For this third edition of Scribes & Scholars the text has been reset to incorporate numerous changes and additions that take into account new discoveries and advances in scholarship. The bibliographical notes (244-294) are especially rich in new material covering the history of the ancient book, the transmission of the classics and the history of classical scholarship. The third edition contains 46 more pages than the second but the increase is due in part to changes in type and page format. The many virtues of this book, which has become a classic of its kind, have been widely recognized and justly praised; in the revision its value is enhanced, though of necessity I will concentrate on what I see as its shortcomings. There follows a collation of the third and second editions: numbers in square brackets [ ] give the pagination of the second.
6.ix, pp.234-237 ‘Fluid forms of transmission: Technical and popular literature’ contains a good discussion of large scale textual alteration, due to editorial revision and updating, in sub-literary genres, the commentary, handbook and scholia. I was startled to see Aratus’s Phaenomena included among ‘texts in which the literary intention is subsidiary or negligible…’.
6.x, pp.237-239 ‘Conventions in the apparatus criticus’ has a limited usefulness. The authors expand abbreviations and explain the meaning of critical signs, but this material does not lend itself to brief treatment, especially when it is intended for the enlightenment of beginners. For example, interested students will want to know the difference between secludere and delere which are both used to mark deletions and are both indicated in the text by square brackets [ ]. Secludere is normally used for text that is regarded as genuine but has somehow landed in the wrong place: it will be found frequently in the apparatus of a text which we know was left unfinished by the author, e.g. Lucretius’s de rerum natura. Delere is normally used for text that is regarded as spurious. It is therefore misleading to say that ‘square brackets [ ] mark portions of the text transmitted in all or at least the great majority of the witnesses but which are deemed by the editor to be spurious later additions’ (p. 238). Compare Mynors’ treatment of the Helen episode at Aeneid 2.567-588 in his OCT and Hollis’s treatment of Metamorphoses 8.597-610. Although orthographical variants should as a rule be excluded from the apparatus, it is worth mentioning that some editors (even ‘the average editor’?) provide an appendix on orthography (among them C.O. Brink, George Goold and Otto Zwierlein). The negative apparatus ought to be called by its name when it is described (238) and some important editions in which it is used might be mentioned (the Snell-Maehler Pindar, Wilamowitz’s Aeschylus and Radermacher’s Quintilian). The abbreviation al. = aliter, alibi, not alias as stated on 238. A reference to A. Delatte and A. Severyns, Emploi des signes critiques 2 (Brussels/Paris 1938), would be helpful. Reading a descriptive treatment of what is found in an apparatus criticus without any accompanying text is like contemplating the pieces of a disassembled engine and imagining their functions without ever having seen them work. Just as the authors point out that every textual problem tends to be sui generis, so also every apparatus will be the unique expression of the editor’s analysis and judgment of the text and its transmission.
14 Lille papyrus of Stesichorus’s Thebaid upsets attribution of lyric colometry to Aristophanes of Byzantium; 18 paragraph on Herculaneum papyri; 51 new sentence on censorship; 56 Arabic versions of Galen; 64 Photius’s interest in textual criticism; 68-69 discussion of Psellus; 75 short paragraph on Planudes and the seven liberal arts; 81 nice observation on infrastructure of books; 96 references to books written in the palace scriptorium; 97 importance of the palace school; 97 observations on the gravitation of books to centers of power and influence; 97-98 conditions affecting the movements of manuscripts; 99-100[88-90] more details on the manuscript holdings of Tours, Bobbio and other monastic centers; 108 more Italian manuscripts that found their way to Germany; 108-109 local revival in the area of Liege with the resurfacing of rare texts; 111 brief discussion of literacy in connection with the 12th century renaissance; 119 early episode in the history of Greek in England; 129-130 importance of Avignon; 203-204 substantial addition to ‘Epilogue’ on impediments to scholarship since the Renaissance.
