homeric catalogue, Oxford: Oxford University Press.
Setaioli, A. (1981) “Dalla narrazione all’ exemplum. episodi erodotei nell’opera senecana,” Atti del convengo internazionale ‘Letterature classiche e narratologia’, Perugia: Istituto di filologia latina, 379–96.
Skidmore, C. (1996) Practical Ethics for Roman Gentlemen. The Work of Valerius Maximus, Exeter: University of Exeter press.
Willcock, M. M. (1964) “Mythological paradeigma in the Iliad,” CQ, n.s. 14: 141–54.