Chapter 6
The Social Relations of the Grammarians

cum . . . Servius a Symmacho rogatus esset quidnam de his existimaret: "licet," inquit, "in hoc coetu non minus doctrina quam nobilitate reverendo magis mihi discendum sit quam docendum, famulabor tamen arbitrio iubentis, et insinuabo."

When Symmachus requested his judgment on these matters, Servius said, "Although it is more fitting for me to learn than to teach in this gathering, to which reverence is due no less for its learning than for its nobility, I shall nonetheless obey the will and bidding of Symmachus, and speak."
—Macrobius, Saturnalia 1.4.4

Within his classroom, the grammarian was the master of his craft, a commanding figure, raised up by his texts, his tradition, and reason, just as he was exalted by the cathedra on which he sat. Still, he could enjoy this mastery for only a fraction of his waking hours, and each day he had to descend from his throne and lift the curtains at the school's threshold.^[1] What happened when he stepped outside?

There was, to begin with, a significant carry-over of honor from the classroom. To be called magister was a mark of respect, a reminder that one was engaged in a splendida ars and was one of the better sort of person ().^[2] The grammarian, set apart from the mob by his expertise, belonged to the Kulturwelt and shared its pride.^[3] As we saw in Chapter 2, that distinction was a target for some Christian writers in

[1] Grammaticarum scholarum liminibus appensa vela : Aug. Conf. 1.13.22; cf. De util. ieiun. 9, Serm. 178.8.

their polemics against the traditional culture. The problem, to all appearances, did not much raze the grammarian, conscious of his prestige and unwilling to swallow embarrassment, whatever his religious affiliations.^[4] Reputation, after all, was a commodity not to be trifled with: it provided the great name that would be boasted on one's epitaph; it was something on which one's native town could preen itself, whether one taught at home or abroad; it might even catch the ear of the emperor.^[5]

Not surprisingly, therefore, grammarians could occasionally serve, like other men of standing, as public spokesmen or conduits of patronage. Early in the reign of Anastasius, probably in 491 or 495, Timotheus of Gaza directed to the emperor a composition concerned with the horrors of the collatio lustralis , an especially hated tax that bore most heavily upon merchants and small tradesmen. (Timotheus's action can be called disinterested, since teachers were not subject to the levy.) We might doubt Cedrenus's bald claim that Anastasius abolished the tax because of Timotheus's composition; other considerations aside, the chronology of his account is obviously confused. But it does appear that the tax was abolished not very long after Timotheus's plea.^[6]

If Timotheus's intercession must remain a bit obscure, a more detailed case study is available from the fourth century in the person of Nicocles.^[7] Having his origins in Sparta, Nicocles chose Constantinople as the place to make his career, and apparently he chose well. By 340 he was successful enough to promise to supply Libanius with forty students, when Libanius was only a budding private teacher of rhetoric in the new capital.^[8] The offer highlights two elements in the relations between men at different levels of the teaching profession: not only the cooperative channeling of students from grammarian to rhetorician,^[9] but also the use of such favors as a club—for Nicocles hoped simultaneously to oblige Libanius and, by establishing Libanius as a rival, to spite a sophist who

[4] Cf. "Prologue" p. 6, on Aetius's teacher at Anazarbus.

[6] See Cedren. 1.627.7-10 Bekker, where Anastasius is also said to have been swayed by a delegation of monks from Jerusalem; Suda T.621; and Part II no. 156.

[7] See, in addition to nn. 8-17 following, the full analysis of the sources in Part II no. 106.

[8] Lib. Or. 1.31.

[9] Cf. also Lib. Ep. 832, to Nicocles, Ep. 398.2, to Acacius (see Part II no. 1).

had wronged him.^[10] But Nicocles' moment as a man of consequence only came two decades later, during the reign of Julian. The stage had been set in late 347 or early 348, when the prince, released from his exile at Macellum, had been allowed by his uncle Constantius to continue his education in relative freedom at Constantinople. There he was taught by Nicocles and the sophist Hecebolius; and when Julian reached the throne thirteen years later and came east, Nicocles was not slow to exploit the friendship he could claim with the emperor for his tuition.^[11] Doubtless because of Nicocles' favor, his brother Sozomenus appears as governor of Lycia early in 363.^[12]

Others, too, were quick to see the potential in the situation. So Libanius evidently managed to forget an old grudge and make friendly overtures to Nicocles, who finds a place in Libanius's correspondence—and so in his network of useful connections—for the first time in 363. Nicocles is thanked for his efforts in spreading Libanius's fame, and Libanius promises to return the favor.^[13] Letters of recommendation make their way from Antioch to Constantinople in behalf of a wandering poet or of men hoping to make a public career.^[14] And perhaps most significant, Libanius attempts to enlist Nicocles' help in dissuading Julian from moving his court to Tarsus from Antioch.^[15] As a known friend of the emperor, however, Nicocles was clearly not confined to serving Libanius's interests: when rioting at Constantinople had set the city prefect and the population against each other, Nicocles helped bring about a reconciliation; and he expected to represent the city in an embassy to Julian at Antioch concerning the same matter.^[16]

But Julian went to Persia and died before the embassy occurred, and Nicocles' influence died with him.^[17] Based on the loyalty of his extraordinary pupil, Nicocles' prominence was as unexpected as Julian's reign,

[10] Lib. Or. 1.31; cf. Ep. 557.1-2, regarding Nicocles' apparent participation in the conspiracy of sophists and grammarians to drive Libanius from Constantinople ca. 343 (Or. 1.44).

[12] Lib. Ep. 1383, with Bull. ép. 1979, 509 no. 4.

[13] Lib. Ep. 810.1-3.

[14] Poet: Lib. Ep. 816, including the none-too-subtle hint that Nicocles should pay him well; for the type, cf. Lib. Ep. 969, in behalf of Diphilus, 632, in behalf of Eudaemon of Pelusium; Aen. Gaz. Ep. 9, 10. Hopes for a career: see esp. Lib. Ep. 832, in behalf of Theodorus, a former pupil of both Nicocles and Libanius; 810, in behalf of Hyparchius (= PLRE I s.v., p. 449); with 1119.1-2, in behalf of Hercul(i)anus (= PLRE I s.v. Herculanus 1, p. 420).

[15] Lib. Ep. 1368.3.

[16] Lib. Ep. 1368.1.

[17] See esp. Lib. Ep. 1265, 1266, 1492, with pp. 214f. below.

and as evanescent. Nor are there other examples to set beside it.^[18] The case of Nicocles is perhaps finally the exception that proves the rule: like the story of the grammarians' origins and social status in Chapter 3, the story of their social relations and of the prestige and influence that they enjoyed in consequence of their profession is largely set within narrow horizons.

The varieties and causes of these restrictions are not far to seek. First and foremost, the grammarian's place on the threshold of liberal disci-plines^[19] matched his liminal position of social mediocrity, the modest good standing beyond which the status and application of his expertise was not by itself sufficient to raise him. As we have already noted and as we shall see further, neither his origins nor his skills were commonly suited to set him squarely before the public eye. The grammarian Philtatius of Athens (if he was in fact a grammarian) might receive a statue as the city's thanks for putting his skills to public use by reestablishing the colometry of texts in one of the city's libraries; but that honor grew out of uncommon circumstances, reconstruction in the wake of barbarian devastation. It is unparalleled.^[20] In more ordinary conditions, the grammarian might claim a moment as a quasi-public figure by stepping out of his professional role to become an encomiast.^[21] But when the city needed the combination of standing and eloquence that made for an effective spokesman or ambassador, the choice fell naturally on the rhetorician.^[22]

Moreover, the grammarian's place on the threshold of liberal culture made it difficult for him to develop a cluster of former students who had gone on to become influential men. The grammarian's role was pro-paedeutic: his students were parvuli ; the rhetorician's were perfecti.^[23] In practical terms this meant that, though some of the grammarian's stu-

[18] Cf. Chap. 3 p. 130, on the status and rewards of imperial tutors in more ordinary circumstances, and p. 132, on Simplicius and Pamprepius and their careers as ex-grammarians dependent on a single powerful patron.

[19] Ennod. Dict. 10.4: liberalium disciplinarum limen. Cf. id. Opusc. 6.11.

[20] Olympiod. frg. 32. For the historical setting, see Thompson, "Athenian Twilight"; Frantz, "Honors." On Philtatius's service and the question of his profession, see Part II no. 119.

[21] See Chap. 3 pp. 123f., with nn. 131, 132.

