New Testament Titles in the Coptic Manuscript Tradition: An Overview

Abstract

This article aims to provide a diachronic overview of the evolution of New Testament Coptic titles, taking into account their textual structure, location inside the manuscript, and ornamental devices. The evolution of a title is not only related to the role of the chronological phase of the book production it belongs to, but also dependent upon the geographic area where its translation—from Greek to Coptic—took place.

1. Introduction

What is the function of a title? What is its temporal relationship with the text it is attributed to? How certain can we be about the contemporaneity of a title and the text with which it is associated? Can we use a title to univocally identify a text? Everybody who has dealt with titles of ancient works is aware that these questions are not easy to solve, and, more importantly, the answers change from literary tradition to literary tradition1. For instance, the response to the last question is certainly negative for the Coptic milieu2.

As is well known, the manuscript tradition of late antique and early mediaeval Egypt is very peculiar in many respects. This is due partly to the fact that the early phase of Coptic literature was translated from Greek and partly to the status of the Coptic language, which never fully supplanted Greek as the leading literary language. Moreover, for a long time, scholars have considered Coptic literature unimportant because the vast majority of the early manuscript tradition for New Testament works is preserved in Greek. But translation—in this case, from Greek into Coptic—is also a crucial aspect of tradition and reception, not only a mechanical or passive activity, and thus the Coptic titles of the New Testament should also be seen as important contributions to this tradition.

On several occasions, I have had the chance to explain that Coptic titles offer a privileged point of observation for analysing the Coptic literary manuscript tradition and the way the Copts interpreted, arranged, and conveyed their own history (Buzi 2017, p. 5). Over time, Coptic titles have changed location in the manuscript3, layout, length, “syntax” (that is, internal structure), and even purpose, thus marking crucial turning points in the manufacture of the writing supports (the shift from roll to codex, and from papyrus codex to parchment codex), important passages of the history of Coptic literature (from the translations from Greek into Coptic to the production of an original literature and, lastly, to the translations from Coptic into Arabic), and changes in the typology of the codex (from single-text codices to multiple-text codices). Titles respond adeptly to literary and technological change.

It is possible to state that the concise titles at the beginnings of Coptic literature (roughly from the 3rd to the 5th century), certainly based on the Greek titles of the works which were by then translated into Coptic, slowly but progressively gave way to longer and longer titles4. In some cases, at least as far as the homiletic and hagiographic original productions (i.e., directly composed in the Coptic language) are concerned, these longer titles became real micro-texts, up to two pages in length, whose narrative thread was sometimes not exactly tied to the content of the works to which they were attached. This phenomenon may justify their re-definition as “paracontent”—instead of the more consolidated definition of “paratext”—that has been recently elaborated upon and proposed to the scholarly community (Ciotti et al. 2018)5.

Because of their nature as textual elements linked to authoritative literature, biblical titles are much more conservative and less subject to a structural evolution compared to other literary genres. In this respect, therefore, they are less useful in tracing changes in tastes, habits, and cultural identity in ancient Coptic society. They are strategic, however, in reconstructing the bibliological and paratextual features of the antigraphs and in shedding light on the level of dependence from the Greek tradition. They also offer insight on the Greek models used for the translation activity into Coptic, a dependence that changes from region to region, from dialect to dialect, and from time to time. This article, therefore, aims to provide an overview of the main features of the New Testament Coptic titles, with particular attention to those of the gospels6.

2. The Problem of Terminology

Before analysing the features of the titles attributed to the Coptic translations of the New Testament, it is necessary to stress the lack of shared terminology used to describe the paratextual category of titles, a fact that makes a systematic study of such a complex phenomenon difficult and even more challenging for a comparative approach involving multiple manuscript traditions.

Titles placed at the beginning or at the end of a text—not only biblical works—are called either initial/final titles, or beginning-/end-titles (Schironi 2010, p. 83), or inscriptiones/subscriptiones (Tischendorf 1843; Buzi 2005, p. 39; Gathercole 2013, pp. 33–76). Opening titles (Gathercole 2013, p. 33), introductory titles (Gathercole 2013, p. 34), and subtitles are also attested. Sometimes, even the term colophon is used to describe a final title, if this contains the sentence ϩⲛ̄ ⲟⲩⲉⲓⲣⲏⲛⲏ (“in peace”) (Gathercole 2013, p. 59). Colophon, however, is not appropriate. Although the definition of what a colophon is remains contested (Maniaci 2022)7, a colophon should contain at least some element related to the act of copying (name of the scribe, place of copy, donor, etc.).

