Printer Friendly

Another chapter in the history of scholia.

The ancient law school about which we have the most information was at Beirut.(1) Editors of legal papyri have occasionally speculated about possible connections between particular ancient texts and the activities of professors of law in that city, but no one has examined the evidence in a body. It is likely, I think, that legal papyri reflect the state of contemporary legal education at Beirut, and that they preserve, moreover, primary evidence for the history of scholarship in general. With so broad a topic, it would be useful first to consider what we know about legal instruction in antiquity, particularly at Beirut, then to review the relevant papyrological evidence, and finally to draw the two subjects together by considering the possible relation between the marginal commentaries of legal papyri and the extensive scholia which fill the broad margins of many medieval manuscripts.


From perhaps the late second century, there slowly developed at Beirut a major school of Roman law. Although the date of its founding is unknown, it seems likely that legal instruction had been available there for a century or more by the time Constantinople's law school was founded in 330.(2) Even from early times Beirut's reputation as a centre for legal studies attracted an international clientele. Fifty-one students who went there to study law are known to us by name. They represent twenty provinces, from Iberia to Armenia and from Europe to Egypt.(3) Relatively few are local. In fact, the earliest reference to legal instruction in the city comes from a native of Cappadocia, Gregory Thaumaturgus. He relates in his Panegyric to Origen of 238 how he prepared for a career in law by taking extensive instruction in Latin and Roman law before leaving for Beirut, a city which he calls [GREEK TEXT NOT REPRODUCIBLE IN ASCII] ... [GREEK TEXT NOT REPRODUCIBLE IN ASCII].(4) Given this `rather Roman' character, Beirut was evidently the best of Eastern cities for solidifying one's knowledge of the law. In the fourth century Libanius, teaching rhetoric in Antioch, sent students to study law there. A letter of 361 describes the departure of his student Hilarinus of Euboea for Beirut, `the mother of the laws': [GREEK TEXT NOT REPRODUCIBLE IN ASCII]. In other correspondence his references to Beirut are similarly honorific: [GREEK TEXT NOT REPRODUCIBLE IN ASCII].(5) The so-called Expositio totius mundi et gentium, composed perhaps in the mid-fourth century with interpolations of the fifth, emphasizes in a single sentence Beirut's beauty and its universal renown as a centre for legal studies:

civitas valde deliciosa et auditoria legum habens, per quam omnia iudicia Romanorum <stare videntur>. [inde enim viri docti in omnem orbem terrarum adsedent iudieibus et scientes leges custodiunt provincias, quibus mittuntur legum ordinationes.](6)

In the fifth century, perhaps the time of the school's greatest renown, the Egyptian epic poet Nonnus celebrated the unshakable wall of laws by which Beirut, and only Beirut, fortified the cities of the world (Dionys. 41.395--98):


Early in the next century, professors from Beirut participated in the development of the Digest. At its publication in 533, Justinian reaffirmed and enhanced the standing of legal studies in that city through special dispensations and lavish praise. He referred to Beirut as the legum nutricem, and stipulated that copies of his new codification should be available only to students there and in the two capitals, Rome and Constantinople.(7) The inclusion of the Western capital reaffirms the specifically Latin basis of instruction in law which Justinian was concerned to preserve and the intended range of influence of the new compilation, He explicitly prohibited the teaching of law at Alexandria or Caesarea, however, although it had been taught there formerly, and he reinforced the prohibition with threats of heavy fines and expulsion.(8)

Justinian also prohibited the writing of commentaries on the Digest: a puzzling ban at first sight, which to all appearances was honoured mainly in the breach.(9) Law always needs explaining, and legal scholars of the sixth century in fact produced abundant commentaries on the Digest and Code, in which they lavished honorifics on their fifth-century predecessors at Beirut. (Some of these later commentators were probably professors at Beirut themselves, most notably the antecessor Thalelaeus, whose praise for the earlier masters is the most generous.(10)) The work of these later scholars is preserved mainly in the Basilica, the massive tenth-century summary, with scholia, of Justinian's codification. The Basilica transmit reliably only the names of five professors of Beirut's golden age of legal studies--Cyril, Domninus, Demosthenes, Eudoxius, and Patricius--but they describe these five in language that is striking even by Byzantine standards.(11) They are, collectively ([GREEK WORD NOT REPRODUCIBLE IN ASCII]), teachers of exceptional distinction ([GREEK TEXT NOT REPRODUCIBLE IN ASCII] ), and, in a phrase to which we shall return in Part III, professors of universal compass ([GREEK TEXT NOT REPRODUCIBLE IN ASCII]). The period of activity for these five is for all practical purposes coextensive with the fifth century.(12) Interestingly, there is no corresponding record of such splendid scholarly activity in contemporary Constantinople, although that city's law school was shortly to be forced into greater prominence. For in 551, not long after this golden age, the history of legal studies at Beirut ended in a destructive earthquake.

Two changes in legal instruction in the Eastern empire took place in or just prior to the fifth century, the era of these [GREEK TEXT NOT REPRODUCIBLE IN ASCII]. Whether the changes first arose in Beirut or elsewhere is unknown.(13) Certainly, however, they determined the nature of the activity of Beirut professors of the fifth century. The first change presumably reflects political realities in the Mediterranean world of the late fourth century. Since the time of Constantine, the Western empire had become progressively weaker politically and socially, and the East more intellectually vibrant. In 395, when Theodosius divided the western and eastern portions of the empire between his two sons, the practical split was confirmed. Attendant upon this division, and probably related to it, Greek replaced Latin at some time in the late fourth or early fifth century as the language of instruction in Eastern law schools.

That Latin had been the language of legal instruction at Beirut in earlier centuries is a natural inference. The language of Roman law is Latin. At Beirut itself, the administrative and social superstructure of the city was emphatically Roman from its founding, by Agrippa in 15 B.C., as a military colony.(14) It enjoyed the ius Italicum and was the home of two legions and of the fleet that policed the eastern Mediterranean. As Fergus Millar has observed, `Of all the Roman coloniae of [the Near East], only one, Berytus,... represented a substantial island of Romanisation, of Latin language and culture and of Roman law, which was to last into the late Empire.'(15) This remained true even as Greek reasserted itself in other Roman coloniae in the Greek East: Latin was the language of a public inscription at Beirut as late as A.D. 344.(16) The city was a [GREEK TEXT NOT REPRODUCIBLE IN ASCII] indeed, and ancient sources confirm this. Gregory Thaumaturgus in 238 speaks of Latin as the language not only of the law but also of secondary literature on the law.(17) In the fourth century Libanius at Antioch repeatedly and ruefully associates the decline of Greek rhetoric with the rage for legal education, particularly at Beirut, and the burgeoning of Latin studies.(18) Liebeschuetz has noted, for the fourth century, `a steady increase in the popularity of Latin and legal studies ... caused by the government's policy of giving preference, when making appointments to the civil service, to applicants who had knowledge of Latin and Roman Law'.(19) By the first decades of the fifth century, however, circumstances had changed. By the sixth, Justinian found it convenient to publish his codification with a preface in both Latin and Greek versions (Tanta-/[GREEK WORD NOT REPRODUCIBLE IN ASCII]), and most of the laws he enacted from 535 on were in Greek. Also significant are surviving fifth- and sixth-century legal texts demonstrating, as we shall see in Part II, that contemporary legal commentaries in the eastern Mediterranean were indeed in Greek.(20)

The second innovation in fifth-century legal education in the East arose from the first, and was pedagogical in nature. Instructional methods changed in order to accommodate ever-increasing numbers of Latinless students of Roman law.(21) These students now began the study of any new text of law by reading an [GREEK WORD NOT REPRODUCIBLE IN ASCII] (index)--a translation of the text into Greek. In the Greek East, these indices were, unavoidably, essential preliminary reading for students of Roman law.(22) After working through the index, when the student finally had in hand the substance of a text, the task was to understand it properly in context, [GREEK WORD NOT REPRODUCIBLE IN ASCII], which began to be produced in the fourth and fifth centuries, served this need. These commentaries follow the model of the commentaries of traditional Greek schooling: they provide glosses of unfamiliar words, give notice of the contents of a passage, highlight parallels in other legal sources and provide cross-references. Discussion follows the text systematically. Lemmata from the Latin source guide the eye.