[22] See esp. Lib. Or. 19-23, on the Riot of the Statues (cf. Or. 1.252-54); and, e.g., PLRE I s.vv. Eusebius 24 (p. 305, with pp. 219-20 below), Firminus 3 (p. 339; on his career, see also Basil Ep. 116, 117, and Kopacek, "Curial Displacements" 327ff.). For the sophist's intercession with the praetorian prefect, cf. Lib. Or. 1.109, Ep. 833.1-2; Theodoret. Ep. LII, to Isocasius.

[23] Pelag. Expos. in Rom. 6.14 (p. 52.1ff. Souter). In the analogy around which the passage is built, the Christians stand in the same relation to the Jews as the rhetorician's students do to the grammarian's.

dents might begin careers with no further training,^[24] most went on to the rhetorician's school. When they subsequently set about their life's business, their loyalty to their more recent teacher would account for a large part of the influence a rhetorician could claim, as Libanius's network of former students shows.^[25] One can surmise that their loyalty to the grammarian, a more distant figure from their earlier years, was more attenuated and remote, less easily mobilized.^[26]

Overshadowed by the rhetorician's inevitably greater visibility and prestige, the grammarian's institutional niche gave him his position of strength but also set his limitations. His expertise was esoteric enough to set him apart from the great majority of the population, but within the charmed world of the litterati it involved no extraordinary distinction, and little mystery. It was obviously not esoteric at all in the way of modern professions, encapsulated and set off by specialized training as they typically are from even the educated lay public. It was instead eminently accessible to all the educated. The grammarian did not, after all, claim to be a charismatic teacher (as the Christian bishop did) and was not by definition a literary artist as well as a teacher (as the rhetorician was), but was fundamentally and simply a man of ratio and memoria. The qualities and attainments that gave the grammarian his authority in the classroom, even if not directly transmissible, were nurtured in his pupils by his teaching and, if he taught successfully, ceased to be distinctive to him. Since the grammarian and the educated layman occupied largely common ground, the grammarian's knowledge was not different in kind—or even necessarily in quantity—from that of any well-bred litteratus. The talented amateur could stand forth as a questioner and critic () of the professionals on their own ground.^[27] Nongrammarians could produce learned works rooted in , whether tricked out in literary elegance like the great works of Macrobius and Martianus Capella or straightforward manuals much like the grammarians' own.^[28]

[24] Cf. Chap. 1 at n. 62.

[25] See Petit, Étudiants esp. 154ff.

[26] Note that Julian was already nearly an adult, with much of his education behind him, when he formed his connection with Nicocles; see the appendix to Part II no. 106. The connection was as extraordinary in this regard as it was in others.

[27] Damasc. V. Isid. frg. 331 Zintzen = epit. Phot. 298, of Agapius.

[28] So the De metris of Fl. Mallius Theodorus, the De differentiis societatibusque Graeci Latinique verbi of Macrobius, or the handbooks of Charisius (cf. Part II no. 200) and Consentius (cf. Part II no. 203). Note, conversely, the works of general learning produced by grammarians on topics ranging from zoology and medicinal herbs (Part II nos. 61, 156) through a survey of the provinces and cities of the eastern empire (Part II no. 76) to humor (Part II no. 117). Cf. also the philosophical or theological writings of grammarians and ex-grammarians: Part II nos. 38, 81, 82, 118, 124, 135.

It is useful to imagine some of the consequences that might follow from this embedding of the grammarian's profession in the shared life of the elite. To begin with, since the grammarian's knowledge was as fundamental to the distinctive literacy and the prestige of the nonprofessional elite as to his own, the conservatism of his doctrine would be reinforced. The common ground would exercise a strong gravitational pull: the profession would be less free than its modern counterpart to follow its own lead (for better or worse, toward new understanding or the merely faddish), to spin off in the direction of esoteric concerns, or to become fragmented (in the manner of many modern professions) in proliferating subspecialties and techniques. Or to put it another way, two kinds of self-interest would converge to stabilize doctrine: if the grammarian within his institutional niche was not about to challenge or renovate radically the tradition that fortified him, as was suggested at the end of the last chapter, neither would he be expected or encouraged to do so by the litterati outside that niche, whose own interests were scarcely less involved. The grammarian's audience did not usually speak the command, Astonish me!

The litterati would instead look for qualities other than searching criticism and innovation. When Augustine wants to illustrate the principle that the Word must be loved before it can be understood, he turns naturally to the example of the grammarian's audience, who expect the grammarian to confirm the greatness of the texts on which he comments and so to confirm the assumptions that they bring to the texts even before they have read them.^[29] When the grammarians are criticized by the learned elite, as they are, for example, by Tiberius Claudius Donatus and by Macrobius, it is not for offering crambe repetita , nor conversely for extending their investigations to a level of expertise beyond the reach of educated laymen, but for superficiality, for falling short of the standards that the educated set.^[30] It is not a conceptually fresh and independent approach that is desired, but more of the same; not brilliance, but application, industry, and affirmation. The expectations thus press perceptibly away from the exercise of purely intellectual gifts, and toward intellectual habits that express ethical qualities—qualities that inform a man's behavior overall, that reveal themselves in his personal relations no less

[29] Aug. De util. cred. 6.13, with which compare Servius's comment at, e.g., Aen. 7.647; and cf. Chap. 5 on the function, in part exculpatory, of figura, antiquitas , and the like.

[30] Ti. Claudius Donatus Interp. Vergil. 1.1.5ff. Georgii; Macrob. Sat. esp. 1.24.12-13, 5.18-22, with Kaster, "Macrobius" 235f., 252ff.

than in his professional activities, and that lend themselves to personal judgments.

We have already seen an example of how such qualities were viewed by the learned elite, in the Saturnalia of Macrobius, where the traits desirable in a grammarian are displayed in a setting of generalized, shared expertise.^[31] There the virtues of modesty (verecundia ) and diligence (diligentia ) are emphasized above all: they are displayed not only by the good grammarian, represented by Servius, but by all the invited participants; they are assumed to have been the controlling virtues of Vergil, as well, whose text is at the center of the discussion. The virtues in fact hold the world of the dialogue together, since they impel men to maintain their contacts with their culture's past at the same time that they allow harmonious and fruitful conversation to proceed. Each member has a special contribution to make out of his own learning; and the dignity of each man's learning corresponds to his place in a hierarchy of birth and age, from the linguistic expertise of the grammarian, the humblest of the group, through the religious knowledge of the great figure, Praetextatus. But at the same time individual knowledge is not stressed, and the express belief is that learning is distributed broadly throughout the gathering, so that one man is prepared to refine or to add to another's contribution without rancor or self-aggrandizement. It is thoroughly appropriate that when analogy is used, one of the conceptual foundations of the grammarian's skill, other members of the group in addition to the grammarian apply it and simultaneously extend it to ratify usages that the grammarian in his own work would not embrace; the sources of linguistic authority are broadened, just as authority in the dialogue as a whole is diffused throughout the group, not focused in one man.

The display of virtues and the diffusion of expertise and authority are part of Macrobius's ideal in his admiring portrait of the aristocracy of birth and letters at Rome. As we shall see, however, the ideal makes contact with facts of life that would affect the professional teacher wherever he taught. Not least important of these was the judgment of professional competence, which rested fundamentally with laymen. A professio at base was no more than a public statement made by men offering "their services . . . at their own discretion."^[32] In a world without specialized training and certification, without formal and rationalized means of establishing standards within the profession comparable to the institutions that serve those purposes today (graduate or professional schools,' learned associations and journals, for example), no mechanisms

[31] See Chap. 2 pp. 60-62; Chap. 5 pp. 171f., 195.

[32] Lewis, Atti 516.

were commonly available for reaching impersonal judgments, or judgments that at least aimed at being objective in principle. Inevitably—and in the circumstances, not unreasonably—one could claim to be competent because the community at large believed one was competent, and especially because respected and prominent members of the community said so. For private teachers, what mattered was the word of mouth passed by fathers willing to send their sons to one's school or by patrons suggesting they should. For teachers aiming at a publicly supported post, what mattered was the approval (probatio , ) of the town council or, correspondingly, of the senate at Rome or Constantinople.^[33]

Even for the formal probatio , the teacher necessarily depended on the good opinion of others outside the profession: in the absence of a formal test, general reputation and the recommendation of well-placed supporters were paramount.^[34] A rhetorician would make his name in part by displaying his literary talents in public contests: such opportunities for star turns and for building an independent reputation were part of the rhetorician's greater visibility;^[35] in this respect as in others, the grammarian's trajectory was lower. His expertise did not lend itself to public displays from which stellar reputations could be won, and in fact

the evidence for such displays of the skills and knowledge specific to his profession is virtually nonexistent.^[36] Instead, his expertise lent itself to displays in private settings and accumulated its reputation less dramatically through contacts made face to face.

