How and Why to Quote a Latin Poem

My analysis of Quintilian's quotational practice begins from the same insight that informs the best modern reassessments of this author: that Quintilian, in writing about rhetoric, is himself employing rhetoric, and that his pointed juxtaposition of theory and practice is a primary compositional principle of the Institutio. Footnote ⁸ Our understanding of Quintilian's use of poetic quotations, therefore, should emerge out of a comparison of Quintilian's articulation of the usefulness of such quotations for the rhetorician with the way he himself uses them. Carlozzo has collected many of the relevant passages.Footnote ⁹ Quintilian makes the memorization of poetry part of his educational program because ‘the ability to recall’ (relatio) material from previous writers is both ‘pleasing in conversation’ (iucunda in sermone) and ‘useful in court’ (in causis utilis, 2.7.4).Footnote ¹⁰ He defines this utility more specifically in an earlier passage, where he observes that ‘the greatest orators’ (summi oratores) use quotations of poetry either to provide ‘support for their arguments’ (ad fidem causarum) or as ‘ornamentation for their eloquence’ (ad ornamentum eloquentiae, 1.8.10); later in the Institutio Quintilian reiterates this link between poetic quotation and rhetorical authority, noting that ‘speeches are full of…reflections from the poets’ by which orators (and, Quintilian says, philosophers) ‘seek authority from many a passage of poetry’ (5.11.39).Footnote ¹¹ Carlozzo himself recognized that Quintilian's own use of quotations often does what Quintilian says it should do: he uses poetry to support his own arguments about how rhetoric should be taught and also adorns his own expressions with poetic phrases. It is true that many of Quintilian's quotations of poetry do not straightforwardly fit either of these uses, most providing examples of stylistic or rhetorical devices that Quintilian is discussing.Footnote ¹² But of the 320 quotations of poetry I examined, about 65 can be reasonably said to provide ornamentation or lend authority to an argument (or both),Footnote ¹³ and it is on examples of those kinds of quotation that I will focus here. These present an opportunity to assess Quintilian's practice in relation to his own theories rather than, as has previously been done, in relation to external standards of what quotation should be or accomplish.Footnote ¹⁴

That Quintilian has followed his own advice, so to speak, is clear from several quotations of poets that appear at significant moments in his work.Footnote ¹⁵ When evaluated in light of his own declarations that such quotations should ‘ornament his eloquence’ and ‘support his arguments’, we discover that the quotations that Quintilian has employed reflect both his care in selecting them and the sophistication of his understanding of the meanings that the passage he quotes can bear. Indeed, Quintilian has foregrounded the significance of poetic quotation in the Institutio by quoting, in the second sentence of the entire treatise, Horace's advice in the Ars Poetica that ‘publication should not be hurried but “kept in store till the ninth year comes round”’ (Ep. ad Tryph. 2, quoting Hor. Ars 388). This passage has been extensively studied, most recently by Ferriss-Hill, as evidence for the name by which Horace's literary critical poem was known in the first century, but its rhetorical force in Quintilian has hardly attracted any critical attention beyond Calboli's attempt to establish Quintilian's attitude toward Horace.Footnote ¹⁶ Peirano Garrison, however, has laid the groundwork for understanding the rhetorical significance of this opening quotation by noting that it ‘claims for the Institutio the literary refinement of poetry’ and ‘situates the Institutio within the prestigious tradition of literary and rhetorical criticism that the Ars Poetica exemplified’, an analysis that, I would add, corresponds to Quintilian's theories: the quotation makes an argument about the status of the Institutio and ornaments it with the expression of a prestigious author.Footnote ¹⁷ But any assessment of the force of this quotation must also note that whatever prestige its inclusion gains for the treatise is immediately complicated by Quintilian's declaration that he has rejected Horace's advice and published his work more quickly, because the public's demands for it had become too insistent to ignore (Ep. ad Tryph. 3 and Inst. 1.pr.2). Quintilian is not merely basking in the glow of Horace's reflected genius but, one senses, asserting a difference between them, as if to say that Horace's advice might make sense for a poet, but that the author of a work on a subject of public significance for which there is widespread demand will not have the luxury to polish it endlessly. Quintilian's citation of this quote and his immediate rejection of its argument is not, as some critics have claimed, apologizing for the Institutio's supposed lack of polish,Footnote ¹⁸ but an assertion of the greater significance of rhetoric as a discursive form of knowledge over and against Horace's attempt, which I discuss in a forthcoming essay, to make rhetorical theory a subcategory of poetics. At the same time, however, Quintilian's rejection of Horace's nine-year waiting period reflects his understanding that, as Ferriss-Hill puts it, Horace was ‘playfully disingenuous in suggesting there is a “magic number” that will suffice to produce a worthy poem.’Footnote ¹⁹ With this, the rhetorician joins Horace in claiming that the consummate artist, whether poetic or rhetorical, knows how, and how quickly, to produce a masterful work.

Two books later, Quintilian begins the first chapter of his third book, which Russell categorizes as ‘the introduction to the main work’ (after two books of preliminary material), with another poetic reference, this time to Lucretius’ famous description of a doctor putting honey on the rim of a cup to disguise the bitter taste of wormwood as a metaphor for the poet disguising the allegedly bitter concepts of philosophy (Inst. 3.1.4, quoting Lucr. 1.936–8/4.11–13).Footnote ²⁰ As with the Ars Poetica quotation, this appears to be an example of poetry providing ornamentation—Lucretius’ lines about honey themselves serve as honey for Quintilian's style—and as an argument that making palatable the technical details of rhetorical theory requires a similar treatment to that which Lucretius used on philosophy. Critical analysis of Quintilian's quote has not progressed much beyond noting this, and that Quintilian has ‘inadvertent[ly]’ altered Lucretius’ text, according to Cole:Footnote ²¹ Quintilian begins with ac (‘and’) where Lucretius, in book one, began with sed (‘but’) and, in book four, with nam (‘because’), and replaces Lucretius’ contingunt (‘they touch’ the rim of the cup with honey) with a different word. Variation in the Institutio's manuscripts makes it difficult to know what word Quintilian used in this latter case, let alone to understand the effect of the change, but Quintilian's alteration of the first word of his quotation may gesture toward the fact that Lucretius himself used two different words in that position the two times he employed this metaphor (which he otherwise repeats word for word). Nethercut has recently argued that any understanding of this much discussed passage must incorporate the fact that although honey and wormwood may appear in book one to be opposites, by the time the metaphor recurs in book four, Lucretius’ readers have come to realize that honey and wormwood both consist of mixtures of the same atoms, and are thus not as distinct as they might first have appeared.Footnote ²² Quintilian's modification of the one word in Lucretius’ text that differentiates Lucretius’ two versions of it signals his understanding of this dimension of Lucretius’ repetition of the metaphor, as does his refinement of its relevance to his text. Immediately after invoking Lucretius as a model Quintilian acknowledges the inseparability of honey and wormwood, writing that ‘I fear my book may appear to have too little honey and too much wormwood, and be more healthy for the student than agreeable’ (3.1.5). And as will emerge in the course of my discussion, Quintilian means for the programmatic dimension of Lucretius’ metaphor to apply to the Institutio as well. According to Nethercut, the metaphor of the honeyed cup is not, as it has usually been understood, a simple statement of the subordination of form to content, but the exemplar of Lucretius’ technique of ‘provisional argumentation’, by which the poet introduces readers to a perspective that, later in the poem, they revise in light of the deepened understanding the poem has given them. This compositional strategy, as I will argue below, is also one that Quintilian employs.