In Eastern law schools, then, index and commentary became the preferred hybrid vehicle for legal instruction in a Latinless context, largely replacing dogmatic interpretatio and instruction based on case studies. Secondary evidence for the rise of the new pedagogy comes largely from references in the Basilica by sixth-century scholars who cite the work of fifth-century scholars at Beirut. They make it clear that the production of [GREEK WORD NOT REPRODUCIBLE IN ASCII] was a major activity of their predecessors.(23) Whether those Beirut professors were in fact the first to produce indices and commentaries is not clear. But if the new approach to legal instruction originated elsewhere, it must have reached Beirut very quickly.

An extensive contemporary example of the kind of commentary that concerns us happily survived to modern times. This is the celebrated pre-Justinian remnant of a commentary on Ulpian known as the Scholia Sinaitica. It is named after its findspot, where it was put to use in the binding of a book.(24) Internal evidence makes clear that the text served educational needs, perhaps at Beirut, perhaps elsewhere. Its likeliest sources are Beirut or Alexandria, since these are the places closest to Sinai where we know that law was taught. But wherever it was composed, we may be sure, on the basis of what we know about Beirut's great teachers from the Basilica, that it is typical of the commentaries they produced.(25)

The Scholia Sinaitica are not, actually, a simple commentary. Rather, they are a chain of excerpts from two separate commentaries on Ulpian's Libri ad Sabinum. The resulting amalgamation consists of brief comments on salient words and phrases with frequent and specific references to legal authorities. The dominant language of discussion is Greek, although introductory lemmata in Latin are common and Latin technical terms appear. There is an oral quality to the text, as if it were the lecture notes of a teacher: the second person addressed to a student is used freely, the first person less so (see [sections] 34, below). While the commentator used essentially the same systematic approach as the composers of ancient hypomnemata on literature, there is a pedagogic immediacy in his work that is lacking there. The `learn this', `skip that', and `important' leave no doubt that these commentaries served an educational purpose. In the sixth century, the same markers distinguish works used in the classroom, such as the Paraphrasis Institutionum of Theophilus, the commentary on the Digest by Stephanus, and that on the Code by Thalelaeus.(26) In the following excerpts from the earlier Scholia Sinaitica, one also notes the repeated use of the didactic [GREEK WORD NOT REPRODUCIBLE IN ASCII] `mark this'--a term not found in earlier literary exegesis:

`Scholia Sinaitica' Ad Ulpiani Libros ad Sabinum:







Bilingual commentaries on this model seem a natural development in the context of legal education in the East in late antiquity. A change so radical as a change in the language of instruction from Latin to Greek--whether it came by decree or by default--would have required the production of a new battery of teaching materials by which teachers could help students find their way from the Latin text of Roman law to the Greek of oral instruction. Commentaries like those preserved in the pre-Justinian Scholia Sinaitica would have answered that need. As Paul Collinet observed, one cause of the fame of Beirut's `universal teachers' may have been that they supplied this need and thus negotiated a smooth transition from instruction in Latin to instruction in Greek. The approach was brought to perfection by their successors, the antecessores of the age of Justinian.

Now, commentary like that found in the Scholia Sinaitica has interesting codicological implications. If students were to follow instructions as precise as those given in [sections] 34 above, they and their instructors alike would have had access to reference texts which corresponded exactly, in their organization and internal demarcations, to the rubrics indicated in such a commentary and to each other. Moreover, since instructors presumably continued to use the same texts and commentaries with successive groups of students, the organization of their books must have remained effectively unchanged as they were recopied in the course of time. This will have called for coordination between scribe and whoever commissioned the copying--particularly if multiple copies were made for multiple students. A passage of Libanius suggests that professors of law at Beirut may have been involved in securing law books for their students. At the very least, of course, they will have concerned themselves with the nature of the texts their students read. In a letter to Hierius, father of Peregrinus, a student of law at Beirut in A.D. 365, Libanius cautions Hierius to pay the instructor the fee he requests, for he is a good teacher. Although his charges may seem high, says Libanius, they are in fact too low if one considers the `quantity of books and how great are the things he procures'.(27) Libanius offers no details about the format of the books to which his students gain access. Papyri, I believe, do.


About two dozen codices of Greek and Latin literature and of Roman law of the fourth, fifth, and sixth centuries display a striking new format.(28) It has three distinctive elements. First, the margins are wide, and sometimes extraordinarily wide in comparison with normal codices. They can range from 6 cm in breadth to well over 10 cm. Second, the marginal annotations of these books tend to be longer than those in conventional codices. Although margins are not actually filled with commentary, nevertheless the length and frequency of the annotations are something new in the Greek East. I consider a marginal note 'long' if it is at least eight words long; notes are frequent if three or more have been added to a passage of at least fifteen lines. These are arbitrary standards and, by comparison with the length and frequency of some scholia, hardly rigorous. Yet in earlier papyri, marginalia that meet even these standards are exceptional. Third, marginalia appear to be the work of professional scribes, sometimes of the same scribe who copied the main text.(29) Informal hands are rare.(30) Notes look planned, not added haphazardly, as in books of traditional formats. They tend to be written in neat blocks, in lines of equal length.(31) Evidence occasionally survives to indicate that they were planned from the moment the book was designed. In the corner of a page of Aratus, for example, a small cross-mark in the upper left-hand corner delimits the space for marginalia.(32) Orderliness was not absolute--annotators did not always trouble themselves to marshal comments precisely below one another--but neat execution is typical of individual notes. Even in inner margins, the marginalia are often neatly copied, with regular margins of their own to separate them from the main text and from the edge of the page. These secondary margins can be as wide as regular margins in a codex of conventional size.(33) The precision of such notes would have been impossible unless they were copied before the codices were assembled. Otherwise, the tightness of the binding and the curve of the writing surface would have introduced irregularities of the kind observable in notes added after binding. We must conclude that they were included by design. These books were ordered from professional scribes who were instructed to make the margins extraordinarily wide and to copy into them prescribed material.

Before the fourth century, nothing of the sort is known. The usual vehicle for literary texts, through the third century, was the roll. It rarely contained notes. Among the surviving Greek and Latin literary texts from Egypt, of which the majority are bookrolls, only a small proportion of rolls have marginal or interlinear comments. Comments that do appear are usually sporadic and brief. Even when they are lengthy, they appear to have been added by readers, not by the professional scribes who in most cases copied the texts.(34) In the fourth century the codex replaced the roll as the standard form of book and for the most part practices in annotation and layout remained the same.(35) Most literary codices--the well preserved Cairo Menander illustrates the norm--have margins too narrow to carry much commentary.(36) They were not intended to: systematic commentary which covered the text line by line continued to circulate, as in the age of the roll, in separate books. With the exception of the unusual codices with which we are concerned here, marginalia were still intermittent.(37)

The most impressive examples of new-format codices containing Greek literature are three texts of Aratus, Pindar, and Callimachus in which the margins are extraordinarily broad and lengthy notes have been written carefully by the primary scribe.(38) Dating of the bookhands of late antiquity is very difficult, of course.(39) The initial editor of the Aratus classified it simply as `Byzantine', but the other two codices are set fairly late by their editors: the Pindar is assigned to the sixth century and the Callimachus, POxy 2258, to the sixth/seventh century. This is the best preserved and the most celebrated. Because of the density of its marginalia and its huge format, it was recognized from the time of its publication as the earliest example of a literary manuscript with scholia of the standard medieval type. Certainly, along with the Aratus and the Pindar, it is important as evidence in the hotly contested history of scholia.

