Greek, Roman, and Byzantine Studies 46 (2006) 101–122 © 2006 GRBS Rummaging in Walz’s Attic: Two Anonymous Opuscula in Rhetores Graeci Thomas M. Conley WO SHORT ANONYMOUS PIECES are published by Walz in volume III of his Rhetores Graeci, Per‹ t«n t°ssarvn m°rvn toË tele¤ou lÒgou (“On the four parts of a complete speech,” pp.570–587) and Per‹ t«n Ùkt∆ m°rvn toË =htorikoË lÒgou (“On the eight parts of the rhetorical art,” 588–609), that bring out some of the peculiar difficulties one encounters when trying to make sense of what are, apparently, examples of Byzantine teachings on rhetoric. Some of the difficulties are due not to the nature of the material but to the way Walz presents it. Granted, his labors were immense, as were those of the printers who set the type and produced the ten volumes of Rhetores Graeci; but his practice of referring the reader to other places to see what he saw in the manuscripts he transcribed is sometimes annoying—as in, for instance, the Anonymous “Ekthesis” at III 725–748, where we are re- peatedly referred to the prolegomena by Maximus Planudes in volume V. Walz does the same sort of thing in his presentation of the two texts under consideration here. Accordingly, we will need to reconstruct the texts he saw; and that requires the “rummaging” alluded to in the title of the present paper. Both pieces are, as published, rather unusual. To summarize: “On the four parts” begins with the reminder that Aphthonios teaches that a finished encomium will have a prooimion, a diêgêsis, an agôn, and an epilogos. Prooimia in general will have four parts and will vary from genre to genre; and there are good examples to be found in the orations of Gregory of Nazianzus (570.8–571.28). There follow some brief comments on nar- rations and the agôn (“contestation” or “argument” in the broadest sense) (571.28–572.6); and some equally brief remarks about epilogoi, which will differ depending on whether the T 102 RUMMAGING IN WALZ’S ATTIC speech is an encomium or a mere lalia (“talk”), as Menander teaches (572.7–24). Here the manuscript contains a short chapter printed as part of the Synopsis rhêtorikês by Joseph Rhakendytes earlier in volume III (at 562.16–564.8), Per‹ toË p«w de› énagign≈skein =htorikoÁw b¤blouw (“How to read rhetorical works”), stressing the necessity of understanding the ÍpÒyesiw developed by each author (563.19f.). Then we find (572.25–573.8) a list of authors to imitate in composing panegyrics, chiefly Gregory of Nazian- zus, but also Aristeides’ Panathenaikos and the speeches of Cho- ricius and Michael Psellos. For deliberative oratory, look to Chrysostom, Basil the Great, Isocrates, Demosthenes, and, again, Gregory of Nazianzus, whose speeches cannot be re- stricted to any one genre. The all-wise (sof≈tatow) Psellos is al- ways useful. Models for letter-writing are next provided (573.9– 25): Gregory of Nazianzus, Basil the Great, Gregory of Nyssa, Synesius, Libanius, and, once more, “the all-wise Psellos.” At this point in the manuscript, Walz tells us in a note, is a chapter on iambic composition that also appears in the Synopsis rhêtorikês by Joseph Rhakendytes at III 559.14–561.15. The section from 560.19 to the end (éretØ st¤xvn pr≈th …) is stylistically consistent with the Anonymous. The models here (562.10–15) are George of Pisidia, Nikolaos Kallikles (a late 11th-century physician-poet), Ptochoprodromos, Gregory of Nazianzus, Sophocles, and “the more eloquent parts” of Lycophron. In certain circumstances, we read where the Anonymous text resumes (at 573.26), dactylic hexameters are recommended, using diction from all sorts of dialects, especially “Ionic” (573.26–574.4). Models can be found in Homer, Oppian, Dionysius the Periegete, Tryphiodorus (in The Fall of Troy), Musaeus, and others. The list of recommended authors is unusual. Even more unusual, however, is the assertion that an iambic verse should contain a complete thought and that enjambment should be avoided (560.19–561.7). Hence, instead of writing efiw tØn §ruyrån ébrÒxoiw pos‹ pãlai par∞lye Mvs∞w, ≤ d¢ t«n Afigupt¤vn fãlagj ÍpobrÊxiow ¶ndon §krÊbh THOMAS M. CONLEY 103 one should write Mvs∞w però yãlassan ébrÒxƒ drÒmƒ, AfigÊptiow d¢ to›w kÊmasin §krÊbh. Moreover, lines that are built around the traditional six peri- statika are to be emulated. Thus, Stauro› P°tron kÊmbaxon §n ÑR≈m˙ N°rvn, which covers, in order, the pragma (“crucified Peter”), chronos (past tense), tropos (“upside down”), topos (“in Rome”), and the prosôpon (“Nero”), the aitia being understood. We will see these peristatika again later. It is clear that the discussion has by this point drifted far from considerations of “the four parts.” Here (574.5) a new title heading appears: EfisagvgikÚn to›w melloËsi grãfein =htorik«w (“Introductory material for those intending to write rhetori- cally”). There is much in the way of ideas (¶nnoia), subjects (Ïlh), and diction (l°jeiw) to be found in the poets, who have much to offer the logographer. The text at 575.12ff. is very close to that of Rhakendytes at 562.18ff. in “How to read rhetorical works”; and Walz tells us in a note at 575.16f. that the manuscript at this point continues the text in Rhakendytes all the way to 564.3. Especially useful, Anonymous tells us, is Homer, from whom examples are provided in prose para- phrase (575.19–576.13). The discussion then turns to dihgÆ- mata, of which there are three kinds: simple (èpl∞), detailed (§ndiãskeua), or very detailed (§gkatãskeua). Examples are provided of each in different accounts of the madness of Ajax (576.13–578.27). At 579.1 the discussion turns to l°jiw, and there follws a disquisition on propriety as determined by subject-matter (Ïlh) and ideas (¶nnoia), with long lists of diction appropriate to, e.g., descriptions of meadows and battles. Diction may be refined or “natural,” inflated or dry (examples are provided); and of particular importance in word-choice is mastery of the Attic dialect, for which Aristophanes and Thucydides provide models (582.19–586.21). For this section, too, we find many parallels (not noted by Walz) in the chapters per‹ l°jevw in Rhakendytes’ Synopsis, at 526.28–534.28: e.g., Anonymous at 580.25ff. and Rhakendytes at 526.28f.; 581.5ff. and 527.10ff.; 582.15ff. and 529.15ff.; 584.13ff. and 533.15–534.5, etc. 104 RUMMAGING IN WALZ’S ATTIC At 586.21–587.16, we see a conclusion and, it appears from didãjomen at 586.24 and 587.13f. and the list of progym- nasmata at 586.24–30, a transition to the piece immediately following, Per‹ t«n Ùkt∆ m°rvn. That title, in actuality, covers only the first section 588.4– 590.5, where the author takes up in turn the traditional eight “parts” familiar from Hermogenes’ On Ideas: ¶nnoia, l°jiw, sx∞- ma, m°yodow, k«lon, sunyÆkh, énãpausiw, =uymÒw. Once these have all been defined and explained, the author turns to “the most serviceable” (xreivdest°rvn, 590.6) progymnasmata. These he takes up in exactly the same order as the list at the end of “Four parts,” depending heavily on Aphthonios but explicitly leaving out xre¤a and gn≈mh (also left out are koinos topos, synkrisis, thesis, and eisphora tou nomou). Accordingly, he be- gins with mËyow, offering an Aphthonian definition and three- way division of species, logikÒw, ±yikÒw, miktÒw (590.8–591.7). The next is diÆghma, for which he also depends on Aphthonios, making his three-way division into dramatikÒn, flstorikÒn, politikÒn, and referring to the Aphrodite story; and, like Aph- thonios, he stresses the role of the six peristatikã: prÒsvpon, prçgma, xrÒnow, trÒpow, tÒpow, afit¤a (591.9–593.11). In the treatment of ±yopoi¤a at 593.13–595.16, Aphthonios is once again the source, in the distinctions introduced between etho- poiia, eidolopoiia, prosopopoiia and in the three-way breakdown into pathetikai, êthikai, miktai. Indeed, 593.13–22 and 594.1–12 are almost verbatim versions of Aphthonios. The advice about style is also consistent with that of Aphthonios (ı texnikÒw at 594.31?). In his treatment of the fourth exercise, ¶kfrasiw (595.18– 598.9), the author begins to depart from Aphthonios. Unlike Aphthonios’ entry, that in “Eight parts” is far fuller; and it uses four of the six peristatika as an armature. First, then, descriptions of a person (prosôpon) should be both detailed and orderly, pro- ceeding from top to bottom, he writes. Then comes description of pragmata, e.g. a battle scene with all the gory details (596.3– 20). And there follow on this entries on description of a kairos, e.g. winter (596.20–30), and of a topos, e.g. a pond or a meadow, along with some general guidelines. The treatments of §gk≈mion and cÒgow (598.11–601.17) are organized in a similar way. Praise, we are told, can be of THOMAS M. CONLEY 105 prÒsvpa, prãgmata, kairo¤, tÒpoi, and “creatures without reason” (êloga). In praising persons, one should speak about their native land, hometown, lineage, upbringing, pursuits and training, virtues both physical and spiritual, and katå tÊxhn prãjeiw. It is also useful to introduce comparisons. “Blame” is the antithesis of praise, but it uses the same headings (kefã- laia). The section on kataskeuÆ and énaskeuÆ (confirmation and refutation) (601.11–609.9), very unlike the treatments in Aphthonios, are also organized around the peristatika: the person, place, time, manner, and reason. After covering those headings, one should introduce arguments from the contrary, amplify, make comparisons (sugkr¤seiw), and provide exam- ples. These can be used in any kind of logograf¤a (602.25– 604.23). It is up to the writer to decide in what order these headings should be put; and every diêgêma—the account of the madness of Ajax, for instance—will cover the earlier discussed headings. There follows a long and exhaustive display on that theme (605.20–608.18) that seems to sum up both this and the previous lesson (reading, with the Paris MS., e‡pvmen at 609.1). If these two pieces can be described as disorganized and strangely proportioned, the same might be said of the late fourteenth-century manuscript Walz used, Parisinus graecus 2918. It is a rather large codex, 191 folio leaves measuring some 30 × 20 cm. “minutissimis scripta,” as Walz puts it (I 140f.).1 In the first 131 fol., we find the traditional Her- mogenean corpus. Then, in a different hand, the following: 132–136v: Nikolaos Sophistes Progymnasmata (cf. I 266–394 Walz) 136v–140: Nikephoros Basilakes Ethopoiiai (I 423–525) 140v: Severus of Alexandria Ethopoiiai (partial) (I 539–548) 141–152: Nikolaos Sophistes Progymnasmata continued 152v–169: Libanius, progymnasmatic meletai (partial) 169–174: Nikephoros Basilakes Progymnasmata continued 174v–176v: Anonymous opusculum beginning pçsa lÒgou fid°a §k mer«n Ùkt∆ sÊgkeitai … 177–179: Per‹ t«n tessãrvn mer«n (III 570–587) 1 Photostats of pages from it can be seen in I. Lana, I Progimnasmi de Elio Teone (Turin 1959) Tav. II, and G. Ballaira, Tiberii De figuris Demosthenicis (Rome 1968) Tab. VII. 106 RUMMAGING IN WALZ’S ATTIC 179v–181v: Per‹ t«n Ùkt∆ mer«n (III 588–609) 182–190: Theon Progymnasmata (I 145–257) 190rv: Rufus Technê rhêtorikê (III 447–460) 190v–191v: Tiberius Peri schêmatôn (VIII 527–577) Parisinus graecus 2918 is certainly not the only disorderly Byzantine manuscript, yet it does give the strong impression of having been rather hastily thrown together without regard for making it intelligible to what one would imagine was its primary audience, the student of rhetoric. It would have made more sense, for instance, to put Rufus and Theon at the be- ginning, along with, perhaps, the material in fols. 174–181 and Tiberius. These would then be followed by an extensive anthol- ogy of exercises for imitation (and perhaps inspiration). There are still problems, however. How does the brief exposition at fols. 174vff. differ from that in 179v–181v, for instance? Why have both Nikolaos and Basilakes? The answer to that question may be that Basilakes uses biblical episodes; but why start with ethopoiiai instead of the traditional first exercises, muthos and diêgêma? Either the scribe who copied Par.gr. 2918 was working at great speed simply to transcribe an older manuscript or he did not know what he was doing—and there is plenty of evi- dence for the latter.2 Nor is Par.gr. 2918 the only place where “Four parts”—or something very like it—can be found. As Walz notes (III 570), an excerpt from Ottobanianus 173, fol. 148ff., printed by Bekker in a note at Anecdota graeca III 1081f., looks very like the text of “Four parts.” And Vat.gr. 883, fol. 220v–223v, published by both A. Kominis and D. Donnet,3 is also quite similar to Walz’s text. Moreover, Laur. LVIII.21, fol. 206f. (the end of the manuscript is missing), and LV.7, fol. 324v–330v, contain what looks like Walz III 570–573.25 plus 559.14–560.16, which Walz prints as part of Rhakendytes’“Synopsis” of rhetoric; and which, as we have seen, appears also in 2918. In Barberinianus 2 See Lana’s collection of scribal errors in the Theon text alone, Pro- gimnasmi 33–38. 