In the same way, the definitions for unit titles and section headings describe a title that introduces a specific and meaningful textual section (an episode inside a cycle of miracles, a paragraph of text segment, maybe marked by a rubric, etc.), while running titles and running headings are alternatively used to define the titles (or a shortened version of them) located in the top margin of a codex and recurring across any opening, often in a gospel codex.

Moreover, there are cases of general titles, which refer to a series of textual units (for instance, the Pauline Epistles) and title-indices or tables of contents, which refer to the content of a book. The definition and adoption of a shared vocabulary that appropriately describes all these phenomena is desideratum that all scholars should see as urgent.

3. The Biblical Translations from Greek into Coptic: When, Where, and How

The whole of the New Testament is preserved in Coptic, both in the Sahidic and Bohairic dialects, but only thanks to the combination of all the extant, often fragmentary, manuscripts. It is necessary to specify, however, that the translations of biblical works from Greek into Coptic are not the product of a centralized operation. The translation process took place in two main phases and independently in different centres at the same time (Takla 2014, pp. 107–8; Funk 2013, pp. 536–46).

The first phase (end of the 3rd/beginning of the 5th century) produced translations into a variety of dialects and sub-dialects, including those that from the 6th century ceased to exist, with the progressive emergence of Sahidic, a dialect that we can define as “standard” when the translations were eventually produced. Unfortunately, we do not know much about the environments which were responsible for these early translations, but it is clear from the codicological point of view that this period is defined by codices, often bilingual, with fluid bibliological, codicological, and paratextual features (the position of titles, for instance, is often irregular, while contemporary non-biblical manuscripts show a preference for final titles). Moreover, it is not easy to define how much the Coptic translations of the New Testament reflect the Greek tradition, since so few Greek manuscripts are datable to that early period8.

The second phase (end of the 7th/beginning of the 8th century) is represented by translations into Sahidic and, above all, into Bohairic, produced in the main monastic centres. They show a different layout, different titles, and sometimes even different textual traditions. As is obvious, other translations were certainly also produced in other periods, but this large-scale two-stage translation operation best describes the evidence that we currently have. What we have, therefore, is heterogeneous from the bibliological point of view and, at the same time, extremely useful for mapping structural (textual) changes over the centuries.

4. The Titles of the Early New Testament Translations

Many of the ancient codices (end of the 3rd–5th centuries)9 that preserve translations of the New Testament are in such a fragmentary state that it is impossible to use them to reconstruct the position of titles. There are, however, a few exceptions, some of which are considered here, that shed light on the main trends and help us to draw some conclusions. A good number of them belong to the hoard known as the “Bodmer Papyri” or “Dishna Papers” (Nos. 1–5 of the following list)10:

(1) The first example, Dublin, Chester Beatty Library 2026 + Oslo, Schøyen collection MS 193, better known as P. Codex Crosby Schøyen 193 (end of the 3rd–early 4th century)11, is a Sahidic multiple-text, single-quire papyrus codex. It includes different literary works, whose unusual combination may be explained either by a specific request of the person or the group who commissioned the book or by the religious identity of the group which was still in fieri. The codex, which is characterized by a non-consistent pagination system, contains biblical works, Melito of Sardis’s De Pascha 47–105 (CC 0222)12, and an unidentified ‘homily’ for Easter morning. All the texts are arranged in two columns per page, apart from the unidentified homily, which is written in a single column. The works are introduced by an initial brief title and concluded by a similar but sometimes non-identical final title, which is normally more emphasized, due to decorative motifs, than the initial one:

pp. 1–6: Remains of flyleaves and small fragments, whose text is not identifiable;

pp. 7 (?)–51: Melito of Sardis, De Pascha;

p. 51, subscriptio: ⲡⲉⲣⲓ ⲡⲁⲥⲭⲁ ⲙⲙⲉⲗⲓⲧⲱⲛ (‘On the Passion, by Melito’).