It has escaped attention, however, that several legal papyri of new format antedate the celebrated Callimachus, the Pindar, and also, perhaps, the Aratus: perhaps seven are dated by editors to the fourth or fourth/fifth centuries, and about the same number to the fifth or fifth/sixth centuries. `New-format' codices, moreover, are proportionately more common among legal papyri than among literary. Books with margins of 6 cm or more account for about half of the twenty-one surviving annotated codices of Roman law, but only about 15% of annotated literary codices, Even among law books with narrower or broken margins, a good half dozen contain neatly written annotations of unconventional length or frequency. The accompanying table (Table 1) of annotated legal papryi presents the evidence. While none of these annotated texts of law appears to have been so copiously annotated as the Callimachus, the legal codices themselves, and also their notes, have the same orderly appearance and the same large format and broad margins. Annotation in informal hands is not entirely absent, but informal notes, for the first time in the history of annotation, are a rarity.(40) Indeed, the combination of distinctive format and script in these texts of law attracted, long ago, the attention of palaeographers, who recognized in them a new form of book emanating--in the words often repeated by E. A. Lowe in his descriptions of these codices--from the scriptorium of `a major eastern school of law': not, that is, from the literary sphere.(41) Perhaps, then, we should look for the beginnings of scholiastic commentary in the context of legal instruction.
TABLE 1. Annotated codices of Roman law from Egypt

Century     Dated        Contents            Publication (with
or date     by (if                             plate or photo)

4th                       Ulpian Ad          PAnt I 22 (CLA
                          Edictum?           Suppl. 1707; photo)

4th          EGT          Ulpian Lib.        PSIXIV 1449
                          XXXII Ad           (CLA Suppl. 1697;
                          Edictum            photo)

a.d. 400     RS           Actio              P.Berol.inv. 16976 +
                          condicticia and    16977 (Seider ii.2.20;
                          Longi temporis     CLA Suppl. 1783)

a.d. 400     RS           Rubrics on         P.Vindob. Lat. 110
                          criminal law       ined. (Seider ii.2.38;
                                             CLA x. 1538)

4th/5th                   Imperial           PRyl III 476
                          constitutiones     (CLA ii.225; photo)

4th/5th                   Legal              PSI XIII 1348:
                           definitions       supplements to text,
                                             not marginalia
                                             (plate; photo)

4th/5th      EGT          Papinian De        P.Berol. inv.
                          Bonorum            6762+6763, P. Louvre
                          Possessione        inv. E7153 (CLA
                                             Suppl. & viii. 1037)

5th          RS           Ulpian Ad          PRyl III 474 (Seider
                          Edictum            ii.2.16; CLA Suppl.
                                             1722; photo)

5th                       Juridical          PRyl III 475 (photo)

5th          RS           Law                PAnt III 153 (Seider
                                             ii.2.19; CLA Suppl.
                                             1789; photo)

5th                       Juridical          MPER N.S. III 38
                          fragment           (photo)

5th                       Law                P. Vindob. inv. L 26
                                             ined. (CLA x. 1524)

A.D.         RS           Law                PAnt III 152
500                                          (Seider ii.2.39;
                                             CLA Suppl. 1711;

A.D.         RS           Gaius              PSI XI 1182
500                       Institutiones      (plate; Seider
                                             ii.2.28; iii.292;

5th/6th                   Legal              RBerol. inv.
                          catechism          11866a, Aegyptus

6th          RS           Digest             P.Heid. Lat. 4,
                                             olim 1272
                                             (Seider ii.2.40;
                                             CLA viii. 1221)

6th                       Codex              PSI XIII 1347
                                             (CLA iii.293;

6th                       Stephanus,         PSI I 55 (photo)
                          Index to Digest

6th                       Digest             P. Reinach 2173
                                             (Seider ii.2.18)

6th                       Law                P.Berol. inv.
                                             6759+6761 ined.
                                             (CLA viii. 1035)

6th                       Law                P.Berol. inv. 6758
                                             ined. (CLA

Century         Catalogue              Width of      Long or
or date         no. ([Pack.sup.2])     broadest      frequent
                                       margin or     notes([dagger])
                                       margin] (in

4th                  2979                2.9, top        (*)

4th                  2960                [1.8],          (*)

a.d. 400             2281                [6.1],          (*)

a.d. 400             2984                 9, side        (*)

4th/5th              2282                 [6.8?],         ?

4th/5th              2982                 7.1,           (*)

4th/5th              2955                 2.5,           (*)

5th                  2974                 5.2, top

5th                  2280                 5.2, side      (*)

5th                  2979.20              6.2, side      (*)

5th                  2286                 [4], outer     (*)

5th                  2993a                11.5?,         (*)

A.D.                 2979.10              10, side       (*)
500                  (Mertens)

A.D.                 2953                 6, outer       (*)

5th/6th              2277                 7, outer

6th                  2966                 8.4, outer     (*)

6th                  2970                 6.6, lower

6th                  2965                 4, lower       (*)

6th                  2971                 [3.7], top     (*)

6th                  2993                 2, outer

6th                  2992                 [4.3], side    (*)

(*) RS: Richard Seider, Palaographie der lateinischen Papyri (Stuttgart, 1981).

EGT: E.G. Turner, The Typology of the Early Codex (Philadelphia, 1976).

([dagger]) `Long:' 8 or more words; `Frequent:' 3 or more per 15 lines of text (see above, p. 276).

About half of the annotated codices of Roman law from Egypt are copies of works that figured in the curriculum of legal studies at Beirut.(42) Works read in the early years, mainly texts of Ulpian and the Institutes, predominate. Longer notes are of two kinds. Some are excerpts from commentaries and explain points of text, characteristically with specific cross-references far more precise than in any literary marginalia of earlier date--[GREEK WORD NOT REPRODUCIBLE IN ASCII] in the literal sense (see, for example, no. 1 below and, on a more modest scale, no. 2). Others, serving the function of indices, translate into Greek the adjacent Latin text (e.g. nos. 3 and, perhaps, 4). Short notes mark useful passages (no. 5), supply Greek glosses for Latin words, and provide subject headings for adjacent text (nos. 6 and 7). The latter correspond to the brief titulorum suptilitatem ([GREEK WORD NOT REPRODUCIBLE IN ASCII]) that Justinian condoned for manuscripts of the Digest.(43) As in the Scholia Sinaitica, the marginalia in legal papyri are almost exclusively in Greek, although Latin lemmata and technical language may be mixed in. Sometimes Latin terms are transliterated as Greek or given Greek inflections; sometimes Greek is written in Roman letters (no. 1: [GREEK TEXT NOT REPRODUCIBLE IN ASCII], [GREEK TEXT NOT REPRODUCIBLE IN ASCII]; no. 4: [GREEK WORD NOT REPRODUCIBLE IN ASCII]). Only one legal text (no. 6 below) contains exclusively Latin marginalia. Across the period, notes make use of the first and second persons, as in the Scholia Sinaitica, and feature the didactic and hortatory [GREEK WORD NOT REPRODUCIBLE IN ASCII] and [GREEK WORD NOT REPRODUCIBLE IN ASCII] (nos. 3, 5). The generally formal script of most additions in these legal codices suggests that they were professionally copied, although students themselves, like readers for centuries before them, must have occasionally added notes to their own books.(44) The selection of the professionally copied additions was perhaps the work of teachers who, as we may infer from Libanius, had a natural interest in their students' access to the texts they needed. Teachers will have seen the advantage of books fitted out with explanations so that a student would not miss important parts and would know where to find supporting documentation. A sampling of these new-format marginalia from legal codices gives a sense of the nature of the material:

1. P. Berol. inv. 16976, Fragment on Longi Temporis Praescriptio (?), suppl. R. Seider, PalLat ii.2.20; [P.sup.2] 2281, A.D. 400.

