The Oxford Handbook of the Second Sophistic

by Daniel S. Richter

  Ε 7: εὐσχολῶοὐδεὶςτῶνπαλαιῶν, ἀλλὰ σχολὴν ἄγω [vocabulary]

  None of the ancients [say] “I am well-leisured,” but “I have leisure.”

  As one can see, Moeris’s primary concern is to provide Attic equivalents for “Greek” words; he is extremely terse, only rarely listing a source, and even less frequently offering a comment (it should also be noted that many of Moeris’s recommendations do not accurately represent the usage of either classical Attic or koinê). Phrynichus, on the other hand, is not afraid to express his opinions: his language not only is insistently prescriptive—“say x,” “x must be said,” “don’t say y,” etc.—but also betrays his purist ideals (on Phrynichus’s life and tastes, see Jones 2008). Consider the following representative examples from his Selection:

  Φάγομαιβάρβαρον· λέγεοὖν ἔδομαι, τοῦτογὰρἈττικόν. (Eclog. 300 Fischer)

  φάγομαι [“I eat”] is barbaric; so say ἔδομαι, for this is Attic.

  Ἐλουόμην, ἐλούου, ἐλούετο . . . λούεσθαι· πάνταοὕτωλεγόμεναἀδόκιμα· εἰ δὲ δόκιμαβούλειαὐτὰ ποιῆσαι, τὸ εκαὶ τὸ ο ἀφαίρεικαὶ λέγελοῦσθαι, ἐλούμην, ἐλοῦτο . . . οὕτωγὰροἱ ἀρχαῖοιλέγουσιν.

  I was bathing, you were bathing, he/she/it was bathing . . . to bathe: all of these things said are unapproved; but if you want to make them approved, remove the ε or the ο and sayλοῦσθαι, ἐλούμην, ἐλοῦτο . . . ; for the ancients speak in this way. (159 Fischer)

  For Phrynichus, Atticism is not merely a refinement of contemporary Greek, but a purification, from which “unapproved” and “barbaric” words are explicitly banned. Moreover, he leaves no doubt about what he thinks of those who fail to adhere to his Atticizing standard: they are “the ignorant” (οἱ ἀμαθεῖς: e.g., 80), “the common folk” (οἱ ἀγοραῖοι: 176), or “the rabble” (οἱ σύρφακες: 407). By contrast, those who Atticize properly are called “the cultured” (οἱ πεπαιδεύμενοι: 176) or “those speaking in an approved manner” (οἱ δοκίμωςδιαλεγόμενοι: 269; on the terminology, see Strobel 2011, 126–168).

  ATTICISM, DIGLOSSIA, AND KOINÊ

  In Phrynichus’s eyes, then, the choice to Atticize is more than just a matter of linguistic taste; it signals membership in a “cultured” elite distinguished from the “ignorant” masses. A sociolinguistic situation of this sort, where the language is split into two distinct registers that correspond to cultural and class divisions, is characteristic of diglossicspeech communities (Ferguson 1959; Niehoff-Panagiotidis 1994, 106–120). Diglossia features, on the one hand, a “low” register, which is learned by everyone “naturally” as they grow up and is used in private, popular, or informal settings; and, on the other, a “high” register, which is considered purer and more elegant than the “low,” learned in school, standardized in grammars, linked to the literary tradition, and reserved for more official and formal settings. On this model, Atticizing Greek, as an artificial, learned variety cultivated by the elite, is the high register, while regular Greek, spoken by everybody else, is the low.

  In imperial Greek society, however, Atticizing Greek was not the only occupant of the “high” register of imperial Greek. Consider this well-known passage by the Skeptic philosopher Sextus Empiricus (Adv. Gramm. 233–235: translation adapted from Blank 1998):

  . . . and in life we shall line up with the customary, unaffected, local usage (τῇ συνηθεστέρᾳκαὶ ἀπερίττῳ καὶ ἐπιχωριαζούσῃ). . . . For example, the same thing is called artophorion and panarion [breadbasket]; again stamnion and amidion [chamber pot] are the same, as are igdis and thuïa [mortar]. But aiming at what is right and clear and at not being laughed at by our slaves (ὑπὸ τῶνδιακονούντων ἡμῖνπαιδαρίων), we shall say panarion, even if it is foreign (βάρβαρόν), not artophoris, stamnion not amis, and thuïa rather than igdis.