As a result, we commonly see the grammarians playing the part of private consultants, a role comparable to that at the beginning of their history, when they were typically attached as slaves or freedmen to the great houses of late-republican Rome. That the role remains, despite the general rise in the grammarians' social status, can be attributed to the nature of their expertise. The idealized picture of Macrobius's Servius, making his modest contribution to the discussion of the learned nobiles in the intimacy of the salon, where his mores are as much on display as his learning, is not far removed from the scenes glimpsed in life. Faustus in Africa receives the liber epigrammaton of Luxurius for vetting. Sidonius Apollinaris in Gaul relies on Domitius for a similar service and invites him to share the contubernium of his estate. In Syria, Libanius recalls the advice he received in conversations with Eudaemon. Diomedes dedicates his compilation to Athanasius, a work written in lieu of conversation (which he also promises when they are together again) and in answer to Athanasius's interest.^[37] So, more generally, there is the typical claim that a work was undertaken at the suggestion or request of the dedicatee.^[38] The stimulus professedly comes primarily from the outside rather than from the writer's desire or duty to make a professional

[36] Only Gell. NA 16.6.1-12 might be so interpreted, concerning quispiam linguae Latinae litterator [= grammaticus ; see ibid. 11], Roma a Brundisinis accersitus , who made himself available for public tests of his expertise (experiundum sese vulgo dabat ). The passage shows, however, that these occasions were later than and independent of his summons to the town. Contrast Pamprepius, given a post at Constantinople after reciting either a poem or a discourse on the soul (Malch. frg. 20 and Damasc. V. Isid. frg. 178 Zintzen, with Part II no. 114); or Ioannes Lydus, given a chair of Latin at Constantinople by Justinian after delivering an encomium of the emperor (De mag. 3.28, with Part II no. 92); or Phoebicius of Bordeaux, gaining his chair through the patronage of a son (Auson. Prof. 10.29-30), probably the rhetorician Attius Patera; and see further below.

[37] Faustus (Part II no. 58): Luxurius Anth. Lat. 1:1.287. Domitius (Part II no. 50): Sidon. Apoll. Carm. 24.10ff., Ep. 2.2. Eudaemon (Part II no. 55): Lib. Ep. 255. Diomedes (Part II no. 47): GL 1.299.2-10, 391.16-17. Cf. also Chap. 4 for the vignettes of consultation and questioning Pompeius inserted in his work, probably as a reflection of his own experiences.

[38] Besides Diomedes, see Don. Epist. praef. 15f. Hardie; Serv. De fin., GL 4.449.3, Cledon. Ars, GL 5.92.2ff.; Prisc. Inst., GL 2.2.24ff. and 194.2ff., De fig. num., De met. Terent., praeex., GL 3.405.8ff.; Eutych. Ars de verb., GL 5.447.8ff. This ancient convention was of course firmly rooted in belles lettres : see the examples collected by Gudeman, Tac. Dial. 41 n. 1; Janson, Prose Prefaces (esp. 116ff. for late-antique texts); Murgia, "Date" 124f.

contribution to his field:^[39] the work itself forms part of the grammarian's social relations, as a token or gift that tightens the bonds between the dedicator and the recipient.^[40]

Several strands in these relations are brought together in the brief story that the grammarian Sacerdos sketches at the beginning of Book 3 of his Artes grammaticae.^[41] Sacerdos dedicated his first book to the vir clarissimus Gaianus, a companion of long standing (contubernalis ), his contemporary and onetime fellow student. Gaianus's father, Uranius, learned of the book and was pleased by it—either, as Sacerdos charmingly remarks, because it was "not put together ineptly" or because it was dedicated to his son—and compelled Sacerdos to undertake a second book. Sacerdos gladly obeyed Uranius's commands and was subsequently commended by him to two other gentlemen of senatorial rank, whose commands Sacerdos then satisfied in his third book. What began as a perhaps well-calculated act of friendship, growing out of the intimacy of contubernium , thus developed into a series of relations that allowed the grammarian to increase the number of his patrons.

For Sacerdos, his long-standing comradeship (contubernium ) with Gaianus was evidently the key: it provided the initial opportunity to display his skills; and it acted as an implicit guarantee, a social code that could assure Gaianus's father and friends that Sacerdos—although certainly not a vir clarissimus himself—was still in some sense a gentleman, one of them, a decent and respectable sort to whom they could deign to give their attention.^[42] It served the purpose, in other words, that is served in different circumstances by stressing mores. Recent research on patronage and public careers under the empire has properly emphasized the importance attached to personal qualities in, say, letters of recommendation, and has noted the weight given good character in appointments and promotions—a weight apparently at least equivalent to specialized competence and merit, as those qualities might be judged in technical or impersonal terms.^[43] In connection with teachers and others in the liberal professions, we have already had occasion to remark the

[39] The only significant exception—and that only partial—is provided by Prisc. Inst. : besides noting the urgings and commands of Iulianus (see n. 38), Priscian speaks of the undertaking (GL 2.2.2f.) as a rem . . . officio professionis non indebitam. For still a different note—the grammarian's work presented as the fruit of gentlemanly leisure—see Serv. De met. Hor., GL 4.468.6, with Chap. 2 p. 66.

[40] Cf. also Chap. 2 p. 68 for comments on teachers' dedications to their pupils.

[41] GL 6.496.5ff. The significance of the passage, and especially the importance of contubernium , are well brought out by Champlin, Fronto 45.

[42] Cf. GL 6.497.1: quoniam iubere dignati estis.

[43] See Pedersen, Public Professionalism 30f.; and esp. Saller, Personal Patronage 95ff.; note also the implications of Campbell, "Who Were the 'Viri Militares'?," and of Brunt, "Administration." Personal or ethical criteria, though almost inevitably subjective in their application, were not therefore arbitrary in themselves; for they made sense (even if it was not our kind of sense) in their social setting, and a coherent account of them can be given. It seems necessary to stress this point, in view of a tendency simply to link or identify the subjective and the arbitrary—so, e.g., Frei-Stolba, reviewing Saller in Gnomon 55 (1983) 142, speaks of "durchaus subjektiven und willkürlichen Kriterien."

importance of such personal criteria: the infrequency with which professional skills are singled out as the sole or primary grounds of commendation, and the regularity with which mores receive equal or greater emphasis, so that the distinction between technical attainments and personal qualities is blurred and the notion of competence embraces both.^[44] This union of ethos and expertise derived from the learned elite's traditional and comforting assumption that not only were the good to be identified with the learned, but true learning must proceed from good character. Gellius's famous definition of humanitas , equating it with learning and education (eruditio institutioque ) on the one hand and with such ethical qualities as devotion and discipline (cura et disciplina ) on the other, generalizes what Ausonius surely had in mind when he described one grammarian of implacidi mores as also (and inevitably) doctrina exiguus .^[45]

In everyday usage, the stress on good mores offered the assurance that the man so described would fit in and not disrupt the stately world of learned gentlemen. Here again the literary description of the good grammarian Servius, "at once admirable in his learning and attractive in his modesty" (Sat . 1.2.15) as he takes his place inconspicuously in the placid microcosm of the symposium, makes contact with the real world of social relations, since that description conforms to the terms used in letters of commendation.^[46] Meeting the need to establish that their subjects are the right sort, such letters run heavily to the praise of personal attributes. Writing in behalf of Eudaemon of Pelusium, Libanius emphasizes his family's respectable standing () and virtuous restraint ().^[47] On another occasion, Libanius refers again to the same characteristics, using them to bracket the observation that Eudaemon is "one of the eloquent," and goes on to add that Eudaemon "is fairer in my eyes than a brother."^[48] In a letter that can stand as a

[44] Chap. 2 pp. 65f., with nn. 137, 138, 140.

[46] See Chap. 2 n. 137.

[47] Lib. Ep . 108.2.

concise example of the genre, written to Alexander of Heliopolis in behalf of an homonymous teacher, Libanius says, "As for the grammarian who shares your name, consider him a good man and one very well versed in literature, and enroll him among those who know how to be a friend."^[49]

When Libanius promises this magnate of Heliopolis that the grammarian Alexander knows how to be a friend, he is promising in effect that Alexander will fit snugly into the man's network of connections and dependents: as a worthy recipient of his favor and protection, one prepared to make what return he could. Alexander was moving from Antioch to Heliopolis, and Libanius here is smoothing the way, enlisting a powerful patron for Alexander, passing him along in much the same way that Gaianus's father passed Sacerdos along to his friends, and probably in much the same terms. The patron could presumably be relied on, for example, to spread the word among his connections in order to guarantee Alexander a healthy number of students in his new position; in return, Alexander could be relied on, if needed, to put his knowledge at the patron's disposal, and in general to play the part of a friend by spreading the word of his patron's loyalty and influence.^[50] The letter would thus ease Alexander into a position of genteel dependence as the lesser member of a patronage relationship—what has recently and well been described as the continuing, reciprocal, but asymmetrical exchange relationship between men of unequal social status.^[51]

The importance of the relationship can be inferred in the case of Sacerdos. It is glimpsed directly in the case of Alexander or of Iulianus Pomerius, who enjoyed the protection of the potens Firminus at Aries late in the fifth century.^[52] It is perhaps attested most warmly by Augustine, when he recalls the patronage of Romanianus at the early stages of his career. A dominant figure of Thagaste, Romanianus gave Augustine his protection and intimacy when Augustine returned home to teach: he entrusted Augustine with the education of his son, and he even helped

[49] Lib. Ep . 1256.3.

[50] Cf. Lib. Ep . 467, thanking a patron for his help at court and promising to make a return through his attentions to the man's son, his pupil; similarly Ep . 491.3.

[51] See Saller, Personal Patronage 1ff., after J. Boissevain. The lesser member of the pair could make his exchange concretely, e.g., by presenting a composition to his patron; or, something intangible but perhaps even more important, by publicizing the greater man's excellence whenever the opportunity arose. For a blending of both types of return, see Lib. Or . 1.111, a panegyric of the PPO Strategius Musonianus explicitly undertaken to repay an earlier favor.

[52] V. Caes . 1.8f., with Part II no. 124.

Augustine in his ambition to seek a more prestigious position at Carthage, although Romanianus would have preferred him to remain in his patria .^[53]

Such relations (and we shall examine other cases below) were crucial to the teacher's professional life—but not only to his professional life, a point that should be emphasized immediately. Just as the line between professional and personal qualities is blurred in the language of approbation and commendation, so the exercise of patronage ignores any distinction between the professional and the personal spheres of the teacher's life: it embraces both and confounds them. In 361 the grammarian Calliopius and his father were teaching Libanius's son at Antioch; Libanius wrote to Iulianus, the praeses Euphratensis , asking him to show favor to Calliopius's brother-in-law, for by doing so, Libanius said, Iulianus would dispose the teachers more favorably toward the son.^[54] In 358, the grammarian Tiberinus, a native of Arabia teaching at Antioch, turned to Libanius when his son was the object of a lawsuit in his native province: Libanius wrote to the governor Maximus, told him that Tiberinus was "a good man and a teacher of the poets," and tried to engage his good will for Tiberinus's son, basing his appeal on Maximus's respect for culture, his respect for Tiberinus's native and adopted cities, and his respect for Libanius himself.^[55]

The same years find Libanius exerting himself in behalf of the grammarian Cleobulus, who had been Libanius's teacher and whom Libanius therefore looked on "as a child does his father."^[56] The extent of Libanius's efforts suggests that this was more than just a form of words. On two occasions we can see Libanius mediating between one or another of Cleobulus's relatives and a pair of imperial officials, seeking a staff position in one case and special favor in the other.^[57] Particularly revealing is a group of eight letters that show the persistence of a good patron at work.^[58] In 359 Cleobulus was being hounded with a suit that threatened a fine he could not afford if it went against him. He turned to a man he regarded as a friend () and protector (), Themistius, who at the time was proconsul of Constantinople, but Themistius was unresponsive.^[59] Libanius then entered the affair, sending a series of letters in the course of the year to Themistius, to Clearchus, who held an official

[53] Aug. C. Acad . 2.2.3.

[54] Lib. Ep . 678; cf. Chap. 2 p. 69; and cf. n. 50 above.

[55] Lib. Ep . 337. Cf. Ep . 431, promising the intervention of Libanius and others in a suit lodged against the grammarian Acacius, who had in the past sent some of his students on to Libanius: Ep . 398.

[56] Lib. Ep . 361.2; cf. Ep . 231.1. The documentation for what follows is discussed fully at Part II no. 32.

[57] Lib. Ep . 361, 82.

[58] Lib. Ep . 52, 67-69, 90, 91, 155, 231.

[59] See esp. Lib. Ep . 67.

position of some sort, and to the physician Hygi(ei)nus, a friend of Themistius. Libanius's repeated petitions had no effect beyond causing his irritation to mount, and the suit finally went against Cleobulus. Changing his strategy to meet Cleobulus's need, Libanius appealed to his own relative Bassianus, who was also a former pupil of Cleobulus. Reminding Bassianus of "the labors of the good Cleobulus" in his behalf, Libanius called on him to reciprocate.^[60] The appeal worked, and Bassianus gave Cleobulus the money to pay his fine, saving him from ruin.

These examples suggest the range of needs served by the teacher's patronal connections; and the case of Cleobulus perhaps best shows the real strength of the conventional pieties through which such connections were expressed: the stress on personal relationship, the pressure toward reciprocity and loyalty, the desire to show oneself a good man by helping, on a variety of fronts, the good man under one's protection. But the case of Cleobulus also clearly suggests the limits of the conventions: although Cleobulus thought he could call on Themistius as friend and protector, Themistius felt free to turn a deaf ear; and Libanius's influence was plainly too weak to move the greater man.

A similar lesson is still more vividly drawn in the case of Nicocles. We have already remarked what prominence Nicocles enjoyed during Julian's reign on the strength of his connection with his former pupil. But the death of his great protector exposed Nicocles to men ready to cut him down to size, or who at least were no longer inclined to pay him respect. Nicocles' isolation is seen in his relations with his former pupil Clearchus, who as vicarius Asiae turned on his old teacher after Julian's death to get his own back for some wrong he believed Nicocles had dealt him in the past.^[61] The precise form of Clearchus's harassment is not clear, but it was sufficiently serious that Nicocles called on Libanius to mediate. Libanius first wrote a letter of rebuke to Clearchus, by turns warning him that Fortune is fickle—"The goddess delights both in raising up the fallen and in bringing down the proud"—and reminding him that he had once been Nicocles' "child" (), that Clearchus's father had entrusted him, as Clearchus had entrusted his brother, to Nicocles.^[62] The latter appeal, of the sort that had worked with Bassianus, had no effect. Although Clearchus's reply is not preserved, one can guess that it put Nicocles and Libanius alike in their places; for the tone of Libanius's second letter is dramatically different, fawning where he had earlier

[61] Lib. Ep . 1265, 1266, 1492. For vague reference to Clearchus's motives, see Ep . 1266.3; and cf. n. 64 below.

been reproachful, and with a sneer for Nicocles: "Granted that Nicocles is otherwise worthless [], he at least deserves respect [] because he 'made you such as you are, godlike Achilles.'"^[63] On the most charitable interpretation, the letter shows Libanius trying to salvage a bad situation by taking the tack least likely to offend Clearchus.^[64]

Nicocles' vulnerability after Julian's death reminds us again of the perils of dependence on a single patron.^[65] It was a lesson Pamprepius had a chance to learn early on, when he fell afoul of the magnate Theagenes at Athens and was forced to quit the city.^[66] It was a lesson, too, that must gradually have impressed itself on Lollianus of Oxyrhynchus when in making his petition to the emperors he pinned his hopes on the only man of consequence he knew, the courtier he calls "Brother": the surviving draft of Lollianus's letter represents his third attempt to enlist "Brother's" help, and the presumption is strong that the man had simply ignored the first two, as Themistius ignored Cleobulus's appeal.^[67]

But even a more differentiated network of relations did not guarantee success or peace of mind; most notably, one risked being squeezed between feuding patrons. So, for example, Diphilus had the misfortune to be the protégé of Libanius when the latter was on wretched terms with another of Diphilus's patrons, Eustathius, governor of Syria. Eustathius promised a favor to Diphilus, who was uncertain whether or not he should accept. (It was less than he had hoped for, and he was a bit put out as a result.) Diphilus turned for advice to Libanius, who persuaded him to take what had been offered. Eustathius thereupon reneged, in order, so Libanius says, to make fools of both of them: "For he thought it would be a terrible thing if Diphilus could spread the word to Palestine [the province where Diphilus taught] that I have enough power to benefit a friend."^[68] Clearly, genteel dependence in any form was not without its drawbacks.

[64] Note that Clearchus, like Nicocles and Libanius, was a pagan; his persecution of Nicocles therefore cannot be ascribed to an antipagan reaction after Julian's death. For Nicocles' paganism and Libanius's allusions to a backlash, see Part II no. 106; for Clearchus, see von Haehling, Religionszugehörigkeit 118f.

[65] Cf. Chap. 2 p. 132, on Simplicius and Pamprepius.

[66] Malch. frg. 20, with Part II no. 114. To judge from his subsequent career (see n. 65), he did not take the lesson to heart.