Text lost; marginal note:


2. PSI XIV 1449, Ulpian, Libri 32 ad Edictum; [P.sup.2]2960, fourth-fifth century.



Marginal note:


3. PSIXI 1182, Gaius, Inst. 3.168; [P.sup.2] 2953, A.D. 500. Text:

tollitur autem obligatio praecipue

solutione eius quod

debeatur. unde quaeritur

si quis consentiente

creditore aliud pro alio soluerit,

utrum ipso iure

liberetur, quod nostris

praeceptoribus placuit, an

ipso iure maneat obligatus:

Marginal note:


4. PAnt III 152, Geek-Latin Work on the Law of Dowry; [P.sup.3] 2971.1 (Mertens), A.D. 500


]retentio <scil. ob res donatas>:

Marginal note (fol. a, right marg.):


5. PRyl III 475, juridical fragments; [P.sup.2] 2280, fifth century

Text fragmentary; marginal note:


6. PVindob L 110, on Roman criminal law, [P.sup.2] 2984, A.D. 400.

Text lost. Marginal note:


catenatus esse debet

non tamen

ut in carcere agat

nisi suspecta

sit persona.

7. PSI 55, Index to Dig. (Basilica 11.1.7 [1.569]), [P.sup.2] 2965, sixth century.





It seems likely that the remarkable format of the books in which these notes appear is a material reflection of the revolution in legal studies in the fifth century and is associated with [GREEK TEXT NOT REPRODUCIBLE IN ASCII] at Beirut.(45) How books of this new style found themselves in Egypt is an interesting question, not easily solved. They could have been acquired in Beirut by students who travelled there from Egypt and took them home when they returned after their studies. Constantinople is another possible source, although less likely: Beirut had greater eclat, in antiquity, as a centre for legal studies, a longer tradition of instruction in the law, was closer to Egypt, and in fact received' students who travelled there from Egyptian cities to study law.(46) Alexandria is a possibility too, for it was also home to a school of Roman law, though its influence is open to question. In Egypt the reception of Roman law seems to have been minimal. Normal legal business was conducted by local law until the fifth or sixth century. Furthermore, Alexandria is ruled out, at least in theory, as a source of texts copied after 533--the year in which Justinian imposed sanctions on legal instruction outside of Beirut and the two capitals.(47) At best we may conclude that the fourth and fifth centuries were a period of monumental change in legal instruction in the East; that Beirut seems to have been at the forefront of the revolution; and that these codices, found in Egypt, reflect the innovations in the form and content of legal textbooks that followed upon that revolution.


The significance of these codices reaches, I think, beyond the ancient school of law, for their annotations bear a close resemblance to the scholia in medieval manuscripts of classical authors--a resemblance closer than anything known from earlier evidence. The selection of the notes, like that of scholia, was clearly premeditated, their execution similarly professional and orderly. It also seems possible that some of these marginalia were compiled, like the best scholia, from multiple sources.(48) The pre-Justinian Sinai text, the oldest known example of a free-standing Greek commentary on Roman law, is certainly a compilation. If such a text is composite in nature, then it is very likely that some contemporary legal marginalia, so close in argumentation and style to the material of the Sinai text, were also copied from composite commentaries.

Explicit compilation such as one finds in the Scholia Sinaitica and in later scholia is not a feature of ancient commentaries or of marginalia in literary papyri (which were copied from commentaries)--despite the fact that ancient commentators also pieced their material together from divergent sources.(49) The conventional method of earlier commentators was to construct syntactically seamless series of comments on the text, citing authorities by name or generically ([GREEK TEXT NOT REPRODUCIBLE IN ASCII]) for the material they chose to use. It was not their habit to copy multiple but discrete comments on individual passages, with labels to mark the start of each new excerpt. This is what one finds in the Sinai commentary, however, where additional comments on particular points are introduced by an indication of source. In medieval scholia, secondary and tertiary material is typically also clearly labelled, often by [GREEK WORD NOT REPRODUCIBLE IN ASCII]. The chief significance of the word, for our purposes, is that it does not appear with this significance in literary papyri. That it is heavily used in Greek scholia and in Biblical catenae suggests that those later compilations, even when rooted in the same Alexandrian commentaries from which ancient papyrus commentaries on literature also derive, passed at some stage through a different tradition. Another by-product of compilation, namely redundancy, was also largely smoothed away by ancient commentators. It is evident for the first time, however, in papyrus marginalia in the very late, new-format Oxyrhynchus Callimachus--another small sign that a new trend had developed in the transmission of commentaries.(50) Annotated legal codices of late antiquity whose marginalia resemble the Scholia Sinaitica in composition and the Callimachus in format are clearly a good place to look for evidence in the continuing debate over the date at which scholia first took form.

I summarize briefly this hotly contested issue. Gunther Zuntz was adamant that scholia on ancient Greek literature were a by-product of the literary renaissance of the Byzantine ninth and tenth centuries, an innovation of `humanistically minded ecclesiastics on the model of the theological catenae marginales with which they were familiar'.(51) These `catenae' on which he lays such stress are commentaries on sacred texts. More precisely, they are collections of excerpts of commentaries chained together (hence the name) in which each excerpt is labelled as to source. In Byzantine manuscripts, catenae characteristically were copied in small scripts into broad margins surrounding the subject text, although in fact they also were transcribed, like the Scholia Sinaitica, as independent texts.(52) Before the ninth century, argued Zuntz, scripts were too bulky and margins too narrow to accommodate such material. With the publication of the remarkable Oxyrhynchus Callimachus, Zuntz slightly modified this position. He saw this codex as a `missing link' in the history of scholia. Still, though, he thought it only a `crude predecessor' of full marginal commentaries of the classical Byzantine type: papyrus, not parchment, with a grossness and bulk in its script which in later manuscripts is much reduced. `The classical balance between text and commentary is not yet within sight; on the contrary, the scholia are sadly squeezed by the--still dominating--text.'(53)

But is this enough to reject an earlier date for the genesis of scholia? Let us reconsider those catenae. The earliest surviving manuscripts containing catenae date to the eighth or ninth century, but tradition says that composite commentaries on scripture originated long before that time: the invention of the catena is traced to Procopius of Gaza in the late fifth or early sixth century.(54) Nigel Wilson has accordingly explored the possibility of a link between these catenae on sacred texts and scholia on classical literature.(55) He asked, usefully, whether catenae might be related, in structure, to surviving compiled commentaries on certain classical authors. He draws attention, specifically, to the obscure Philargyrius, who lived perhaps at Milan in the fifth century and wrote a commentary on Vergil's Eclogues. His commentary is noteworthy, for our purposes, because in it he uses the word aliter whenever he wants to separate alternate comments on a single passage. Such a usage is exactly parallel to the use, in Greek scholia and catenae, of [GREEK WORD NOT REPRODUCIBLE IN ASCII]. Wilson assumes that Philargyrius did not invent the device himself but copied it, rather, from a Greek model. And that model, presumably, was composite scholia on a Greek text. If his hunch is right, Philargyrius's little aliter also supports the theory that composite scholia were in existence in Greek by the fifth century.

The accumulated evidence for the early invention of scholia, when put all together, may not be conclusive, but it is highly suggestive. It derives from three very different kinds of source. First there is the literary evidence, including those fifth-century commentaries by Philargyrius on the Eclogues and the Oxyrhynchus Callimachus of the sixth or seventh century with its dense, carefully copied, and composite marginalia. Second is the tradition that claims that catenae were invented by Procopius, who lived in Gaza in the late fifth and early sixth centuries. Third, there is the evidence from the law schools, most conspicuously the composite fifth-century Scholia Sinaitica and the long and detailed marginalia in contemporary legal codices--marginalia for all practical purposes the same in nature as the Sinai text, and very similar in execution to scholia. Cumulatively, the evidence suggests strongly that we should set back by at least four centuries, from the ninth to the fifth, Zuntz's date for the `invention' of scholia. This is not to say that the invention was applied, thenceforth, in manuscripts of every classical author. As Wilson observes, it is a plausible assumption that it affected only certain authors at first. What does seem incontrovertible, however, is that scholiastic commentary did indeed exist in the fifth century for a variety of authors and texts.