  Again in discussion we shall consider those present and avoid common words, seeking out a more urbane and literary usage (<τὰς> μὲν ἰδιωτικὰςλέξειςτὴνδὲ ἀστειοτέρανκαὶ φιλολόγονσυνήθειαν), for as literary usage is mocked by laymen, so is lay usage ridiculed among the literary set (ὡςγὰρ ἡ φιλολόγοςγελᾶταιπαρὰ τοῖς ἰδιώταιςοὕτως ἡ ἰδιωτικὴ παρὰ τοῖςφιλολόγοις). So, deftly responding to each occasion with just the right word, we shall seem to speak faultless Greek (δόξ-ομεν ἀμέμπτωςἑλληνίζειν).

  This passage is an excellent example of the distinct registers in imperial Greek—an “urbane and literary” opposed to a “lay” usage, each with its own separate, marked vocabulary (Niehoff-Panagiotidis 1994, 117–118). But note that Sextus nowhere refers to Attic or Atticism, preferring ἑλληνίζειν (“to speak proper Greek”) and συνήθεια (“the standard [language]”), both customary Hellenistic and imperial grammatical terms. Moreover, his advice does not always match Atticist orthodoxy: thuïa, which Sextus considers colloquial, is recommended as Attic by Phrynichus (Eclog. 136 Fischer; cf. Pollux 10.103), and neither of Sextus’s terms for “breadbasket” (panarion, artophoris) is attested in classical Attic (Blank 1998, 252). Sextus, who was probably writing in the second or early third century CE, is thus describing a diglossic system that features lay vs. intellectual usage (ἡ ἰδιωτικὴ vs. ἡ φιλολόγοςσυνήθεια), but one in which the “high” register is not “Attic,” but simply “urbane” Greek, and the “low” is what laymen or slaves speak.

  In fact, this sort of diglossia, featuring a “high,” non-Atticizing register and a “low” colloquial, had probably been in place since the rise of the koinê in the fourth century BCE. While “low” koinê Greek displayed considerable regional and social variations and underwent significant linguistic development over time (Brixhe and Hodot 1993), the “high” register, as is typical in diglossic situations, remained remarkably stable (this is the koinê described above). Throughout the Hellenistic and imperial Greek worlds this literary koinê, as I will refer to it, continued to be the language that was used by the elite, taught in schools, and in which any prose with artistic ambitions was composed (Blomqvist 2010; Colvin 2009; Jannaris 1903; Versteegh 2002).

  In the Hellenistic period, the only writers of literary koinê whose work survives in significant volume are the historians Polybius of Megalopolis (ca. 180–120 BCE) and Diodorus of Sicily (ca. 100–40 BCE).2 But in the early imperial era (under Augustus and Tiberius: 27 BCE to 37 CE), we have the rhetorician and historian Dionysius of Halicarnassus, the geographer and historian Strabo of Amasia, and the Jewish scholar Philo of Alexandria. After a gap of nearly thirty-five years, another crop of authors, flourishing under the Flavians, Nerva, and Trajan (69–117 CE), employs this brand of literary language: the Jewish historian Josephus, the orator Dio of Prusa, the biographer and essayist Plutarch of Chaeronea, and (if current dating is correct) the novelist Chariton of Aphrodisias and the anonymous author of On the Sublime. Despite the variation in language use among these texts, they can be clearly differentiated as a group from koinê texts aimed at a less educated audience, such as Tobit, the Life of Aesop, the Septuagint, and the New Testament (an excellent overview in Kaczko 1993).

  By the second century CE, however, certain Atticizing critics no longer deemed the “high” register of written koinê practiced by these writers as sufficiently “pure” or “cultured.” Consider Phrynichus’s criticism of Plutarch’s use of the word σύγκρισις (Eclog. 243 Fischer):

  Among his works, Plutarch wrote a certain “Comparison (σύγκρισις) of Aristophanes and Menander.” And I am amazed at his use of unapproved language (ἀδοκίμῳ φωνῇ), given that he had reached the heights of philosophy and was clearly aware of what σ�
�γκρισις once meant [i.e., “combination”]. And he similarly misuses συγκρίνειν [as meaning “to compare” instead of “to combine”] and συνέκρινεν [“he/she/it compared” instead of “he/she/it combined”]. So one should say ἀντεξετάζειν and παραβάλλειν [“to compare”].