The preface to Institutio book four is also marked as a significant passage by a dedication to the emperor Domitian. This passage, too, invokes poetic authority, albeit in a less specific way than the passages so far analyzed. Quintilian justifies the postponement of this dedication until so late in the work (book one had begun with a dedication to Marcus Vitorius) by reference to ‘the greatest poets’, who invoke the muses ‘not only at the beginning of their works, but also later on’ (4.pr.4). This bald appeal to poetic precedent, complete with superlative maximi, does not, however, specify what poets or passages Quintilian has in mind. Russell suggests that Quintilian means the re-invocation of the muses of the beginning of the catalogue of ships in the Iliad or in the second half of the Aeneid, but these are not really parallel to Quintilian's postponement of the emperor because Homer and Vergil begin with the muses and re-invoke them later, but Domitian is not mentioned at all until the preface to Quintilian's fourth book. Traditional critics of Quintilian might attribute this vagueness to his shaky grasp of literature; more recently the fulsome praise of Domitian that precedes it has prompted Roche to suspect the presence of irony.Footnote ²³ More may be said, however, about how this passage reflects on Quintilian as a rhetorician.

The irony that Roche identifies is counterbalanced, if not quite eliminated, by the juxtaposition of the preface to book four with Quintilian's presentation of the theory of how oratorical prefaces should be composed in the first chapter of the very same book this preface introduces. Just as Lucretius’ philosophical teachings produced students who could better assess the atomic basis of his own honeyed cup metaphor, this juxtaposition constitutes Quintilian's invitation to assess his rhetorical practice in this book's opening passage in light of his advice about oratorical prefaces. The best prefaces, Quintilian specifies, will do what he does with Domitian in the present passage: they will ‘address primarily the persons whom we wish to win over to our side’ (4.1.63); they will introduce topics calculated to ‘prepare the hearer to be more favorably inclined towards us for the rest of the proceedings’ (4.1.5), and will imply, as Quintilian's reference to his obligation to the emperor's nephews does, that the speaker has ‘undertaken the case out of a sense of duty to a relative or a friend or (best of all if possible) to his country’ (4.1.7). Quintilian's postponement of his dedication to the emperor has allowed him to locate it in a place where Domitian can take pride of place in a preface recognizably constructed to conform to many of the features of the ideal, or at least textbook-perfect, preface. But the student of those instructions who has also found hints of irony in Quintilian's praise of the emperor will also discover that Quintilian advises the orator that ‘it would be foolish of me to warn against saying anything explicitly or even hinting at anything unfavorable to [the judge], were it not that this does happen’ (4.1.11) and that ‘nowhere else is it more necessary [than in the preface] to be careful to avoid suspicion’ (4.1.56). Such a reader may also note that the vagueness of Quintilian's poetic justification for the placement of this dedication finds a corresponding vagueness in Quintilian's praise of Domitian's eloquentia, which, as Roche has noticed, lacks a specific example even though elsewhere Quintilian shows himself perfectly capable of citing examples of other emperors’ rhetorical skill.Footnote ²⁴ The preface to book four reveals Quintilian as a master rhetorician, simultaneously conversant in all the textbook-perfect techniques of the oratorical preface and yet willing to break those rules in spectacularly daring ways.

Since these dimensions of Quintilian's rhetoric emerge most distinctly to the student who examines Quintilian's rhetorical performances in light of how Quintilian teaches, it should be unsurprising that the final book of the Institutio Oratoria signals in various ways the need to reassess everything that has come before, just at the point when the reader has absorbed most of the author's teaching. The portrait of the orator in retirement that appears at the beginning of the final chapter of the treatise (12.11), so similar to Quintilian's own self-representation as a retired orator in the preface to its first book (1.pr.1–5), is one such signal that Quintilian the rhetorician has been present during the reader's instruction every bit as much as Quintilian the praeceptor has,Footnote ²⁵ Quintilian's employment of the term doctus (‘learned’ but also ‘one who has been taught’) in the middle of book twelve, another.Footnote ²⁶ But Quintilian also signals this fact in the preface to book twelve, where he introduces the final book of his treatise as ‘the hardest part of the task I set myself’ because it treats topics for which he has ‘no predecessor to follow.’ He conveys his sense of being lost at sea with a quotation from Vergil's Aeneid: ‘sky all around, and all around the deep’ (3.193, quoted at Inst. 12.pr.4). On first glance it is an apt quotation, not only because these words are spoken by Aeneas in a moment of uncertainty at sea but because, as Peirano Garrison notes, it echoes the seafaring metaphors with which Quintilian had begun the treatise in the preface to book one.Footnote ²⁷ But a closer inspection of the Vergilian context reveals that this quotation does not just look ahead to the closing book, but also back at the preceding eleven. Aeneas speaks this line as the Trojans depart from Crete, which they visited because they misinterpreted Apollo's prophecy telling them to ‘seek [their] ancient mother’ (Aen. 3.96). That is, the line that Quintilian quotes comes not, as he seems to quote it to mean, at the end of a successful journey through familiar waters before boldly setting out into uncharted territory, but after a foolhardy wild-goose chase that forced the Trojans to endure a ‘deadly year’ (letifer annus, 3.139) of plague and starvation before they realized their error and finally put themselves on the right course. It comes, that is, at a moment when the Trojans must reinterpret information that they had believed they had understood. Quintilian's use of this quotation near the end of the Institutio therefore raises the question of how much of the previous books should be reassessed from the perspective the reader has gained by the time they reach the last one. This compositional technique has been observed in other imperial Roman authors,Footnote ²⁸ but it is particularly apt for a didactic work like the Institutio in which the reader does not fully develop whatever skills the work is teaching until reaching the end of its final book. Here Quintilian notifies his reader that if they wish to achieve a true understanding of what has come before, they must return to the beginning and see what their training allows them to see.