There was probably also an initial title, but the first pages of the codex are mostly illegible.

pp. 52–74 (?): On Jewish Martyrs;

pp. 75–107 (?): Epistle of Peter;

p. 75, initial title, after a white column: ⲧⲉⲡⲓⲥⲧⲟⲗⲏ ⲙⲡⲉⲧⲣⲟⲥ (‘The Epistle of Peter’);

p. 107, final title, more prominent: ⲧⲉⲡⲓⲥⲧⲟⲗⲏ ⲙⲡⲉⲧⲣⲟⲥ (‘The Epistle of Peter’);

pp. 107–24: Book of Jonah;

p. 107, initial title, hardly readable: ⲓⲱⲛⲁⲥ ⲡⲉⲡⲣⲟⲫ<ⲏⲧⲏⲥ> (‘Jonah the prophet’);

p. 124, final title: ⲓⲱⲛⲁⲥ ⲡⲉⲡⲣⲟⲫⲏⲧⲏⲥ (‘Jonah the prophet’);

pp. 124–28: Unidentified work on Easter morning;

p. 52, initial title, after a white column: ⲙ<ⲙ>ⲁⲣⲧⲩⲣⲟⲥ ⲛⲓⲟⲩⲇⲁⲓ ⲉⲛⲧⲁⲩϣⲱⲡⲉ ϩⲓ ⲁⲛⲧⲓⲟⲭⲟⲥ ⲡⲡⲣⲟ (‘The Hebrew martyrs under the kingdom of Antiocus’);

p. 74, final title, more explicitly marked as such by paratextual elements such as non-typographical decorative features compared to the initial one: ⲙ<ⲙ>ⲁⲣⲧⲩⲣⲟⲥ ⲛⲓⲟⲩⲇⲁⲓ (‘The Hebrew martyrs’).

(2) Cologny, P. Bodmer III (4th century)13 is a multiple-quire codex in early Bohairic (B4), which contains the Gospel of John and the Book of Genesis. It was originally composed of 84 leaves, whose first 12 (corresponding to the first quire) are only partially preserved. The text is arranged in one column; the pagination is irregular14.

f. 1: used as pastedown;

f. 2: blank;

ff. 3–73r: Gospel of John;

f. 73r, final title: ⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ|ⲕⲁⲧⲁ ⲓⲱⲁⲛⲛⲏⲥ (‘Gospel according to John’);

f. 73v: blank;

ff. 74–83r: Genesis;

f. 74r, initial title: ⲅⲉⲛⲉⲥⲓⲥ (‘Genesis’);

f. 83v: blank;

f. 84: used as pastedown.

(3) P.Bodmer XIX (second half of 4th century/first half of the 5th century)15 is a parchment codex, written in Sahidic, which contains, in its current form, the second part of the Gospel of Matthew and the beginning of the Epistle to the Romans. The text is written in two columns, and the pagination is irregular16. The Gospel of Matthew is closed by a final title, standing alone in the middle of the right-hand column, while the Epistle to the Romans is not introduced by any title:

ff. 1–4: blank, except for an initial title, which was certainly added by a later untidy hand on f. 3r ⲡϩ̄ⲁⲏ ⲛ̄ⲙⲉⲣⲟⲥ ⲙⲁⲑⲉⲟⲥ (‘the last part of Matthew’)17;

pp. 77–166: Gospel of Matthew, 14, 28–28, 20 (end);

p. 166, final title, ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ|ⲡⲕⲁⲧⲁ ⲙⲁⲑⲑⲁⲓⲟⲥ (‘the Gospel according to Matthew’);

pp. 1–6: Rom. 1, 1–2, 318.

(4) Codex Scheide (5th century)19 is a complete parchment codex (236 ff.)20 of uncertain origin (al-Mudil? Dishna?), written in Middle-Egyptian (dialect M), which contains the Gospel of Matthew in Coptic and a doxology in Greek and Coptic. The Gospel of Matthew is closed by a prominent and well-preserved final title on p. 455, in red and black ink: ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ | ⲡⲕⲁⲧⲁ ⲙⲁⲑ⳿ⲑⲁⲓⲟⲥ | ϩⲛ̄ ⲟⲩⲉⲓⲣⲏⲛⲏ. This is followed by the number 1518 (ⲁⲫⲓⲏ) that Schenke interpreted as the amount of στίχοι, a datum useful to calculate the payment of the scribe21. After the conclusion of the gospel, the pagination begins anew, and a different hand has written the doxology.

