by Federica Nicolardi
From Defining Authorship, Debating Authenticity: Problems of Authority from Classical Antiquity to the Renaissance
Edited by Roberta Berardi, Martina Filosa, and Davide Massimo
© 2021 Walter de Gruyter GmbH, Berlin/Boston
NOTICE: THIS WORK MAY BE PROTECTED BY COPYRIGHT
YOU ARE REQUIRED TO READ THE COPYRIGHT NOTICE AT THIS LINK BEFORE YOU READ THE FOLLOWING WORK, THAT IS AVAILABLE SOLELY FOR PRIVATE STUDY, SCHOLARSHIP OR RESEARCH PURSUANT TO 17 U.S.C. SECTION 107 AND 108. IN THE EVENT THAT THE LIBRARY DETERMINES THAT UNLAWFUL COPYING OF THIS WORK HAS OCCURRED, THE LIBRARY HAS THE RIGHT TO BLOCK THE I.P. ADDRESS AT WHICH THE UNLAWFUL COPYING APPEARED TO HAVE OCCURRED. THANK YOU FOR RESPECTING THE RIGHTS OF COPYRIGHT OWNERS.
The particular composition of the library in the Villa dei Papiri at Herculaneum led scholars in the past to believe in the existence of autographs among the carbonized scrolls.1 In particular, the presence of copies reflecting different editorial stages of Philodemus' works suggested the identification of his personal library in that in the Villa, as well as the identification of the manus Philodemi in some provisional editions of his works.2 Although these theories have been gradually put aside, modern scholars still question the differences between provisional and fair copies in the collection. Double copies of the same work in the library in the Villa usually differ in their subscriptio: Philodemus' On Rhetoric Book 2, for example, has been preserved in two different rolls, whose final parts -- the so-called midolli -- can be found in P. Herc. 1672 and 1674; in the end title of P. Herc. 1674, after the genitive [x] we simply find the title of the work, [x], whereas in the subscriptio to P. Herc. 1672 after the author's name we can read the expression [x]. The meaning of the term [x] in the end titles of the Herculaneum papyri is still debated.3 A milestone in this question is represented by the palaeographical survey conducted on the Herculaneum papyri by G. Cavallo. Following the analysis of formal features of the rolls, especially handwriting and text layout, he concluded that the term [x], which we find in rolls where he detected inelegant handwriting and irregular layout, can only refer to a provisional copy, intended for a limited circulation.4
The greatest difficulty in the interpretation of this term arises from the re-reading of the subscriptio to P. Herc. 1427, which preserves the last ten columns of the only surviving copy of Philodemus' On Rhetoric Book I. The end title of this papyrus had been long read as follows: [x]. In line with the absence of the term [x] and in the light of the 'tecniche librarie e grafiche accurate' which can be observed in P. Herc. 1427, Cavallo did not spot any substantial differences between this papyrus and those who preserve definitive editions.' In 1995 D. Delattre, by rereading the subscriptio in the original papyrus, correctly identified the term [x] instead of the genitive [x]; the correct reading of the end title of this roll being [x].6 Nevertheless, the non-inelegant handwriting and the regular layout of the text in this papyrus would seem to be in conflict with the identification of a provisional copy. From this difficulty some new interpretations of the term [x] have arisen. To mention just one of the most interesting views, T. Dorandi, partially in line with Cavallo, has claimed that the previous editions of On Rhetoric Books 2 and 3 (and presumably also the previous edition of Book 1) were intended for limited circulation in the school, whereas the definitive editions were intended for publication. Furthermore, according to Dorandi, the provisional copies would already have in themselves a definitive text and they would exclusively differ from the fair copies with respect to the formal characteristics of the rolls.7
Throughout the study of the text in On Rhetoric Book 1, the analysis of the corrective interventions has turned out to be particularly interesting in order to shed some light on the editorial stage of the surviving copy. It is well worth making some clarifications about this roll; every remark regarding formal features such as handwriting, text layout, corrective interventions, must necessarily derive from the observation of the final part of the roll, the midollo, P. Herc. 1427, that is the only extant part obtained by continuous unrolling by means of Piaggio's machine. By the time of this mechanical opening, in 1786, most of the original volumen, i.e. circa 95% of the total extent, had already been removed by the so-called scorzatura. This opening method, when used as a preliminary step to prepare a roll for Piaggio's machine, consisted of making cuts along the length of the volumen, in order to free the so-called midollo from its external parts, which were the most damaged and difficult to open. The scorze were then put aside: this caused the loss of connection between the midollo and the corresponding scorze, which were inventoried separately and later opened by scraping off the layers.8 Moreover, this scraping led to the complete loss of each layer, except for the ultimo foglio, that is to say the last leaf: once the 'unrollers' (svolgitori) had transcribed the text they could see, by drawing it they scraped it off in order to reveal the layer below. As a consequence, most of the roll has been lost or only preserved in drawings, with the only exception of the midollo P. Herc. 1427 which, in spite of its small extent (just over 80 cm), allows for some interesting remarks. In the midollo I have identified more than 40 corrective interventions. This number is particularly relevant if we take into account the small amount of text preserved in this papyrus (8 entire and 2 partial columns).