The Master at Work

What they see, at least when they begin examining the ways that Quintilian has used poetry to ornament his text and support his arguments, is a rhetorician deeply versed in literary history and dazzlingly virtuosic in his appropriation of the authority of that tradition in support of his stature. Consider, for example, another passage in which critical study has not advanced beyond noting the differences between Quintilian's text and that of his source material. At Institutio 2.13.8, Quintilian signals the culmination of his arguments about the proper balance between doing ‘what will help one win’ (quid expediat) with ‘what is honorable’ (quid deceat) by quoting a Vergilian phrase, ‘I shall enjoin, repeat, and urge again, again’ (praecipiam ac repetens iterumque iterumque monebo). Almost the same line appears at Aeneid 3.436, although Quintilian has praecipiam (‘I shall enjoin’) where Vergil has praedicam (‘I shall foretell’). It is a textbook (so to say) example of using a poetic quotation to ornament a text and bolster an argument because with it Quintilian emphasizes the point he is making. As for the alteration to Vergil's text, an explanation for what Cole called ‘the most common type of discrepancy’ to be found in Quintilian's quotations of VergilFootnote ²⁹ is not difficult to produce: the speaker in the Aeneid is Helenus, a prophet, for whom praedicam is appropriate, whereas Quintilian is a ‘teacher’ (praeceptor), for whom praecipiam fits better. But this superficial explanation belies the way that Quintilian's modification does not just suit his argument that the expedient should take precedent over the honorable, it actually exemplifies it: changing praedicam to praecipiam is an expedient change for an ‘instructor’ (praeceptor) to make in quoting the language of a prophet.Footnote ³⁰ And if we move beyond a focus on textual matters, we find a suggestive resonance in Quintilian's choice of this particular passage to authorize his claim about the relationship of expedience to honor: Helenus, as Papaioannou has noted, can be understood as a metapoetic figure, his prophecy about the events of Aeneas’ life, many of them known from preexisting mythological tradition, a kind of in-text representation of Vergil's own process of ‘critical selectivity and re-writing of preexisting material.’Footnote ³¹ And if Helenus serves to highlight Vergil's own power to modify the epic tradition to suit his poetic needs, then Quintilian has similarly employed this passage to highlight his power to do the same with the poetic tradition. Quintilian's choice of passages is anything but unconsidered and his modification of the text of the passage he chooses to quote is germane to his larger argument in the passage in which it appears.

When examined in the context of ancient quotation practices, Quintilian's introduction of changes to a quoted text to serve his rhetorical purposes does not appear to be unusual. Cicero reports (De Or. 3.141) that Aristotle mocked Isocrates with a parody of a line from Euripides’ Philoctetes (a play fragmentary to us but extant in complete copies at least until the time of Dio Chrysostom) that is preserved by none other than Quintilian himself: according to him Aristotle used to say ‘shame to keep quiet and let Isocrates speak’, replacing Euripides’ ‘barbarians’ with the name of his rival teacher (Inst. 3.1.14).Footnote ³² Nor were pointed alterations the province only of rhetoricians:Footnote ³³ Plutarch, advising on how to promote virtue through the reading of poetry, endorses examples of philosophers interpolating and rewriting lines of Euripidean and Sophoclean tragedies in order to salvage supposedly immoral lines (Mor. 33e–d), and Whittaker identifies a ‘persistent inclination of the scholars and writers of the ancient world to introduce into their quotations deliberate alteration’ based on the quotations from Plato, Aristotle, and the Stoics in the Didascalicus of Alcinous, a text composed, roughly speaking, in the same period in which Quintilian wrote.Footnote ³⁴

These intentional modifications of texts must be understood as taking place in a reading culture in which, as Zetzel has argued, ‘reading and transcribing were far more interactive processes in antiquity than should make us comfortable.’Footnote ³⁵ Citing examples of intentional intervention such as ‘hyperarchaism and analogical leveling’ in the text of Sallust, interpolations in Ovid's Metamorphoses and Juvenal's Satires (the result, according to Tarrant, not of deliberate or accidental error but of imitation),Footnote ³⁶ the expurgation of Martial's epigrams in some manuscripts and the omission of prefaces in others, as well as the existence of an alternate ending to Terence's Andria known already to Donatus, Zetzel concludes that ‘scribes…did not feel any particular compunction about making alterations and additions’ to the texts they copied. Thus for ancient readers encountering these scribal copies, variation was commonplace yet meaningful, and at least as likely to result from conscious alteration as from error.

This would be particularly true for the examples discussed here, which involve primarily well-known authors whose texts were probably more stable than obscure ones, rendering deviations from familiar versions all the more obvious.Footnote ³⁷ That Quintilian quotes Vergil more than any other poet has been understood as nothing more than ‘a genuine tribute to the poet…whom Quintilian wishes Roman students of rhetoric to read’,Footnote ³⁸ but the care with which Quintilian has selected Helenus’ prophecy, as well as that he exhibits in examples I will discuss below, suggests that Quintilian may prefer Vergil because alterations to his text are more noticeable than those of less well-established authors. That is, the relative textual stability of Vergil was most suited to provoking the kind of scrutiny that I have been arguing Quintilian expected his readers to apply.Footnote ³⁹

The analyses I am proposing, however, require not only that Quintilian's readers expect and attend to these textual changes but also consider the relevance of the quotation's original context to that of the passage in which it is quoted. Suetonius’ descriptions of emperors quoting poetry for persuasive ends suggests that this, too, was a common enough practice in Quintilian's day. Mitchell argues that Suetonius uses examples of emperors quoting poetry as a ‘metatextual device’ to invite his audience ‘to sympathize with the emperor-character as a fellow enthusiast for literature.’Footnote ⁴⁰ That such quotations, even from emperors, sometimes failed to produce the desired effects should only strengthen our confidence that audiences attended closely to the appropriateness of such quotations; according to Power, the portrait of Claudius in Suetonius, Tacitus, and Cassius Dio suggests that he ‘was well known for quoting passages of Homer that could be turned against him for humorous effect.’Footnote ⁴¹ The examples given all require that the audience recognized that there was, or could be, more than meets the eye in the quotation of poetry. Quintilian assumes that his readers, especially those who are attuned to the rhetoricality of his performance throughout the Institutio, will apply this kind of scrutiny to his quotations as well.

