Bryn Mawr Classical Review

Bryn Mawr Classical Review 2001.09.16

Brian A. Krostenko, Cicero, Catullus, and the Language of Social Performance.   Chicago and London:  University of Chicago Press, 2001.  Pp. xvi + 360.  ISBN 0-226-45444-4.  $85.  $40.  

Reviewed by Andrew R. Dyck, University of California, Los Angeles (
Word count: 4146 words

As he explains in the Introduction to this revised Harvard dissertation, K(rostenko) has set his sights on a group of words and their derivatives often used by Catullus to describe persons or situations in his polymetric poems: bellus, elegans, facetus, festivus, lepidus, venustus--"the argot of Catullus's circle for describing stylish behavior" (K., p.21). Interpreters have been inclined either to capitulate that the "nuances must elude us" or, by the New Critical method of reading the poems as self-contained vessels, to try to deduce the meaning from the context of the given poem (K., pp.12-14). Dissatisfied with these approaches and following up on a suggestion of V. Buchheit, K. proposes to set these terms into the larger context of their general usage at Rome in the course of the Republic and (in the final chapter) beyond. Data from comparative and sociolinguistics are thus to be placed ultimately in the service of literary criticism. K. makes his work accessible by equipping almost all of the quoted Greek and Latin with his own stylish translations.1

The first problem K. faced was that of designating the analysand. He has decided to call his group of words "the language of social performance." This is a bit opaque and may not greatly help the browser examining a shrink-wrapped copy in a bookstore (my first encounter with the book) to know what to expect. The title might have suggested a study of such expressions as convivium ornare, salutare, cum aliquo sermonem conferre, or the like, i.e., descriptions of various social gestures performed by the Romans. But what K. is studying is, in fact, the language used to appraise people who appear in society or actions performed there; it might perhaps be more accurately called "the language of socio-aesthetic appraisal." But even if either "the language of socio-aesthetic appraisal" or "the language of social performance" or tertium quid is agreed as the topic, what K. offers is not a full account of it even in the Republic, for there are other words, e.g., bonus, callidus, decorus, honestus, praeclarus, pulcher, suavis, etc., which fall within this semantic sphere but which, while they may be glanced at in passing, do not receive full discussion by K. because they were not buzz-words in Catullus' circle. K., while aware, of course, of the existence of these other words, nonetheless regularly speaks as if "the language of social performance" comprised only the words he has chosen to study. It should be made clear to potential readers, then, that what K. is studying is a subset of the words used by Romans to express aesthetic judgments of persons or behavior encountered in society, a subset chosen for its prominence in the polymetric poems of Catullus.

A foundational claim of K. (p.15) is that semantic shifts in his words do not occur piecemeal but as part of a sea-change in Roman attitudes toward aestheticism in the 2nd century BCE. This is a linchpin in K.'s overall argument since the semantic shifts he finds are mostly assigned (without direct evidence) on this basis to the 2nd century. In arguing this case, however, K. has not cited all relevant evidence and has fitted the evidence he has used a bit roughly to his thesis. On p.26 K. claims "there is good evidence that all manner of aesthetic forms began to be used in the second century not to express Roman civic ideals, as seems to be true in the earlier period, but also to advertise the status of a particular social class." This is based on three pieces of evidence: (1) a description from a speech of Cato of someone singing, reciting Greek verses, etc., (2) a riddle posed by Lucilius, perhaps intended for a convivium, (3) Scipio Aemilianus' mention of noble youths learning to dance and to play musical instruments. Let us examine each of these items: (1) It is not clear who was the target of Cato's strictures; if it was M. Caelius (as K. is inclined, p.25 n. 15, to believe), as a tribunus plebis (in 184?: the name is given with a query at MRR 1, 375) and first known office-holder of his family, he might not have been seeking to "advertise the status of a particular social class," but rather his personal qualities; the same may be said of the noble youths mentioned by Scipio (3): they presumably had no need to advertise their class but cultivated such arts in order to set themselves apart from others of their kind (cf. K.'s own comments on individualism on pp.83 and 180). As to the convivium (2), there is evidence it existed in the third century (Pl. Mil. 165, usually dated to 206), and Plautus already has riddles (e.g., Merc. 361). K. (p.33 and n.41) is keen to separate benigne as used of a dinner by Lucilius (1269M) and as used at Pl. Mil. 739, but what is the difference? In both passages the iunctura is accipere benigne, and--what K. does not mention--Plautus' speaker Periplectomenus adds that he will receive his guests lepide et lepidis victibus: what could sound more Catullan? It is possible that K. is right in referring the semantic changes he has studied to the second century, but the evidence he adduces is not "good," as he claims, but slight and needs to be balanced against other evidence. Possibly there is in fact more continuity of usage than K. allows, and class distinctions may factor in more subtly; possibly, too, semantic change does not advance smoothly pari passu with social change but in fits and starts.