(5) Barcelona, Arxiu Històric de la Companya de Jesús a Catalunya, P.Palau Rib. Inv.-Nr. 181–8322 (second half of the 5th century) is an incomplete parchment codex, written in Sahidic, which contains the gospels of Mark, Luke, and John. The text is arranged in two columns. The identification of different hands, the presence of a double set of quire signatures, and the blank pages between the end of Luke and the beginning of Matthew (ff. 91v–94v) suggest the hypothesis of two originally independent circulation units, which were later bound together to form a new larger circulation unit, according to the model of transformation A4 proposed by Patrick Andrist, Paul Canart, and Marilena Maniaci23.

ff. 1r–91r (=P.Palau Ribes inv. 181): Gospel of Luke (numbered from ⲁ to ⲣϥ [sic] and last page left unpaginated);

f. 1r: initial title, ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ ⲛⲕⲁⲧⲁ ⲗⲟⲩⲕⲁⲥ (‘the Gospel according to Luke’);

f. 91r: final title, ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ|ⲛⲕⲁⲧⲁ ⲗⲟⲩⲕⲁⲥ (‘the Gospel according to Luke’);

ff. 91v–94v: blank and unpaginated;

f. 94v: final title, ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ|ⲛ|ⲕⲁⲧⲁ ⲓ̈ⲱⲁⲛⲛⲏⲥ (‘the Gospel according to John’);

ff. 95–169 (=P.Palau Ribes inv. 183): Gospel of John (all paginated from ⲁ to ⲣⲛ);

ff. 170r–228r (=P.Palau Ribes inv. 182): Gospel of Mark (paginated from ⲁ to ⲣⲓⲍ);

f. 170r: initial title, ⲙⲁⲣⲕⲟⲥ (‘Mark’);

f. 228r: final title, ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ|ⲛⲕⲁⲧⲁ ⲙⲁⲣⲕⲟⲥ (‘the Gospel according to Mark’);

ff. 228v–230v: blank and unpaginated.

(6) Codex Schøyen (first half of the 4th century)24, in the Mesokemic dialect, housed in a private collection25, is supposed to have come from a monastery in the Oxyrhynchite region. It does not belong to the same library of the Dishna Papers. Of the original 45 leaves, only 39 are preserved. The codex is the earliest attestation of the Gospel of Matthew in any Coptic dialect. Its main feature is that it does not correspond to the text of any other Coptic nor Greek manuscripts containing Matthew. Hans-Martin Schenke in his editio princeps stated that the Codex Schøyen represents a translation of an entirely different Gospel of Matthew26. The last page of the codex has a final title (which does not reflect the unusualness of the text), separated from the text by a stylized vegetal motif: ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ|ⲛⲕⲁⲧⲁ ⲙⲁ|ⲑⲉⲟⲥ. It cannot be excluded that there was also an initial title.

(7) Oslo, University of Oslo Library, P.Osl. inv. 1661 (4th century)27 consists of little fragments of a papyrus codex that contain Matthew 11:25–30 (in Greek and Akhmimic dialect) and Daniel 3:51–53 (in Greek). An initial title is located in the upper part of the bilingual page containing the gospel: ⲉⲩⲁ[ⲅ]ⲅ̣ⲉⲗ[ⲓ]ⲟⲛ|[ⲡⲕⲁ]ⲧ̣ⲁ ⲙⲁⲑⲁⲓⲟⲥ. Due to the extremely fragmentary state of preservation of the manuscript, it cannot be excluded that there was also a final title28.

Despite the limited number of samples of early codices that preserve one or more titles29, it is possible to draw some preliminary conclusions:

-

The co-presence of initial and final titles is frequent and probably was perceived by then as the conventional way to present works30.

-

When both types of titles are present, final titles are more emphasized by means of ornamental devices than the initial ones.

-

When only one typology of title is present, final titles are more frequent.

-

Running titles, which will become rather common in later New Testament codices, have not yet been introduced31.

It is important to stress that similar features are observable in contemporary manuscripts that preserve apocryphal works, such as the Codex Tchacos32, which conveys the Letter of Peter to Phillip (pp. 1–9), the Apocalypse of James (pp. 10–32), the Gospel of Judas (pp. 33–58), and the Book of Allogenes (pp. 59–66), all written in Sahidic dialect. The first three works are closed by final titles:

[ⲧ]ⲉⲡⲓⲥⲧⲟⲗⲏ ⲙ̄ⲡⲉⲧⲣⲟⲥ ϣⲁ ⲫⲓⲗⲓⲡⲡⲟⲥ (‘the Letter of Peter to Philip’);

ⲓ̈ⲁⲕⲱⲃ[ⲟⲥ] (‘James’);

ⲡⲉⲩⲁⲅ̣ⲅ̣ⲉⲗⲓⲟⲛ ⲛ̄ⲓ̣̈ⲟⲩⲇⲁⲥ (‘the Gospel of Judas’).