Corrections have been mostly made by the main scribe, sometimes by an editor of the text, whom I identified due to his different -- sloping and sometimes more cursive -- handwriting.9 The most frequent errors are mechanical ones; the scribe has often skipped some text portions, which have been restored -- by him or by the editor -- supra lineam, in the upper or lower margin, or in the intercolumnium; we can also find numerous incorrect word endings, since adjectives, participles, and nouns often erroneously agree with the preceding word. Most of the time the scribe has immediately noticed these oversights and corrected them in scribendo: it is clear from the fact that the correct sequence appears within the line, immediately after the mistaken one, which is deleted.10 Just to mention an evident example of correction in scribendo, in col. 3 ll. 11-13 Sudhaus11 (Fig.
1) the scribe, instead of writing the adverb [x] with which a new sentence begins, has written the participle [x], probably because of the presence of the accusative [x] right before it:12
[x]
As soon as he realized the error, he did not just replace the wrong sequence [x] with [x], but he decided to delete the entire word (by means of dots above the letters and a horizontal stroke) and to leave a vacuum before writing the correct word, in order to mark the end of the sentence before the adverb. This shows that the scribe is very interested in clarity, even if making comprehensible corrections may represent an uneconomic use of space.
As for the editorial history of the text, some interventions, which can hardly be recognized as corrections of non-mechanical errors, have turned out to be the most interesting.
1. In the penultimate column of the text (col. 6 1. 13 Sudhaus, Fig. 2),13 it is possible to read the infinitive [x], in which the sigma is added supra lineam and the theta is written over a pre-existing tau: the text ante correctionem must have been [x]. The correction is required by the presence of a verbum dicendi at 1. 10, which governs [x]. Because of the form of the letters and the corrective modalities used, the intervention seems attributable to the main scribe and so it was probably made not long after the first copying of the text. In this particular case, we may perhaps think of a mechanical error; however -- unless we imagine that the scribe controlled the text on the antigraph, possibility which obviously cannot be excluded for this case and for the following ones -- the correction would have required significant attention to the syntax, which seems hardly attributable to the awareness of the scribe, although he is materially responsible for the intervention.
2. Another -- albeit more complex -- syntactic modification is in coll. 21. 39-2 l. 2 Sudhaus (Fig. 3),14 where the expression [x] occurs. By the observation both of the papyrus and of the Oxonian drawing (disegno) in which the column is reproduced, the presence of many deleted letters is immediately evident. While the text post correctionem is easily readable, the text ante correctionem is harder to recognize, since the traces of ink by means of which the letters are deleted cover them up almost totally. The first deleted letter is easily readable: it is an epsilon, clearly due to diplography ([x] instead of [x]); this mechanical error is not relevant to the present analysis. By comparing the papyrus to the disegno it is possible to notice that in the former we can partly glimpse the form of the deleted letters and that the disegnatore was not always able to distinguish the ink used by the scribe to trace the pre-existing letters from the ink used to delete them. After the eta two other letters are deleted, which in the disegno may seem [x]. By looking at the papyrus under the microscope it is possible to identify the lower parts of these two letters, which do not coincide with those drawn with uncertainty by the disegnatore: a vertical trace followed by an oblique descending rightwards from mid-height; a round outline at the bottom left of the following letter. These traces are compatible with [x]. After alpha, an ascending trace slightly oblique is what is left from another deleted letter; this trace is covered up with an ink circle, that the disegnatore confused with the body of a phi (which would be too small and high, though). It is possible, therefore, that the text ante correctionem was [x] and that the scribe was actually writing [x]. The my, only partially traced, would be the last letter written by him before he stopped and made the correction ([x]). In this section of Book 1 Philodemus is talking about the possibility, claimed by some, to consider as a perfect expert of rhetorical art ([x]) the one who has practised all the rhetorical [x], i.e. all the rhetorical genres. If my hypothesis on the intervention is correct, it will be remarkable that the expression ante correctionem, with the active verb [x] governing an object ([x]), does not substantially differ in meaning from the middle form [x] taking the dative [x]. The transition from one form to the other seems to represent a stylistic improvement rather than a mere correction and this does not enable to consider this intervention as conceived by the scribe.