3 A. Kominis, Gregorios Pardos (Testi e studi bizantino-neoellenici 2 [Rome 1960]) 127–129; D. Donnet, Le traité PERI SUNTAJEVS LOGOU de Grégoire de Corinthe (Brussels/Rome 1967) 319.137–323.282. THOMAS M. CONLEY 107 240, we find “Four parts” attributed to Psellos! And at Laur. LVIII.21, fol. 193–205, we find a work entitled Per‹ logogra- f¤aw that has the same opening as 2918, fol. 174f. All these manuscripts, like 2918, have been dated at mid- to late- fourteenth century. But things are even messier. We noted earlier that various sections of “Four parts” appear also in Rhakendytes. Some but not all of the overlaps with his “Synopsis” are noted by Walz. There are many apparent parallels, especially in the section on lexis (at III 579ff.): e.g., Anon. 572.25–573.8 ≈ Rhak. 521.7–27; 580.1ff. ≈ 528.9–13; 580.25f. ≈ 526.28f.; 581.5f. ≈ 527.10f.; 581.20f. ≈ 527.24f.; 582.15f. ≈ 529.15f.; 583.6f. ≈ 532.18f.; 583.27f. ≈ 533.7f.; and 584.13ff. ≈ 533.15–534.5. Moreover, the lists of recommended writers in the alleged text of “Four parts” printed by Bekker seems closer to Rhakendytes at, e.g., 521.15ff. and 526.13ff. than to the text printed by Walz at III 572.25–573.25. Rhakendytes, of course, made no pretense to originality (indeed, he is quite open about his borrowings: see III 471.11–17). His Encyclopedia, from which the “Synopsis” is drawn, “borrowed” extensively from Nikephoros Blemmydes in the sections on physics and logic; and in the “Synopsis” itself, the chapter on the basilikos logos (III 547–558) is cribbed almost verbatim from Menander Rhetor (368–378 Spengel). Could it be that he also lifted the section on iambics at 559.14– 562.15 and part of the chapter on “How to read rhetorical books” at 562.19–564.8 from “Four parts”? Or did Anon- ymous consult Rhakendytes’ “Synopsis”? There is probably no definitive answer to this; but it is clear that there is more in “Four parts” that is not in Rhakendytes than there is that is in the “Synopsis.” It is also clear that those two sections are the only ones in the “Synopsis” that are stylistically consistent with the style of “Four parts”—but more on that below. Turning now to “Eight parts,” this piece’s title, like that of “Four parts,” covers only the first section (588.4–590.5), which ends with an awkward transition to the sections on the various progymnasmata exercises that make up the rest of the treatise and which, as we have seen, owe their greatest debt to Aphthonios. And like that of “Four parts,” the first section of “Eight parts” has nothing to do with what follows; but, once the progymnasmata are taken up, it is far better organized and 108 RUMMAGING IN WALZ’S ATTIC methodical than “Four parts”—particularly in its systematic deployment of the peristatika in its explanations of how to develop each exercise. This, incidentally, seems to be unpar- alleled in the progymnasmatic literature. One might be tempted to see overlaps with Rhakendytes here at the beginning of “Eight parts,” too. The opening sen- tence (588.4–7), for instance, is very close to what we read in the “Synopsis” at III 516.12f.; and the definitions of énãpausiw and =uymÒw at 589.25–30 are very close to Rhakendytes’ at 545.9f. and 545.19ff. Both, however, go back (at least) to Her- mogenes Peri ideôn 1 (220.6ff. Rabe), where they are introduced as the basic elements of each of the stylistic ideai; and that list can be found throughout the Byzantine rhetorical tradition. So there is no need to bring Rhakendytes into the picture at all. “Eight parts” appears also in Laur. LV.7 (which we saw earlier) at fol. 331–334, beginning pçsa grafØ =htorikÆ, kín me¤zvn …, and ending at 605.19 in Walz’s edition. Just before it is a short treatise that begins with the same words as the piece in Par.gr. 2918 at fol. 174v, pçsa lÒgou fid°a §k mer«n Ùkt∆ sÊgkeitai. In the superscription, however, the contents are listed as efisagvgikÚn per‹ logograf¤aw, per‹ §pistol«n, per‹ st¤xvn—which would seem to match the contents of the early parts of “Four parts” at 574.5–575.16+562.17–564.3 and 573.9–25+559.14–562.15. The superscription also attributes this piece to Gregory of Corinth; but Bandini notes that Greg- ory’s name seems a later addition to the text4 and in his view the treatises at fol. 324ff. and 331ff. constitute a single work. Whatever Bandini’s reasons were, the same might be said of the two pieces printed by Walz from Par.gr. 2918. I suggested earlier that the end of “Four parts” might be construed as a transition to “Eight parts”; but that, too, might be a scribal contrivance. Nevertheless, the two share a conspicuous stylistic feature. In “Four parts,” we see almost 70 occurrences of second-person singular verb constructions or pronouns in just under seventeen pages, as against only 10 first-person verbs, most with se or soi as their objects, for instance: 4 And see further on this attribution D. Donnet, “Précisions sur les oeuvres profanes de Grégoire de Corinthe,” BIHBelge 37 (1966) 89–91. THOMAS M. CONLEY 109 ka‹ tÚ §n toÊtoiw sofÚn filokrinÆseiw, ka‹ ényologÆseiw ka‹ épotami- eÊseiw t“ logism“ (575.6f.); éllå dØ ka‹ tå t«n tragik«n ka‹ tå loipå poihtik≈teron metabãl˙w ka‹ tÚ frãsevw kãllow diabl°pvn prosoikeioËn ßjeiw tª ofike¤& frãsei §k parafrãsevw prÚw tØn prokeim°nhn ÍpÒyesin. éllå f°re ka‹ aÈto‹ mikrÒn ti suneisen°gkvm°n soi ka‹ efiw ¶nnoian ka‹ efiw frãsin (576.10– 14); ıròw ˜pvw kayarã §stin ≤ frãsiw: §ån d¢ tapeinot°ra soi prÒkeitai ¶nnoia, tuxÚn ˜ti guna›ka ı basileÁw ¶ghmen §p‹ t“ teknopoi∞sai, épagge¤l˙w Íchl«w (582.26–29). Likewise, in “Eight parts,” second-person verbs and pronouns appear over 50 times in 20+ pages, as against 10 first-person constructions, for instance ıròw ˜pvw p°fuke taËta pãnta gen°syai. ka‹ oÈd¢n t«n lexy°ntvn parå fÊsin §st‹ ka‹ édÊnaton (592.6–8); épÚ t«n aÈt«n går peristatik«n §pixeirÆseiw, énatr°pvn µ aÎjvn §n t“ c°gein tÚ kakÒn, ka‹ épÚ t«n loip«n tÒpvn t«n §ktey°ntvn soi (605.16–19); shme¤vsai d¢ ka‹ toËto, ˜ti §pån §gkvmiãz˙w tinã, ka‹ tå katoryvy°nta toÊtƒ katå m°row di°rx˙, xrØ poie›sya¤ se kayÉ ßkaston aÈt«n sÊnkrisin (609.1–4). In each, imperatives appear five times. There is, in short, a very pronounced “addressed” quality to both—and indeed, at 590.6, we see f¤le. Is this featuring of the second-person a mere literary device, or is f¤le addressed to a real person? It is in any event a feature not often found elsewhere in the rhetorical literature, the most conspicuous example being the treatise printed by Russell and Wilson1 as Treatise II (pp.76–225 = 368–446 Spengel), which exhibits a very similar stylistic turn. It has parallels also, of course, in advice literature, such as the Consilia et narrationes attributed to Kekaumenos (late 11th cent.);5 and in paraenetic discourse, e.g. Basil I’s advice to Leo (PG 107.XXI–LVI). But 5 D. A. Russell and N. G. Wilson, Menander Rhetor (Oxford 1981). 6 Ed. B. Wassiliewsky and V. Jernstedt (St. Petersburg 1896); and see C. Roueché, “The Rhetoric of Kekaumenos,” in E. Jeffreys (ed.), Rhetoric in Byzantium (Aldershot 2003) 23–37. 110 RUMMAGING IN WALZ’S ATTIC these were surely not models for whoever composed the two treatises we have been considering. There are, as it happens, some parallels from more closely related literature. In Psellos’s “synopsis” of rhetoric, composed in the popular fifteen-syllable “political” verse around 1060 and dedicated to young Michael Doukas,2 we see at the very beginning Efi mãyoiw t∞w =htorik∞w tØn t°xnhn, stefhfÒre, ßjeiw ka‹ lÒgou dÊnamin, ßjeiw ka‹ lÒgou xãrin, ßjeiw ka‹ piyanÒthta t«n §pixeirhmãtvn (1–3) and later éfÉ •nÒw pou ÙnÒmatow prÒfasin ßjeiw lÒgvn, oÂon går e‡poiw ériymÒn, sÊlleg° moi tosoËton §j flstor¤aw, d°spota, t“ lÒgƒ proshkoÊshw (266–268) and sÁ dÉ ¶xe moi tØn sÊnocin, e‰t §r≈ta yarroÊntvw, kég≈ soi tØn diãlusin l°jv toË zhtoum°nou. e‰tÉ oÈ yaumãzeiw, d°spota, toË grãfontow tØn t°xnhn, ín ¶x˙w efiliktãrion braxÁ t∞w ˜lhw t°xnhw (287–290). It might be added that the epithet stefhfÒre shows up three more times; and d°spota ten times in all. Although not nearly as “addressed” as Psellos’s synopsis, John Tzetzes’ “Epitome of Rhetoric” (III 670–686 Walz), com- posed about 1160, occasionally slips in an imperative (e.g. ka‹ diãgnvsin aÈt«n ékoÊvn mãye at 679.14) and several second- person constructions and pronouns (e.g. zhtÆmata gin≈skeiw d¢ tåw dekatre›w tåw stãseiw, 679.27; nËn d¢ ka‹ parade¤gmasin §ke›na pareisãjeiw, 680.1; and, at the very end, §g∆ d¢ toËto oÏtv soi tanËn §pejhgoËmai, 686.6). It is not known who the addressee was, but Tzetzes’ “cast of characters” includes mis- chievous elves (bãskanoi telx›new, 675.10), adulterers and pro- fligates (moixo¤, êsvtoi, 675.32), whoremasters (pornobosko¤, 681.30, 683.2 [of Socrates!]); and his examples of exotic cus- toms at 670.11–13 include polygamy among the Scythians, sex with mother among the Persians, and between siblings in the Caucasus. All of this suggests a young man as the intended 7 Poemata pp.103–122 Westerink; also in Walz III 687–703. THOMAS M. CONLEY 111 audience, probably from an elite family, for composing poems for such families is how Tzetzes made the little money he did. Nor are the parallels limited to treatments of rhetoric. The Per‹ suntãjevw lÒgou ≥toi per‹ toË mØ soloik¤zein ka‹ per‹ barbarismoË (“on the syntax of the sentence and how to avoid solecisms and barbarisms”) by Gregory of Corinth, also com- posed in the mid-twelfth century, has a distinct “addressed” quality, as it reads like a transcript of a teacher’s lessons to students. After each “lesson,” we hear the teacher saying ka‹ oÏtv suntãjeiw Ùry«w.6 The treatise has five basic parts: (1) On the parts of speech (lines 1–218); (2) On the cases (219–300); (3) On prepositions (301–403); (4) On the syntax of the verb (404– 504); and (5) On barbarisms (505–631) No second-person con- structions are found in sections 3 and 4; but in the other sec- tions we find second-person constructions almost too numerous to inventory. Examples include ıròw, fidoÁ tå Ùkt∆ m°rh toË lÒgou, éllå tå m¢n dÊo énagkaiÒtata, tÚ ˆnomã fhmi ka‹ tÚ =∞ma (12f.) fid¢ ˜pvw tå sumplhrvtikå t∞w §nno¤aw =Æmata épedÒyhsan kãtv metå pollã. prÒsexe oÔn ka¤, kín efiw tÚ l¤an §ktetam°non ≤ ¶nnoia ßlk˙, tÚ =∞ma zÆtei tÚ épodotikÚn ka‹ mhd°pote stªw, efi mØ eÍrÆseiw aÈtÒ (133– 136) prÒsexe oÔn ka¤, katå toÁw ériymoÊw … §ntaËya går to›w plh- yuntiko›w, …w ıròw (192–195) prÚw taËta goËn tå érx°tupa épeikÒnize ka‹ épeÊyene tåw émfiballo- m°naw soi l°jeiw ka‹ oÈ barbar¤seiw (520–522) oÏtv grãfvn §n pçsin oÈ barbar¤seiw … diå toËto oÈd¢ grãfom°n soi per‹ aÈt«n (626–629). It may be worth noting that the verbs and the pronouns are consistently singular, except for Gregory’s occasional editorial “we,” which suggests that it is not a class that is being ad- dressed but an individual. Not much is known about Gregory’s 8 E.g. line 36 and repeatedly; references are to line numbers in Donnet’s edition. Gregory (ca. 1070–1156), metropolitan of Corinth, also composed a commentary on Hermogenes’ Per‹ meyÒdou deinÒthtow (Walz VII.2 1090– 1352) and a capacious treatise on Greek dialects (Libri de dialectis linguae Graeci, ed. G. Schaefer [Leipzig 1811]). Some MSS. attribute to him the brief Per‹ trÒpvn at Walz VIII 763–778, as well. 112 RUMMAGING IN WALZ’S ATTIC career beyond the fact that he was a rhetorician and gram- marian who became a bishop. While he may have occupied a chair in the Patriarchal School at some time, there is no ex- plicit record of that or of any social circle he may have been close