  This entry shows that, in addition to avoiding non-Attic words and forms, the Atticist had to beware of legitimate classical Attic words that had changed their meaning. As far as we can tell, Phrynichus is right: συγκρίνειν means “to combine” in classical Attic prose and is used mostly in philosophical treatises (hence Phrynichus’s comment on Plutarch’s reputation in that field). But συγκρίνειν as “compare” had been standard, “high” koinê for centuries, and had never been considered “colloquial” or “unapproved” (e.g., Polyb. 6.47.10; Dion. Hal. Dem. 33). For Phrynichus, however, this is no longer sufficient; now, any deviation from Classical Attic is attacked as a vulgarism or barbarism. Plutarch may have written an elegant, literary koinê, inspired by the great works of the classical past, but his failure to replicate “pure” Attic, or even worse, his lack of interest in doing so, has indelibly tarnished his status as a pepaideumenos (cf. Eclog. 160 Fischer).

  Atticism can thus be seen primarily as a reaction against the perceived “impurities” of literary koinê, which it is trying to displace at the top of the sociolinguistic hierarchy, and only secondarily against the “low” colloquial speech of the masses, which had always been considered inferior and inappropriate for literary discourse. When Galen ridicules the Atticists and defends “contemporary” Greek, he is obviously not talking about the Greek spoken on the streets (cf. De differ. puls. 583, 1–2k, where he differentiates the Greek of “sailors, merchants, and traders” from his own) but the educated standard (Sextus’s συνήθεια) that he and other elite writers employ in writing and speaking. Whereas before it had been enough to avoid inappropriate words and constructions from the lower registers to write literary Greek, now one was required, at least in the opinion of strict Atticists, to only use terms, forms, and syntax “approved” by or “attested” in, classical Attic texts. It is essential to stress the amount of effort this required. Any speaker could sound Attic relatively easily, by incorporating some unmistakably Attic forms and words into his speech or by alluding to a well-known Platonic or Demosthenic passage; we might call this “positive” Atticism (cf. Luc. Rhet. praec. 16; Lucilius, Anth. Pal. 11.142; Ammianus, Anth. Pal. 11.157). But “negative” Atticism—which stipulated avoiding words, forms, and even meanings unattested in classical Attic—required either a prodigious memory or grammatical and lexical aids. To sum up: the Atticists did not introduce a “high” register into an otherwise undifferentiated imperial Greek; they added an even higher register on top of the one that already existed. Knowledge of the classical Attic dialect now became an essential prerequisite for entry into elite literary and rhetorical circles.

  ATTICISM AND CLASSICISM

  When did Atticism usurp the position of literary koinê? Our primary testimony for negative Atticism derives from writers active after 140 CE—Lucian, Galen, Philostratus, Phrynichus, and Moeris—and the first sophist whom Philostratus refers to as “Atticizing” (VS 568) is Herodes Atticus (fl. 130–177 CE). The authors whose prose shows evidence of the conscious avoidance of koinê vocabulary and forms—e.g., Arrian of Nicomedia, Aelius Aristides, Lucian, Maximus of Tyre, Aelian of Praeneste, and Philostratus—were all writing in the period ca. 140–250 CE. According to this evidence, it appears that a full-fledged linguistic Atticism began to first take shape under Hadrian and subsequently flourished under the Antonines and the Severans (Schmid 1887–1897, 1:210).

  The origins and practice of positive Atticism, however, are more obscure. The earliest references implying that the use of “Attic” language was thought to enhance one’s style occur in the late first and early second centuries CE: Plutarch criticizes those who “use Plato and Xenophon only because of their style, and pluck off only what is pure and Attic” (διὰ τὴνλέξιν . . . τὸ καθαρόντεκαὶ Ἀττικόν: De prof. virt. 79D; cf. also Quaest. Plat. 1010C, but De aud. 42D refers to Attic style, not language) and ca. 100 CE, Pliny the Younger refers to the “Greek, or rather Attic language” (sermo Graecus, immo Atticus: Epist. 2.3) of the sophist Isaeus of Syria. The first examples of “positive” Atticizing practice, in which recognizably “Attic” words, phrases, and forms are ostentatiously strewn throughout a work that is otherwise written in literary koinê, are found in the surviving speeches of the sophists Favorinus of Arelate and Polemon of Laodicea (both active in the first half of the second century CE). That the former’s Atticism was deliberate is confirmed by his boast in his Corinthian Oration, probably delivered in the 120s CE, that he deserved a statue at Athens because of his “Atticizing” prowess (ὅτι ἀττικίζειτῇ φωνῇ: [Dio] Or. 37.26; cf. the criticisms by Galen, Opt. Doctr. 1.41–42K, and Phrynichus, at Fischer 1974, 139). Polemon’s interest in Atticizing is suggested by the anecdote that the grammarian Secundus had helped him with the language of his speeches (Phryn. Eclog. 236; cf. 140, 396). In rough outline, then, there are some indications that using the classical Attic dialect lent prestige to one’s prose around the turn of the second century CE; this “positive” Atticism was most likely well established by the 120s at the latest, but only in the 130s or 140s would it have been allied with a purist “negative” variety that condemned literary koinê.