28-29[25-27] re-evaluation of Valerius Probus with references on 249 to the important articles by Jocelyn in CQ; 33 a more positive assessment of the role played by handbooks and commentaries; 39-43[35-37] more cautious attitude toward the significance of subscriptiones; 70 Eustathius’s knowledge of Sophocles’Antigone 1165-1168 derived from quotation in Athenaeus; 108-109 classical books in Britain; 121 concluding paragraph on quality of medieval translations from the Greek promoted from bibliography on ; 125 greater scepticism about the breadth of Lovato’s classical reading; 130-131 the vetus Carnotensis discarded as Petrarch’s source for the Spirensian tradition of Livy; 133 manuscripts of Tacitus and Apuleius from Montecassino; 134 more thoughtful evaluation of Coluccio Salutati; 138 Poggio no longer the inventor of humanistic script; 140 expansion of passage on manuscripts that followed the one-way road to the printing house; 145 Matritensis 3678 the source of the variants found in Politian’s collation of the Silvae; 218 change in illustration of recentiores, non deteriores.
4 a different illustration of the capacity of the papyrus roll; 22 a new quotation from De lingua Latina 9.106; 25 library now correctly located in Porticus Octaviae; 90 Virgil of Salzburg deleted; 143 new sentence on Valla; 167 the younger Estienne’s edition of the Anacreontea; 177 Casaubon’s unfinished commentary on Aeschylus; 183 concluding sentence on Nicolaus Heinsius deleted; 198 discovery of the Homeric Hymn to Demeter; 201 epigraphical evidence for the text of Crito 46b4-6; 210 date of Scaliger’s Catullus corrected to 1577; 219 new example adduced, Knights 254, to illustrate importance of indirect tradition.
New texts: 14 Lille papyrus of Stesichorus’s Thebaid; 36 Gallus papyrus; 196 the Mani codex; 197 fragment of Livy, Book 11; 245-246 Lille papyrus of Callimachus containing text and notes. On 4 and 54 inventory numbers for papyri are now supplied. Was there no room, even in the notes, to mention the papyrus codex containing the Alcestis Barcinonensis [edition by M. Marcovich (Leiden/New York 1988)] and Cicero’s In Catilinam I.6-9, 13-33 and II [ Ciceró. Catilineries. Papyri Barcinonenses, ed. R. Roca-Puig (Barcelona 1977), not a new text but a noteworthy discovery?
The reference to ad Atticum 12.6a.1 ( 24) should be accompanied by a reference to ad Atticum 6.2.3 for Cicero’s correction of Phliuntii to Phliasii. On 227 a confusion has resulted from rearrangement of the text. ‘The omission of Sophocles Antigone 1167 in all manuscripts has been mentioned already in connection with Eustathius and the importance of the secondary tradition.’ This is true of the second edition  but not of the third (219) where mention of Eustathius has been deleted.
The reader is carried through the chapters on transmission by the momentum of discovery: it is a tale told with learning and élan. The great virtue of this section is the authors’ perception of the human element, the collective effort of many generations that triumphed over accident, ignorance and neglect. They delineate clearly the interactions of books and readers in their historical settings and thus present more than a picture of how a text survived and migrated from point A to point B: transmission becomes a lifeline between cultures, stimulating, invigorating, rejuvenating the minds of those who made the effort to seek, find and preserve. The chapter on textual criticism, however, comes as a disappointment precisely because it is lacking in this human element. Textual criticism is divorced from the critic and the history of transmission. The student finds himself either adrift in the abstractions of stemmatic theory presented as if it were a problem in Euclidean geometry or stifled by the apparent certainties of the most empirical branch of textual criticism, emendatio. (F.W. Hall’s treatment of textual criticism, pp.108-198 in A Companion to Classical Texts [Oxford 1913] is still instructive, especially for its wealth of quotation.)