Similarly, Cologny, P. Bodmer XLI33, which contains the Acta Pauli, is closed by a final title: ⲡⲡⲣⲁⲝⲓⲥ ⲡⲁⲩⲗⲟⲩ (‘the Act [sic] of Paul’).

Another useful comparison, at least for what concerns the fluidity of the position of titles in the early Coptic codices (with a predominance of the final ones), is with the titles of the Nag Hammadi codices, whose presence can be summarized as follows34:

-

Nine works have only initial titles;

-

Twenty-two works have only final titles;

-

Eleven works have both initial and final titles;

-

Nine works do not have titles, but incipit and/or desinit which in some way have the function of a title;

-

Seven works have no titles or other paratexts which could substitute for their function;

-

Eleven works have incipit and/or desinit which do not substitute the title (but in some way recall their content);

-

Six works have internal titles;

-

Lastly, for 7 works, it is impossible to say if they had any titles because the manuscripts that transmit them are very fragmentary.

If we compare this situation with that of the titles of the more or less contemporary Greek codices, we observe that in the Greek tradition, initial titles, although not standard, are more frequent, and running titles are also present35. We can deduce that the Coptic tradition may have applied some local initiatives in the act of translation of New Testament works or at least the use of different models.

5. The New Testament Titles in 6th–8th-Century Manuscripts

Unfortunately, biblical codices datable to the 6th–8th centuries are not numerous. This is the moment when, slowly but surely, parchment began to replace papyrus as the exclusive material support used to make codices in Egypt. Until the end of the 7th century, however, both materials were still used there. Those made of papyrus are normally of smaller dimensions compared to the parchment ones, but in general, all of them have a layout characterized by one column.

An example of the book production of this period is a group of manuscripts from the Monastery of Apa Jeremias at Saqqara. In the winter of 1924–1925, five small papyrus codices, with their original bindings, were seen at a dealer’s in Cairo. The three volumes of the group which were in the finest conditions36 passed into Mr. Chester Beatty’s possession, while the other two reached the Michigan University Library. The writing and the general format of the five volumes are sufficient to prove that they were not only found together, but must have originated in the same scriptorium37.

Dublin, Chester Beatty Library, Cpt 81338 (=CLM 64) contains the Gospel of John and the Pauline Epistles. When titles are preserved, the codex shows a more or less regular alternation of initial/final titles, which have a substantially identical structure. At f. 29r, for instance, one reads:

Final title: ⲧⲉ ⲡⲣⲟⲥ ϩⲣⲱⲙⲁⲓⲟⲩⲥ (‘The [Epistle] to the Romans’);

Initial title: ⲧⲉ ⲡⲣⲟⲥ ⲕⲟⲣⲓⲛⲑⲓⲟⲩⲥ (‘The [Epistle] to the Corinthians’).

The two titles are preceded and closed by a line of fish bone-shaped signs. Final titles still exist in this period, a fact that is confirmed also by Dublin, Chester Beatty Library, Cpt 814, ff. 2–16739, which contains the Acts of the Apostles and the Gospel of John, both of which are concluded by a final title (f. 98r and f.165r) that is included between bands of dashes and diplai, and sided by two zeta-shaped coronides.

Unfortunately, the New Testament manuscripts from Deir el-Balai’zah, which are datable between the end of the 7th and the 8th centuries, are very fragmentary, and it is therefore hard to draw reliable conclusions about their titles, although the co-existent use of initial and final titles seems to be again the common practice.

For instance, the fragment Oxford, Bodleian Library, Copt. d 1640, which contains the Pauline Epistles, preserves the following titles:

Final title: ⲧⲉ [ⲡⲣ]ⲟⲥ ⲧⲓⲙⲟⲑ[ⲉⲟⲥ ⲃ] (‘The second [Epistle] to Timotheus’);

Initial title: ⲧⲉ [ⲡⲣ]ⲟⲥ ⲧⲓⲧⲟⲥ (‘The [Epistle] to Titus’).