Such critical attention does not stop at the question of whether Quintilian's quotation is accurate or not but rather attempts to discern the significance of any deviation from putative sources. An example of how revelatory the latter approach can be is Quintilian's quotation of a phrase that appears nowhere in Latin poetry but that appears to conflate two well-known Vergilian lines. In his discussion of vowel quantity, Quintilian notes that ‘a short syllable becomes long when it is followed by another, even a short one, which begins with two consonants’ (9.4.85). He illustrates this phenomenon with the phrase agrestem tenui musam in which, he says, ‘a is short, and gre is short, but will make the preceding syllable long.’ Halm, in his edition of 1868, identified in the phrase agrestem tenui musam a ‘confusion’ (confudit) of Eclogues 1.2 (siluestrem tenui musam) and 6.8 (agrestem tenui…musam) arising out of Quintilian's supposed penchant for ‘quoting from memory’ (‘ex memoria citans’).Footnote ⁴² But confusion and a faulty memory begin to seem inadequate explanations for the line when one realizes that the two Vergilian passages that Quintilian has combined themselves repeat two separate Lucretian phrases (siluestris musa at 4.589 and agrestis musa at 5.1398). No one has doubted that Vergil's imitation of these two lines reflects his exquisite craft, and Quintilian's later combination of those same two lines reveals not his faulty memory but his deep understanding of the literary history of Vergil's imitation.Footnote ⁴³

Quotations without differences from the quoted texts have attracted even less attention than those that exhibit them, yet these also provide Quintilian opportunities to display his literary and rhetorical savvy. When Quintilian criticizes ‘harsh metaphors derived from distant resemblances’ ([tralationes] durae a longinqua similitudine ductae, 8.6.17), he gives two unattributed examples, both featuring metaphors involving snow. The first is a familiar enough line, the phrase capitis niues, ‘snows of the head’, which comes from Horace's Odes 4.13.12, where the poet mocks the gray hair of an aging woman. The second is a hexameter in which ‘Juppiter spat white snow upon the wintry Alps’ (Iuppiter hibernas cana niue conspuit Alpes, Inst. 8.6.17). Porphyrio attributes this line to the poet Furius Bibaculus (fr. 15 Courtney/fr. 80 Hollis) in his commentary on Horace's Satires 2.5.41 where Horace appears to parody this line, replacing Iuppiter with Furius as the spitter and describing him in the previous line as ‘swollen with fat tripe’. Much more, however, links these lines then their provocative metaphors of snow. Bibaculus, whose poetry survives only in fragments but whose poems attacking Octavian seem to have been available at least into the time of Tiberius,Footnote ⁴⁴ seems to have been a favorite target of Horace, especially for this line about the Alps; elsewhere in the Satires, Horace contrasts his own poetic style with that of ‘the pompous poet of the Alps’ (turgidus Alpinus, 1.10.36), which nearly all critics since Bentley have understood as an unflattering reference to Bibaculus.Footnote ⁴⁵ So Quintilian has strikingly combined two passages that are implicated with each other not just by the imagery of snow but by literary history: one of the poets whose style Quintilian criticizes (Horace) had criticized the other (Bibaculus) on stylistic grounds.Footnote ⁴⁶

If it strikes a dissonant note for modern readers to find so admired a poet as Horace lambasted for the same stylistic failings as so obscure and maligned a writer as Bibaculus, then perhaps this is Quintilian's point: that any author may be liable to stylistic faults and, in any case, that such judgments are arbitrary. In fact, this is a position that Horace himself, however strident his attacks on Bibaculus may have been, would probably have agreed with, since he represents himself throughout the Satires as being liable to the same poetic faults as those found in the poets he criticizes. Satires 2.5 in particular, the poem in which the parody of Bibaculus appears, has been called a ‘mock-didactic, pseudo-instructional treatise’; it ends with a cynical assessment of the motives of those who give, and receive, gifts such as homes and farms, which sits uneasily with the opening of the next poem in the book that begins with Horace giving thanks to Maecenas for precisely such a gift.Footnote ⁴⁷ Quintilian's deployment of this quotation shows off his knowledge of literary history and Horace's satiric technique, while representing himself as a critic unafraid to challenge critical orthodoxies by grouping together a canonical author like Horace with an acknowledged hack like Bibaculus.Footnote ⁴⁸

Productive Dissonance

An even deeper, more intricate knowledge of literary history can be found in Quintilian's quotation of a fragment of a Republican drama that adapts a line from Attic tragedy, although in this case the resonances of the quoted material threaten to complicate Quintilian's claims to rhetorical mastery as much as they confirm them. The quotation comes as Quintilian is defending the extensive course of study that he recommends (Inst. 1.12), one that includes a whole range of disciplines, including music, geometry, basic acting, and gymnastics (1.10f.). Quintilian's defense is itself a rhetorical performance that ends (1.12.16–18) with a turn to invective against his critics who argue that his program is too arduous. ‘The excuse of “difficulty” is a cloak for our idleness’, declares Quintilian; his orator, by contrast, ‘will easily persuade himself to spend the time which is wasted [by others] in the theater or the Campus, in gaming or idle talk, not to say sleep and long drawn-out dinners, in listening to the geometrician and the teacher of music.’ And the reason that Quintilian's orator will be willing to commit to such extensive study, ignoring the appeal of ‘mercenary ends and filthy lucre’ (uilem usum et sordidum lucrum) that leads others to take shortcuts, is because he has, according to Quintilian, ‘formed a real concept of eloquence’ that leads him to ‘set before his eyes that “speech, queen of the world” (reginam rerum orationem) of which the famous tragic poet speaks’ as his lofty and honorable objective in undertaking this study.