Other interventions reveal lexical modifications, of which I offer here two examples.
I. In fr. 4 ll. 3-5 Sudhaus (Fig, 4), Philodemus is introducing a distinction between [x] based on their different need for natural ability, practice and method: [x] ... 15 The sequence [x] is written supra lineam and replaces some letters deleted by means of a horizontal stroke. By observing the disegno, the deleted sequence may seem [x], but if we look at the papyrus it is possible to understand that the disegnatore was not able to detect the horizontal stroke and that he reproduced only its final part, which looks like the middle stroke of eta: the correct deleted sequence is [x] ([x]). Therefore, the text ante correctionem was [x], which is unlikely to be considered as a mechanical error, unless we imagine a confusing state of this passage in the antigraph, perhaps with variants or unclear corrections. The term [x] is attested in Phld. Poem. I. P. Herc. 460 col. 125 l. 3 Janko, and here it would not be out of context, since it refers to a 'division', 'classification'.
2. In col. 61. 34 Sudhaus (Fig. 5),16 the multispectral images and the observation of the papyrus under modern microscopes make it possible to read more than what is reproduced in the disegno and published in the previous editions. As for the text resulting from the correction, in 1872 Gomperz had already correctly identified the genitive [x].17 Part of this word, the sequence [x] is written supra lineam above deleted letters. Philodemus is talking about weak argumentations and, before leaving this theme in order to conclude the book, he claims that many other defects spread out through deductive argumentations ([x]), but then adds that there is no point in insisting on their weaknesses, since they are immediately evident. As for the text ante correctionem, the first deleted letter is certainly a my; looking at the second one, only its right part survives, an oblique trace descending from left to right, compatible with a my: after this incomplete letter the sequence ov is deleted. Therefore, the text ante correctionem was probably a form of the term [x], with the meaning of 'coherence', 'permanence' ([x]) . As in the previous cases, neither the error (the term [x] instead of [x]) nor the emendation can be easily attributed to the scribe's full awareness. Plural forms of [x] are not usual and it is not possible to immediately clarify what the text ante correctionem could have meant. The reference, which I have mentioned, to 'insistence', 'perseverance' in the sequent line may be of interest, although the term used there is [x].
The corrective interventions presented above have a relevant common characteristic, which I have already underlined: although the scribe actually and concretely made the corrections, the conceptual responsibility for the interventions cannot be attributed to him. We find syntactic modifications as well as stylistic improvements, which were carried out during the drafting of the copy or not much later and which reveal in some ways the intention of the author of the text.18
As a result, I believe it is possible to conclude that the first book of Philodemus' rhetorical treatise is preserved in a provisional copy, whose text has not been fixed yet and which might have been at least partially written under Philodemus' supervision.19 In this scroll, some changes and the way these changes have been made by the scribe reveal the intention to facilitate the preparation of the future definitive copy. From this perspective, also the scribe's care to make clear corrections rather than save space and intervene discreetly and 'noninvasively' makes more sense.20
Another element related to the stichometric total recorded in the subscriptio of P. Herc. 1427 leads back to the same conclusion. After the name of the author, the title of the work and the indication of the book number, we can read the total of stichoi copied in the roll: [z] XXXX, 'Total number (scil. of stichoi) 4000'. As we know, the stichos is the unit of measurement for ancient works and it usually corresponds to a Homeric hexameter (circa 15-16 syllables). However, the equivalence between stichos and real line is actually not mathematically certain and can vary from roll to roll.21 Determining the actual ratio between stichos and real line in the roll of On Rhetoric Book 1 is the key to understanding stichometric annotations. In this book, besides total stichometry which we find in the end title, there are annotations referring to partial stichometry too: in addition to numerals written in the margins for every 100th stichos,22 in P. Herc. 1427 I have also detected dots traced to the left of the columns every 20 (real) lines. They are certainly stichometric dots which were used in order to