This requires us to accept, then, that there was significant intercourse, in that earlier period, between the sacred and the profane spheres--for how otherwise could the idea of scholiastic commentary have been transferred apparently so effortlessly between scriptoria that produced texts as different in nature as those represented by the evidence? The legal evidence is especially tantalizing, given the relatively early date of the Scholia Sinaitica and the evidence of half a dozen fourth- and fifth-century codices of law. Innovations in the format of the codex have been traced already, by palaeographers, to legal scriptoria. Is it possible that innovations in the form of legal commentaries inspired Biblical catenae and literary scholia? Are the scriptoria of some ancient law school, perhaps at Beirut, the source to which we should trace the invention of scholia? The answer, I think, is `Yes, but...'. To speak of the three spheres in which commentaries of a new kind suddenly appeared in the fifth century--the ecclesiastic, the literary, the legal--as if they were mutually independent is misleading. There was, in the East of the fifth century, an interdependence among the three strong enough to suggest that a good idea arising in one could easily have spread to the others.

In the first place, members of each of the three professions--whether teachers, experts on the law, or men of the church--will have been immersed in their youth in the traditional texts of the ancient literary tradition. In Christian centuries the grammatical curriculum remained essentially classical, and this classical heritage is something ecclesiastics and persons trained in law will have had in common. One can also cite specific linkages between the legal sphere and the church. By the fifth century, for example, the accumulated body of Roman law included ecclesiastical elements.(56) Lawyers were now Christians, and Christians could properly take up legal studies. Even a cursory review of the history of legal studies at Beirut, furthermore, makes clear that its symbiosis with the church was close. Some of our chief sources are ecclesiastical authors. Procopius himself, commentator on scripture and `inventor' of the catena, sent students the relatively short distance from Gaza to study law at Beirut. Most telling is Justinian's charge to the governor, the bishop, and the professors of law to oversee, for the city of Beirut, the enforcement of his ordinances governing the proper use of the Digest.(57) That Eustathius, bishop at Beirut from about 545, may even have provided auditoria for legal instruction on the grounds of his new cathedral therefore seems, under the circumstances, perfectly fitting.(58)

Commentaries from late antiquity, moreover--whether they explicate scripture, law, or literature--share an approach to the subject text which is closer in character to later scholia than to earlier commentaries. The deliberate compilation of information from multiple sources, with those sources labelled, is a new development in this age, and it crosses the lines of genre. Legal works like the fifth-century Scholia Sinaitica and the Digest of the sixth century take this form. They are catenae no less than the catenae on scripture which Procopius is credited with inventing at about the same time.

Vocabulary characteristic of the ancient legal classroom also found a place, eventually, both in church offices and in classical libraries. We have seen that the command [GREEK WORD NOT REPRODUCIBLE IN ASCII], sometimes supplemented by the judgmental [GREEK WORD NOT REPRODUCIBLE IN ASCII], is common in fifth- and sixth-century texts of Roman law. It catches on. In the tenth century, it can be found even in the scholia of a manuscript of canon law.(59) In the literary sphere, [GREEK WORD NOT REPRODUCIBLE IN ASCII] is unknown in papyrus commentaries from Egypt--[GREEK WORD NOT REPRODUCIBLE IN ASCII] is the word conventionally used to draw the reader's attention to something significant. In later centuries, however, it figures commonly, often accompanied by [GREEK WORD NOT REPRODUCIBLE IN ASCII], in scholia to certain classical texts.(60) I suspect the two terms entered the technical vocabulary of commentators late, and elsewhere than at Alexandria, for both are as rare in the scholia to poetry as they are common in scholia to certain prose authors. Scholia on the best poets derive ultimately, by common consent, from the extensive commentaries produced by Alexandrian scholars of the Ptolemaic and early Roman centuries--commentaries which were far more numerous than Alexandrian commentaries on prose. [GREEK WORD NOT REPRODUCIBLE IN ASCII] is found only once, for example, in the scholia to Sophocles, twice in those to Theocritus, and three times each in the scholia to Nicander and to Pindar--all compilations built largely on an Alexandrian base. It is exceptionally common, however, in the scholia on Plato, Oppian, Thucydides, and, above all, Lucian.(61)

Homer supplies an interesting illustration of the very limited appearance of the term in poetic scholia: despite the great bulk of scholia surviving for the Iliad, [GREEK WORD NOT REPRODUCIBLE IN ASCII] turns up there only sixty-four times. Typically it figures in formulations like [GREEK TEXT NOT REPRODUCIBLE IN ASCII] ... or [GREEK TEXT NOT REPRODUCIBLE IN ASCII] N ..., which refer to critical sigla applied to the text according to an Aristarchan or other critical system. Often, though, it is free-standing and syntactically unrelated to its surroundings, as often in earlier legal comments:

Scholia in Iliadem 3.11 (scholia vetera) [GREEK TEXT NOT REPRODUCIBLE IN ASCII]. (Genevensis 44, ed. J. Nicole)

Scholia in Iliadem 3.188 (scholia vetera) [GREEK TEXT NOT REPRODUCIBLE IN ASCII] (Genevensis 44, ed. J. Nicole)

Scholia in Iliadem 9.443c (scholia vetera) ex. [GREEK TEXT NOT REPRODUCIBLE IN ASCII], AT (Marcianus 454 [`Venetus A'], ed. H. Erbse)

Scholia in Iliadem 23.36a, (scholia vetera) Ariston. [GREEK TEXT NOT REPRODUCIBLE IN ASCII]. Aint (Marcianus 454 [`Venetus A'], ed. H. Erbse)

What is also significant about these sixty-four occurrences is that each one of them (with only a single possible exception) is to be found among the exegetic scholia, not in the Aristarchan scholia of the `Viermannerkommentar'.(62) I do not argue that the exegetic scholia, whose date and provenance are disputed, are a product of late antiquity. We know otherwise from the commentary on Iliad 21 in POxy II 221, a scholarly hypomnema of exegetic character copied in the second century.(63) Nor would I claim that the exegetic scholia to Homer, or any other scholia that employ these distinctive terms, were composed in conscious imitation of commentaries produced in schools of law. I simply draw attention to the striking use, in the very different contexts of literary and legal studies, of technical terminology which is alien to the assuredly Alexandrian scholia. It suggests that the pedagogy represented by ancient legal commentaries and by certain grammatikoi on literature diverged markedly from pedagogy based on the Aristarchan tradition. A closer look at distinctive terms like these may lead to a better understanding of the circumstances in which the exegetic scholia to Homer developed.

I close with one final piece of evidence of the symbiosis between church and law school in late antiquity. This is the honorific [GREEK TEXT NOT REPRODUCIBLE IN ASCII] applied in the Basilica to fifth-century teachers of law. The term had currency in earlier centuries in hyperbolic expressions of range: [GREEK WORD NOT REPRODUCIBLE IN ASCII] was used of the games; the emperor was [GREEK TEXT NOT REPRODUCIBLE IN ASCII]. In ecclesiastical usage contemporary with our papyri, the term referred to the ecumenical patriarchate.(64) When it is applied in legal commentaries to professors of law, it not only indicates the grand scope of the law but may also suggest a link between teachers and clerics. Beirut's great teachers of law, that is, like the patriarch, are the dominant authorities in their sphere. Surely the same construct also served clerics in the assertion of authority. It is perhaps no coincidence, then, that contemporary papyri preserving fragments of legal commentaries like those written by the 'universal teachers' have the same format as sacred catenae. Given the tightness of links between church and legal instruction, the question of whether sacred or legal catenae were invented first may even be moot. Commentaries of identical format, both sacred and legal, may have been produced simultaneously in Gaza and in centres of legal study. Conceivably there were individual scribes who worked in both spheres. Whatever the subject text, the marginal annotations which they transcribed are the earliest examples of the scholia which we know from Byzantine manuscripts.