[67] PCollYoutie 2.66 = POxy . 47.3366, B23. For the level of Lollianus's other connections, see Chap. 3 n. 138.

[68] Lib. Or . 54.56-57, with Part II no. 49. The incident concerns Diphilus qua poet, but that should not alter the general point made here.

In the discussion so far we have been concerned almost exclusively with private patronage and mediation. But the presence of the governor Eustathius in the incident just noted and the desirability of an extended network of patrons raise another issue: the place of official patronage and state intervention in the lives of our grammarians. It is time, then, to shift our attention from private to ostensibly public connections: in the rest of the chapter we shall consider the involvement of imperial officials in the schools of the provinces, the apparent blurring of the distinction between municipal and imperial authority in scholastic matters, and the implications for the kinds of relationship we have examined in the pages above.^[69]

Evidence for such involvement is not lacking; we should note first, however, that the most notorious piece of evidence is in fact of little consequence. I refer, of course, to Julian's law of 17 June 362, which was intended, as he subsequently made clear, to drive Christian teachers from the schools.^[70] By this law, Julian commanded that all teachers be approved first by a decree of the town council, with the agreement of the best men of the town, and that the decree then be sent for his own review and judgment. The demand was unprecedented in two regards. First, it asserted explicitly that in the matter of education the local council merely acted as the emperor's proxy.^[71] Second, and perhaps more striking, it extended the scope of the council's probatio to include not only teachers seeking municipally funded positions and immunities^[72] but all teachers, public and private.^[73] This amounted to a clear restriction of what had hitherto been a free professio .

But neither this broad limitation nor the principle of imperial review of local probationes (in any application of the term) appears to have outlived Julian's equally unprecedented sectarian purposes. The law, it is true, was incorporated in the Code; but in 438 it could only have been a dead letter, with little more meaning for the conduct of affairs than the old command of Constantine (CTh 16.10.1) that if the palace or any other public building were struck by lightning the haruspices should be consulted retento more veteris observantiae . In fact, we do not know of a single case before or after Julian's reign in which the normal probatio of a provincial

[69] The role of the state in late-antique education has often been discussed. In the notes below I have limited my citations to the most recent or most important secondary works. See also the survey in Walden, Universities 130ff.

[70] CTh 13.3.5, with Julian. Ep . 61c Bidez.

[71] Thus the initial decretum curialium and subsequent dispatch to the emperor is justified quia singulis civitatibus adesse ipse non possum .

[72] As was already customary; see above, n. 33.

[73] Plainly implied by the generalized command iubeo, quisquis docere vult, non repente nec temere prosiliat ad hoc munus, sed iudicio ordinis probatus decretum curialium mereatur optimorum conspirante consensu .

curia was reviewed by the emperor or in which a private teacher needed to undergo such probatio .^[74] And for private teachers' we have a text more eloquent than this silence. In 425 private teachers in Constantinople were prohibited from using the auditoria reserved for those the city senate appointed as public teachers and were compelled to do all their teaching in private houses.^[75] Intended to limit the competition that publicly appointed teachers faced, the law is the most explicit attempt (beyond Julian's) to regulate private teachers;^[76] but it shows no interest whatever in determining who might teach, merely where. If even so stringent a measure at the heart of the empire did not impose greater control than this, then we should assume a fortiori that there was still less regulation in the provinces. Private teachers, who probably made up the majority of the profession, would thus have stood entirely outside the state's concern.

The case of public teachers in the municipalities is, however, more complex; for although we find no general iudicium by the emperor of the sort Julian envisioned, there are a number of specific episodes that reveal the hand of imperial authority. Governors and other officials play important roles in appointments, either making recommendations or approving curial decrees.^[77] Teachers at Athens are removed from their chairs by the governor, and others are installed in their places.^[78] The emperor himself is called on to approve immunities already decreed to a rhetorician of Antioch by the town council.^[79] Salaries, too, show the reach of the central government. Down to the end of the third century salaries drawn from imperial funds are attested only for the chairs of Greek and Latin rhetoric established by Vespasian at Rome and for the chairs of rhetoric and philosophy created at Athens through the benefaction of Marcus

[74] Note that the clause concerning imperial review (hoc enim decretum . . . accedant ) was dropped from the version of the law entered at CJ 10.53.7; and with the clause quoted in n. 73, contrast CTh 13.3.6 (an. 364), si qui erudiendis adulescentibus vita pariter et facundia idoneus erit, vel novum institual auditorium vel repetat intermissum , intended to cancel the effects of Julian's law.

[75] CTh 14.9.3.

[76] Restraint of competition: So correctly Speck in his review of Lemerle, BZ 67 (1974) 386-87, also noting that it is incorrect to speak of a monopoly given to the public teachers (against Lemerle, Premier humanisme 364; cf. also Jones, LRE 999).

[77] Lib. Or . 1.48, 82-83, 106 (cf. Ep . 1255.2-3); Aug. Conf . 5.13.23. These instances, and the others remarked in nn. 78, 79, 81, and 82, will be discussed below.

[78] Lib. Or . 1.25.

Aurelius.^[80] Beginning in 299, however, we find more such salaries in the provinces: they are received by Eumenius at Autun, by Eudaemon of Pelusium at Elusa, by Libanius and one of his rivals at Antioch; supplementary payments to teachers at Ancyra are arranged by the governor Maximus; salaries from the fisc are ordained by Gratian for grammarians and rhetoricians at the capital cities of the northern Gallic diocese, as they are a century and a half later by Justinian at Carthage.^[81] Finally, state intervention in salaries could take still another form, as we see at Antioch around 360: there the teachers drawing municipal salaries were paid from the city's funds, but the funds were administered by imperial officials.^[82]

Such evidence—combined with the continuing, indeed expanded, immunities granted to professores —has been taken to demonstrate imperial encroachment on a sphere of activity formerly belonging to the cities, an "étatisation" of education, in fact, that was part of a general characteristic of late antiquity, the increased centralization of power.^[83] Now I do not propose here to deny that imperial authority seems to have been more intrusive in our period than previously; it is difficult to believe that it was not, but the range of evidence available from early and late empire alike does not provide the grounds for solid proof. I would, however, like to suggest that it is mistaken to think that any effective "étatisation" occurred, and I would like to show that such notions as encroachment and intrusion must be set in their historical context and must be modified.

[80] Suet. Vesp . 18, in a general account of Vespasian's benefactions; Dio Cassius 72.31.3, similarly for Marcus; Philost. V. soph . 2.2 (p. 566); Lucian Eun . 3, 8; with the comments of Parsons, PCollYoutie 2, pp. 445f. Note esp. the observation of Parsons that the notices for this earlier period in the Historia Augusta , when they are not uselessly vague (Hadr . 16.8), are either probably (Pius 11.3) or certainly (Alex. Sev . 44.4) anachronistic; on the last passage see also Nutton, "Archiatri " 216. Chairs of rhetoric, at least, seem to have been established at Constantinople soon after its founding, with pay from the emperor, presumably on the model of Rome; see Lib. Or . 1.35, 37, with Kaster, "Salaries" 39ff.

[81] Eumenius: Pan. Lat . 5.14.14 Galletier; see Chap. 3 n. 87. Eudaemon: Lib. Ep . 132. Libanius: Ep . 28, 740, 800. Libanius's rival: Or . 1.110. Maximus at Ancyra: Lib. Ep . 1230.2. Gratian: CTh 13.3.11; see Chap. 3 pp. 116f. Justinian: CJ 1.27.1, 42; see Chap. 3 pp. 117f.

[82] Lib. Or . 31.19; see Liebeschuetz, Antioch 152f., though caution is needed concerning some details of his discussion and the scope of his conclusions; Kaster, "Salaries" 57f. Note, however, that there is no evidence that the emperors or their subordinates ever attempted to set the amount that a city could pay its teachers from municipal funds; cf. Kaster, "Reconsideration" 102 n. 9.

[83] See esp. Marrou, Histoire 434ff., esp. 441f., with, e.g., Wolf, Schulwesen 41f., Riché, Education 7, Kirsch, "Cura " 284-86.

When we look more closely at our evidence, we see, first, that some distinctions are needed, since not all the evidence points in the same direction. Take, for example, perhaps the most drastic and commonly noted case, Libanius's report (Or . 1.25) that in 339 or 340 the provincial governor at Athens simply removed the three sophists who held the public chairs in the city and saw to their replacement by three other professors. This episode plainly has little to do with the control of education as such, but it has a great deal to do with the maintenance of public order, which was a governor's oldest and most critical responsibility: in their self-absorbed rivalries the teachers had allowed their students to run riot, and the governor decided to control the disturbances by removing their most conspicuous cause.^[84] Other evidence clearly involves extraordinary circumstances. Thus, it is true that the of Nicomedia asked the governor of Bithynia to confirm their decree inviting Libanius to assume the town's chair of rhetoric. But it is also true that the governor's authority met a special need at that time; for Libanius was removing himself from Constantinople, where he had been brought under a charge of magic by his rivals, and Limenius, the proconsul of Constantinople, had written to Nicomedia in an attempt to deter the town from making the appointment. The council evidently thought it prudent to meet that attempt by adding the governor's support to their decree; and it is noteworthy that when the town council of Nicaea passed a similar decree for Libanius, at just about the same time and in the same circumstances, it had not bothered to involve the governor.^[85]

Similarly, it is true that the immunities voted by the town council for Eusebius, a teacher of rhetoric at Antioch, required the confirmation of the emperor,^[86] although the regular approval of immune status had been established as a council's responsibility since the reign of Antoninus Pius, more than two centuries earlier. But again, the case of Eusebius was clearly not at all regular. Eusebius was evidently not a full-fledged sophist of the city, qualified to stand among the statutory number of teachers who might receive immunity; rather, he was merely the assistant of Libanius, who as a sophist of the city of course enjoyed immunities of his own. In other words, Eusebius stood extra numerum : in voting him immunities the council had gone beyond its customary powers, and the

[85] Lib. Or . 1.48, with Martin and Petit, Libanios 1.218f. ad loc. ; and cf. Norman, Autobiography 161 ad loc .

[86] See n. 79 above.

grant of extraordinary immunities required imperial confirmation.^[87] In fact, the entire episode in which Eusebius was involved under Theodosius recalls nothing so much as the case of the sophist Claudius Rufinus nearly two hundred years earlier, under Septimius Severus and Caracalla. Both teachers—Eusebius certainly, Rufinus probably—received extraordinary grants of immunity to begin with; both voluntarily undertook a single liturgy for their cities (Eusebius an embassy) after they had received their grants; in both cases the towns subsequently attempted to impose further liturgies once a chink in the armor of immunity had been opened; and both men succeeded (Eusebius with Libanius's patronage) in having the immunity reaffirmed by the emperors.^[88]

Against these special cases there is much evidence that attests both the cities' ordinary freedom to appoint or depose their public teachers^[89] and the laxity of imperial control. A governor might be asked to lend his authority to a council's decree in delicate circumstances, as in the case of Nicomedia's invitation to Libanius; but then again he might not, as in the case of Nicaea. A governor might be called upon to confirm a special honor for a sophist^[90] or to intervene so that a dilatory council would pass its decree inviting a sophist to assume a chair.^[91] But when a sophist

[88] For Rufinus (= PIR C.998), see CIG 3178 = IGR 4.1402 = Syll .^[3] 876 (Smyrna), with Bowersock, Greek Sophists 41, and Nutton, "Two Notes" 54. Eusebius's tale is most easily followed in Lib. Ep . 907.3-5. For Libanius's patronage, see Ep . 870 and 904-9, enlisting the aid of officials and other influential men at Constantinople; Ep . 918-21 and 960, thanking some of the same men, and others, for their help.

[89] See the references at n. 33 above.

[90] Lib. Or . 54.48; cf. n. 79 above.

can reject a praetorian prefect's request that he move from one city to another and can selectively respect letters from the emperor himself according to his own convenience or will, we are clearly not witnessing the iron-fisted exercise of central authority.^[92]

This is obviously not to say that the provincial governors and other officials were of no consequence; the instances already noted suggest quite the opposite. But these and other cases available for inspection also suggest the need for a more delicate and differentiated view. First, we might note that what has broadly been regarded as encroachment can often on closer examination be seen to be the result of invitation, a teacher's or a town's willing exploitation (whether benign or malign) of an official's formal authority or informal influence. For instance, a teacher knew that he could appeal to the governor if an action of the council displeased him.^[93] He might even petition to bring the emperor's authority directly to bear on the .^[94] In their squabbles with one another, too, the teachers were not slow to make an ally of the local imperial authority. So Eunapius retails at length how the proconsul of Achaea was drawn into the rivalries that centered on Prohaeresius, after the sophist's competitors at Athens had bribed an earlier governor to secure his exile;^[95] Prohaeresius's enemies probably arranged his conviction on some trumped-up charge, in much the same way as Libanius's rivals at Constantinople tried to suppress him by bringing a charge of magic with the

[92] Lib. Or . 1.74, Libanius refuses the prefect's invitation to return to Constantinople from Nicomedia but accedes to an invitation from the emperor; Ep . 405, 432, 438, 439, Libanius later refuses to heed letters from the emperor demanding his return to Constantinople when he wishes to remain in Antioch; cf. Or . 1.100; and see Kaster, "Salaries" 41ff.

[93] See the generalization at Lib. Or . 25.49, a passage that also plainly stresses the council's freedom in its ordinary dealings with its appointed teachers. Such intervention in council business as Libanius posits here Was hardly new; cf. Bowersock, Greek Sophists 38; Burton, "Proconsuls" 104f.; and esp. C. P. Jones, Roman World 99ff., 111ff., with the references at p. 112, nn. 69, 70.

[94] So Lollianus of Oxyrhynchus, PCollYoutie 2.66 = POxy . 47.3366, with Part II no. 90. As Parsons correctly concluded (PCollYoutie 2, pp. 441ff.), there would seem to have been no precise precedent in law or in fact for Lollianus's specific request that the emperors compel the council of Oxyrhynchus to make the arrangements necessary for the payment of his salary. For appeals to the emperor, see also the cases of Eusebius and Claudius Rufinus at n. 88 above.

aid of the proconsul Limenius.^[96] As we have just seen, the latter case caused the town council of Nicomedia to take the unusual step of seeking a governor's approval for their decree of appointment; for a similar initiative by a town council—but undertaken in more placid circumstances and more informally—we can recall an episode in the career of Augustine, the request by the curia of Milan that Symmachus, then prefect of Rome, provide a man to fill their chair of rhetoric.^[97]

Little in these dealings is fundamentally new, or peculiar to late antiquity; and this touches on a second point worth emphasizing, the continuity apparent from the early empire to the late. Consider the example just mentioned. It is true that in his official capacity as city prefect Symmachus would have been unusually well informed about the teachers of Rome, in consequence of the law of 370 that made the prefect and the magister census responsible for keeping track of all students arriving in the city.^[98] But his action apparently had little to do with his official powers, which of course had no connection with Milan; further, it must be remembered that, his official position aside, Symmachus was at the time one of the most prominent orators and men of letters at Rome—just the sort of man whose judgment would be valued in such a matter. This personal standing is surely adequate in itself to explain the request of Milan's council; the episode, in fact, is reminiscent of the request Pliny once made of Tacitus to send a suitable teacher from Rome for the position that Pliny was trying to establish at Comum.^[99]

In other cases, of course, the official status of the man making a recommendation was evidently more pertinent to the matter at hand, but here too it is difficult to find a clear break from precedent to set the late-antique experience apart. Libanius, for instance, recalls that the proconsul of Achaea, Strategius Musonianus, was responsible for a decree of the at Athens inviting Libanius to take up a chair of rhetoric in the city.^[100] It is not clear whether Strategius formally introduced the motion for curial action or informally (but no less influentially) recommended Libanius to the council's attention. From what Libanius says,

[96] Lib. Or . 1.43ff. For the governor's involvement in teachers' quarrels, cf. perhaps also Himer. Or . 46.1-2, in a speech "against his enemies and against the proconsul Basilius," produced in the midst of what Himerius calls a war that the entire people is waging against him at Athens.

[97] Aug. Conf . 5.13.23.

[98] CTh 14.9.1.

[99] Plin. Ep . 4.13.10. Note, however, that Pliny, writing as a private person, offers no guarantee that the man Tacitus sends will get the position, but stresses that he is leaving the choice open to the parents.

[100] Lib. Or . 1.82-83, 106.

the former seems marginally more likely,^[101] but the distinction does not much matter for the point being made here; a governor's ability to make formal proposals for curial vote is attested already in the reign of Plus.^[102] And with Libanius's anecdote we may compare a letter of Fronto to his son-in-law, Aufidius Victorinus, governor of Germania Superior at the time, in which Fronto asks him to use his influence to secure a public appointment for a rhetorician in one of the cities in his province: although Fronto is careful to note that he has not heard the man speak but is relying on the judgment of others, he says nothing to suggest that the request itself is extraordinary, and he takes it for granted that Victorinus would simply have this sort of favor within his power.^[103] In another connection, we have already seen that the emperor might grant extraordinary immunities to a favored teacher; here again, the record extends back to the second century.^[104]

Such continuity is of course most apparent in the formal measures that fixed the regular immunities of teachers from the end of the first century onward. These measures, moreover, raise a final point concerning the actions of the state; for where those actions can be seen to take on a broad pattern, as opposed to the particular actions of individual officials, or where they sketch a general tendency that might be termed "policy," their chief characteristics are reticence and a rather spotty internal consistency.

As they are first attested for the principate in Vespasian's grant of A.D. 74, these immunities grant freedom from taxation and from the quartering of soldiers.^[105] They were, however, considerably expanded within

[102] Implied by Forsch. Eph . 2 no. 19 = Abbott and Johnson, Municipal Administration no. 98 = Inschr. Eph . 1a (IGSK 11:1) 21; see Burton, "Proconsuls" 105.

[103] Fronto Ad amic . 1.7 (p. 169 van den Hout): velim, domine, ut adiuves eum, quo facilius in civitate aliqua istius provinciae publice instituendis adulescentibus adsciscatur . Compare Lib. Ep . 1366.1-3, quoted at n. 91 above.

[104] See above, pp. 219f., on Eusebius of Antioch, with Philost. V. soph . 2.10 (p. 589), 2.30 (p. 623) (Marcus Aurelius to the sophist Hadrian; Caracalla to Philostratus of Lemnos).

the next two generations, so that by the time Hadrian early in his reign confirmed his predecessors' acts, the list included a wide range of exemptions from potentially time-consuming or expensive offices and munera .^[106] The grants were continued throughout the third century^[107] and con-Armed under Constantine, who declared that the favored teachers and their possessions in their cities were exempt from all public service (omnis functio ) and extended the exemption to their wives and children.^[108] They were later reconfirmed under Gratian, Valentinian II, and Theodosius in 382,^[109] under Honorius and Theodosius II in 414,^[110] and again under Theodosius II and Valentinian III in 427.^[111]

Throughout their history these immunities were regarded as the favors (beneficia ) grateful emperors bestowed as a return for and to facilitate the exercise of the teachers' serviceable skills.^[112] As practitioners of necessariae artes , the teachers were useful () to the cities in which they worked, like the physicians with whom they are commonly paired in the laws or like the navicularii and negotiatores who sustained the grain supply

[107] Implied by CJ 10.53.2 (Gordian), 10.47.1 (Diocletian and Maximian), beyond the writings of Ulpian and Modestinus.

[108] CTh 13.3.1 (an. 321/124) + 13.3.3 (an. 333) = CJ 10.53.6; cf. CTh 13.3.10 (an. 373 [Seeck]) for medici et magistri urbis Romae and their wives. Teachers' sons of curial status were declared liable for curialia munera in 383; thus CTh 12.1.98, ipsos quin etiam filios magistrorum, qui ex curiali stirpe descendunt, simili modo obnoxios esse decernat (if magistri here = "teachers"); but see also CTh 13.3.16 (an. 414) ad fin .

[109] CTh 11.16.15 (on the exemption of certain functionaries and officials from sordida munera ), which includes the clause circa ecclesias [om. CJ 10.48.12], rhetores atque grammaticos eruditionis utriusque vetusto more durante ; cf. CTh 11.16.18 (an. 390).

[110] CTh 13.3.16-17.

[111] CTh 13.3.18.

[112] Beneficia divorum retro principum : see CTh 13.3.3. The grounding of immunities in the recipients' utility is especially well emphasized by Kuhn, Städtische und bürgerliche Verfassung 1.83, the introduction to what is in many ways still the best general survey of the subject; and by Herzog, "Urkunden" 981ff.

of Rome.^[113] Indeed, the teachers could speak of their very profession as a or munus , and reasonably enough.^[114] Immunity from further personal and financial obligations in theory left them free to devote themselves to maintaining the stable, civil life of their cities by producing a class of men honesti and docti . Secondarily—a point emphasized far more in modern accounts than by the emperors and the jurists—they might contribute to the well-being of the empire as a whole by educating the men who would become worthy members of the bureaucracy.^[115]

The granting of these immunities is commonly regarded as the centerpiece of the emperors' educational policy—Hochschulpolitik, politique universitaire , or the like—and there may in fact be little harm in speaking of "policy" (rather than, say, "the expression of partially articulated attitudes and impulses"), as long as we recognize that this policy ran heavily to laissez-faire and involved some noteworthy contradictions. For example, despite the fact that the immunities touched both imperial and municipal burdens,^[116] the choice of immune teachers was left to the cities; in this respect, the occasional involvement of provincial governors is perhaps less surprising than the fact that their involvement was not regular and formal. Or again, the exemptions were meant to encourage teachers to devote themselves to their own civitates , since their enjoyment of immune status was conditioned on teaching there;^[117] but exception was made for those who came to teach at Rome, encouraging the best or

[115] On the former point, see Chap. 1 above. The latter motive (on which see, e.g., Walden, Universities 265ff.; Marrou, Histoire^[6] 446f.; Nellen, Viri 15ff.; Klein, "Kaiser Julians Rhetoren- und Unterrichtsgesetz" 90ff.; Nixon, "Latin Panegyric" 95f.) is mentioned in a measure in support of teachers for the first time in the sixth century, and then concerning only the teachers at Rome (Cassiod. Var . 9.21.8). Although the emperors, their ministers, and their panegyrists were hardly unaware of the schools' usefulness in this regard (cf. CTh 14.9.1 [an. 370] ad fin ., Eumen. Pan. Lat . 5.5.3f. Galletier, Anon. Pan. Lat . 7.23.2, Prisc. De laud. Anast . 248ff.), the modern emphasis seems to confuse results with intentions; it must certainly be wrong to project the intention back, for example, to Vespasian's funding of chairs of rhetoric at Rome (so, e.g., Steinmetz, Untersuchungen 85). The emphasis in the codes and jurists falls on the teachers' services to their own localities, through which they gain the beneficia of the emperors.

[116] See Neesen, "Entwicklung" 204f.

[117] See esp. Dig . 27.1.6 (Septimius Severus and Caracalla ap . Mod.), 9.

at least the most ambitious teachers to leave their homes for the capital and its greater prestige.^[118]

Perhaps most important, the policy might jeopardize the cities it was meant to help, by removing from the list of potential liturgists members of the upper classes whose contributions would matter most for the cities' well-being. As G. W. Bowersock has rightly emphasized, it was for this reason that Pius circumscribed his predecessors' more extravagant gestures by establishing that each city's council could designate only a small number of teachers and physicians to receive immunity.^[119] Yet even this impulse toward consistency and restraint is joined by a contradictory measure, for which Pius himself was responsible: those who were exceptionally learned () would be entitled to immunity even beyond the number of teachers with ordinary immune status.^[120] Since a claim of exceptional learning would be based on the testimony of influential friends,^[121] and since those best placed to secure such testimony would be the most prominent and wealthy among the learned, the measure would provide a loophole for the very men whose services would most benefit their cities.^[122] It is therefore not surprising to find from the second century onward that cities could be altogether reluctant or inconsistent in recognizing immune status—especially of those extra numerum —and that immunity could be revoked and at times might drive a wedge between the teachers and their towns.^[123] Nor is it surprising that the provision of immunity needed to be repeated and reconfirmed so often by the emperors.^[124]

The immunities, an indirect subvention of teachers by the state, represent something less than a consistent and dearly thought-out policy;

[118] So another ruling by Septimius Severus and Caracalla; see Dig . 27.1.6 (Mod.), 11, with Nutton, "Two Notes" 61ff.

[120] Dig . 27.1.6 (Mod.), 10.

[121] See the case of Aelius Aristides (well analyzed by Bowersock, Greek Sophists 36ff.), who rounded up letters from Pius, Marcus Aurelius, and a former prefect of Egypt while making his case before a succession of provincial governors.

[122] On this exemption, see esp. Nutton, "Two Notes" 52ff.

[123] See Bowersock, Greek Sophists 34ff.; the case of Eusebius, pp. 219f. above; and Lib. Ep . 293, 723.