to.7 In other words, the identity of the intended audience of this work will probably never be known. Our last example is found in the Per‹ ÙryÒthtow suntãjevw (“On correct syntax”) by John Glykys,8 a student of Gregory of Cyprus (of whom more below) who in 1315 was named patri- arch of Constantinople. Glykys’s grammar was composed for his son George (so the note at Cod.Laur. LV.7, fol. 416 [pp. XI– XII: poiÆsato d¢ tÚn lÒgon prÚw tÚn •autoË uflÒn, kÊrin Ge≈rgion); thus the several places where he introduces second- person formations, e.g. ka‹ oÏtv diå pãntvn eÍrÆseiw aÈt∞w tØn shmas¤an … tå parade¤gmata d¢ katvt°rv §ktey°nta saf°steron ı l°gomen poiÆsei. e‡ ti går ka‹ doke› pros¤stasyai, toÊtou soi tØn lÊsin parejÒmeya (p.9) taËta soi ka‹ per‹ t∞w tr¤thw ¶stv suzug¤aw toË mØ diafuge›n se xãrin mhdÉ §p‹ toÊtoiw tØn ékr¤beian … ‡syi …w ≤ dotikØ dÊnam¤n tina fid¤vw ¶xei (27). éllå går §peidØ tØn t«n =hmãtvn te ka‹ Ùnomãtvn sÊntajin katå tÚ dunatÒn soi dihryr≈samen … ka‹ per‹ taÊthw doËna¤ soi, …w dunatÒn, t∞w ÙryÒthtow toÁw tÊpouw (34–35). And he ends with kín m¢n oÔn ka‹ to›w êlloiw dÒj˙ ti ka¤rion ±nÊsyai, xãrien ín e‡h, efi soË khdÒmenow, to›w êlloiw eÍreye¤hmen tå m°tria §n toÊtƒ xarizÒmenoi, ka‹ oÈ mãthn oÈdÉ efik∞ tÚn noËn §j≈rvw §p‹ taËta tr°cantew: efi d¢ t∞w boulÆsevw époleify∞nai dÒjomen, éllå to›w ge dØn ufl°si parå t«n pat°rvn édelfå ka¤, ˜ph potÉ ín ¶xoi, poyhtå dika¤vw ín aÈto›w ka‹ e‡h ka‹ nom¤zoito (59–60). And if it seems to others that something worthwhile has been ac- complished, it would be pleasant if, in our care for you, we might be found to have gratified others with a modest contribu- tion to the field and not to have attended to it in vain, carelessly, 9 For what little is known of Gregory’s career, see R. Browning, “The Patriarchal School of Constantinople,” Byzantion 33 (1963) 11–40, at 19–20. 10 Citations are to the edition of Albert Jahn, Joannis Glycae opus De vera syntaxeos ratione (Bern 1849). THOMAS M. CONLEY 113 and irrelevantly. But if we seem to have fallen short of your wishes, yet it comes with the love of fathers for sons, and how- ever it turns out, it would be and be thought to be something that was rightly to be sought. Even when we bear in mind that this is addressed to his son, Glykys here goes well beyond the conventional boundaries of sentiment in treatises on correct grammar. Putting “Four parts” and “Eight parts” side by side with these epitomes of rhetoric and grammar suggests that we may be looking at a sort of sub-genre of instructional literature. They are in any event quite unlike the more discursive and de- tached works of Hermogenes, for instance, or even Doxapatres’ “Lectures on Aphthonios” (published by Walz at II 81–563) or the “catechetical” introductions to treatment of stasis issues such as those published by Rabe in his Prolegomenôn Syllogê (218– 228). The elementary nature of the material, the ample supply of examples, the distinct teacher-student quality of the instruc- tion, and the pronounced “addressed” quality of the language may be rhetorical indications that, like Psellos’s Rhetorica, they may all have been composed on special order. Of course, that invites speculation about who did the ordering and, in the cases of “Four parts” and “Eight parts,” who complied. I will address this question in due course. Another striking stylistic feature of “Four parts” and “Eight parts” is the frequent occurrence of rare, obscure, even mystify- ing diction. Some of these words can be traced back to Clas- sical or Late Classical usage: §ndiãskeuow and §gkatãskeuow at 576.22, for instance, are terms used by Hermogenes at Peri heureseôs 2.7 (122.15–124.15 Rabe) (and also by Eustathios pas- sim and by Tzetzes Chil. 11.270, for instance). Also attested in antiquity are, e.g., pro&dom°noiw (571.17), temãxion (573.13), xorhg¤a (574.16), kelarÊzei (579.28), bukãnh (580.5: “ox horn”), §festr¤w (582.14), and ékrvthriasmÒw (596.13). Others are attested in Byzantine sources: e.g. flstori«dew (571.10; see Tzetzes Chil. 4.781 et al.), §pistatik≈teron (573.6), xasmƒd¤a (573.29; frequently in Eustathios), parasÊla (575.25 and George Pachymeres Hist. 5.2, II 439.4 Failler), Ùrukt¤w (580.6, protective screen formed of interlocked shields; see Tzetzes Chil. 11.609), èlusidvtÒw (596.11, which, with y≈raj, seems to refer to chain mail; see Eustath. ad Il. 5.13), and ê#low (602.1; 114 RUMMAGING IN WALZ’S ATTIC Tzetzes Chil. 10.496 and frequently in patristic texts). Some seem to be hapax legomena: proasxole›n (571.25), perittolej¤a (562.2, instead of the more usual perittolog¤a, as at 583.9); purgose›stai (580.6: “towershakers,” devices used by sappers?); and traxulekte›n (580.18), traxulej¤ai (600.23), and traxu- lej¤a (601.5). Others are puzzling. What does bãptein (570.11, bãptonta at 571.23) mean in a rhetorical context (of exordia)? Or eÈktiko¤ (572.7), which in grammar designates the optative? And is simnot°raw at 594.29 a scribal slip? In an age that valued pure Attic diction so highly (and see, e.g., the optative prosxo¤hw at 575.5 and the duals in the Ajax narrative: xero›n 577.23, blefãroin 577.27—not to mention the excursus on Atticism at 583.6ff.), all of this is rather unusual. The excursus on Atticism is interesting for at least two reasons. First, as obscure as many of the words might be (see, e.g., the catalogue of things to be named in a description of a battle at 580.1–8), all but a few of them can be found in the Lexicon compiled by one “Zonaras” some time between about 1190 and 1253 (the year our earliest manuscript was copied). “Zonaras” of course draws mainly on old sources; but he cites Psellos 70 times, and has one citation each to Tzetzes and The- odore Prodromos. Second, the list of approved Attic diction at 585.12–586.18 seems to draw heavily from the De dialectis of Gregory of Corinth: e.g., Anon. 585.12–13 ≈ Gregory p.52; 585.15–16 ≈ 62–63; 585.22 ≈ 110; 585.23 ≈ 123–124; 585.25 ≈ 159; 585.28 ≈ 171–172; 586.2–3 ≈ 172; 586.10–11 ≈ 59–60; 586.12 ≈ 38–39; 586.15–16 ≈ 40.9 Perhaps such diction makes up the l°jiw t«n nevt°rvn that so many “among us”—as Anonymous puts it—are so enthusi- astic about. Who are these ne≈teroi? Chief among them seems to be ı sof≈tatow CellÒw (573.8, 24; mentioned also at 572.29), along with Ptochoprodromos and Nikolaos Kallikles among the poets; and, evidently, “Zonaras” and Gregory of Corinth—i.e., late eleventh- to early thirteenth-century writers. References to proper comportment during speeches made in praise of the emperor (571.23f., 600.16ff.), moreover, bring to 11 I owe this information to K. Alpers, Das attizistische Lexicon des Oros (New York/Berlin 1981) 23–36 and 129. THOMAS M. CONLEY 115 mind the orations delivered by the ma˝stvr t«n =htÒrvn, an office that did not exist before 1081.10 And the references to encomiastic praise by comparison at 599.23ff.