If the discussion of stemmatics (211-216) had been founded on a text, then the authors could have used real manuscripts and their readings to illustrate how recensio works. The textual traditions of Catullus and Lucretius are obvious examples. Such a procedure would also have provided excellent materials for giving greater precision to concepts like separative error and significant error and for explaining how to interpret the apparatus criticus (237-239). Moreover, a purely theoretical approach to stemmatics obscures the most important function of the stemma: it makes historical statements about the documents with which it is concerned. The abstract nature of the stemma tends to trap students into thinking that all it can reveal is that manuscript B was copied from manuscript A, a notion that will be instantly dispelled by reading any one of Michael Reeve’s articles on the history of a text.
Although conjectural emendation has long had the lion’s share of attention, it is in many ways the least part of textual criticism. This is a simple matter of arithmetic and probability: the number of correctly transmitted words is greater than the number of transmitted errors and in cases of serious corruption the probability of conjecturing the author’s words is very small. The former preeminence of emendation was established by the difficulty of the enterprise, the certainty of the result obtainable in particular cases of corruption, the literary prestige of the classical authors, the rhetorical virtuosity of its greatest practitioners and by the assiduous cultivation of prose and verse composition in Greek and Latin. In keeping with this tradition the authors devote most of their attention in this chapter to the emendation of well documented types of error (222-233) and eschew questions of a theoretical nature, including a formal definition of textual criticism, and ignore (apart from the definition on 208) the process, dubbed examinatio by Paul Maas, which determines the correctness of the text. Although some have thought examinatio redundant because it is a necessary preliminary to emendatio, it deserves separate recognition as the process by which unfairly impugned texts are vindicated. As Maas observed, this is a critical act on a par with emendation [ Textual Criticism (Oxford 1958) 23 bottom]. Accordingly the authors might have mentioned, for example, Housman’s repunctuation of Catullus 64.324 or Timpanaro’s defense of the interjection o in Lucretius 3.1. Although I understand and sympathize with the authors’ need to be brief and to stick to tried and tested techniques (239), their selection of corruptions and cures makes textual criticism seem dry and mechanical, rather than, as it often is, work requiring insight and imagination.
Surely it comes within the scope of ‘a simple introduction for beginners’ (p.v) to broach significant issues and at least to introduce them to the two great antinomies of the textual criticism of classical texts: (1) theory and method vs experience and common sense; and (2) textual criticism vs literary criticism. The following comments are directed to ‘The development of the theory of textual criticism’ (208-211) which ends abruptly with Lachmann’s edition of Lucretius.
(1) Classical textual criticism has long been hostile to theory and method because theory is normative and method is mechanical and both are inimical to solving textual problems each of which presents its own unique set of circumstances and is not amenable to a set of prescribed rules. True enough in the matter of solving individual problems. But in the larger areas of educating the critic and of developing a rationale for authenticating texts and for determining the strategies that can successfully be applied to correcting corrupt texts, a comprehensive theory of textual criticism will define precisely the nature of the task and provide (not prescribe) methods of analysis and argument that embrace the text in its various historical aspects, linguistic, literary, cultural and transmissional.
A greater openness to theoretical discussion can only help to improve the conceptual framework within which the critic works. Such discussion would have closed the book on false problems like manuscript authority vs critical judgment, objective vs subjective criteria for the evaluation of variants and conjectures, and conservative vs conjectural criticism. It would help us to understand better fundamental working assumptions. For example, why does the textual critic so easily assume parity with the author in constructing his text? If the answer is that the texts are filled with demonstrable errors, one may pose the further question: How effective is conjectural emendation in restoring an author’s text? Estimates are extremely pessimistic. E.J. Kenney has written, ‘If he [i.e. the critic] has studied the history of textual criticism he will know as a matter of demonstrable fact nearly all conjectures are wrong, and he will accept that many of his solutions are in the nature of things provisional.’ If the rate of success is so small, then we must face the fact that textual criticism itself provides no foundation for the elevation of the critic to his position of authority but rather that the literary and cultural value placed on classical texts and the deficiencies of their transmission have conspired to put the critic on a level with the author. What will readers do with a Catullus denuded of the hundreds of emendations that help to make him lepidus et venustus or an Aeschylus robbed of the labors of Auratus, Turnebus and Hermann? The impulse for practising textual criticism is an aesthetic one and the activity itself is a normal process in the socialization of the text.