  It is often asserted, however, that Atticism had its beginnings much earlier, in the Augustan period, under the auspices of Dionysius of Halicarnassus, who championed the imitation of “ancient,” or what we call “classical,” authors in his works On Ancient Orators and On Composition (e.g., Dihle 2011, 47–48; Horrocks 1997, 51 and 73; Kennedy 1972, 242–243 and 553–556).3 Because Dionysius’s primary models were Attic writers such as Demosthenes, Thucydides, and Plato, his program is often referred to as “Atticism,” occasionally with the modifier rhetorical or stylistic added to differentiate it from the linguistic or lexico-grammatical variety of the second century CE (Swain 1996, 20–27). The potential for confusion here is obvious. The emulation of the style of Attic authors should be clearly distinguished from the use of the forms, vocabulary, and characteristic constructions of the Attic dialect. Despite Dionysius’s insistence on following “Attic” authors as stylistic models, he does not extend this mimesis to morphology or vocabulary like the later Atticists. In fact, even “stylistic Atticism” is something of a misnomer: Dionysius does not restrict his models to Attic writers—for him Homer, Herodotus, and the lyric poets are also worthy of emulation. For these reasons, we should refer to his program as classicism (on which, see Hidber 1996 and Wiater 2011), rather than the potentially misleading “Atticism.”

  Dionysius’s classicism, that is, his concern to “imitate” the style of the great prose writers and poets of the pre-Hellenistic past, would become an essential feature of imperial education and paideia. The linguistic Atticism of the mid to late second century CE can be seen as a particular outgrowth of this broader classicism, but the two movements should not be equated: for this reason it is best to reserve the term “Atticism” exclusively for the practice of speaking and writing in the Classical Attic dialect. While Dionysius’s style may owe something to that of various classical authors, and his prose may display an increase in certain Attic trademark features (such as optative use: Anlauf 1960; cf. Wahlgren 1995, 30n1 for other scholarship) and a concomitant reduction in the use of some koinê terms (Usher 1960, 364–365), his language, like that of his contemporaries Strabo and Philo, is still basically the literary koinê employed by his Hellenistic predecessors Polybius and Diodorus (Palm 1955, 206).4

  The same goes, more or less, for the prose of late first-century CE writers, such as Chariton, Plutarch, and Dio of Prusa. While each of them, in the course of emulating or alluding to particular literary models, may have on occasion used certain Attic words or ph
rases associated with classical authors (e.g., Wegehaupt 1896 on Dio), there is little conclusive evidence for either the systematic “revival” of classical Attic forms or words that had disappeared from literary koinê or the consistent, conscious rejection of koinê features. The surviving evidence thus points to the deliberate pursuit of Atticism as a second- and third-century CE phenomenon: an initial stage, represented by Favorinus and Polemon in the first part of the second century CE, that featured a “positive” Atticism, followed by the flowering of Atticism proper, personified by figures such as Phrynichus and Herodes Atticus.

  VARIETIES OF ATTICISM

  Even after the advent of purist Atticism, however, the levels of strictness in Atticist practice could vary considerably. Many Greek texts from this period continue to be written in a koinê similar to that of Dionysius and Plutarch, despite some concessions to Atticizing tastes: e.g., Pausanias’s Periegesis, Marcus Aurelius’s Meditations, the work of Galen, and significantly, treatises by the grammarians of the period (like Herodian or Apollonius Dyscolus). But one also encounters figures such as the orator Aelius Aristides (117–181 CE), undoubtedly the most rigorous extant Atticizing author. His declamations, particularly those on classical themes, are prodigious feats of prose composition, in which he not only reproduces the Attic dialect, assiduously avoiding classically unattested words and forms (although even he is not perfect), but also imitates the style of Demosthenes and Isocrates (Boulanger 1923, 395–412; Pernot 1981, 117–146). Aristides’s adherence to Attic purity is often taken as the norm for the Antonine era, but the only other surviving works that display a similar level of rigor are three short declamations by Lesbonax (Kiehr 1907, 8–18) and one by Herodes Atticus (Albini 1968, 12–16; Schmid 1887–1897, 1:195–200).