It is particularly important to trace the evolution of titles in this period affixed to the Pauline Epistles, since later translations, conveyed by codices from the White Monastery and the Monastery of St. Macarius (9th–11th centuries), seem to have followed two different paths and traditions41. Unfortunately, the extant evidence does not allow a full comparison.

6. The New Testament Titles from the 8th Century Onward

We lastly concentrate our attention on the New Testament titles produced between the end of the 8th and the beginning of the 11th centuries, a period in which many meaningful changes occur in the manufacture of the codex, from the definitive emergence of parchment over papyrus, to the increase in the dimensions of codices (no miniature codices are produced in this period, for instance), to the shift of titles to the beginning of the text (a phenomenon whose reasons are still to be fully explored). Several exceptions also exist in the case of codices of biblical content.

New York, Morgan Library and Museum M56942, from the Monastery of the Archangel Michael at Phantoou (modern Hamuli), contains the four Gospels and the following related titles43:

f. 3r, initial title: ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ ⲉⲧⲟⲩⲁⲁⲃ ⲛⲕⲁⲧⲁ ⲙⲁⲑⲑⲁⲓⲟⲥ ⲥⲩⲛ ⲑ(ⲉⲱ) (‘The holy Gospel according to Matthew. With God’);

f. 38r, final title: ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ ⲛⲕⲁⲧⲁ ⲙⲁⲑⲑⲁⲓⲟⲥ (‘The Gospel according to Matthew’); the rest of the leaf is blank;

f. 39r, initial title: ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ ⲉⲧⲟⲩⲁⲁⲃ ⲛⲕⲁⲧⲁ ⲙⲁⲣⲕⲟⲥ (‘The Gospel according to Mark’);

f. 60r, final title: ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ ⲉⲧⲟⲩⲁⲁⲃ ⲛⲕⲁⲧⲁ ⲙⲁⲣⲕⲟⲥ (‘The Gospel according to Mark’);

f. 60v blank;

f. 62r, initial title: ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ ⲉⲧⲟⲩⲁⲁⲃ ⲛⲕⲁⲧⲁ ⲗⲟⲩⲕⲁⲥ ⲥⲩⲛ ⲑ(ⲉⲱ) (‘The holy Gospel according to Luke. With God’);

f. 84r, final title: ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ ⲛⲕⲁⲧⲁ ⲗⲟⲩⲕⲁⲥ (‘The Gospel according to Luke’)

f. 84v: (almost) blank;

f. 85r, initial title: ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ ⲉⲧⲟⲩⲁⲁⲃ ⲛⲕⲁⲧⲁ ⲓⲱϩⲁⲛⲛⲏⲥ (‘The holy Gospel according to John’);

f. 113v, final title: ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ ⲉⲧⲟⲩⲁⲁⲃ ⲛⲕⲁⲧⲁ ⲓⲱϩⲁⲛⲛⲏⲥ (‘The holy Gospel according to John’).

Despite the fact that the text is arranged in two columns, initial titles are two columns wide and are much more emphasized than the final ones, being decorated with elaborated frames of interlaced ropes. Final titles, on the other hand, are simply marked by dividers and occupy only one column, whose remaining part is left blank44. There are no running titles.

New York, Morgan Library and Museum M57045 contains the Pauline Epistles and presents the following title:

f. 1r, hybrid title, both general, referring to the whole work, and introducing the Epistle to the Romans: ⲡⲁⲩⲗⲟⲩ ⲁⲡⲟⲥⲧⲟⲗⲟⲩ ⲧⲉⲡⲣⲟⲥ ϩⲣⲱⲙⲁⲓⲟⲥ ⲉⲡⲓⲥⲧⲟⲗⲏ ⲓ̄ⲇ̄ (‘Of the Apostle Paul. To the Romans. 14 Epistles).

Then, the initial titles of the single epistles follow, with no blank page between the end of an epistle and the following one:

f. 17r, initial title: ⲧⲉ ⲡⲣⲟⲥ ⲕⲟⲣⲓⲛⲑⲓⲟⲩⲥ ⲁ̄ (‘The first [Epistle] to the Corinthians’);

f. 31r, initial title: ⲧⲉ ⲡⲣⲟⲥ ⲕⲟⲣⲓⲛⲑⲓⲟⲩⲥ ⲃ̄ (‘The second [Epistle] to the Corinthians’);

f. 40v., initial title: ⲧⲉ ⲡⲣⲟⲥ ϩⲉⲃⲣⲁⲓⲟⲥ (‘The [Epistle] to the Hebrews’);

…

…

f. 83v: final general title: ⲧⲟⲩ ⲁⲅⲓⲟⲩ ⲁⲡⲟⲥⲧⲟⲗⲩ ⲉⲡⲓⲥⲧⲟⲗⲏ ⲓ̄ⲇ̄ ⲥⲧⲓⲭⲟⲥ ⲉ̄ⲫ̄ⲟ̄ⲉ̄ (‘The forteen Epistles of the holy Apostle. 5575 stichoi’).