If I must assign priority for the invention, however, I suppose I should assign it to the schools of law, given the unparalleled, and very tangible, innovations in book format in contemporary legal scriptoria, and given the new need to provide students of Roman law in the fifth century with explanations of that law in Greek. Professors in contemporary law schools had the motivation to develop scholiastic commentary, and they found the means.

(*) I wish to thank the referees of Classical Quarterly and also Michael Haslam, Linda Jones Hall, Joel Itzkowitz, Constant Matheeussen, and Nigel Wilson (whose 1967 article in this journal suggested my title) for reading and improving this paper. I hope it has profited also from the comments of audiences which heard earlier versions, both at the 1995 meeting of the American Philological Association and at the University of Illinois. I also am thankful to Paul Mertens for graciously providing, in advance of full publication, the catalogue numbers of the third edition of R. A. Pack, The Greek and Latin Literary Texts from Graeco-Roman Egypt(2) (Ann Arbor, 1965) (hereafter [P.sup.2]).

(1) P. Collinet, Histoire de l'ecole de droit de Beyrouth (Paris, 1925) (hereafter `Collinet'). Reviewed critically, updated, and mainly upheld by L. Wenger, Die Quellen des romische Rests (Vienna, 1953), esp. pp. 619-32.

(2) Collinet, pp. 16-20, canvasses and rejects arguments for an Augustan or a Hadrianic foundation; he settles on the early third century or the late second as the time the school was established. Whether there was a single institution or more than one is open to question. For the earlier period of legal studies at Beirut, Fergus Millar opts for the latter: `It is essential to stress that all our evidence up to the fourth century does indeed present the teaching and learning of Roman law as an aspect of the "rather more Roman" city that Berytus was; and that we should speak of "law schools" rather than of "the Law School".' `The Roman colonies of the Near East. A study of cultural relations', in H. Solin and M. Kajava (edd.), Roman Eastern Policy and Other Studies (Helsinki, 1990), pp. 7-58, at p. 23.

(3) Collinet, pp. 114-15 (a table).

(4) Gregory Thaumaturgus, orat. panegyr. ad Origenem 5.62, H. Crouzel (ed.), Remerciement d Origene, suivi de la lettre d'Origene a Gregoire (Paris, 1969); PG 10.1065-6: [GREEK WORD NOT REPRODUCIBLE IN ASCII] (scil. a guardian angel) [GREEK TEXT NOT REPRODUCIBLE IN ASCII]. In fact Gregory never reached Beirut but was detained in Caesarea, where his life took a new direction as a result of his encounter with Origen.

(5) [GREEK TEXT NOT REPRODUCIBLE IN ASCII]: Libanius, ep. 652. I (a.d. 361, to Anatolius, consularis Phoenices); [GREEK WORD NOT REPRODUCIBLE IN ASCII] ep. 438.5 (A.D. 355); [GREEK TEXT NOT REPRODUCIBLE IN ASCII]: ep. 1529.1 (A.D. 365).

Also from the fifth century is a papyrus, from Hermupolis in Egypt, containing panegyrics on teachers at Beirut ([GREEK TEXT NOT REPRODUCIBLE IN ASCII] are the terms used: P.Berol. inv. 10558+10559, E. Heitsch, Die griechischen Dichterfragmente der romischen Kaiserzeit [Gottingen, 1961], nos. xxx and xxxi, [P.sup.2]1851). They may have been associated with the teaching of law: [Tau] ..... [GREEK TEXT NOT REPRODUCIBLE IN ASCII][ (no. xxx, line 75);].. co[.] .... [GREEK TEXT NOT REPRODUCIBLE IN ASCII] (no. xxxi, line 35).

(6) Geogr. gr. min. II, p. 517, [sections] 25.

(7) On legal studies at Rome see D. Liebs, Die Jurisprudenz im spatantiken Italien (260-640 n. Chr.) (Berlin, 1987); H. J. Scheltema, L'enseignement de droit des antecesseurs (Leiden, 1970) (hereafter 'Scheltema, L'enseignement'), p. 3.

(8) The Digest of Justinian, T. Mommsen, P. Krueger (edd.), English tr. by Alan Watson (Philadelphia, 1985). Const. Omnem 7: Haec autem tria volumina a nobis composita tradi eis (scil. to students) tam in regiis urbibus quam in Berytiensium pulcherrima civitate, quam et legum nutricem bene quis appellet, tantummodo volumus, quod iam et a retro principibus constitutum est, et non in aliis locis quae a maioribus tale non meruerint privilegium: quia audivimus etiam in Alexandrina splendidissima civitate et in Caesariensium et in aliis quosdam imperitos homines devagare et doctrinam discipulis adulterinam tradere: quos sub hac interminatione ab hoc conamine repellimus ut, si ausi fuerint in posterum hoc perpetrare et extra urbes regias et Berytiensium metropolim hoc facere denature librarum auri poena plectantur et reiciantur ab ea civitate, in qua non leges docent, sed in leges committunt.

(9) Const. Deo Auctore [sections] 12 (Ideoque iubemus...) nostram autem consummationem, quae a vobis deo adnuente componetur, digestorum vel pandectarum nomen habere sancimus, nullis iuris peritis in posterum audentibus commentarios illi applicare et verbositate sua supra dicti codicis compendium confundere: quemadmodum et in antiquioribus temporibus factum est, cum per contrarias interpretantium sententias totum ius paene conturbatum est: sed sufficiat per indices tantummodo et titulorum suptilitatem quaedam admonitoria eius facere, nullo ex interpretatione eorum vitio oriundo.

Cf. Const. Tanta-/[GREEK WORD NOT REPRODUCIBLE IN ASCII] 21 Hoc autem, quod et ab initio nobis visum est, cum hoc opus fieri deo adnuente mandabamus, tempestivum nobis videtur et in praesenti sancire, ut nemo neque eorum, qui in praesenti iuris peritiam habent, nec qui postea fuerint audeat commentarios isdem legibus adnectere: nisi tantum si velit eas in Graecam vocem transformare sub eodem ordine eaque consequentia, sub qua et voces Romanae positae sunt (hoc quod Graeci [GREEK TEXT NOT REPRODUCIBLE IN ASCII]), et si qui forsitan per titulorum suptilitatem adnotare maluerint et ea quae [GREEK WORD NOT REPRODUCIBLE IN ASCII] nuncupantur componere. alias autem legum interpretationes, immo magis perversiones eos iactare non concedimus, ne verbositas eorum aliquid legibus nostris adferat ex confusione dedecus.

Scheltema, L'enseignement, p. 16, lists possible explanations for the contradiction, none satisfactory to him: the threat was an empty threat; it was revoked at once but we have no notice of the revocation; `interpretatio' carried a meaning different from what we understand as `commentary'. Matheeussen (below, n. 21) has proposed that the prohibition applied only to the official copies of the Digest prepared for each of the three recipient cities. Honore (ZSS 110 [1993], 766-7) offers a simple solution; namely, that this impractical and almost unenforceable restriction was simply allowed, quietly, to lapse almost as soon as it was promulgated. It was omitted from the publication in 534 of Cordi.

(10) Thalelaeus: RE 9 (1934), cols. 1208-10.

(11) H. J. Scheltema, N. van der Wal et al. (edd.), Basilicorum libri LX (Groningen, 1953-85). Wenger (above, n. 1), pp. 622ff. provides several illustrative quotations. Collinet, pp. 117-206, with specific references for the particular epithets at p. 125, adds Amblichus and Leontius.

(12) Wenger (above, n. 1), p. 622 n. 261, lays out the evidence for setting the termini post quem and ante quem as, respectively, the first decade of the fifth century and the year 500.

(13) Wenger's cautious assessment of the evidence for the achievements of the Beirut professors (above, n. 1, pp. 626-9) is a useful antidote to Collinet's sometimes exuberant analysis. For the present argument, however, the fact that the fifth century was a period of significant pedagogical and codicological innovations in the East is the critical fact.