The phrase that Quintilian quotes here—‘speech, the queen of the world’—is known from a passage of Cicero (De Or. 2.187)Footnote ⁴⁹ to come from Pacuvius’ Hermiona (fr. 187 Warmington). Given the fragmentary state of that text it is impossible to be certain what meaning Pacuvius attached to that line, but it looks like an adaptation of a line of Euripides, not from that poet's Hermione but from his Hecuba (816), where the deposed queen of Troy describes ‘rhetoric’ as the ‘sole ruler where mortals are concerned’ (πειθὼ δὲ τὴν τύραννον ἀνθρώποις μόνην). Quintilian's brief citation of Pacuvius’ quote does not indicate whether he had both versions in mind, although his refusal to say which of these two authors he is actually quoting is not evidence for his lack of familiarity with his sources, but part of the effect: by saying that the line comes from ‘the famous tragic poet’ (non ignobilis tragicus) and not from Pacuvius specifically, even though elsewhere he is perfectly happy to name Pacuvius as his source,Footnote ⁵⁰ he allows both the Pacuvian and Euripidean versions to be operative.

But any reader who is familiar with the Euripidean context in which Hecuba describes rhetoric in this way—as I have argued both Quintilian and his audience likely would have been given the Roman penchant for scrutinizing quotations—will notice that Hecuba's argument is exactly the opposite of what Quintilian has cited the line of Pacuvius to support. Hecuba, newly enslaved after the fall of Troy and begging Agamemnon to help her take revenge on the man who had killed her son Polydorus, reflects that, in her powerlessness, she has recourse only to persuasion and wonders ‘why is it that we mortals take pains to study all other branches of knowledge as we ought, yet we take no further pains, by paying a fee, to learn thoroughly the art of persuasive speaking, sole ruler where mortals are concerned, so that we might be able to persuade people of whatever we wish and gain our ends?’ (Eur. Hec. 814–19)Footnote ⁵¹ For Hecuba, the fact that persuasion is sole ‘ruler’ (τὴν τύραννον, Pacuvius’ regina) means that students should not waste their time studying anything else, but Quintilian has cited this line to mean exactly the opposite: that rhetoric, as ‘queen’, encompasses all other forms of knowledge. There can be no question that Quintilian was unaware of the Euripidean context: the diametrical contradiction between Quintilian's claim and the argument of the source of the fragment he quotes could hardly be the product of coincidence, especially given the care that I have demonstrated that Quintilian takes in selecting poetic passages. Quintilian knew both Pacuvius (whatever he said about rhetoric) and Euripides. The question is why he would quote them in a way that so blatantly contradicts not only his specific argument in this passage but his general claim that the utility of poetry for orators is as a source of authority for their arguments.

But this is not the only example of Quintilian quoting texts that, when measured against Quintilian's own theory that quotations of poetry should ornament a text and lend support to arguments, appear to refute or complicate those arguments.Footnote ⁵² For example, when Quintilian argues that teachers should be strict with young children because it is hard to change bad habits once they are formed (Inst. 1.3.12f.), he cites a line from Vergil's Georgics in support of his argument: ‘so strong is habit in the tender plant’ (2.272). Peirano Garrison is surely right that with this quote Quintilian signals the significance of poetry to his work by ‘filtering’ agricultural metaphors from earlier rhetorical treatises ‘through the didactic tradition of Vergil's Georgics’,Footnote ⁵³ but this process also involves an inversion, not to say outright distortion, of Vergil's meaning. In the passage Quintilian is quoting, Vergil argues exactly the opposite of Quintilian, namely that young plants need special care and attention because they are fragile. For example, he says we should never shock them by transplanting them suddenly into unfamiliar soil, because such sudden changes will kill them, ‘so strong is habit in the tender plant.’ Anyone who knows the Georgics well, or who takes the time to investigate, will have reason to question Quintilian's claim.

Or consider several of the Vergilian quotations discussed above. Quintilian's invocation of the figure of Helenus as a model for his modification of the poetic traditions he quotes may serve, as I claimed above, as a confident declaration of Quintilian's literary power and autonomy, but could equally be understood as a bald admission of his distortion of the material he quotes. The plausibility of such an interpretation increases when one notes that elsewhere in the passage Quintilian recommends that ‘some things should be covered up in a speech’ (2.13.12), and that later in the Institutio Quintilian advises his students that ‘lines [of poetry] can be invented resembling well-known ones’ (ficti notis uersibus similes, 6.3.97). These more cynical perspectives on the orator's ability to alter the meaning of quoted material find resonance in the Vergilian passage as well, because Helenus’ prophecy, which fails to notify Aeneas that he will visit Carthage and predicts routes of travel different from those he will actually follow, is recognizably partial and inaccurate.Footnote ⁵⁴ And although Quintilian's understanding of the Lucretian background to the two lines of the Eclogues that Quintilian combines (9.4.85) seems to me to establish definitively the depth of his literary knowledge, the line he has produced only imperfectly supports the metrical point he is ostensibly making, that two consonants will lengthen a preceding short vowel. The first syllable of the word agrestis will not, in fact, necessarily be lengthened by the following syllable because the cluster gr, being made up of a mute followed by a liquid, does not always have this effect.Footnote ⁵⁵ In fact, the word agrestis in Vergil scans with a short first syllable more often than it does with a long one, a problem that Quintilian could have avoided if he had just confined his quotation to Eclogues 1.2, which begins with the metrically unambiguous siluestrem. Footnote ⁵⁶

Quintilian wanted, as I have shown, to combine the lines to signal his literary sophistication, but engaging the literary history of the lines also produces problems for the orator attempting to present himself as an authority on poetry. The Lucretian lines in which the two phrases appear, as Buchheit has shown, stand in contradictory relationship to each other, the one (4.589) describing how rustic people (genus agricolum) mistake echoes for music performed by satyrs and nymphs (siluestris musa), the other (5.1398) describing the origin of music in a kind of golden age when ‘the “rustic muse” (agrestis musa) was in its prime.’Footnote ⁵⁷ That is, the first passage debunks rustic beliefs about such music, and the second purports to provide a true origin story for it, providing an unstable basis for the kind of authority Quintilian is using these passages to authorize. This ambivalence in Lucretius had already been recognized by Vergil, who referred, as Breed has observed, to these contradictory lines (along with Theocritus) in programmatic positions in the Eclogues in order to invoke, simultaneously, two sources of authority for pastoral poetry, the one primal and originary (Lucretius’ agrestis musa), the other imitative and adaptive (his siluestris musa).Footnote ⁵⁸ Such a multidimensional and self-contradictory gesture of authorization might make sense in a book of pastoral poems, but to introduce the same complexity into a technical didactic passage of a rhetorical treatise where Quintilian expects his audience to recognize and accept his authority should provoke surprise, especially when Quintilian gives that same audience further reason to question his expertise by providing a flawed example of the technical point he is asking them to learn.