K. goes on to discuss his individual words and their semantic fields (pp.34 ff.). Here are found a number of good individual observations,2 but there are problems, mostly again related to claims for the second century. Take the case of venustus: we are told (p.40) that "the meaning of venustus changed in the second century as a result of social shifts" and that one of these changes involved a change in its semantics to encompass an "erotic element." While admitting that the erotic component cannot be found in references of venustus to art or artful speech, he suggests that "it is possible that it figured in the original application of venust(us) to dancing and artwork" (p.49); but, in fact, K. has offered no evidence of the application of venustus to dancing and will offer none. His case rests wholly on a connection construed between the cinaedus and dancing (pp.49-50), but he seems unaware that he has thereby shown nothing for venustus; he goes on to claim that venustus has been de-eroticized by the late Republic, possibly under the influence of χάρις (pp.50-51). But this, too, is wrong: Cicero and Catullus both have venustus with an erotic valence, as K. himself later shows (pp.259 and 267); given the etymological connection with Venus this is likely to have been present in the word from the beginning; and Greek χάρις itself could be eroticized.3

Again K. writes (p.75): "It is possible to view the narrowing of lep(idus), facet(us), and festiv(us) to describe verbal wit not as a function of real semantic development but as a trick of the sources: a more prominent role is played by 'tricking' or 'jollity' in comedy, the context for most of the earlier attestations, and by verbal wit in Cicero, source of most of the late Republican attestations." Here K. has found a very apt interpretation, given the state of our evidence; yet he immediately dismisses it, claiming instead that "the narrowing of lep(idus), facet(us), and festiv(us), especially beside the developments in bellus and elegans, is better attributed to their appropriation by a particular social class." There are, in other words, two possible explanations, appropriation by a social elite and a trick of the sources. But in order to settle the question one would need a control, i.e., a non-elite author of the late Republic whose usage would either show that the earlier sense persisted on another level of society or that the new usage was by now all-pervasive. As things stand, K. cannot exclude the alternative hypothesis that he has himself so clearly formulated.

The classic situation is that of bellus and elegans: here there is an observed change in usage between Plautus and Cicero (pp.54 and 35-38). Another hypothesis that could explain the phenomena is that the semantic changes K. studies were already underway in Plautus' time but concealed by the genre in which he wrote; but in that case, K. argues, "one might have expected them to be parodied in Plautus," whereas "the language of social performance appears in ways that might parody elite speech only in Terence" (p.75). An argumentum ex silentio is, of course, hazardous in general, especially so, I would argue, in this case: Terence has an ear attuned to the subtleties of elite speech, whereas Plautus does not.4

Chapters 3 and 4 deal with K.'s lexemes in relation to the vocabulary of rhetoric. Chapter 3 usefully discusses the semantics of his words in relation to Greek counterparts and leads to the conclusion that "in general our terms translate Greek terms that belong" to the same semantic fields (p.128). In chapter 4 K. observes (p.133) that there are two possible ways in which a Latin rhetorical language could have formed, namely (a) simple semantic extension of approximately equivalent Latin words and (b) an ideological attempt to justify aestheticism as socially valuable. Chapter 3 might have suggested (a) as the sufficient explanation; but K. insists on (b).