Despite the fact that this codex is very likely contemporary to the previous one—in this case, the colophon precisely dates it to 897/898, and even the ornamental devices to mark the presence of titles are very similar—it shows some meaningful differences. If we exclude the general final title, all the epistles are introduced by a title but not concluded by a final one. Moreover, the first title introduces the Epistle to the Romans and offers a general title for the whole set of letters. It is necessary to stress that the Epistle to the Hebrews is located immediately after 2 Corinthians, and therefore is in a sequence which is not that of the oldest Greek witnesses (Codices Sinaiticus, Alexandrinus, Vaticanus, Papyrus 46, and Ephraemi Rescriptus), where it is located after the letters to the Thessalonians, as the last of the letters addressed to groups and not to individuals46.

New York, Pierpont Library and Museum, M57247, which contains the Catholic Epistles, presents a similar situation, including the following titles:

f. 1r, hybrid title, general—that is referred to the whole work—and at the same time introducing the Epistles of Peter: ⲛⲕⲁⲑⲟⲗⲏⲕⲏ ⲛⲛⲉⲛⲉⲓ[ⲟⲧⲉ] ⲛⲁⲡⲟⲥⲧⲟⲗⲟⲥ ⲡⲉⲧⲣⲟⲥ (‘Catholic [Epistles] of our fathers the Apostle Peter’);

f. 6r, final title: ⲧⲉⲡⲓⲥⲧⲟⲗⲏ ⲙⲡⲉⲧⲣⲟⲥ ⲁ̄ (‘The first Epistle of Peter’);

f. 6r, initial title: ⲧⲉⲡⲓⲥⲧⲟⲗⲏ ⲙⲡⲉⲧⲣⲟⲥ ⲃ̄ (‘The second Epistle of Peter’);

f. 9 r, initial title: ⲧⲉⲡⲓⲥⲧⲟⲗⲏ ⲛⲓⲱϩⲁⲛⲛⲏⲥ ⲁ̄ (‘The first Epistle of John’);

f. 14v, initial title: ⲧⲉⲡⲓⲥⲧⲟⲗⲏ ⲛⲓⲱϩⲁⲛⲛⲏⲥ ⲃ̄ (‘The second Epistle of John’);

f. 15r, initial title: ⲧⲉⲡⲓⲥⲧⲟⲗⲏ ⲛⲓⲱ(ϩⲁ)ⲛⲛⲏⲥ ⲅ̄ (‘The third Epistle of John’);

f. 15v, initial title: ⲧⲉⲡⲓⲥⲧⲟⲗⲏ ⲛⲓⲁⲕⲱⲃⲟⲥ (‘The Epistle of James’).

The Cairene complementary fragments, a portion of the same now-disassembled codex that is for the most part in New York48, preserves the Epistle of Jude and the related title:

f. 4r, initial title: ⲧⲉⲡⲓⲥⲧⲟⲗ[ⲏ ⲛ]ⲟⲩⲇ[ⲁ] (‘The Epistle of Jude’).

We can deduce that the final title of the first Epistle of Peter is residual and used to compensate for the fact that the initial title was also used to introduce all the letters. Thus, the Pauline Epistles and the Catholic Epistles, as distinct sub-corpora, were dealt with differently in this particular case in the Coptic manuscript tradition49.

At this point of the manuscript production created for the library of the Monastery of the Archangel Michael50, final titles seem to disappear, while none of the New Testament codices from Hamuli have running titles.