facilitate the final counting of the lines on the part of the scribe and which were always written every 10 sichoi.23) The presence of these points in every 20th line is a very relevant element, allowing to precisely determine that in the roll of On Rhetoric Book 1 a stichos corresponds to two real lines (approximately 12 to 15 syllables).24 However, the total number of stichoi registered in the subscriptio and the ratio 1:2 between stichos and real line seem to be in contradiction with some objective data. In particular, 50 lines (= 25 stichoi) before the end of the text which is marked by a coronis, there is a stichometric ny in the intercolumnium, which indicates the 3700th stichos (= the 7400th real line): if the stichoi were really 4000, we would expect to count 300 more stichoi (= 600 real lines) after the ny, instead of the actual 25 stichoi. This probably suggests that the total number indicated in the end title was rounded up. The approximation from 3725 to 4000 might have been meant to compensate for the presence of insertions outside the columns (which were probably not included in the line counting), and perhaps also to allow for the possibility of future interventions by the author.25 This might turn out to be particularly relevant with reference to a provisional copy and to the necessity of subsequently recopying the text into a fair copy.
In light of these considerations, it can be concluded that the provisional nature of Philodemus' On Rhetoric Book 1 is not to be understood with regard to the material and external state of the book, but rather to the state of its text. Whether Philodemus was actually and constantly present in the Villa dei Papiri during the constitution of most of the library or not, his authoriality emerges -- occasionally and yet clearly -- from the editorial activity in full swing in those texts which have not yet progressed to their definitive editorial state.
Fig. 1: P. Herc. 1427, col. 3 ll. 11-13 © Biblioteca Nazionale. Napoli -- Brigham Young University. Provo. USA; duplication by any means is forbidden).
Fig. 2: P. Herc. 1427, col. 6 l. 13 (© Biblioteca Nazionale, Napoli -- Brigham Young University, Provo, USA: duplication by any means is forbidden).
Fig. 3: P. Herc. 1427. col1. 3 ll. 1-2 (lBiblioteca Nazionale, Napoli -- Brigham Young University, Provo, USA; duplication by any means is forbidden).
Fig. 4: P. Herc. 1427, fr. 4 ll. 3-5 (© Biblioteca Nazionale, Napoli -- Brigham Young University, Provo, USA; duplication by any means is forbidden).
Fig. 5: P. Herc. 1427, col. 6 l. 34 (© Biblioteca Nazionale, Napoli -- Brigham Young University, Provo, USA; duplication by any means is forbidden).
_______________
Notes:
1 The images of P. Herc. 1427 (MSI) and of its disegni are reproduced by courtesy of the Ministero dei Beni e delle Attivita Culturali e del Turismo © Biblioteca Nationale. Napoli-Brigham Young University. Provo. USA; duplication by any means is forbidden. I am very thankful to Prof. G. Abbamonte. Prof. P. Fioretti. and Prof. A. Stramaglia for the useful advice and the stimulating discussion on the topic of this paper.
2 The idea of the manus Philodemi is already in Comparetti (1880) 162. See Cavallo (1983) 26-27, who refused this idea, with reference to previous literature.
3 See Del Mastro (2014) 30-33, who has recently taken stock of the question; for an exhaustive overview see now D'Angelo (2018) 128-9. At the present stage of our knowledge, the term occurs in the end titles of P. Herc. 168 (Phld. Op. inc.). 1427 (Phld. Rh. I), 1674 (Phld. Rh. II), 1506 (Phld. Rh. III) and, according to a recent study by D' Angelo (2018), in the subscriptio to the theological work preserved in P. Herc. 89/1383. Regarding the presence of the term in Philodemus' rhetorical works, Sudhaus ([1892] xv, Sudhaus [1895a) 44, Sudhaus [1895b] 70-85, and Sudhaus [1896]) viii- xi), who published almost all the Herculaneum rhetorical papyri, claimed that the two different titles referred to two different works, since Philodemus would have written a first work, [x], and then the proper [x]. Starting from Comparetti's reflections (Comparetti [1910] 124), scholars have come to consider the necessity to identify different copies of the same work instead of different works. https:// doi.org/10.1515/9783110684629-009
4 Cavallo (1983) 63-64.
5 Cavallo (1984) 19. He ascribed the handwriting of the roll of On Rhetoric Book 1 to the first Philodemean age (75-50 BC) and placed its scribe (whom he named 'Anonimo XX') in a group ('Gruppo N') characterized by the peculiar bending of certain strokes ('contorsione o almeno curvatura di certi tratti'. Cavallo [1983] 38-39). One of the most representative examples of this group is P. Herc. 1426, which preserves the fair copy of Philodemus' On Rhetoric Book 3.