—comparisons with David, Aaron, Moses, and Joshua—may recollect a speech in praise of Isaak II Angelos by John Kamateros in 1193, who does just that:3 o‰mai går … s¢ metå toË Dab‹d êndra katå tØn kard¤an aÈtoË. éllÉ eÔ poi«n ı lÒgow efiw mnÆmhn ≤negk° me Dab¤d (p.249.18–21); k¤nei prÚw pÒlemon tåw xe›raw, …w ı Mvs∞w, ka‹ otow ı m°gaw érxiereÁw ÉAar∆n taÊtaw ékamãntvw … g¤nou moi tå prÚw pÒlemon strathg«n, …w ÉIhsoËw toË Nau∞ (254.8–11). All of this supposes that Par.gr. 2918 is an apograph of an older manuscript, not an original compilation. But if we are on the right track, we may put the terminus post quem for our anon- ymous opuscula in the last decade of the twelfth century. If that is plausible, we should probably date the composition of the two treatises to the period of the Nicaean exile or shortly thereafter, i.e., between 1204 and about 1300. While it is true that the “elements” of each stylistic idea enumerated at the beginning of “Eight parts” and the doctrine of six peristatika so pervasive there go back to Hermogenes,11 it is Aphthonios and Menander who are explicitly mentioned as authorities (570.5, 572.23f.); and, as we saw, the progymnasmata sections of “Eight parts” rely heavily on Aphthonios. The influence of Menander is perhaps most evident in the emphasis in both pieces on encomiastic matters; and the second-person con- structions we discussed earlier might also suggest Menander as the model for whoever wrote the works under consideration. This in turn suggests that they were intended for a student (and not, as in much of the literature, for teachers of rhetoric) whose future mature rhetorical performances might take place in the 12 It is generally agreed that this office was created during the reign of Alexios I Komnenos (1081–1118). See A. Kazhdan, ODB II 1269. 13 V. Regel, Fontes rerum byzantinarum II (St. Peterburg 1917). 14 See Peri ideôn 1 (pp.218–220 Rabe) and Peri heureseôs 3.5 (140.16ff.). Of course, these had become so widespread in the literature that it is hard to say with confidence that our author(s?) actually had Hermogenes in mind. 116 RUMMAGING IN WALZ’S ATTIC setting of the court, although the reader is assured that many of the same principles will apply in deliberative or even forensic settings. We might also point out that the best manuscripts we have of Menander—Baroc. 131, Vat.gr. 306, and Par.gr. 2423 (which stops at p.390.30 Spengel, in the laliã chapter)—all date from the second half of the thirteenth century, evidence perhaps of the renewed interest in epideictic in a period when scholars were intent on recovering all they were able of Classi- cal, Patristic, and, indeed, Byzantine literature. What we have, then, may be dated to the Palaiologan Revival, a full genera- tion before Rhakendytes.12 If it is possible to date our opuscula to the Palaiologan Re- vival, the place of composition is clearly Constantinople, the only place where the works of the many authors recommended would be available—both to the author(s) and to the person to whom they seem to be addressed, the f¤le. Moreover, those works would not have been available to just anyone. Only someone of high social rank or someone not far removed could be expected to put his hands on texts of Demosthenes, Gregory of Nazianzus, George of Pisidia, Michael Psellos, and the rest. The list of recommended readings should also suggest that these works were probably not composed at the Nicaean court before 1261—nor, for that matter, in Thessalonica or Ephesus, where some rhetorical instruction was available—as there is every indication that books were in short supply at the Nicaean court during the 57 years of exile after the taking of Constan- tinople by the Latin crusaders. After the restoration, however, a great effort was made to recover the treasures of antiquity, the words of the Fathers, and the cultural glories of pre-1204 Byzantium.13 15 Might the quotation at 585.6–8 also point to this conclusion? Walz sees in éllÉ ≤me›w ge tÚn Pa¤ona ka‹ ÉIlluriÚn paraskeuasm°non §p‹ pÒlemon ka‹ tå toiaËta an allusion to Dem. 1.15, but this is far from what Demosthenes says. Who are ≤me›w? The Athenians certainly were not preparing for war with the Paionians and Illyrians; but in the 1260’s and 1270’s, the Palai- ologoi were. 16 See e.g. E. Fryde, The Early Palaeologan Renaissance (1261–c.1360) (Lei- den 2000); N. G. Wilson, Scholars of Byzantium (London 1996) 218–264. Of THOMAS M. CONLEY 117 If we are right to date the composition of the opuscula published by Walz to some time after 1261 and before Rhaken- dytes compiled his Encyclopedia (probably ca. 1320), then perhaps we are in a position to suggest some candidates for authorship. Of course, with what evidence we have, we shall never know with any certainty; but we should bear in mind that the selection of which works to preserve in manuscript and which manuscripts to copy was based largely on the reputation of the author. The anonymous works preserved in Par.gr. 2918 and the others we glanced at earlier must originally have been ascribed to recognized authorities. A number of names from our period come to mind. Manuel Holobolos, ma˝stvr t«n =htÒrvn (an office reestablished after 1261) from about 1265 until his death in 1284, is one such name; Nikephoros Choumnos (ca. 1250–1327) is another. There is no evidence that Holobolos wrote anything like the treatises we have been