To avoid the nettles of theory one asserts that the critic is born, not made or recites the parable of the dog hunting for fleas or aphorisms about ratio and Constantinopolitanus (cf. the comments on 233). These may stimulate the adrenalin when it comes to conjectural emendation but they do not provide a sturdy foundation for textual criticism as a whole, which is concerned with the history and authenticity of texts. I do not want to speculate on the role of DNA in the formation of the critic: I do believe, however, that the rigorous program prescribed by Madvig for the education of the critic has more to commend it than exhortations to the employment of common sense and good taste [ Adversaria Critica ad Scriptores Graecos et Latinos (Hauniae, i.e. Copenhagen, 1871) vol. 1, 95-96]. With regard to the power of ratio it must be admitted that our way of thinking is not commensurate with that of the ancients and that a purely logical analysis unencumbered by predispositions and unclouded by aesthetic perceptions, if such a thing is possible in dealing with literary texts, will not necessarily produce the truth. In Lucretius 6.179 Lachmann’s conjecture calescit seems a logical necessity, but parallels from other sources are decisive in showing that it is wrong. Bentley’s nitedula for the transmitted vulpecula in Horace’s Epistles 1.7.29 has the whole arsenal of ratio to support it: everything favors it except the logic of the fabulist and the poet. Although in my opinion it must be rejected, it is more instructive than the authors allow (‘the amusing change’ 186). Suppose for the moment that nitedula and vulpecula were transmitted variants. How would one argue the case? Therein lies the rub. An ‘insistence on logic’ (186) is not peculiar to Bentley’s textual criticism as is demonstrated by the adoption of cornicula at Epistles 1.7.29 in a recent edition of Horace.
These remarks, however, are not intended to suggest that conjectural emendation must be the product of mere guesswork or intuition. An emendation that is congruent with the known facts of transmission, with linguistic, and in poetry, metrical usage and with the author’s usus scribendi is part of a hypothetical construct that postulates a series of unverifiable events leading up to the introduction of an error into the text and then supplies as authentic an unverifiable reading. The more cogent the arguments, the greater the probability that at the very least the textual problem has been correctly diagnosed. In a very few cases the merits of an emendation will be self-evident as is the case with Bentley’s Paris at Aeneid 10.705 and as a result the fact that it is no more than a hypothesis seems immaterial. But in the majority of cases the complexity of the issues involved and the possibility of alternate hypotheses which can explain and correct the same error make one realize that the evidence and arguments adduced in support of an emendation, if they are properly presented, will be of greater value than the emendation itself. As long as critics are determined to establish correct texts and as long as this activity takes them beyond the extant documentary evidence, theory and method will guide their steps.
(2) If the impulse for practising textual criticism is aesthetic, then it is pointless to maintain a strict separation of textual and literary criticism. False is the notion that by some superhuman act of aesthetic asceticism the critic can disown the literary culture (and its habits of thought) into which he has been born and read his text in a purely clinical fashion. Every critic, whether he avows it or not, approaches his text with a head full of assumptions, often unexamined, about the many aspects of literary composition; unity, structure, levels of style, poetic and unpoetic language, decorum; and assumptions about the creative faculty of his author and the nature of literary production. Bentley’s edition of Paradise Lost showed on a grand scale what happens when the literary critical assumptions of the critic are allowed to determine what an author ought to have said. Indeed one scholar has recently described Bentley’s edition as a kind of literary criticism written in the wrong idiom (see section on bibliography below). On a much smaller scale, Ribbeck’s emendation of moritura to monitura at Aeneid 12.55 is not a solution to a textual problem but a revelation of what Ribbeck wanted his Vergil to say. Better to identify and analyze one’s assumptions than to be unaware of them or to pretend that they do not exist. To persist in maintaining the unnatural dichotomy between textual and literary criticism is to deprive the critic of his chief faculty, percipient reading, and to claim for textual criticism a greater degree of objectivity than it can achieve.