It is not even necessary to probe so deeply into the intricacies of Quintilian's quotational practice to find examples of quotations that are flawed in this way. Sometimes such incongruities are evident even on casual examination, as can be seen in an example of a quotation taken from Ovid. As an illustration of how ‘place’ is ‘a powerful means of giving a favorable or invidious turn to a case’, Quintilian quotes two lines from the speech given by Ajax in Metamorphoses book thirteen, in which the warrior argues that he, and not Ulysses, should receive the arms of Achilles: ‘Is it before the ships that we plead our case, and here that I'm confronted with Ulysses?’ (13.5f., quoted at Inst. 5.10.41) Ajax's reference to where he is speaking does indeed seem effective, since it calls attention to the fact that Ajax is being forced to defend his claim to Achilles’ arms in the very place where he saved the Greek fleet from destruction while Ulysses hid. But Ajax is nevertheless a strange choice to hold up as a rhetorical model because, as everyone would have known,Footnote ⁵⁹ he loses the contest of words to Ulysses, who, in most versions of the story, uses his superior rhetorical ability to produce an unjust verdict against the honorable but ineloquent Ajax.Footnote ⁶⁰ It could be argued that Quintilian endorses Ajax as a model because he provides a kind of mythological prototype of the uir bonus dicendi peritus, the ‘good man skilled in speaking’, that appears as the ideal orator throughout the Institutio, but, although Ovid portrays Ajax as a competent speaker, if not quite dicendi peritus, this explanation of Quintilian's choice, by linking the rhetorically ineffective Ajax with the uir bonus, serves only to weaken Quintilian's promotion of a concept of moral oratory that was already counterintuitive, as the laboriousness of Quintilian's definition of the uir bonus in Inst. 12.1 reveals. If the preceding analysis of Quintilian's quotational practice has demonstrated anything, it is that Quintilian knows exactly what he is doing whenever he quotes poetic sources, and whatever interpretation we propose of a passage where Quintilian appears to endorse a failed speaker as a model must reflect that.

Again it proves useful to consider the larger context of the passage that Quintilian has quoted. Tarrant has argued that Ovid, throughout his career, programmatically articulates an ‘ironic and sceptical view of formal rhetoric’.Footnote ⁶¹ Ajax's speech is only one in Tarrant's catalogue of Ovidian speakers whose attempts at persuasion produce a ‘striking lack of success’ and, although Tarrant understands this representation of rhetoric as a generic reflex of elegy's requirement that the amator be powerless, he notes in the case of the contest over the arms of Achilles that Ovid seems to be ‘showing how dishonest rhetoric can extort an unjust victory from an audience wanting in discernment’ by ‘award[ing] victory to the speaker more adept (or unscrupulous) in using the tricks of the orator's trade.’Footnote ⁶² Such a view of formal rhetoric might be unsurprising in a poet known as a provocateur and who, according to his self-representation, rejected an oratorical career in favor of a poetic one (Tr. 4.10.17–30), but this perspective would seem to strike a more discordant note in the work of a teacher of rhetoric like Quintilian, especially one famous for his commitment to formulating a moral basis for rhetoric. But as Halsall has reminded us, quoting an unpublished lecture delivered by Vickers in 1993, Quintilian's reputation as a moralist is as much a misrepresentation of him as the dismissive criticisms with which I began this article: if one attends, Halsall writes, the ‘flagrant disregard for ethical principles’ that can be found throughout the Institutio (just not in the passages promoting the uir bonus), it becomes ‘difficult to see Quintilian as the pillar of honest rhetorical practice’ that he is usually assumed to be.Footnote ⁶³ It is beyond the scope of this paper to demonstrate this fully; my argument elsewhere that Quintilian's representation of the uir bonus itself should be understood not as his most deeply held oratorical ideal but as a flagrantly self-undermining parody of attempts to develop a moral rhetoric, complete with a clichéd call-back to the mythologized father figure of Roman eloquence Cato the Elder, may serve as an example of what such a demonstration might entail.Footnote ⁶⁴ For the present discussion it is sufficient to note that throughout the Institutio Quintilian represents the quotation of poetry as a rhetorical technique much more worthy of a Ulysses than an Ajax.