discussing, however. As for Choumnos, we know that he personally provided instruction in rhetoric to his daughter, Eirene-Eulogia Choumnaina.14 This might ex- plain both the tone of the two works and the use of f¤le that we noted earlier. But here again, there is no evidence to sup- port this possibility. There is another possibility that may have some support. As mentioned earlier, the treatise at fol. 174vff. of Laur. LV.7 is attributed to Gregory of Corinth. Could it in fact be attributed, along with “Eight parts” at fol. 331ff., to another Gregory, Gregory of Cyprus? There are several extant progymnasmata exercises composed by this Gregory, a noted teacher in the service of the emperor Michael VIII and, from 1283 to 1289, ___ 37 manuscripts held by G. Prato to have been produced during the Nicaean exile, only six are secular: three containing rhetorica (Hermogenes and Aphthonios in the main) and three the Lexicon of “Zonaras”: G. Prato, “La produzione libraria in area Greco-orientale nel periodo del regno latino di Costantinopoli (1204–1261),” Scrittura e civiltà 5 (1981) 105–147. 17 On Choumnaina see A. Hero, “Irene-Eulogia Choumnaina Palaiolo- gina, Abbess of the Convent of Philanthropos-Soter in Constantinople,” Byzantinische Forschungen 9 (1985) 119–147. 118 RUMMAGING IN WALZ’S ATTIC patriarch of Constantinople.15 There are also two encomia attributed to him, one to Michael VIII and one to Andronikos II.4 And it is this Gregory of whom Nikephoros Gregoras wrote several years later (using Gregory’s given name, George), ka‹ ∑n thnikaËta énØr §n lÒgoiw §p¤shmow t“ basikik“ sug- kateilegm°now klÆrƒ Ge≈rgiow ı §k KÊprou, ˘w tÚn §n ta›w grafa›w eÈgen∞ t∞w ÑEllãdow =uymÚn ka‹ tØn ÉAttik¤zousan gl«ssan §ke¤nhn, pãlai polÁn ≥dh xrÒnon lÆyhw krub°nta buyo›w (PG 148.308C). He was, that is to say, an outstanding rhetor in the imperial clergy who “rescued from the depths of the distant past” the noble fashion of the Greek language and the Atticizing tongue. As a teacher, Gregory could claim some other outstanding rhetors as his students, among them Nikephoros Choumnos, who distinguished himself both as a scholar and as a high palace official; possibly Konstantinos Akropolites (d. 1324) and Maximus Planudes (1255–1305), the latter one of the greatest scholars of Byzantium; and, as we noted earlier, John Glykys.16 As we also noted, Glykys shared with his teacher an intense interest in Attic purity; and the style of his “On correct syntax” shares many features with that of our anonymous opuscula. It is tempting, then, to see Glykys looking to Gregory as a model in writing a treatise on Attic syntax for his son, and to young Glykys as the f¤le to whom the opuscula seem to be addressed. There is another possibility that is perhaps more intriguing. Yet another pupil of Gregory’s was the protovestiarissa Theodora Rhaoulaina (ca. 1242–1300), the daughter of John Kanta- kouzenos and Eirene Palaiologina who became an important patron of letters and an accomplished writer in her own right.17 18 See e.g. the chreia published by J. F. Boissonade, Anecdota graeca (Paris 1829–30) II 269–273; and his speech in praise of the sea at PG 142.433– 444. 19 Boissonade, Anecdota graeca I 313–358, 359–393. 20 On all these figures the best comprehensive survey is still C. Con- stantinides, Higher Education in Byzantium in the Thirteenth and Early Fourteenth Centuries (1204–ca. 1310) (Nicosia 1982). 21 A useful profile of this extraordinary woman can be found in D. M. Nicol, The Byzantine Lady: Ten Portraits (Cambridge 1994) 33–47; see also A.- THOMAS M. CONLEY 119 Among her other accomplishments, she founded the monastery of St. Andrew in Constantinople and renovated a small build- ing nearby where she provided a residence for the aged Greg- ory and his extensive library after his resignation as patriarch in 1289. Theodora and Gregory were long-time correspondents, as is shown by the inclusion of more than a dozen letters to her in a collection of Gregory’s letters originally copied by none other than John Glykys, which has still not been properly edited.18 The letters show that the relationship between the two was a warm one, as we find Gregory addressing her in one letter as t°knon §mÚn sÁ ka‹ monogen¢w t°knon … §g∆ d¢ patØr toiòde (letter 194, p.597 Kugéas); and in another as yÊgater §mÆ (letter 212, p.600). So it would not surprise us if he were to address her as f¤le in an introduction to rhetoric. Terms of endearment do not of course bring us very close to a plausible identification of author and addressee. Let us turn from Gregory, then, and ask, Is there anything in Theodora’s writing that suggest a connection with the opuscula in Walz? We might begin by looking at two manuscripts, one actually copied by her, Vat.gr. 1899, containing the orations of Aelius Aristeides; and the other produced under her supervision con- taining Simplicius’ commentary on Aristotle’s Physics, no. 3649 in the Historical Museum in Moscow.19 ___ M. Talbot, “Bluestocking Nuns: Intellectual Life in the Convents of Late Byzantium,” in C. Mango and O. Pritsak (eds.), Okeanos: Essays Presented to Ihor Sevcenko (Cambridge [Mass.] 1983) 604–618. 22 S. Eustratiades published many of them, on the basis of two manu- scripts in Vienna: Ekklesiastikos Pharos 1 (1908), 2 (1908), 3 (1909, 4 (1909), 5 (1910). A fuller conspectus of the sources can be found in W. Lameere, La tradition manuscrite de la correspondence de Grégoire de Chypre (Brussels 1937). Excerpts from his letters to Theodora were published and commented on by S. Kugéas, “Zur Geschichte der Münchener Thukydideshandschrift Augustanus F,” BZ 16 (1907) 592–603. On the collection copied by Glykys see S. Kourousis, “ÑO lÒgiow ofikoumenikÚw patriãrxhw ÉIvãnnhw IG ı GlukÊw,” EpetByz 41 (1974) 309–311. 