Despite the extreme fragmentary conditions of the codices from the White Monastery at Atripe (modern Sohag), which do not allow a systematic analysis of their titles, we can observe some unusual practices, including the more or less systematic use of running titles for the gospels. These appear in:

CLM 467 = MONB.JV ([ⲡⲕⲁⲧⲁ/] ⲙⲁⲑⲁⲓⲟⲥ, [ⲡⲕⲁⲧⲁ/] ⲙⲁⲣⲕⲟⲥ, ⲡⲕⲁⲧⲁ/ⲗⲟⲩⲕⲁⲥ) (‘the [gospel] according to Matthew’, ‘the [gospel] according to Mark’, ‘the [gospel] according to Luke’)51;

CLM 473 = MONB.KE (ⲡⲕⲁⲧⲁ/ⲗⲟⲩⲕⲁⲥ) (\the [gospel] according to Luke’)52;

CLM 478 = MONB.KJ (ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ/ⲛⲕⲁⲧⲁ/ⲗⲟⲩⲕⲁⲥ; [ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ ⲛⲕⲁⲧⲁ] ⲓⲱⲁⲛⲛⲏⲥ) (‘the gospel according to Luke’, ‘the gospel according to John53;

CLM 488 = MONB.KU + MONB.LO (ⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ…) (‘gospel…’)54;

CLM 484 = MONB.KQ attest (ⲡⲕⲁⲧⲁ/ⲗⲟⲩⲕⲁⲥ) (‘the [gospel] according to Luke’)55.

The Pauline Epistles, at least in some cases, also present running titles. In Vatican Library Borg.Copt.109.XXII.86, ff. 1–4 r/v = CLM 514, for instance, the following running title is readable in the upper margin: ⲧⲉ ⲡⲣⲟⲥ ⲉⲫⲉⲥⲓⲟⲩⲥ (‘the [Epistle] to the Ephesians’).

As for final and initial titles, they are both in use in the biblical manuscripts of the White Monastery, although it is difficult to evaluate if they share an equal weight of importance or if initial titles are more frequent. Here, some examples of final titles are listed:

CLM 471 = MONB.KS (Paris, BnF 129.5, f. 165v): ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ ⲛⲕⲁⲧⲁ ⲙⲁⲑⲑⲁⲓⲟⲥ (‘the Gospel according to Matthew’);

CLM 514 (Borg.Copt.109.XXII.85 f.8r): ⲧⲉ ⲡⲣⲟⲥ ⲑⲉⲥⲥⲁⲗⲗⲟⲛⲓⲕⲉⲩⲥ ⲁ (‘the first [Epistle] to the Thessalonians’); the title immediately precedes that attributed to the second epistle to the Thessalonians.

The following are examples of initial titles:

CLM 497 = MONB.LG: ⲡⲉⲩⲁⲅⲅⲉⲗⲓⲟⲛ ⲛ̄ⲕⲁⲧⲁ ⲙⲁⲣⲕⲟⲥ (‘the Gospel according to Mark’);

CLM 514 (Borg.Copt.109.XXII.85 f.8r): ⲧⲉ ⲡⲣⲟⲥ ⲑⲉⲥⲥⲁⲗⲗⲟⲛⲓⲕⲉⲩⲥ ⲃ (‘the second [Epistle] to the Thessalonians’); the title immediately follows that attributed to the first epistle to the Thessalonians.

Decorative elements are normally minimal in both cases. As for the gospels, a new gospel does not necessarily start on the recto, as it happens in Hamuli, since the parchment seems to be preserved as much as possible, leaving very little blank space in the manuscripts from the White Monastery.

7. Conclusions

To conclude this brief and provisional overview of the New Testament Coptic titles, we can note that, unlike all other literary genres which saw the slow but progressive extension and re-elaboration of titles, the New Testament titular traditions were approached with a more conservative attitude, which maintained their original structure even in light of different textual traditions and models that varied from place to place and from period to period. Subscriptiones sometimes appear even in very late codices, while running titles are not in use until the mid-9th century and seem to belong to specific productive milieux (White Monastery, Monastery of Mercurius at Edfu)56. Initial titles are always at the beginning of a new leaf in Hamuli, but the same cannot be said for the White Monastery. Lastly, in the production of the Monastery of the Archangel Michael, biblical works are in two columns and in a bimodular script, while those of the White Monastery prefer the one-column layout and the unimodular script. This evidence offers another clue to the existence of local traditions, behind which we have to imagine different Greek models and specific aesthetic practices.

The identification of the cultural centres which were active in the early phase of the Christian Egyptian manuscript production—not only of biblical content—is probably one the main challenges that Coptic studies should address in the future. This would shed new light on the paths and networks through which models and antigraphs circulated, being adapted to local necessities and cultural inclinations, a process that remains virtually unknown until now.