(14) Foundation by Augustus: CIL III 161, 165, 166, 6041, 153; Ulpian lib. I de censibus (Dig. See also Strabo, Geog. 16.2.19 (756), with the discussion of Millar, `The Roman Colonies' (above, n. 2), pp. 7-58, esp. p. 12. See also Millar, The Roman Near East 31 B.C.-A.D. 337 (Cambridge, Mass., 1993), esp. pp. 279-85 and 527-8.

(15) Millar, `The Roman Colonies' (above, n. 2), p. 8.

(16) Ibid., p. 14.

(17) Gregory Thaumat., orat. panegyr, ad Origenem (1.7) (above, n. 4), [GREEK TEXT NOT REPRODUCIBLE IN ASCII]. He also expresses concern that his extensive study of Latin will be an impediment to effective expression in Greek.

(18) Libanius at or. 2.43 complains of the decline of Greek rhetoric in his day: [GREEK TEXT NOT REPRODUCIBLE IN ASCII]. In orat. 2.44 (A.D. 381 or 382) he links the decline specifically with the current privileged position of Latin over Greek, and of the craze for legal studies: [GREEK TEXT NOT REPRODUCIBLE IN ASCII] and at orat. 48.22 (n.b. 384/5) he laments the annual exodus of the youth of the best families in Antioch for Beirut and Rome: [GREEK TEXT NOT REPRODUCIBLE IN ASCII].

(19) J. H. W. G. Liebeschuetz, Antioch: City and Imperial Administration in the Later Roman Empire (Oxford, 1972), p. 246.

(20) An argument for instruction in Greek now to be discarded (Collinet, p. 212) held that Cyril, the earliest Beirut scholar included among [GREEK TEXT NOT REPRODUCIBLE IN ASCII] (he probably began to teach in the first or second decade of the fifth century), wrote a work entitled [GREEK TEXT NOT REPRODUCIBLE IN ASCII]. The passage on which the claim is based, however (BS 314-21, Heimbach I 646), is corrupt, as Scheltema, L'enseignement, pp. 9-10 n. 35, has pointed out: schol. [GREEK WORD NOT REPRODUCIBLE IN ASCII] (Thalelaeus) on Cod lust. 2.3.6: [GREEK TEXT NOT REPRODUCIBLE IN ASCII]. Even if the source were uncorrupted, however, this would be no guarantee that the work on definitions was in Greek, for Latin in the Basilica was uniformly transliterated into Greek.

(21) In the East it was not only students of law who lacked Latin: Scheltema, L'enseignement, pp. 11-2. I am indebted for the analysis that follows to the clear and patient exposition provided by Constant Matheeussen of the Universite catholique de Bruxelles in a lecture he gave at Wayne State University in January 1996.

(22) Scheltema, L'enseignement, esp. ch. 2, `L'enseignement avant et apres la codification'. For an example of an index and the passage it translates, see C. Matheeussen, `L'interpretation de la gratuite du mandat dans les scholies des Basiliques et la reductibilite du "salaire" du mandataire', Subseciva Groningana III: it is apparent that the surviving index to Dig. (Paul) is a translation of the Latin into Greek and not, as often averred, a summary (pp. 50 and 58).

(23) The scholars who most frequently provide these references are Thalelaeus, Theodore, and Stephanus: Collinet, pp. 243ff.; RE s.v. Thalelaeus (above, n. 10).

(24) But it since has been lost. J. Baviera, J. Furlani (edd.), FIRA pars altera (Florence, 1940), pp. 637-52; [P.sup.2] 2958, pre-Justinian: 5th cent.?, E. G. Turner, The Typology of the Early Codex (Philadelphia, 1976), p. 126.

(25) Collinet, pp. 272-9; relevant citations in the Basilica are collected at pp. 311-12.

(26) Scheltema, L'enseignement, p. 11.

(27) Also, the man can use the money: Libanius, ep. 1539:... [GREEK TEXT NOT REPRODUCIBLE IN ASCII]. The `quantity of books' refers to those to which the professor has access, whether he wrote them himself or owns them. A library of good quality would presumably recommend a professor to a prospective student.

(28) K. McNamee, `An innovation in annotated codices', Proceedings of the XXI International Congress of Papyrology, Berlin (August 1995), forthcoming.

(29) Texts of new format in which the same scribe added both notes and commentary (references marked with an asterisk, hem and in subsequent notes, are texts of law; see below): [P.sup.2] 145, 186, 1487, *2281, *2282, *2286, *2982; perh. *2277, [P.sup.3] 2362.3, *2979.1, *2979.2 (Mertens).

(30) Annotations in relatively informal hands in codices of new design: [P.sup.2] 2866, 2867, *2971, *2974. In layout, however, these books resemble other texts of new format.

(31) Annotations are written in neat blocks at the outer edge of the page of [P.sup.2] 145, 1356, [P.sup.3] 2861.21 (Mertens); at the inner edge of the page of [P.sup.2] 119; at both inner and outer edges of [P.sup.2] 142, 186, 1487 (esp. e.g. fol. 4r, 5r, v), *2953. Texts in which it is unclear whether notes are in the inner or outer margin are: [P.sup.2] *2280, *2282, 2866, 2867, [P.sup.3] *2979.1, *2979.2 (Mertens). Both the left and right edges of notes are preserved in [P.sup.2] 145, *2280, 2866, 2867, *2984, *2966, [P.sup.3] *2979.1, *2979.2(?) (Mertens).

(32) Aratus Phaenom: P.Berol. inv. 5865, M. Maehler, APF 27 (1980), 19-32 ([P.sup.2] 119, `Byzantine').

(33) The widest such, at the extreme edge of a page, is 3 cm ([P.sup.2] 145: Aristophanes and 1356: Pindar, discussed below). The scribe of `2979.1 (Mertens) maintained a border of similar dimensions between text and commentary. The widest margin I have seen between a block of marginal commentary and the inner edge of the page is 2.5 cm (in [P.sup.2] 186, the Oxyrhynchus Callimachus); elsewhere it runs to about 1-1.5 cm, e.g. in [P.sup.2] 119 (Aratus, above, n. 32) and 1487 (the `Antinoe Theocritus').

Martin Irvine, The Making of Textual Culture (Cambridge, 1994), p. 385, draws attention to the regularity in the apportionment of space in medieval manuscripts: `The page was typically ruled in three unequal columns, a large central column for main text and two outer columns or margins of varying width for glosses [i.e. scholia]' (emphasis added).

(34) Bookrolls with long annotations in informal hands include: [P.sup.2] 361 (Epicharmus), 1237 (Ibycus), 1360 (Pind.). The layout of these bookrolls is conventional. No special steps were taken to accommodate marginalia.

(35) C. H. Roberts and T. C. Skeat, Birth of the Codex (Oxford, 1987), p. 37. For an earlier, but still useful, study of the relative frequency of rolls and codices, see W. H. Willis, `A census of literary papyri', GRBS 9 (1968), 220; Willis's material is updated, for codices, by E. G. Turner, The Birth of the Codex (Philadelphia, 1977), pp. 102-34.

(36) H. Riad, A. el-Kadr Selim, L. Koenen, The Cairo Codex of Menander (PCairJ 43227) (London, 1978); [P.sup.2] 1301, second half of the fifth century (G. Cavallo and H. Maehler, Greek Bookhands of the Early Byzantine Period: AD. 300-800 [BICS Suppl. 47, 1987], pl. 16b).

(37) Informally annotated codices of conventional design include: [P.sup.2] 141 (Aristophanes), 201 and 215 (Callimachus), 406 (Euripides).

(38) Aratus, Phaenom.: above, n. 32; Pindar, P. 1: MPER N.S. I 23, K. McNamee, Proceedings of the XX International Congress of Papyrology (Copenhagen, 1993), pp. 177-84 ([P.sup.2] 1356, 6th cent.); Callimachus, varia: POxy XX 2258 ([P.sup.2] 186, 6th or 7th cent.).

(39) G. Cavallo and H. Maehler, Greek Bookhands (above, n. 36).

(40) The marginalia of PReinach 2173 ([P.sup.2] 2971) look, for example, as if they may have been added by a reader rather than a professional scribe, but there can be no certainty.

