There are many mistakes in B: groups of letters that make no sense, words without breathing marks or accents, and, above all, numerous omissions indicated by a blank space that more or less correspond to a part of the text the scribe was unable to decipher. Some lacunae have been filled in by a second copyist (B2), who corrected many mistakes or readings he considered erroneous by writing above the lines, in the margins, or by erasing. B2 can be dated paleographically to the first half of the twelfth century, shortly after B. As for the text consulted by B2, an overall examination of its contents leads to the conclusion that it was Ω, the same one used for B. We can presume the anonymous corrector had accurately collated B with Ω, and had been able to read and understand much more than his predecessor. Evidence supporting this hypothesis can be found in the cases in which B2 was unable to decipher the text completely and where he leaves the same blank spaces to indicate that a word or phrase was obscure or incomprehensible to him. The anonymous B2 was a person with a good knowledge of the Greek language and philosophy and was, at the same time, a meticulous reader of Diogenes’ Lives.
Manuscript P (eleventh/twelfth centuries) was also directly copied from Ω. The occasional omission of a phrase is the result of a saut du mēme au mēme1 and is not sufficient in itself to prove that P descended from Ω through a lost intermediary text. P, even in its early phase (before it had been corrected), transmits the text of Ω in a form that is less faithful than B, revising it frequently based on conjecture.
P was corrected on various occasions by several successive hands (perhaps six), indicated by the sigla P2 to P7. The main job of correction was done by P4. This anonymous scholar (whose work can be placed, based on paleographic evidence, in the first decades of the fourteenth century) made many corrections, erasing previous readings and noting alternative readings in the margins.
P has been the most influential manuscript in the Laertian tradition. It is closely related to at least eight other codices, dating from the beginning of the fourteenth century to the first decades of the sixteenth. All of these manuscripts descend directly or indirectly from P and bear witness to the different stages of P’s transmission.
The position of F (thirteenth century) is more controversial. One peculiarity is its frequent omission of more or less extensive passages, often filled in by a later corrector (F2) thanks to a subsequent collation with another, identifiable version of the Greek text. Some omissions in F come undeniably from sauts du mēme au mēme. Elsewhere, beginning in Book 5, and particularly in Book 7, the names of the authors and titles of the works quoted by Diogenes are left out. I would not exclude the possibility that this last particularity mirrors a personal choice made by the anonymous copyist of F, a scribe who may have been indifferent to the expertise of scholars. Some omissions of words or phrases may also have occurred at the whim of F’s copyist. Another specific characteristic of F is the frequent alteration of the word order in an attempt to render the text more fluent by cleaning up real or perceived roughnesses or anomalies. The corrections and adjustments of vocabulary and style respond to similar promptings. We are dealing with conjectural readings done by a copyist. Yet F preserves many readings superior to those transmitted by B and P (and Φ). Not all of them can be explained as lucky conjectures; at times it is necessary to concede that F is the only codex to have preserved traces of the original text that were corrupted in B and P (and Φ).
I conclude that F was not copied directly from Ω, but through an intermediary text (γ) that was copied from Ω. This source (γ) was contaminated by readings from the vulgate tradition (α), which limits its importance for establishing a definitive Greek text of Lives.
The moment has come to say something about the process of formation of the vulgate tradition (α) and to present some thoughts on its initial diffusion. This tradition is partially represented today by codex V (the oldest descendant of α, preserved up to 6.66). It was produced, perhaps as early as the mid-twelfth century, from a fusion of the P tradition and the lost model of F (γ), a manuscript (now lost) of hybrid character, the bearer of a vulgate text (α) that was extensively revised, interpolated, and corrected, but which can nevertheless be traced back to Ω, the common ancestor of the oldest extant continuous manuscripts. We know that V, at least, was modeled on α; but we cannot exclude that α is the direct or indirect source for other lost manuscripts, or of manuscripts that have been preserved but insufficiently researched. In any case, V’s contribution to the reconstruction of the original Greek text of Lives is extremely limited.
A handful of more recent manuscripts (fifteenth and sixteenth centuries), not yet systematically explored, transmits a version of the Greek Lives that is extremely contaminated and adulterated. Among these I will only point out the codex of Prague, Národní Knihovna, Raudnitzianus Lobkowicensis VI Fc 38, late fifteenth century (siglum Z), as it served as the basis of the editio princeps, the first printed edition of the Greek text of Lives, edited by Hieronymus Frobenius and Nicolaus Episcopius, and published in Basel in 1533. This edition therefore disseminated a Greek text of Lives that is entirely unreliable, often corrupted and interpolated, and which subsequent editors (at least until Cobet) have tried to correct through more or less plausible personal conjectures, or by a nonsystematic recourse to other manuscripts.2
I now move to the group of manuscripts that transmits one or both collections of the excerpts; these are usually known as Excerpta Vaticana, because their common ancestor is a manuscript now in the Biblioteca Apostolica Vaticana (Vaticanus gr. 96 = Φ).
The presence of numerous errors common to Ω and Φ proves that they descend from a common ancestor (Χ). Their lineage was not direct, but rather passed through two lost links: a transliterated transcript of Χ, χ (dated around the tenth century), and the autographon excerptoris (dated around the eleventh century). The indirect descent explains why the text of Φ is often inferior to Ω (at least in its first phase, reflected by B and P1). It is also probable that Φ preserved inferior readings in Χ, which were corrected in Ω. Still, among Φ’s variants there are many that offer a text superior to Ω, a phenomenon that can be explained by presuming that Φ conveyed Χ more faithfully than Ω did. Finally, we cannot exclude the possibility that some conjectures or revisions were made by the anonymous scholar who organized the two collections of the Excerpta Vaticana (one of which, Φh, was put together using, in addition to the Lives, many passages from the Suda), or that some variant readings reflect contamination from external sources (parallel gnomological and lexicographic traditions, etc.).
Finally I come to the excerpts in the Suda, the Greek Anthology (Pal), and those preserved in codex Vi.
Extensive anonymous extracts from the doxographic sections of Diogenes’ Lives reached the Suda (c. 975–980) through a lost anonymous collection of Philosophica (Σ) with an uncertain date. The biographic sections of Lives were transcribed by the editors of the Suda principally from a summary (made between 829 and 857) of a work of Hesychius of Miletus (sixth century) entitled Biographical Dictionary of Learned Men. The undeniable similarities between the biographic entries in the Suda and those in Lives could be explained by supposing that Diogenes and Hesychius had access to a common source. There are sporadic cases, however, in which the Suda (or its source) supplemented Hesychius’ text with additional information taken from Diogenes Laertius.
Thanks to the use of Constantinos Cephalas’s Anthology (c. 900) as an intermediary source, the Greek Anthology includes an ample selection of poetry composed by Diogenes Laertius himself or by other authors quoted in Lives. This anthology is preserved in the codex of Heidelberg, Palatinus gr. 23+Paris, Parisinus suppl. gr. 384 (mid-tenth century). Parts of these poems were used in the Planudean Anthology, put together by Maximus Planudes (c. 1260–1305). The original is preserved in the codex of Venice, Marcianus gr. 481 from 1299, as well as two authentic originals corrected in 1300 or 1305 under Planudes’ supervision: London, British Library Add. 16409, and Paris, Parisinus gr. 2744.
The numerous common errors between Σ and Ω in this case also derive from a common original source, namely X. We can also speculate that the same source was used for the poetic compositions extracted from Lives that reached Pal by way of the codex of Constantinos Cephalas’s Anthology.
Suda and Pal can thus substantially contribute to the reconstruction of text Χ, from which they derive by way of distinct intermediary manuscripts (Σ for Suda, and the codex Cephalae for Pal).
Vi, the most ancient (and brief) document of the Laertian tradition, transmits a text that does not present significant changes with respect to the variations we have seen in Ω and Φ.
I will conclude with a few words about the codices that transmit the Vita Platonis. The third book of Lives has enjoyed its own independent dissemination in four manuscripts that preserve it in its entirety, and in at least three others that preserve excerpts. The continuous manuscripts, copied in Constantinople in the middle of the first half of the fourteenth century, include the corpus Platonicum (more or less complete) accompanied not only by Diogenes’ Book 3 (Plato’s life), but also by other introductory texts to the Platonic Dialogues. All of these codices directly or indirectly derive from P, and are thus not useful for reconstructing the Greek text of Book 3.
Modern scholars have struggled for more than two hundred years to establish a reliable edition of the Greek text. This has proved difficult because there has been a sharp disagreement over the relative value of the many different manuscripts that have been preserved.
In his 1887 introduction to his Epicurea, the German classicist Hermann Usener attempted the first systematic investigation of the manuscript tradition of Diogenes’ Lives. He concluded that the three oldest continuous manuscripts (B, P, and F) were by themselves sufficient to establish the text.
Edgar Martini disagreed with Usener’s depreciation of the value of more recent manuscript sources. We can summarize Martini’s conclusions as follows: two types of manuscripts are independently derived from a common archetype. The first class (α) includes more recent manuscript sources that must, in turn, be subdivided in two subclasses (γ and δ); the second class (β) is made up of the oldest manuscripts. Since P and F contain common errors they share with B (the more genuine representative of class β), or with the testimonies from class α, the two manuscripts P and F are considered by Martini to be representatives of a “mixed” text. The contamination was explained by supposing that the (lost) progenitor of class α was created at an ancient date (eleventh century), preceding all of the preserved manuscripts.
Despite severe criticism from the German scholar Alfred Gercke, who questioned the need to postulate the existence of a class α, Martini defended his reconstruction, insisting that the entire textual tradition depends on two exemplars (α and β) that are descendants of a single ancestor. Gercke remained skeptical, and reiterated his objections against the existence of the two classes α and β, maintaining, with new arguments, that only codices B, P, and F are necessary for establishing the Greek text of Lives.
The Italian scholar G. Donzelli shared Gercke’s disparaging judgment of the later, vulgate manuscript sources for Diogenes. She showed that the contribution of these later manuscripts to establishing the text of Lives is negligible, and that the few good readings they contain are nothing more than the conjectures of Byzantine and Renaissance scholars.
Armand Delatte and Ingemar Düring have also declared themselves in favor of the oldest tradition. Delatte called attention to the excerpts presented by the Suda and the extracts preserved in Φ that had been discovered in the meantime by Martini. According to Delatte, Φ and the Suda were the only two witnesses to the oldest, archetypal text (what has come to be called X). Düring, on the other hand, minimizes the importance of the Suda, and maintains that Φ comes from the same ancestor as B, P, and F, but that it constitutes an independent branch within the tradition.
After Martini, the study of the excerpta Vaticana was continued by Long, Biedl, and Tartaglia. Biedl, in particular, has proved that Φ is the progenitor of all the other sources of the excerpta Vaticana, and that it represents a branch distinct from that of the continuous manuscripts B, P, and F.
In the preface to his edition of the Greek text (1964, 1966), Herbert Long acknowledges the superiority, when compared to the later vulgate, of the three oldest manuscripts, B, P, and F, but he places them all on the same plane, without investigating the reciprocal relationships among them and without considering the excerpta Vaticana (Φ). According to Long, the Laertian manuscript tradition derived from an archetype that contained variant readings and was irremediably contaminated. This contamination made it impossible to ascertain the relationships between the manuscripts.
On the basis of new criteria, Denis Knoepfler reexamined the question and arrived at a different and more plausible conclusion.
It was he who first suggested that the three most ancient continuous manuscripts (B, P, and F) had a singular model (Ω). Φ, though related to Ω, is not one of its descendants; Φ and Ω do, however, refer back to a common model, Χ. Ω thus assumes the role of subarchetype when compared with the actual archetype Χ. Certain divergences between Φ and Ω allow us to establish that while Ω directly resembles Χ, for Φ we must presume the existence of at least a lost intermediary model (χ). B occupies a significant position among the continuous manuscripts because it derives directly from Ω. Despite its numerous minor errors, B transmits a text that is not yet contaminated (apart from the corrections of B2) by the revisions that F and P were subject to (through the two lost models ω and ω’). B is superior not only to the so-called vulgate manuscripts (α), which are contaminated by the readings of ω and ω’, but also to the two other continuous manuscripts: P (to which it is closely related), and F, a manuscript with a text that is more problematic.
***
Knoepfler for the first time also hypothesized that there were two different branches of the manuscript tradition: an Italo-Greek branch, represented by the manuscripts B, P, and F (and their descendants), and an Eastern branch, represented by codex Φ copied in Constantinople before the dissemination of the vulgate. Φ (like B, P, and F) descends from Χ, but through a lost continuous manuscript transliterated and corrected in numerous places (χ). χ in turn indirectly becomes a basis of Φ around 1100. The Eastern tradition would not have been entirely unknown in the West, where it had an indirect influence on the vulgate. The vulgate in turn would be disseminated in the East through P, after the codex was corrected based on α (version P4). From the East, P was brought again to Italy by Guarino da Verona (c. 1370–1460) and produced (at Florence) H, one of the Greek texts used by Camaldolesi Ambrose Traversari for his Latin translation.3
My own philological analysis, however, has proven that Ω and Φ, as well as Vi, Suda, and P, can all be traced (directly or indirectly) to a single textual source, Χ, preserved, it would seem, in the area of Constantinople. This idea finds confirmation in a study of the paleographic and codicologic characteristics of the older descendants of Ω.
The oriental or Constantinopolitan origin of Φ, Vi, Pal, and Suda is undeniable. The hypothesis that Ω was an old codex preserved in a library in southern Italy depends on the assumption that B, P, F, and V were all written by Italo-Greek hands. New studies of the written items attributed to southern Italy prove that B alone is of Italian origin (or at least copied by a scribe trained in southern Italy). The paleographical and codicological peculiarities of P, F, and V lead to the conclusion that all three were written in Constantinople, or at least in the oriental area.
Since the text of B does not altogether differ from that of P and F, or from their common ancestor (Ω), we may further deduce that B does not transmit a distinctively Italo-Greek branch of the manuscript tradition. Two hypotheses could explain the presence of B in southern Italy (or Sicily). First, that B was produced by an Italian copyist in Constantinople and then brought to Sicily. Second, that after P and γ (F’s model) were copied, Ω traveled from the Eas
t to the West and remained in Sicily at least for the time necessary for B to be transcribed and, a bit later, for Ω to be used by the anonymous B2 to carry out numerous corrections and supplements to the earlier B. If one considers, in particular, the material in B2, the second hypothesis seems more plausible. In both cases, however, the terminus ante quem for model B’s arrival in Italy is the date of B2, the first half of the twelfth century.
The dissemination of Lives is thus unitary and fundamentally oriental. From the same textual source (Χ) there derives, on the one hand, the ancestor of the continuous manuscripts (Ω) and, on the other, the ancestor of the manuscripts of excerpts, now by Φ.
***
For my new edition of the Greek text of Diogenes’ Lives, I reexamined the transmission and provenance of the various manuscripts. I collected the principal ancient manuscripts (B, P, F, Φ, and Vi) and the probable (partial) progenitor of the vulgate (Vaticanus gr. 1302, early fourteenth century: siglum V). I considered the readings in some recent manuscripts, and consulted previous studies and editions. Finally, I considered the evidence in the indirect tradition, in particular the Suda and the Greek Anthology.
I established my Greek text of Diogenes’ Lives on the basis of the oldest continuous manuscripts (B, P, and F), the excerpta Vaticana (Φ), and the other Byzantine excerpts in the Suda, Pal and Vi.
B and P (before correction) are the two codices integri antiquiores most worthy of trust. B often provides us with a text closest to Ω because its scribe had copied Ω in a mechanical way. P, despite being contemporaneous with (or a bit earlier than) B and also directly derived from Ω, transmits a less pure text because it had already been deliberately altered. F is a source that should be used with more caution because of its peculiarities and its contamination by the vulgate tradition. In its original state (F1), however, it can contribute, along with B1 and P1, to the reconstruction of the archetype of the continuous manuscripts (Ω). I have taken into account the readings in the manuscripts derived from the vulgate (α) and from other more recent sources sporadically, when their text seemed to me superior (primarily on grounds of conjecture) to that transmitted by B, P, F, Φ, and the other ancient sources (Suda, Pal, and Vi).
Lives of the Eminent Philosophers Page 75