These are just two of the many issues that can profitably engage the minds of interested students who might otherwise assume that the methods of textual criticism are in no need of internal examination and that the activity itself is drudgery or, as one author provocatively phrased it, cultural vandalism.
(The preface is dated January 1990. It will be obvious what items appeared too late to be noticed by the authors.)
In their treatment of the ancient book (1.i) the authors do not mention illustration: see Kurt Weitzmann, Ancient Book Illumination (Cambridge, Mass. 1959) and Illustrations in Roll and Codex (Princeton 1970). T. C. Skeat has published a follow-up to the paper referred to on 244: ‘Roll vs codex—a new approach?’ZPE 84(1990) 297-298. Michelle P. Brown’s A Guide to Western Historical Scripts from Antiquity to 1600 (Toronto/Buffalo 1990) provides an affordable companion volume of plates for Bischoff’s Paläographie (262): see also Barbara A. Shailor, The Medieval Book (Toronto 1991, reprint of 1988 edition). A basic bibliography on literacy in the Middle Ages (267) should mention Rosamond McKitterick’s The Carolingians and the Written Word (Cambridge 1989) and M. T. Clanchy’s From Memory to Written Record. England, 1066-1307 (Cambridge, Mass. 1979). A. Grafton’s paper, ‘On the scholarship of Politian and its context’ (276), is now reprinted in his Defenders of the Text (Cambridge, Mass./London 1991). On the German humanist and associate of Erasmus, Beatus Rhenanus, see John F. D’Amico, Theory and Practice in Renaissance Textual Criticism. Beatus Rhenanus, between Conjecture and History (Berkeley/Los Angeles/London 1988). Joseph M. Levine has written an excellent, constructive analysis of Bentley’s foray into Miltonic textual criticism, ‘Bentley’s Paradise Lost,’JHI 50 (1989) 549-568. The same author has also written a marvelous account of the quarrel between the ancients and the moderns, The Battle of the Books (Ithaca/London 1991). The statement (217) that papyri ‘do not as a whole offer markedly superior texts’ should be accompanied by a reference to a noteworthy case where they do provide valuable readings: Michael W. Haslam, ‘Apollonius Rhodius and the papyri,’ICS 3 (1978) 47-73. The reference to B. L. Ullman’s ‘Petrarch’s favorite books,’TAPA 54 (1923) 21-38 given in the 2nd edition  should not have been omitted in the 3rd (273). In the select list of books and articles on textual criticism (293) there is room for E. J. Kenney’s article ‘Textual Criticism’ in the Encyclopedia Britannica, 15th ed. (1974) vol. 20, 614-620, noteworthy for its conciseness, its sagacity and its interdisciplinary approach; and for G. Pasquali’s article ‘Edizione’ in the Enciclopedia Italiana (1932) vol. 13, 477-480. The Oxford University Press has announced a 3rd edition of Metzger’s The Text of the New Testament.
The reader who peruses Paul Maas’s Kleine Schriften will find a review of Sir W. W. Greg’s The Editorial Problem in Shakespeare, notes on two passages in Richard III and similar items. The lesson is there for those who want to take it: the textual critic cannot devote his attention exclusively to one type of transmission or to one epoch of literature. In a time of exciting and controversial editorial projects outside the field of Classical Studies—the Oxford Shakespeare edited by S. Wells and G. Taylor with its monumental shelf-mate William Shakespeare: A Textual Companion; the Kane-Donaldson Piers Plowman; and the most controversial, the corrected text of James Joyce’s Ulysses edited by H. W. Gabler—it is regrettable that the authors did not cast a glance, however brief, at modern editing and textual criticism. This is now easily accomplished by referring the student to G. Thomas Tanselle’s masterly survey, ‘Classical, biblical, and medieval textual criticism and modern editing,’Studies in Bibliography 36 (1983) 21-68. He concludes with these words: ‘Editing ancient texts and editing modern ones are not simply related fields; they are essentially the same field. The differences between them are in details; the similarities are in fundamentals.’