[124] Remarked by Marrou, Histoire 435.

but the other, direct, state support—salaries granted from the imperial fisc to provincial teachers—was still less coherent. As was noted above, such salaries are more visible in the late empire; but the central government appears at the same time to have been unsystematic in its grants, which at least throughout the fourth century seem to have been treated as an exceptional privilege. This point is perhaps especially to be emphasized in view of a common belief that imperial officials normally had a voice in selecting teachers precisely because they administered a generalized system of state salaries.^[125]

If we look at the known cases, we see two circumstances in which imperial funds were applied to provincial teachers. First, there were occasions when emperors wished to support teachers in a specific region: thus, Gratian's measure for the provincial capitals in the northern diocese of Gaul when the imperial court was resident at Trier;^[126] or Justinian's for Carthage when teachers were made part of the imperial apparatus after the reconquest.^[127] Beyond that, we find what amount to ad hoc or ad hominem grants. Eumenius at Autun receives a salary from the fisc as a sign of the favor he enjoyed with the imperial masters he had lately served as magister epistularum and with the understanding that he was to apply the money to revive the school of Autun as his own benefaction to the city;^[128] the salary thus provided a way of rewarding a faithful client—ut intelleges meritis tuis etiam nostram consuluisse clementiam —and of channeling imperial funds to the city through a prominent citizen who would gain in prestige thereby. At Ancyra, the governor Maximus provides supplements for the teachers' (presumably local) income; the supplements, it is important to note, are mentioned as extraordinary benefactions deserving special praise, along with the literary contests, prizes, and public works Maximus also provided.^[129] Imperial salaries in kind () were granted to Eudaemon at Elusa and Libanius at Antioch; the latter case, which is the better known, shows clearly that the grant depended on the personal relations between the

[125] On this Mitsprachrecht see, e.g., Wolf, Schulwesen 42; Kirsch, "Cura " 284.

[126] CTh 13.3.11 (an. 376). Even this measure does not seem to suppose that the selection of teachers would be taken out of the cities' hands, and the clause ut singulis civitatibus . . . nobilium professorum electio celebretur (see Kaster, "Reconsideration" 107f.) may well suppose the opposite.

[127] CJ 1.27.1, 42 (an. 534). With these localized measures contrast the generalized grant of imperial salaria to churches, confirmed at CJ 1.2.12, 2 (an. 451); cf. Theodoret. HE 1.11, with Jones, LRE 898f.

[128] Pan. Lat . 5.14.14 and 11. 3 Galletier. There is no suggestion that other cities in the region enjoyed similar benefits.

recipient and the official in charge (in this instance, the praetorian prefect) and that the salary could be bestowed, diminished, and restored according to the favor or hostility of the official of the moment.^[130]

It is possible, of course, that such subsidies became more widespread over time. Nearly all our precise evidence dates from the fourth century; and if, as has sometimes been argued, the cities exerted ever less control of their revenues and contributed less to the maintenance of their services,^[131] the imperial government may have picked up more of the slack. Some indirect evidence points in this direction. Procopius reviles Justinian for abolishing all the imperial salaries for teachers and physicians that his predecessors had established;^[132] although this is a patent distortion by a notoriously hostile source, the charge should suggest that a significant number of such salaries did exist.^[133] Much the same seems to be implied by a law of 531 that grants a privilege to memoriales, agentes in rebus , and all others qui salaria vel stipendia percipiunt publica , mentioning teachers of liberal studies in that number.^[134] Still, it would certainly be mistaken to suppose that the cities withered, and the state extended its support, uniformly and universally. In this regard, as in virtually all aspects of life in antiquity, the fact of local variation must be kept in mind. We know that even as some cities did decline, others maintained a good measure of independent prosperity in the late fifth century and well into the sixth.^[135] Just so, we know that some cities continued to provide their teachers' salaries in the same period.^[136]

Much of the aid that the imperial government provided was no doubt well intentioned, and a good bit of it no doubt gave welcome additional

[131] See Jones, LRE 733f.; and esp. Liebeschuetz, Antioch 153ff.

[132] Anecd . 26.5.

[133] Procopius's statement plainly contradicts the known directives of Justinian, for teachers at Carthage in 534 (CJ 1.27.1, 42) and at Rome in 552 (Nov . "App." 7.22; the conflict with the latter was already noted by Nutton, "Archiatri " 211 n. 144). If it has any basis in fact, the charge probably grows out of some attempt to limit such salaries for economy's sake.

[134] CJ 3.28.37, le-f.

[135] See Roueché, "New Inscription" esp. 183ff.

support to local institutions: so, for example, A. H. M. Jones may well have been correct to suggest that Gratian's measure of 376 enabled some towns of northern Gaul to boast endowed chairs where they had had none before.^[137] But imperial intervention cannot have resulted in imperial control in any substantive sense. There was certainly no control of curriculum; nor did the state attempt to make the professions of grammar and rhetoric hereditary, as it did for so many occupations and statuses. And whatever form the imperial presence took—making itself felt, say, by subsidizing teachers or in a governor's influence—and whether that presence was invited or provoked, novel or traceable through long-standing precedent, it was typically no more than sporadic or indecisive or formless.^[138] It is difficult to discover in the evidence examined above an éatatisation of education newly developed in late antiquity, and it is still more difficult to think that such éatisation could have come about in the absence of a bureaucratic structure—a specialized, formalized, rationalized mechanism of the central administration—through which the state could effectively assume responsibility for organized education. No such structure existed at any time in antiquity. In late antiquity, when a governor spent vast amounts of his time hearing cases and overseeing tax collection, and when a governor typically held office for no more than a year, he could scarcely have spared much time to supervise education, and any supervision that a given governor might have cared to exercise could have had little long-term impact. The question whether imperial officials became more intrusive is therefore finally secondary.

It is secondary because in the circumstances such involvement could not help but be—as to all appearances it was—haphazard and particular, circumscribed by the same kinds of personal connections as those discussed earlier in this chapter. Whether one was dealing with local or imperial authorities, with a principalis in the town council or with the aloof and powerful governor, those dealings would not go far unless one were well supplied with patrons. Our analysis of the state's role thus brings us back full circle. One can surely speak of a mingling of authority, municipal and imperial, formal and informal. But as virtually all the evidence surveyed in the preceding pages shows or implies, it was an authority exercised unpredictably or ad hoc , harnessed or combatted by mobilizing personal relations.

[137] See Jones, LRE 998, with Kaster, "Reconsideration" 108f.

[138] Cf. the remark of Speck in his review of Lemerle, BZ 67 (1974) 392, concluding his discussion of the so-called university of Constantinople: "Staatlich institutionalisierte und gefördete Bildung aber blieb in der Spätantike ein Versuch, der sich nur partiell und vorübergehend durchsetzen konnte."

A teacher receiving an imperial salary in kind at Elusa desires the privilege of commuting the salary to cash: we find his case being put by his friend Libanius, who approaches an influential private citizen of the town and asks him to do what he can to secure the favor.^[139] A teacher about to move from Antioch to Heliopolis has the way prepared, again by Libanius: letters solicit the favor of Domninus, the governor of Phoenice, and the local potens Alexander, himself a former governor of Syria now living in retirement in his home town.^[140] The differences in the formal status of one's supporters are finally less significant than the general and enduring patterns of patronage and personal connections, which remain the same regardless of the circumstances in which they are found. In the fourth century, Augustine begins his career at Thagaste, and then moves on to Carthage, with the help of his local supporter, Romanianus, as he later makes his way from Rome to Milan with the support of the city prefect Symmachus.^[141] In the fifth century, Pamprepius's lot at Athens is linked to his relations with the magnate Theagenes; when he is driven from Athens to Constantinople after falling out with Theagenes, he gains a public salary as a grammarian at the capital through the patronage of Illus, whose official career thereafter largely determines the course of Pamprepius's fortunes.^[142] In the sixth century, Ioannes Lydus begins his career by being drawn into the imperial service by a cousin and by a home-town acquaintance who is praetorian prefect at the time; when he later grows dissatisfied with the service, he is appointed to a chair of Latin grammar by Justinian himself—a connection perhaps mediated by another patron, Gabriel, to whom Lydus presently dedicates his first works when Gabriel is the city prefect.^[143] This last detail is worth savoring. It calls to mind another small piece of information that we owe to Lydus: the notice that Suetonius dedicated his Caesares to his patron Septicius Clarus when the latter was prefect of the praetorian guard.^[144] These two dedications to patrons by sometime imperial servants and antiquarians bridge more than four hundred years in the life of the empire, reminding us how much that life remained cast in the same mold.

[139] Lib. Ep . 132, to Eutocius of Elusa, in behalf of Eudaemon of Pelusium.

[140] Lib. Ep . 1255, to Domninus; 1256, to Alexander. On the latter, see also above at n. 49; and for Alexander's governorship, see PLRE I s.v. Alexander 5, pp. 40f.

[141] See nn. 53 and 97 above. Note esp. the role of Augustine's Manichaean friends in arranging a display before Symmachus, Conf . 5.13.23: ego ipse ambivi per eos ipsos manichaeis vanitatibus ebrios . . . at dictione proposita me probatum . . . Symmachus mitteret .

[142] See Part II no. 114.

[143] See Part II no. 92.

[144] Ioan. Lyd. De mag . 2.6. On Suetonius and Septicius, see Syme, Tacitus 501, 778f.; Wallace-Hadrill, Suetonius 6; and more guardedly Baldwin, Suetonius 36ff.