K. constructs a debate in 2nd century Rome which pits Cato, the Stoics, and Terence on one side against those who applied this "language of social performance" to rhetorical theory. This picture is unlikely to be correct. Let us begin with the Stoics. K. outlines Stoic rhetorical theory, which did not have much use for "decoration," only for plain-speaking (pp.135-36). He then admits that the ideal of plain speech could also have non-Stoic sources and goes on to discuss the Elder Cato, who is not claimed as a Stoic. He remarks that those hostile to the embellishments encouraged by Greek theory "might frame their resistance in terms of Stoic rhetorical theory." They "might," but did they in fact? K. has not shown that they did. By p.145, however, K. is speaking of "Cato's Stoic response to rhetoric" and cross-referencing the discussion at 135-36. By p.145 Terence, too, has been enlisted in the cause ("Terence and others influenced by Stoic rhetoric"; sim. 153).

Now the influence of the Stoics (and Greek philosophers generally) at Rome is usually dated from the famous philosophers' embassy of 156/5 (see the evidence collected at Carneades T 7 Mette), whereas Cato's attack on M. Caelius (which once again serves as K.'s evidence) may date to 184 (see above) and Terence's plays to the 160s. It is only figures of the succeeding generation, Q. Mucius Scaevola "the Augur," P. Rutilius Rufus, and Q. Aelius Tubero, whom Posidonius singled out as Romans who followed the Stoic way of life (fr. 81 Th.); it was then, too, when Romans are first known to have studied with Stoic teachers. In spite of Cicero's apologetic claim that the Elder Cato studied Greek in old age (Sen. 3), there can be no question of a "Stoic response to rhetoric" on his part; his views coincided with those of the Stoics, insofar as they did, by accident.

What about Terence? K. cites several passages (Ad. 803-6, Eu. 419-29, 286-88, 454-57) in which one character uses terms such as facetus, lepidus, lautus, and venustus to put down another chanracter. K. claims (p.143) that "in these passages Terence is taking advantage of the nascent application of the language of social performance to rhetoric, using it ironically of rhetorical failures...These passages permit the cautious suggestion that the language of social performance at the time of Terence was on its way to being used as part of the technical language of rhetoric..."

Was K.'s "language of social performance...on its way to being used as part of the technical language of rhetoric" in the time of Terence? K. cites some evidence for rhetorical instruction at Rome, referring on p.137 to an edict of "the censors C. Fannius Strabo and M. Valerius Messalla" expelling rhetoricians from Rome; sim. p.143 and n. 37. But in fact, a careful reading of Suet. Rhet. 1 shows that these two were the consuls (for 161) and the decree in question that of the senate (the censorial decree was from 92 and is subjoined by Suetonius). K. says nothing in either passage about the language of instruction of the rhetoricians expelled in 161, but it was surely Greek (see Kaster ad loc.). The evolution of Latin as a language of instruction for rhetoric comes later; the first evidence for it is the censorial edict of 92, specifically banning those teaching rhetoric in Latin.5 Greek rhetorical instruction would evidently have been available to Terence, and he was, of course, skilled at rendering from Greek to Latin; hence it will not have been difficult for him to render προοίμιον as principium (if that is what he is doing at Eun. 457; see K., p.143) or to have written himself a defense speech in accord with Greek rhetorical precepts (prologue to the Andria , discussed by K., pp.144-45). If instruction in rhetoric in the Latin language was beginning at Rome only ca. 100, there is not likely to have been a settled Latin rhetorical vocabulary in the 160s, let alone an incipient use of "the [Latin] language of social performance" as a technical language of rhetoric. Nor does it follow from K.'s evidence that Terence was somehow aligned with Cato: his use of lepidus, facetus etc. to deride certain characters need not show that he opposes such values per se, only that those characters' claim to them is absurd.

One last project of K. in chapter 4 is to elicit some evidence for the other side of his 2nd century debate. He focuses on the pair suavis gravis and argues that it "was coined in the second century as a kind of slogan" to deal with the paradox that " 'pleasant inessentiality,' in social practice no less than in rhetoric, was important to many Romans" (p.145). K. discusses the phrase at some length, including its shape as a rhyming pair, its possible Greek antecedents, etc. (pp.145-50). The phrase is, K. acknowledges, first attested in Cicero (Inv. 1.3) and often used by him. Yet K. denies that it was created by Rome's greatest phrasemaker and wants to claim it for some unknown author of the 2nd century. On what basis? The author of Rhet. Her. (4.69) has the sequence et gravitatem et dignitatem et suavitatem; K. infers (p.151) that this "is meant to recall the pair suavis gravis." Now it is true that Rhet. Her. elsewhere prefers the form suavitudo to suavitas; but this divergence is easily explained, for, as K. himself shows (pp.150-51), the latter "is necessary to secure the rhyme." But if, when Rhet. Her. was written, suavis gravis was already a fixed phrase with the expressive power K. attributes to it, the author would hardly have broken it with the insertion et dignitatem. K. claims (p.152) that suavis gravis is "at 'full strength' in the Rhet. Her., a little farther from its origins in Cicero"; I suspect the relations may be just the reverse.

I doubt, then, that many will believe in K.'s postulated 2nd century philhellene coterie consciously reacting to conservative trends in Roman culture and forging the semantics of certain words as their weapon (even Cicero had difficulty plausibly claiming much Greek erudition for the circle of Crassus at the beginning of the first century: de Orat. 2.1-4); K. realistically admits that Catullus himself did not have that kind of impact on semantics (see below).

The second half of the book, devoted to tracking K.'s lexemes in specific corpora, is chockablock with sensitive and illuminating analyses of nuance and context. It is a bit disappointing, however, that K. moves directly from Cicero's speeches to de Orat. to Catullus to the later history of his lexemes and thus offers no connected treatment of Cicero's philosophical essays and especially his letters, which are our richest source of information about social interactions in the Republic. In the chapter on the speeches I found particularly good K.'s analysis (pp.192-93) of the deflating of prosecution witnesses (not necessarily "Clodia's retinue," though Cicero strives to make them look like such) at Cael. 67.6 I would have welcomed some discussion of whether Cicero's different approaches may be related to the different audiences addressed: are the senatorial jurors of pro Quinctio and the Verrines treated differently from the later mixed juries, or are the speeches delivered before the people such as Cat. 2 different from Pis. and other senate speeches? The material from the speeches leads K. to the conclusion that aestheticism has by the late Republic become the "dominant paradigm" in Roman culture (p.201). Though I cannot pursue the point at length here, K.'s evidence seems to me rather to argue the indispensability of an aesthetic component in the mix of qualities that make up the individual: the lack of it can lead to censure, as in the case of Piso, but so too can an excessive amount (cf. Verres).

The chapter on Cicero's rhetorica focuses chiefly on de Oratore and in particular the so-called excursus de ridiculis (2.216-90), since this is where K. finds the most innovative use of his terms. He has a good grip on the nostalgic atmosphere of the work in general ("almost...a dialogue of men under whom things would have been different": p.231) and his analyses issue in a number of insights (e.g., "whenever venustus refers to humor in de refers to spur-of-the-moment humor": p.214; or the deployment of lepos for humor as an equivalent of χάρις: p.211). K. is right to think that a particular concern of Cicero is to defend his own use of humor in his speeches, but against whom? Surely it is against members of the nobility cut out along traditional lines, men like Cato, who drily remarked at Murena's trial "what a witty consul we have" (Plut, Cat. min. 21; cf. K., pp.224-25). K. perhaps offers one explanation too many when he tries also to bring in the Atticist controversy here, since Atticism was "at most a marginal concern" in de Orat.,7 and it is not clear that the Atticists disliked or discountenanced the use of humor (in spite of K., pp.226-27).8

K.'s labors on Cicero and the rhetorical tradition come to fruition in the chapter on Catullus (no. 7). By setting Catullus into this broader context, he is able to show that his words of praise are not merely an argot of his circle taken up by society at large (like the "jive talk" borrowed from jazz clubs in the 20s) but rather a borrowing by Catullus (and, no doubt, his circle) from the elite speech of his day (K., p.287). Thus even the latently erotic valence of venustus is shared with Cicero (p.267), and K. is able to isolate what is special about Catullus' application of venustus to Sirmio (pp.264-65, n. 82). The Catullan usage of lepidus emerges in sharper focus when K. has shown when Plautus is or is not a good precedent (pp. 259 and 267 n.) or the difference in Lucretius' usage (p.260, n. 69). This vantage point also enables K. to detect the different register of venustus and bellus in the polymetrics vs. the elegies (pp. 282-87). I found particularly telling the analysis of the complex c. 81, issuing in the conclusion: "Unlike Cicero, who attacks the 'trumping' use of the language of social performance petulantly, from the outside, Catullus attacks from the outside and the inside" (p.276). All in all, K. enriches Catullan studies with a series of sensitive, highly nuanced readings.

My reservation here is about K.'s claim that "the close correspondence of the ideological contexts of the social performance in both Cicero and Catullus suggests that Catullus has made a conceptual, and not a lexical semantic extension" (p.288). Elsewhere K. argues (p.287) that Catullus "takes the language in circulation in the society at large and heats it to piquancy" (p.287). This seems exactly right; if so, then Catullus' contribution lies not in semantic extension but in a new value-system according to which (in the polymetrics) K.'s "language of social performance" is the dominant social value. The difference is typified by K.'s comment (p.254): "Unlike [Cicero's] Crassus, Catullus wants to hear belle et festive often, to the exclusion of bene et praeclare"--a difference of values, not of semantics.9

Next follows a brief coda (chapter 8) sketching, with some admitted omissions, the subsequent fate of K.'s lexemes, to which the designator "language of social performance," as he admits (p.307), no longer applies. What they become, in fact, beginning with Cicero's late rhetorica, is the conventional language of literary criticism most fully attested in Gellius. K. struggles halfheartedly to find some residual political potency for his words in the career of Marc Antony, who was certainly an extravagant aesthetic performer, but, in fact, there is no indication that he took such terms as a political program, and indeed there was no incentive for him to do so and powerful disincentive, since Cicero and later Octavian were pummeling him with propagandistic claims that he had abandoned traditional Roman values. The volume closes with a list of works cited and a helpful set of indices rerum, verborum, and locorum.

I have compiled a list of minor difficulties and misprints, which is available upon request. Here is a sample: the Arcanum was the name of one of Quintus Cicero's properties near Arpinum, not its location (see Shackleton Bailey on Att. 2.2 of his Camb. edn.; K. p.37, n. 59); on p.52 K. refers to "Atticus' sons"; in fact, T. Pomponius Atticus is not known to have had any son, only a daughter, Caecilia Attica, by his wife Pilia; cf. Shackleton Bailey on Att. 78.2 of the Camb. edn.; the context shows that Cicero is referring to his own son and nephew (Cicerones nostros Deiotarus filius...secum in regnum: Att. 5.17.3); on p.81 K. cites, translates, and discusses the interpretation of a difficult fragment of Lucilius (15-16MW = 16-17K), but he makes what is itself obscurum even obscurius by adopting Warmington's interpretation but not his text and punctuation: Warmington prints anti and takes this (as his note explains) as the Greek preposition, which can mean "instead of" (LSJ s.v. III); K., however, prints "ante" with a comma immediately following; he translates this as "instead of" and takes it, despite the comma, as a preposition governing the immediately following words; it is very dubious that Latin ante can be taken in this sense (cf. OLD s.v.); more plausible is the interpretation of Puelma Piwonka cited by K. at n. 8; de Orat. 2.28 does not necessarily indicate "the superior attractiveness of Greek" (pace K., p.97, n. 25), merely that it is remarkable for the Greeks to admit that a Roman excels them in the use of their own language; p.158, n. 9 misinterprets the proverbial phrase about the lame man and the ball: cf. A. Otto, Die Sprichwörter und sprichwörtlichen Redensarten der Römer, p.85; p.184: Sulla's dominate and the concomitant villa-building is oddly retrojected to the 90s.