23 For the former, see A. Turyn, Codices Graeci Vaticani saeculis XIII et XIV scripti (Vatican City 1964) 63–65; for the latter, B. Fonkic, “Zametki o gre- 120 RUMMAGING IN WALZ’S ATTIC The first leaf of Vat.gr. 1899 bears the following inscription composed by Theodora: Ka‹ tØn ÉAriste¤dou d¢ tÆnde tØn b¤blon grafe›san ‡syi parå t∞w Yeod≈raw kal«w efiw êkron gnhs¤vw §skemm°nhn ÑR≈mhw n°aw ênaktow édelf∞w t°kow Kantakouzhn∞w §j énãktvn ÉAgg°lvn Douk«n fue¤shw PalaiolÒgvn fÊtlhw ÑRaoÁl dãmartow DoÊka xaritvnÊmou KomnhnofuoËw prvtobestiar¤ou. And at the beginning of the Moscow manuscript, we read another inscription by Theodora: Ka‹ d°lton aÈtØn toË sofoË Sumplik¤ou tØn t«n fusik«n §kdidãskousa b¤blon ÉAristot°louw o sofoË polÁw lÒgow kalligrafe› te ka‹ m°thsi suntÒnv édelfidØ kratoËntow ≤ yeoË dÒron Douk«n Komnhn«n PalaiolÒgvn fÊsa dãmar ¶xontow toË ÑRaoÁl kl¤sin xãri trÒpoiw ér¤stou ka‹ krat¤stou t“ g°nei. While it is true that these iambs hardly qualify as poetry, the absence of enjambment (more pronounced in the first than in the second) brings to mind the éretØ st¤xvn pr≈th laid down in the passage on iambics that Rhakendytes evidently lifted from “Four parts” (III 560.19; and see Walz’s n.11 on p.573). The absence of enjambment in Theodora’s iambs is, of course, no evidence that she had “Four parts” in mind when she composed them; so we had better look elsewhere. Much more substantial evidence of Theodora’s rhetorical skill can be found in her Life of saints Theodore and Theophanes, victims of persecution and torture in the ninth century for their refusal to comply with the iconoclasm mandated by the emperors Leo the Armenian and Theophilos.20 This is a curious and complex ___ ceskich rukopisjach Sovietskich chralinisc,” Vizantjskij Vremmenik 36 (1974) 134–138, with a plate at 137. 24 Ed. A. Papadopoulos-Kerameus, ÉAnãlekta ÑIerosumitik∞w Staxuo- log¤aw IV (Jerusalem 1897) 185–223. There is a brief account of the career THOMAS M. CONLEY 121 work that has attracted little attention from scholars beyond oc- casional notice of its implications for ecclesiastical controversies in which Theodora and Gregory were involved between about 1275 and 1289, having mainly to do with the status of the patriarch Arsenios. This is not the place to go into those con- troversies, nor to provide a detailed analysis of the Life com- posed by Theodora. But there are some rhetorical features in it that hint at the extent of Gregory’s influence on Theodora and that are in line with principles laid down in “Four parts” and “Eight parts.” Like much Byzantine literature, Theodora’s Life is packed with literary allusions, Classical and, particularly, Scriptural, as in passages that are virtual centos of passages from the prophets (at e.g. pp.200.4–27, 202.28–203.25, 215.29–216.17). Theodora also works into her account a variety of familiar adages, or paroimia; and it is interesting that of eight instances, almost all can be found in the collection of proverbs compiled by her master, Gregory of Cyprus (most accessible in PG 142.445–470): e.g., Life 189.1f. jurÚw efiw ékÒnhn and dalÚw prÚw tÚ pËr = PG 142.464A, 465A; 199.3 pãnta kãlvn §k¤nei = 465B; 203.30 prÚw k°ntra lakt¤zein = 465A; 204.14 boËn §p‹ gl≈tthw = 453B. Theodora’s diction, furthermore, is scrupulously Attic, re- flecting perhaps the training she received from Gregory and reminiscent of the importance of attikismos in “Four parts” (e.g. III 583.6ff.). Aside from a few instances of Patristic or Byzan- tine items (e.g., sÊgkellow at 194.7, of Michael Synkellos), the diction of the Life is almost obsessively Attic—to the point, one might add, of obscurity. Theodora’s lexicon was enormous, and full of rare usages of the sort that Hesychios and Eustathios labored to explain, e.g. skai≈rhma at 198.25 or katalimpãnvn at 209.12. One unusual expression occurs at 205.6f. in refer- ence to Theodore and Theophanes, oÈdÉ efiw noËn bãcantew, which seems to mean something like “not immersing in con- templation.” This is not an expression one finds attested—at least, not in the standard references—in either Classical or ___ of Theophanes by D. Turner in Encyclopedia of Greece and the Hellenic Tradition, ed. G. Speake (London 2000) 1630–1632. 122 RUMMAGING IN WALZ’S ATTIC post-Classical Greek, Attic or otherwise. As we noted earlier, however, there is something similar in the opening section of “Four parts”: at 570.10ff. we read éretØ d¢ prooim¤ou tÚ efiw tØn prokeim°nhn ÈpÒyesin bãptein, and at 571.23 another reference to tØn ÍpÒyesin bãptonta. In the end, of course, it is hard to say what to make of this. A stronger argument could be made if one could show, for instance, that Theodora follows the advice in “Eight parts” to organize one’s narrative or descriptions around the six peristatika (III 592.14ff., 602.2–25). The only indication that Theodora had any such thing in mind, however, is when she explicitly brings up the heading of a‡tion at 201.2. But in general, looking for the peristatika pattern in the Life would not advance an argument linking Theodora and “Eight parts,” in any event. In the first place, on the one hand, “Eight parts” is a set of rather elementary lessons; and Theodora, by the time she composed the Life, was certainly sophisticated enough not to need such mechanical rules. And on the other hand, it seems obvious that any reasonably rounded account of a saint’s life would address the “who, what, where, when, how, and why” topics at some level; and so finding them in play, as it were, would prove very little, if anything. Our speculations have, at this point, carried us well beyond the confines of Rhetores Graeci; and we have found ourselves rummaging in attics other than Walz’s. The results have not been very satisfying, either. But if identifying the author and audience of “Four parts” and “Eight parts” is not possible, after all, at least we have been able to situate those short pieces in a literary/rhetorical context and suggest a plausible time-frame for their composition. Whoever composed them probably did have a particular recipient in mind (as opposed to a fictional “you”), someone in the early stages of rhetorical education, but already well-versed in the works that made up the canon looked to by Byzantine students of literature. In short, there is more to them than meets the proverbial eye. August, 2005 244 Lincoln Hall University of Illinois 702 S. Wright St. Urbana, IL 61801 THOMAS M. CONLEY 123 t-conley@uiuc.edu