6 Delattre (1995). As for the different readings of the subscriptio and its complete transcription, see Del Mastro (2014) 87-88.
7 Dorandi (2007b) 70-77.
8 For the reconstruction of the roll, see Nicolardi (2018) 125-62.
9 See Nicolardi (2017) for a detailed analysis of all the corrective interventions in P. Herc. 1427.
10 Several corrections in scribendo of mechanical errors in P. Herc. 1427 are described in Nicolardi (2017), now also in Nicolardi (2018) 102-21.
11 I hereinafter refer to the numbering of the fragments and the columns as provided by the indications written on the 19th-century cardboard, on which the papyrus is glued, and as reproduced in Sudhaus (1892).
12 The reading of the adverb post correctionem is mine: for the previous editions, see the critical apparatus in Nicolardi (2018) col. 234 ll. 11-13.
13 See col. 237 l. 13, in Nicolardi (2018).
14 See col. 233 l. 37-234 l. 2, in Nicolardi (2018).
15 See col. 230 ll. 27-29, in Nicolardi (2018), with comment ad loc., and the introduction to this edition on Philodemus' classification of arts.
16 See col. 237 l. 34. in Nicolardi (2018).
17 See Gomperz (1872) 29 (= Dorandi [1993] 74).
18 This is both true if the changes occur for the first time in this papyrus and if they derive from a previous provisional copy used as antigraph, whose only trace we would thus find in our papyrus. As a matter of fact, it is not possible to completely exclude that a misleading and confused state of the passages in question in the antigraph caused the text ante correctionem and then led the scribe to pay more attention, to verify and to correct. Essler (2017) 71-75 has recently identified in P. Herc. 152/157 (Phld. De dis.) and in other Herculaneum papyri interventions which he has firmly attributed to Philodemus' intention.
19 If this hypothesis is correct and the changes are made for the first time in this copy, this book could be considered to be, as P. Fioretti usefully suggested to me, an author's idiograph.
20 Another possibility, for which I am thankful to the anonymous reader of this paper, might be that this roll had been firstly conceived as a definitive copy. The uneconomic intervention in col. 3 ll. 11-13. Sudhaus might thus be explained with the intention not to ruin the formality of the copy. Then, only as a result of the first revision by the main scribe and of the multiple interventions made by him, the book might have been 'downgraded' to provisional copy, with consequences on the subscriptio and on the final stichometric counting (on these two aspects see below), which were written at the end of the roll after this first revision.
21 In Herculaneum papyri 100 stichoi usually correspond to circa 180-200 real lines. See Bassi (1909) 326-32.
22 With this aim the alphabetical series of 24 letters (not enriched with stigma, coppa and sampi) is generally used; a new series can begin after omega (alpha = 100; beta = 200; gamma = 300; ... omega = 2400: alpha = 2500: beta = 2600). On the use and the importance of stichometric numerals see Del Mastro (2011) 38; Del Mastro (2012) 41-43; Essler (2008) 299-305, with reference to previous literature.
23 Stichometric dots also occur in other Herrulaneum papyri: see Bassi (1909) 70-71; Philippson (1920) 235, 364-5: Ohly (1924) 191; Obbink (1996) 63-64; Del Mastro (2012) 41-42.
24 The lack of precise correspondence between standard stichos and the actual number of syllables and letters per stichos is also attested in other Herculaneum rolls: in the roll of Philodemus' De pietate, e.g., one stichos corresponds to two lines with just 12 syllables: see Ohly (1928) 56.
25 The rounding up to 4000 is particularly interesting, since this number, as Del Mastro (2012) 35-38 showed on the basis of Gal., De indolentia 28 Brodersen, might have been a limit beyond which it could have been necessary to divide the work into two rolls.