For example, in the same passage where Quintilian recommends that students learn poetry by heart because it will be ‘useful in court’, he emphasizes that part of this utility derives from the fact that ‘things which “have been said well by someone else” (bene a quoque dicta)…have “more authority” (plus auctoritatis) and often win “more praise” (laudem maiorem) than if they were our own’, because they ‘have not been contrived for the sake of the case in hand’ (2.7.4). Quintilian reiterates this point several times, including in his discussion of sources of rhetorical authority, where he argues that such quotations are more persuasive because the orator's own compositions will be assumed to be biased, but ‘opinions which can be attributed to nations, people, wise men, distinguished citizens, or “famous poets” (inlustribus poetis)’ are ‘in a sense testimonies’ (testimonia) that ‘are actually “all the more effective” (potentiora) because they are not given to suit particular cases’ (5.11.36f.). The comparative language Quintilian uses in both passages (maiorem, potentiora) makes clear his position that quotations from poets are more authoritative, more effective, and, in a word, more persuasive than the orator's own formulations. In book twelve he goes even further, saying that references to historical exempla can ‘take the place of evidence or even legal precedents’, but that ‘the fictions of the great poets’ are just as good, because they ‘are either sanctioned by the guarantee of antiquity or believed to have been invented by great men as moral lessons’ (12.4.1f.). Quotations of poetry, for Quintilian, are what maxims were for Aristotle, who said that such sententious expressions ‘have one great benefit in speeches that comes from a lack of sophistication in their hearers’ (Rhet. 1395b1f.). Quintilian argues that audiences will accept the evidence that quotations offer without questioning it: they ‘are believed’ (creduntur, 12.4.2) because they seem to be ‘spoken or given by minds free of prejudice and favor for the simple reason that they seemed either very honorable or very true’ (5.11.36f.). Inherent in Quintilian's recommendation that orators use poetry for ornamentation and support for arguments, then, is a recognition that the speaker should attach such quotations to his more tendentious or questionable arguments, because such quotations are not subject to the same scrutiny that other kinds of evidence will be.Footnote ⁶⁵ Quintilian leaves no doubt that he understands this by illustrating the power of poetic quotation with a notorious case of fabrication: ‘[a] famous example [of seeking authority in a quotation of poetry]’, writes Quintilian, ‘is the way in which the Megarians were defeated by the Athenians in their dispute over Salamis by means of a line of Homer (not in fact found in every edition) which showed that Ajax had united his fleet with the Athenians’ (5.11.40). Already in the time of Aristotle the lines in question (Il. 2.557f.) were known to have been interpolated (Rhet. 1375.b29), and Quintilian makes clear that he knows this by mentioning the weak attestation of the line.

I have argued throughout that Quintilian intends his readers to learn how to use poetry in oratorical contexts from him and then to analyze Quintilian's own use of poetry in light of his theories. If I am correct, this means that such readers will recognize not only Quintilian's virtuosity but the dissonances and incongruities that such quotations sometimes produce. It means, that is, that Quintilian has composed such passages in the full knowledge of the dissonances and incongruities they contain.Footnote ⁶⁶

‘Personally, I should be happy simply to accept the view of antiquity’: A Case Study

Critics who have recognized these dissonances have not typically adopted this position, resorting instead to positing interpolation or textual corruption, for example, when a quotation appears to fail to illustrate the figure Quintilian quoted it to illustrate.Footnote ⁶⁷ In what follows, I present one of the most citationally dense passages of the Institutio Oratoria as a case study in how simply assuming that Quintilian knows what he is doing, however perverse or dissonant his deployment of quotations may at first appear, can be understood to be a fundamental component of his pedagogical purpose.

In the twelve paragraphs from Institutio 1.10.9 to 1.10.21 Quintilian refers to Orpheus, Linus, Timagenes, Homer, Vergil, Pythagoras, Plato, Socrates, Sparta, Stoicism, Lycurgus, Archytas, Euenus, Sophron, Eupolis, Aristophanes, Menander, Cicero, the Salian hymns, and a Greek proverb, all in support of the idea that music should be part of the orator's education. Just the sheer number of references (not all with quotations) indicates that this is a section in which Quintilian feels the need to exert considerable rhetorical pressure on those he seeks to persuade. As in some of the other passages I have considered, problems with Quintilian's argumentation arise as soon as one scratches the surface of many of the authorities he names (both poetic and otherwise), as Colson had already observed in his 1924 commentary on this section.Footnote ⁶⁸ Quintilian's claim that music is important because Socrates learned to play the lyre as an old man (1.10.13) distorts the focus of a well-known story that is usually cited to emphasize the value of learning late in life, not the value of learning music specifically.Footnote ⁶⁹ When Quintilian says the ‘“greatest generals” (maximi duces) played on the lyre and the pipe’ (1.10.14) the sources he may have in mind fail to support him in various ways: Cicero describes only one general, Epaminondas, who was musical (Tusc. 1.4), and if Quintilian has Alcibiades in mind, Plutarch agrees he learned the lyre but says he never played the pipe (Alc. 192D). Colson cannot find any passage in Plato where the philosopher says music is ‘essential for his statesman, or politikos as he calls him’, as Quintilian claims he does (1.10.15), nor can he find any Stoic saying what Quintilian says they said, namely, ‘that some of their wise men might give some attention to [music]’. There is an overarching problem, too, as Colson points out, which is that in his use of Greek sources Quintilian seems not to distinguish the broad sense of ‘music’ as ‘everything associated with the muses’ (thus including all the arts) from the narrower sense (of producing and understanding melodies) in which Quintilian clearly means it. Colson's overall assessment of these ‘dubious and half-hearted’ quotations is that Quintilian ‘shows little of the conviction that appears in the Republic, the Laws, and the Politics, that music really affects the character permanently.’Footnote ⁷⁰

On my reading, however, this is not a problem of conviction, but a problem of our interpretive frame. If we believe that Quintilian sincerely hopes to convince us that the poetic tradition supports the inclusion of music in rhetorical training, then the passage is indeed unconvincing. But if Quintilian's purpose is to dramatize, in a way that a well-trained orator cannot fail to recognize, how an orator might need to employ the authority of poetic quotations to sustain what Kennedy calls the rhetorical ‘strain’ of such a ‘farfetched’ argument,Footnote ⁷¹ then we find a bevy of such quotations that seem calculated to highlight just how far such an orator would have to be willing to go to make his case. Quintilian claims that Aristophanes ‘in more than one work, shows that boys were brought up in music in the old times’ (1.10.18); he does not name a passage specifically but ‘in the old times’ (antiquitus) suggests he is thinking of the scene in Clouds where ‘Better Argument’ describes ‘old education’ (τὴν ἀρχαίαν παιδείαν), including the education in music ‘that their fathers bequeathed them’ (ἁρμονίαν ἣν πατέρες παρέδωκαν, 961–8). But the Greek education that Quintilian is trying to make a model for his proposed system is pretty clearly mocked in the play; Better Argument's speech contains, as Dover has pointed out, ‘a good deal of nonsense’, including anachronism and nostalgia, that mark it as parody.Footnote ⁷² Quintilian then cites Menander's ‘Foundling’ (Hypobolimaeus) as further evidence of the antiquity of musical education, since in that play, Quintilian says, the old man, in giving an account to the boy's real father (who is claiming him back) of the expenses he has incurred on his education, notes that he has paid large sums to ‘teachers of the lyre, and teachers of geometry’ (1.10.18). I do not agree with Colson that the ‘situation alluded to seems very uncertain.’Footnote ⁷³ Quintilian explains just enough to make us doubt the appropriateness of his reference: an old man has raised a foundling—Quintilian uses the title Hypobolimaeus and not the other attested title, Agroecus (‘The Country-Dweller’) in order to emphasize this relationship—and is demanding that the biological father pay for the boy's education if he wants to reclaim him. That education, the old man says, included expensive music lessons, so the reference does in a narrow sense support Quintilian's contention that music was part of Greek education, but it hardly supports his overall argument that music should be taught. For the scene to make sense, the old man must be complaining about the expense (and possibly the uselessness) of the education he was forced to give the boy. Again and again, the poetic, philosophical, and historical material Quintilian quotes in this section as evidence that music should be part of the orator's education fails to support his argument or even contradicts it.