(41) Cf. B. Bischoff, Latin Palaeography, tr. Daibhi O Croinin and David Ganz (Cambridge, 1990), p. 74, speaking of the `older (eastern) half-uncial' manifest in a set of Latin manuscripts of the third to fifth centuries, mainly from Egypt: `Given the very considerable component of legal texts transmitted in this kind of writing, the Latin law school of Berytos (Beirut) probably played a role, if not already in its formation then certainly in its use from the third to the fifth century.'

(42) Scheltema, L'enseignement, p. 8, provides a table of texts read in each of the five years of study in pre-Justinian and post-Justinian Beirut; cf. Collinet, p. 230.

(43) Const. Deo Auctore 12 and Tanta-[GREEK WORD NOT REPRODUCIBLE IN ASCII] 21 (see above, n. 9).

(44) Zacharias the Scholastic describes how Severus, for example, who began law studies at Beirut in 486 or 487, systematically added comments to the margins of his books: Zacharias Scholasticus, Vita Severi 25, M. A. Kugener (ed.), Patrologia Orientalis II fasc. 1 (Paris, 1903), pp. 52ff. The text is preserved in Syriac translation, difficult to interpret. Kugener translates: Severe etudia les lois autant qu'on peut le faire, examina et approfondit tous les edits imperiaux y compris ceux de son temps, compara ensemble les commentaires contenus dans les precis des lois, nota dans les cahiers des racines auxiliaires de l'oubli et du souvenir et laissa, comme des [GREEK WORD NOT REPRODUCIBLE IN ASCII] a ceux qui viendraient apres lui, ses livres et ses notes' Scheltema, L'enseignement, pp. 14-15 n. 45, interprets: `Severe consulta les constitutions, auxquelles on s'etait refere dans les [GREEK WORD NOT REPRODUCIBLE IN ASCII]; puis il copia les gloses marginales faites par d'autres etudiants et munit ces gloses de lemmes latins [i.e. Kugener's `racines auxiliaires de l'oubli et du souvenir']. Il preta a des etudiants plus jeunes les cahiers de cours ([GREEK WORD NOT REPRODUCIBLE IN ASCII]) ainsi formes.'

(45) In any case, the new and careful format is in keeping with the obsession of the Roman administration to avoid alteration to the original text: J.-L. Mourgues, `Ecrire en deux langues. Bilinguisme et pratique de chancellerie sous le haut-empire romain', Dialogues d' histoire ancienne 21 (1995), 124.

(46) Collinet, p. 114, names four, from Alexandria and Heliopolis. Legal papyri whose specific Egyptian provenance is known were excavated at sites stretching from the Fayum to, perhaps, the Thebaid: Crocodilopolis ([P.sup.2] 2283), Theadelphia ([P.sup.2] 2961), the Fayum ([P.sup.2] 2286, ?2983, ?2984), Oxyrhynchus ([P.sup.2] 2954, 2963, 2965, 2968, 2969, 2975, 2982?), Antinoopolis ([P.sup.2] 2979, ?2953, 2988, [P.sup.3] 2979.1, 2979.2 (Mertens), PAnt III 155), Hermupolis Magna ([P.sup.2] 2989, ?2990; Hermupolis is also the provenance of the Berlin fragments of epikedeia on professors at Beirut, [P.sup.2] 1851: see above, n. 5), and ?the Thebaid ([P.sup.2] 2280, 2967, 2974).

(47) On the assumption that the prohibition was observed, PHeidLat 4 (Digest, [P.sup.2] 2966), PSI XIII 1347 (Codex, [P.sup.2] 2970), PSI I 55 (Stephanus, Index to the Digest, [P.sup.2] 2965), and PReinach 2173 (Digest, [P.sup.2] 2971) are not from Alexandria.

(48) The most striking example of compiled scholia from the Byzantine period are the scholia vetera to Homer's Iliad in the margins of Venetus A (Marcianus 454), which combine versions of learned commentaries by four Alexandrian scholars: H. Erbse, Scholia graeca in Homeri Iliadem (Stuttgart, 1969-87), vol. i, introduction.

(49) Didymus, at the time of Augustus, produced commentaries on Greek poetry which collected excerpts from the works of his Alexandrian predecessors: R. Pfeiffer, A History of Classical Scholarship i (Oxford, 1968), p. 273.

(50) On the evidence for compilation in its marginalia, see K. McNamee, `Missing links in the development of scholia', GRBS 36 (1995), 399-414.

(51) G. Zuntz, An Inquiry into the Transmission of the Plays of Euripides (Cambridge, 1965), pp. 272-5.

(52) N. G. Wilson, `A chapter in the history of annotation', CQ n.s. 17 (1967), 244-56, esp. 252-4.

(53) Zuntz, An Inquiry (above, n. 51), p. 274n., with the argument enlarged at Zuntz, Die Aristophanes-Scholien der Papyri(2) (Berlin, 1975), Nachwort.

(54) Wilson CQ (above, n. 52), cites R. Devreesse, Dictionnaire de la Bible, `Chaines', (Paris, 1907-), B. Altaner, Patrology, tr. H. Graef (Freiburg, 1960), pp. 622-3, O. Bardenhewer, Geschichte der altchristlichen Literatur 4 (Freiburg, 1924), p. 13.

(55) N. G. Wilson, CQ (above, n. 52), with the argument developed in Scholars of Byzantium (Baltimore, 1983), pp. 31-3, and `The relation of text and commentary in Greek books', in C. Questa and R. Raffaelli (edd.), Il Libro e il Testo (Urbino, 1984), pp. 105-110.

(56) W. K. Boyd, The Ecclesiastical Edicts of the Theodosian Code (New York, 1905, repr. 1969).

(57) Const. Omnem [sections] 10: Et haec omnia in hac quidem florentissima civitate [i.e. Constantinople] vir excelsus praefectus huius almae urbis tam observare quam vindicare, prout delicti tam iuvenum quam seriptorum qualitas exegerit, curae habebit: in Berytiensium autem civitate tam vir clarissimus praeses Poenicae maritimae quam beatissimus eiusdem civitatis episcopus et legum professores.

(58) Collinet, pp. 63-70.

(59) [GREEK WORD NOT REPRODUCIBLE IN ASCII] in a manuscript of Canon Law: MS. Laud. Gr. 39 (X; N. G. Wilson, Mediaeval Greek Bookhands [Cambridge, Mass., 1973], pl. 18).

(60) In a manuscript of Photius's letters it introduces a marginal note in Letter 8 that refers to Thucydides and Demosthenes: Ms. Barocci 217 (IX; Wilson, Mediaeval Greek Bookhands [above, n. 59], pl. 15). It also appears relatively frequently in the later scholia to Aristophanes.

(61) The collocation [GREEK TEXT NOT REPRODUCIBLE IN ASCII] occurs a remarkable ninety-six times in the scholia to Lucian.

(62) The single possible exception (at Il. 23.36a) is a passage of the scholia vetera traced by Erbse to Aristonicus but copied, I think, from an exegetic commentary. It is not a lengthy comment such as fill the margins of Marcianus 454 (`Venetus A'). Rather, it is a shorter note written in an inner margin, the position in which two other exegetic notes employing the term [GREEK WORD NOT REPRODUCIBLE IN ASCII] also appear. Erbse notes the absence of a diple beside the affected line and comments `fort. neglegentia scribae'; but perhaps the diple is lacking because this note, like many others written in this position, is in fact from the exegetic, not the Aristarchan scholia.

(63) Where, however, the term does not appear.

(64) Collinet, pp. 167-76.
COPYRIGHT 1998 Cambridge University Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1998 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:McNamee, Kathleen
Publication:The Classical Quarterly
Date:Jan 1, 1998
Previous Article:Salpe's PAIGNIA: Athenaeus 322A and Plin. 'H.N.' 28.38.
Next Article:Two notes on Greek dithyrambic poetry.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters