[In the following essay, Stout offers a publication history of Pliny's Letters and notes that textual criticism on the work from the 1800s sheds important light on the authoritativeness of the source of the manuscripts used by scholars.]

For about 375 years after the death of Pliny the Younger, which probably occurred a.d. 113, his Epistulae were circulated in two corpuses, one containing the nine books of the Letters to His Friends and the other containing in a single book the correspondence between him and the Emperor Trajan.

In a prefatory letter introducing Book i of the Letters to His Friends Pliny says that he has put together for publication some of his letters that had been written with especial care, and that if this venture meets with favor from the public he will look up some that have been omitted and publish them along with others that he may write from time to time. This introduction leads us to think it probable that the first publication of his epistles contained only a part of the nine books of these Letters and that other portions of the collection, as we now have it, followed at intervals. This inference receives some support from his statement in ix.19 that his correspondent Ruso had seen the letter that is published as vi.10 and wished to discuss a statement made in it. When these Letters first come to our knowledge, however, the nine books are together in one corpus.

Since this collection contains no letters that were written after 108 or 109, and since Pliny was not burdened with administrative duties in the years 108-110 and was awaiting assignment as governor of a province, it seems not unreasonable to assume that if these letters had been published in units of one or more books at a time, he may have brought them together and published them as a collection about 109 or 110. Since he did not include in this corpus any of the first fourteen letters of Book x, in spite of the fact that Trajan's name and evident personal friendship for him would have added luster to the collection, we may perhaps also infer that at the time of this publication Pliny already had in mind the idea of publishing his correspondence with the emperor as a separate work and was reserving these letters for inclusion in it.

Pliny was governor of Bithynia and Pontus probably in a.d. 111-113, somewhat less than three years.¹ In the 52nd letter and again in the 102nd letter of the correspondence with the emperor, Pliny informs Trajan that the soldiers and civilians in the province had joined him in offering thanks to the gods on the anniversary of the emperor's accession to the headship of the state and had prayed for his preservation and continued success. These religious observances were on the dies imperi of Trajan, January 27, 112 and 113. It may be regarded as certain that similar sollemnia would have been observed on this anniversary in 114 and reported in another letter if Pliny had remained as governor until that date. There are no letters of governor or emperor in the collection that have in view the termination of Pliny's unusually personal representation of the emperor in the province or the arrival of a successor. This, combined with the fact that no trace of Pliny is found anywhere after this time, makes it highly probable that he died in service in the province before January 27, 114, and leads to the further conclusion that his correspondence with Trajan was published by someone else after Pliny's death. The manuscript may have been found among his papers, ready for publication up to the point where it breaks off, but awaiting additions that were never made because death cut short his administration and the correspondence. Pliny can hardly have planned to have the collection close as we have it at present.

These two works, the nine-book collection of the Letters to His Friends and the one-book volume of the correspondence with the emperor, were preserved separately until near the end of the fifth century. We have no evidence that anyone ever saw the nine-book corpus in that interval, and during the same period the only person that gives any evidence of having seen the other is Tertullian, about a.d. 200. Directly or indirectly, he knew the contents of x.96 and 97, in which the problem presented by the presence of the Christians in Pliny's province is discussed. It is probable on the whole, therefore, that the letters of Pliny were not much read in the late second, the third, and the fourth centuries, and that the making of manuscripts of them was infrequent or that it ceased entirely in these centuries.

In the fifth century we find Apollinaris Sidonius (c. 430-480), toward the end of his life an enthusiastic admirer and imitator of Pliny, intimately acquainted with the Panegyric and the nine books of the Letters to His Friends, but without any knowledge of the correspondence with Trajan.

Sidonius was thoroughly trained in his youth in the schools of southern Gaul, in which a wide range of literature was studied, chiefly from the rhetorical standpoint. His family belonged to the nobility, both his grandfather and his father having been prefects of Gaul. In his early manhood he gave himself to the writing of poetry but did not give up entirely an ambition for political power. His father-in-law, Avitus, was emperor at Rome for a brief time (455-456). Sidonius gained great reputation by the poetic panegyric which he delivered when Avitus assumed the consulship at Rome, January 1, 456.

After the defeat of Avitus, Sidonius continued in favor at times with others who reached the principate and welcomed the support of his pen. He delivered a poetic panegyric when Maiorianus assumed the consulship, January 1, 458. Again, on January 1, 468, he honored Anthemius by a similar tribute, winning such favor that he was at once appointed Praefectus Urbi at Rome for that year. At the end of his year as Praefectus Urbi, apparently discouraged by the rough state of turmoil of the time, he renounced political life for service in the church. His great reputation and his connections gained him a bishopric at once, in spite of his total lack of training for that office. It may be noted that Avitus also was made a bishop at once when he gave up the throne, thirteen years before this. Sidonius was assigned to the see of Clermont, whose territory included his home of many years at Auvergne. From 469 until his death in about 480 he did excellent and courageous service for his people here as the pressure of the Goths in that portion of the empire continually increased.

At the beginning of his bishopric he gave up the writing of poetry. His poetry had been steeped in pagan mythology rather than in Christian lore and imagery. Induced evidently by his enthusiasm for Pliny's Letters, he began in 469 to collect and revise his own letters for publication. These Letters constituted his main contribution to literature from that time until his death. He repeatedly lets it be known that he is modeling his work on that of Pliny. As in Pliny, each of his letters as published is practically confined to one subject. In vocabulary and phrasing, these letters of Sidonius show many reminiscences of Pliny. Many of these Plinian touches may have been added as the letters were reworked for publication. Geisler's list of parallels in Sidonius to earlier writers² notes 337 echoes of Plinian diction and ideas, drawn almost evenly from each of the nine books of the Letters to His Friends, but the list contains not one such parallel with the language of the correspondence with Trajan, which Sidonius clearly did not know. Geisler lists ten possible reminiscences of Pliny in the carmina of Sidonius. This is a very small number, and in none of them is the parallelism close. In most of them Sidonius seems to be echoing Horace or Ovid or Persius or Statius, all of whom influenced him at this period of his life, rather than to be borrowing from Pliny. This suggests that Sidonius may have picked up his nine-book manuscript of Pliny's Letters in Italy late in his life and that he did not know them in his earlier years in Gaul. His late enthusiasm for them and his high standing in literary, political, and ecclesiastical circles were certain to stimulate interest in Pliny as his own Letters were published.

The commercial reproduction of the Epistulae of Pliny was almost certainly given up early in the second century, not to be revived until many centuries later. There is no evidence that they were ever again in demand in the market before the Renaissance. Even in Pliny's lifetime they did not rank with his orations in the book mart nor in the author's esteem. Pliny nowhere seems to think of his Epistulae as literary productions, or to look to them for recognition as an author during his lifetime or for fame in the future. His hope for this was in his orations, although he was tempted at times, without however seriously yielding to the temptation, to venture into poetry or history. Epp. [Epistualae] ix.11 has been thought by some to show that his Epistulae were on sale in Lugdunum about 108 or 109, but libellos in this passage, as again in iv.26, has no reference to Epistulae. In vii.2 Pliny himself sets up a contrast between epistulae and his scripta, for which his correspondent Iustus had asked.

Most of the existing copies would gradually disappear after commercial publication of the Epistulae ceased. Their preservation to the later world depended on the chance preservation of a copy or a few copies that might rest undisturbed in libraries until some Sidonius stumbled upon one of them, found it interesting, and brought it to the attention of others. This would lead to the making of copies that would get distributed more or less widely, and they would thus once more enter the stream of current intellectual life. How many precious works of antiquity failed to come to the notice of an understanding and appreciative discoverer before the loss of the last copy that had been thus laid away!

Of the nine-book corpus of the Letters to His Friends, each of two copies, both apparently in Italy, left a descendant before suffering destruction. From these two copies are derived all the manuscripts of the first nine books of the Letters which either now exist or have left traces of their past existence in descendants known to us. One of these two, which I shall refer to as W, became the immediate or the remote mother of what we now call the eight-book and the nine-book families of the manuscripts of the Letters to His Friends; the other became the immediate or the remote mother of the nine-book manuscript of these Letters which was incorporated with a manuscript of the correspondence with Trajan by a late fifth-century editor, probably in Italy, to form the mother manuscript, Z, of the ten-book family of the manuscripts of the Letters.

The studies of Keil on the sources available for the constitution of the text of the Letters are fundamental. These were published in 1853, 1865, 1866, and 1870. Good scholars, some of them keen critics, had worked at the problem of the text at times for centuries before he began its study. Especially in Italy, in the fifteenth century and the early years of the sixteenth, these critical efforts engaged the interest of many excellent scholars. But their results, although valuable in some respects, serve on the whole to make the problem more complex and difficult for the modern critic. The effort to perfect the text of the Letters continued to interest at intervals some of the best scholars between 1518 and 1853, but they brought the problem no nearer solution. The best text developed through these unscientific efforts before the editions of Keil is found in the edition of Cortius and Longolius in Amsterdam in 1734. This is due to the fact that these editors had at hand the largest collection of manuscripts and collations ever brought together before the editions of Keil and Merrill. It included readings from M and B, important manuscripts that had not been carefully studied before. All the critics and editors before Keil were baffled by insufficient command of the manuscript evidence, and even more by their inability to evaluate scientifically the sources which they had.

The methods of text criticism had been reduced to a rudimentary science by Keil's time, and this enabled him to bring order into the study of the text of the Letters. He was the first to discover that all the manuscripts that have been preserved to our time descend from three originals and therefore fall into three families. This is the greatest contribution ever made to the study of the text. These families he called the nine-book, the eight-book, and the ten-book families, names which still serve well enough in spite of some inaccuracy in the name of the second family.

Since Keil's day several other manuscripts of portions of the Letters have been found and identified, and their places in Keil's three families determined by Merrill, Miss Johnson, and Rand. Merrill has also made available in his edition of 1922 the readings of most of the important manuscripts. His contribution to the study of the problems of the text are of prime importance, second only to that of Keil. The discovery of what is now called the Morgan Fragment and the searching study of it by Lowe and Rand in 1922 have helped to clarify the course of development of the modern text. The student of the text of the Letters today has at his service more and better fundamental information and expert guidance than was ever available before.

In an article in TAPA [Transactions of the American Philological Association], 55 (1924), 62-72, the present writer made the claim that the nine-book and the eight-book families of Keil derive from a common parent. This statement presents the fundamental thesis of the present study, in which the evidence of the manuscripts in many disputed readings is newly evaluated under rules of interpretation that necessarily follow if a common origin for the nine-book and the eight-book families is accepted.

One of the best evidences of a common ancestor for two manuscripts is the omission in both of them of identical portions of the text. The omission of iv.26 and of all the text after v.6 in both B and F shows them to have descended from a common ancestor in which these losses had already occurred. The omission of epistles 8, 12, 23, and 24 of Book i in any manuscript identifies it at once as descending from the common ancestor of the eight-book manuscripts. Frequent instances of the omission of an identical word or group of words of the correct text is sufficient to prove a common ancestor for the manuscripts in which the identical omissions occur. The following five identical omissions in one book of the Letters first suggested to me that X and Y had a common ancestor:

iv.5.4 (94.18)³ quo sit excusatius quod ipsum librum non tamen ultra causae amplitudinem extendimus Z om. XY

iv.7.7 (96.2) risum magis possit exprimere quam gemitum Z om. XY

iv.10.3 (100.11) neque enim minus apud nos honestas quam apud alios necessitas valet Z om. XY

iv.12.5 (103.9) ut ego statim feci Z om. XY

iv.15.11 (107.27) quia votis suis amor plerumque praecurrit Z om. XY.

All modern editors⁴ accept from the Z tradition the added part of the text in the first four of the above readings. The identical omission of exactly the same words in X and Y in each of these four cases, without similar words or endings to induce the omissions and to limit their extent, can not be plausibly explained except on the assumption that X and Y are from the same ancestor, in which these omissions had already been made. The added text in the fifth reading cited above is accepted by all modern editors except Keil. For critics who accept as original the words omitted in both X and Y in this reading, they are added proof that X and Y are from the same parent manuscript, in which the omission had already been made.

As stated above, my first conviction that X and Y are merely two branches of a single tradition was based on their identical omissions in these five cases. Further study of the character of the Z tradition, however, has convinced me that the added words in each of these five readings are editorial additions in the ten-book tradition and do not belong in the original text of the Letters.⁵ If I am right in this, these common omissions are not omissions at all; they illustrate the reliability of the united testimony of X and Y to the text and fit in perfectly with the assumption of a common origin for X and Y, but do not prove it. That can be proved only by errors and peculiarities that are traceable to a common ancestor. To establish the common origin of the X and Y families it will now be necessary to examine other readings in which X and Y have an identical incorrect reading.

The three following apparent omissions deserve consideration:

ii.17.17 (57.3) cryptoporticum xystus violis odoratus; teporem solis infusi repercussu cryptoporticus Z om. XY

ii.19.5. (60.18) cum labore etiam cum labore Z om. XY

iv.12.3 (103.4) quaestio sed tamen quaestio Z om. XY

The words omitted by X and Y clearly belong in the text. The fact that all three omissions are due to jumps made from a word to the same word further along makes the error in each case an easy one. In any one case the assumption that the jump was independently made by two different scribes need not be felt to be inherently impossible, but that such an identical jump should be made in three different cases by two scribes copying the same text independently is highly improbable. No such list could be made of identical jumps by scribes of X and Z or of Y and Z. I doubt whether a parallel could be found in the work of independent scribes in any author in an equal amount of text. It seems necessary to assume that the three omissions go back to a common ancestor in the history of the X and the Y traditions in which the loss had already occurred.

The omission of single words, especially of short ones, occurs occasionally in the work of scribes, but if two scribes agree frequently in the omission of identical words at the same point in the text, the omission must be assumed to go back to a common ancestor which did not have the omitted words.

The X and the Y families agree in omitting many single words that are found in the Z sources, but in practically all these cases the consensus of critics has been gradually recognizing that these words do not belong in the original text but have been contributed by editors in the Z tradition. For example, in the ten-book manuscripts an editor has frequently completed the construction for Pliny by adding connectives such as et or -que or ut, pronominal forms such as hic or suus or meus, prepositions such as a or ex, and various forms of sum. The fact that X and Y agree in not having these words that Pliny did not place in the text but which have been supplied editorially in the ten-book manuscripts furnishes evidence of the high percentage of validity for readings which are attested by both X and Y, and is consistent with, but does not prove, the assumption that X and Y are from a common source. For editors who accept many of these “improvements” into the text, however, they do prove that X and Y have a common source. This is also true of the following three readings:

v.6.40 (130.23) fistulis om. XY add. Z

iv.7.2 (95.12) librum om. XY add. Z

iv.11.9 (101.18) cubiculum om. XY add. Z

The first of these is accepted into the text by all modern editors, the second and third by all except Keil and Schuster. For those who accept them, their omission in both X and Y practically proves a common origin for X and Y; but I think that all three words were supplied editorially in the ten-book tradition, probably by the original editor of the first manuscript of the tradition. If I am right in this, their omission in X and Y is but another proof of the faithful witness to the text found in W, their original source.

Another suggestion of the common source of X and Y is found in the form of the names of the addresses in the Letters.⁶ The parent manuscript, Z, of the ten-book family evidently gave both the nomen and the cognomen of the persons addressed in the Letters where the editor had been able to determine them. The nine-book and the eight-book manuscripts regularly give one name only, sometimes the nomen, sometimes the cognomen; but whichever is used, they uniformly have the same. Their practice in this represents the usage which we should expect of Pliny himself in the original correspondence and in his own publication of the Letters. In the 212 letters in which we have readings of both the X and the Y traditions, the only exception to the broad general statement as to the agreement of X and Y in the superscriptions that can not readily be seen to be the result of scribal error is in the superscription of v.10, where Y preserves the reading of W in using Tranquillus, while some scribe of the X tradition has consciously or unconsciously substituted the nomen Suetonius of this well-known correspondent. In all the other letters addressed to Suetonius both X and Y have only the superscription Tranquillus.

The final and crucial test of the hypothesis that X and Y are derived from the same text, W, of the Letters comes when the evaluation of the manuscript evidence which this hypothesis makes necessary is applied to the variant readings throughout the Letters. If the application of this necessary rule for evaluating the manuscript evidence gives correct readings, conviction that it is correct must grow as we proceed, and the hypothesis must be accepted. …

If X and Y are independent derivatives from W, and Z is independent of both X and Y, (1) where X and Z agree and Y differs from them, X must be assumed to have preserved the text of W, and there is thus no real manuscript evidence for the reading of Y; the reading preserved in both X and Z carries the only ultimate manuscript authority for the text; and (2) likewise, where Y and Z agree and X differs from them, the concurrent reading of Y and Z carries the only ultimate manuscript authority for the text.

The first of these two rules for evaluating the manuscript readings at any point has been observed from the time of Keil's studies in 1865, 1866, and 1870, resulting in great improvement of the text of the Letters. The second has never been applied in the determination of the text, although it has generally been recognized that the agreement of Y and Z against X brought the reading of X into some doubt and required careful consideration. (See Keil, 1870, Praefatio XXVII.) The full acceptance of the rule is decisive for the text in hundreds of places.

Since the publication of Keil's text, additional manuscript evidence has been made available, but a faulty interpretation of this evidence, based upon Otto's unscientific article⁷ in Hermes (1886), carried the editors Müller (1903), Kukula (1908), and Merrill (1922) progressively farther astray, introducing about 400 unjustified changes into the text of Keil. Throughout this period good scholars have published careful studies of a total of hundreds of readings of the Letters. By the methods of logical criticism, applied with constantly improved tests for checking their results, such as a knowledge of Pliny's clausulae and new studies of idiosyncrasies of his style, many of the errors of the editors already mentioned have been convincingly corrected. So many of these now generally accepted corrections are verified by the application of our rule 2 as to give strong proof of its correctness. This in turn gives support to, amounting practically to proof of, the hypothesis upon which the rule is founded, namely that X and Y are merely two branches of one tradition, W. …

Notes

1. His administration covered almost three years. From Epp. x.42, 61, 62, and 77 it is known that Calpurnius Macer was governor of Moesia Inferior during Pliny's first and second years in Bithynia. From CIL III 777 Macer is known to have been in Moesia in 112, and from CIL III 12470 that he had a successor in Moesia by 116. See Stout, The Governors of Moesia, Princeton, 1910, pp. 62-63. The dates assumed for Pliny's governorship in Bithynia, 111-113, may not be entirely accurate, but they are not far wrong.

2. Geisler's list is published at the end of Luitjohn's edition of the poems and letters of Sidonius in Monumenta Germaniae Historica VIII 351 ff.

3. References in parentheses are by page and line to the text of Merrill's editio maior of the Letters, 1922.

4. “All modern editors”: In the last eighty-one years six critical editions of the complete correspondence of Pliny have been published in which reexamination of the text has been a prime purpose of the editor. They are the four Teubner editions by Keil, 1870, Müller, 1903, Kukula, 1908, Schuster, 1933, and the editions of Merrill, 1922, and Mlle Guillemin, 1927. Throughout this study the expression “all modern editors” is to be interpreted as referring to these six critical editions. Reference will occasionally be made to other editions, complete or partial, published in this same period, in which a careful restudy of the problems of the text seems to have been only a secondary interest of the editor.

5. The discussion of the genuineness of these readings is deferred until further preparation for it has been made by some study of the history of the text. All the readings referred to in this introductory statement will be discussed in detail later.

6. An interesting detailed study of the names of the addresses in the manuscripts is given in Barwick, “Zwei antike Ausgaben der Pliniusbriefe?” Philol. 91 (1937) 423-448, and in an earlier article by Robbins, “Tables of Contents in the MSS of Pliny's Letters,” CP [Classical Philology] 5 (1910) 476-480. My own study of the matter discussed in this paragraph was made in 1926.

7. For a discussion of the methods and conclusions of Otto see Stout 1926.5-31.

Bibliographical References

Barwick, Karl. “Zwei antike Ausgaben der Plinius Briefe?” Philol., 91 (1936), 443-448.

Guillemin, Anne-Marie. “Quelques remarques sur la critique du texte de Pline le Jeune,” Rev. de Philol., 49 (1925), 93-100.

Keil, H. C. Plini Caecili Epistularum libri Novem, Epistularum ad Traianum liber, Panegyricus, editio minor, Leipzig, 1853; and editio maior, Leipzig, 1870.

Merrill, E. T. C. Plini Caecili Secundi Epistularum Libri Decem. Leipzig, 1922.

Schuster, M. Studien zur Textkritik des jüngeren Plinius. Vienna, 1919.

———. “The Mind of the Scribe,” CJ [Classical Journal] 22 (1926-1927), 405-417.