To sum up: Like many a (revised) dissertation, K.'s book needed a more thorough rethinking before publication; in particular the first half needed to be rethought in light of the second, the individual chapters in light of one another. The goal of the study, namely to contextualize Catullus' terms for the things he particularly values, is, despite terminological problems, a worthy one. K.'s strength lies in a good grasp of the principles of linguistics and an ear finely attuned to semantic nuance; and as a sensitive reader of literature he sheds bright light on a number of Catullus' polymetric and elegiac poems and some passages of Cicero. In the process he also clarifies the origin of a peculiarly Roman contribution to rhetorical terminology. His conclusions would have been further strengthened had he been able to include a systematic study of the usage of Cicero's letters. K. is less cogent when he moves from pure semantics and textual interpretation into an attempt to construct explanatory models on the basis of cultural history: here his theses needed to be more carefully calibrated to the evidence. As things stand, K. moves too facilely from social to linguistic change, relies on a small and partial pool of evidence which is not interpreted carefully enough, and does not pay sufficient attention to such chronological markers as we have.

Finally a note to the University of Chicago Press: one hopes that the thwacking price ($85.00 hb./$40.00 pb.) will not any time soon become standard for ca. 350-page academic books without plates printed in the U.S.


1.   There are occasional problems, though, e.g., Aen. 4.218 famam "reputation," not "belief" (K. p.29, n. 27); Lucil. fr. 1120M habearis "are appraised at," not "have" (ibid., n. 28).
2.   E.g., that "elegans in Plautus always has to do with personal comportment and is always pejorative" (p.35) or the discussion of the application of the different words for "light" (p.85).
3.   Cf. Eur. Hipp. 527 and K.'s own remark apropos the Greek Anthology (p.288).
4.   Cf. the interesting comparison of diction by J.S.T. Hanssen, Latin Diminutives: A Semantic Study (Bergen, 1952), 61-71, where one would now substitute "elite speech" for "the Scipionic circle."
5.   See further A. Manfredini, "L'editto 'De coercendis rhetoribus Latinis' del 92 a.C.," SDHI 42 (1976), 99-148; J.-M. David, "Promotion civique et droit à la parole: L. Licinius Crassus, les accusateurs et les rhéteurs latins," MEFRA 90 (1979), 135-80.
6.   This should, however, be assigned to the second part of the argumentatio, not the peroratio, which begins in 70; cf. Austin on 56-58 and 70-80. K. might have noted the specifically erotic connotations of deliciae and ineptiae.
7.   So Jakob Wisse, "Greeks, Romans, and the Rise of Atticism" in Greek Literary Theory after Aristotle, ed. J.G.J. Abbenes et al. (Amsterdam, 1995), 69, to whose discussion of the history of Atticism K. appeals (p.226, n. 78); Wisse finds a "leichte anti-attizistische Polemik" only at 3.25-37a (so in the commentary of Leeman et al., vol. 4, p.150).
8.   Epicurus' alleged lack of lepos (N.D. 2.74, cited by K., p.213; sim. content also at 1.123 and 2.46) was not specifically a lack of "the wit, grace, and humor of Socratic dialectic" (though Epicurus rejected dialectic) but part of the canard that Epicurus was a grim and unscrupulous polemicist that seems to go back ultimately to Timocrates; cf. D.N. Sedley, "Epicurus and His Professional Rivals" in Etudes sur l'Epicurisme antique, ed. J. Bollack-A. Laks (Lille, 1976), 121-59.
9.   K. argues (p.251, n. 31) for the MS reading dis(s)ertus at c. 12.9 rather than Passerat's differtus accepted by virtually all recent editors. One understands his keenness to have disertus here, but doubts remain since he has not found a satisfactory solution for the syntax of the genitives leporum...facetiarum.

Read Latest
Index for 2001
Change Greek Display
Books Available for Review

HTML generated at 13:27:30, Friday, 03 April 2009