But this is not because Quintilian has a bad memory, or is insufficiently familiar with Greek literature (as Odgers or Goodyear might say), or just lacks conviction in his claim (so Colson). It is because Quintilian himself is providing his students with examples of the way rhetoricians, in his own telling (2.7.4, 5.11.3638, 12.4.1f., discussed above), use quotations from poets to support their arguments, namely, when they want their audience to accept their arguments unquestioningly and without scrutiny. He all but announces this in the heavy-handed remark at the beginning of his discussion that ‘[p]ersonally, I “should be happy” (contentus esse) simply to accept the view of antiquity’ that music should be part of rhetorical training (1.10.9), a line that baldly invites his audience not to question the authorities he cites, and he continues to strike this note when he claims that a passage from Vergil should be taken as ‘open confirmation of a “very great writer” (auctor eminentissimus) that music is connected also with the knowledge of things divine’, even though the passage he refers to—Iopas’ song in the first book of the Aeneid—‘omits the gods altogether’ and so fails to provide support for the divine link with music that Quintilian is attempting to make.Footnote ⁷⁴ These references to the assumed persuasiveness of poets, here articulated as part of Quintilian's rhetorical performance and not as a lesson in technique, only throw into relief how unpersuasive the quotations he uses actually are.

But in failing to provide proof that music should be taught, Quintilian has in fact succeeded in what I believe his true goal to be, both here and throughout the Institutio Oratoria: not (only) to train budding rhetoricians in the use of rhetorical techniques, but (also) to teach them to recognize when others are using rhetoric on them. Aristotle had already advocated something like this as the purpose of rhetorical training, writing that ‘one needs to be capable of being persuasive about opposite things…so that the way things are might not go unnoticed, and in order that, if someone else uses arguments unjustly, we ourselves might have the means to refute them’ (Rhet. 1355a29–33).Footnote ⁷⁵ And closer in time to Quintilian is the example of Aulus Gellius, who, according to Tischer, employs ‘a rhetorical strategy to provoke critical consideration of quotations.’ For example, Gellius’ ‘omission of an explanation’ for an error of attribution in Cicero's De Gloria produces a ‘blank space that prompts the reader to supplement what is missing.’Footnote ⁷⁶ Quintilian's performance of the rhetoric of poetic quotation is a case study in these dimensions of his pedagogy. Sometimes the analysis of Quintilian's quotations redounds to his credit, by showcasing the intricacy and sophistication of his selection and his productive modification of the Latin poetic tradition. To return to his schema of quotations as sources of ornamentation and argument, this type of passage illustrates the former, providing a masterclass in the adaptation of poetry for rhetorical purposes. But sometimes the scrutiny Quintilian invites catches him in the act of deploying poetry in coercive and even dishonest ways in order to demand deference to his authority, because that is, by Quintilian's own reckoning, one of the primary reasons an orator might use such quotations. No authority is absolute, Quintilian seems to say. Even the best rhetorician's constructions are susceptible to analysis and deconstruction.

Conclusion: A Quintilian for the First (and Twenty-first) Century

Calling attention to his own coercive deceptiveness may seem to undermine Quintilian's claim to rhetorical mastery and his authority as a teacher, and perhaps it does from the point of view of modern expectations—some of them derived from a misunderstanding of Quintilian!—that a teacher must possess exemplary morality.Footnote ⁷⁷ I would prefer, however, to recognize in Quintilian a realistic assessment of the needs of his students not just in their careers as orators but more generally in the rhetorical landscape of late first-century Rome, where all Romans were subject to a continuous onslaught of rhetorically sophisticated, imperial propaganda in support of the emperor's vision of the world.Footnote ⁷⁸ It should go without saying that the agents of imperial propaganda did not concern themselves with honor or morality, but with naturalizing and maintaining the regime's power by any means necessary. For Roman men seeking political advancement, or even survival, in such an environment, the ability to recognize, understand, deconstruct, and, if they chose, to reproduce such rhetoric was paramount. Quintilian gives them a training ground to develop their acumen for this work in the Institutio, replete as it is with rhetorical performances, including those that attune them to the elisions and fractures that are inherent in any deployment of rhetorical force.Footnote ⁷⁹

This is a strikingly different Quintilian than the traditional conception of him as an anachronistic and somewhat naïve thinker, a ‘Pollyanna about Flavian oratory’ who ‘doubtless looked the other way more than was good for his conscience.’Footnote ⁸⁰ If the sophisticated, utterly up-to-date, and provocative Quintilian I have been describing has escaped our notice for so long it is at least in part because the traditional Quintilian, more than any other ancient author, looks like the people who study him: in this Quintilian we find a professor of Latin language and literature, advocating the importance of his subject as part of a broad educational program in a world in which, if we believe his (patently rhetorical)Footnote ⁸¹ complaints about it, few people value that type of education as much as they should. There is a part of us that needs Quintilian to be who we have always thought he is. Advocates for the traditional understanding of Quintilian as a last defender of moral rhetoric looking back on a lost golden age of oratory from the vantage point of a degenerate present argue that ‘[m]odern readers…will find [Quintilian's] ideas as useful today as they have been in many centuries for nearly two thousand years.’Footnote ⁸² But such nostalgia as has been ascribed to Quintilian was as ill-suited to the political landscape of the high Roman Empire as it is to ours, where instantaneous and multimedia communication across vast distances allows an infinite number of often coercive and deceptive arguments to reach mass audiences. Quintilian's self-aware, deconstructive, and even subversive understanding of, and performance of, rhetoric does indeed remain useful. We have only to recognize it for what it is.