Introduction

Contents

• Overview
• Importance of the Sahidic Version
• Brief Overview of Greek-Coptic Diglots
• Christology in the Coptic Versions of John
• Kasser's Timeline for Coptic Dialects
• Why a New Edition
• PHI Version Basis
• PHI Version Methodology
• Analysis of the Resultant Text
• The Sahidica Text and eMss
• The Sahidica Version Basis
• The Sahidica Version Goal: Usability
• The Sahidica Methodology
• Notes on Textual Changes
• Notes on the Parallel Greek Text
• Limitations of the Sahidica Text
• Additional Collation
• References

Overview

The New Testament was translated into Sahidic late in the second century (again, roughly).[2] This paper explains the motivation and methodology underlying a new version of the Sahidic New Testament. A few terms should be kept clear in the reader's mind.

• Sahidica.org is the name of the organization behind this new version.
• The Sahidica Project is the effort focusing on the gathering, collating, and publishing of both (a) editions of the new version and (b) the various electronic versions of the manuscripts used to create this new version.
• The Sahidic text is the name used specifically to describe the electronic Sahidic text that is the basis for the various published forms of this new version.
• Sahidica is an abbreviated form of name of the new version itself. Technically, the full name of the version is Sahidica: The Sahidic New Testament.

Importance of the Sahidic Version

Several factors make the Sahidic New Testament important. These were outlined by the current editor in 1989 in a brief document, which was part of a software package that compared variant readings in many translations. That document is reproduced here (in an updated form).

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      IMPORTANCE OF THE SAHIDIC LANGUAGE: IN RESEARCH AND TRANSLATION
         Copyright (c) 1989-2008 by J. Wells. All rights reserved.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

It is noteworthy that the New Testament was translated into three languages
during the Koine period (b.330 AD), that is, while the Koine (common) Greek
was still the lingua franca of the Roman world and universally understood.
This gives us an opportunity to see how people who read, wrote, spoke, and
thought in New Testament Greek rendered the divine Word into their mother
tongues. These three languages were Latin, Syriac (spoken in Palestine), and 
various dialects of Coptic (spoken in Egypt).

To date, the vast majority of research has been done on the first two. This
is unfortunate since the Coptic (especially the Sahidic dialect) surpasses 
both Latin and Syriac in several ways. First, it represents a textual 
tradition generally considered to be superior. Second, it is represented by 
more, and older, manuscripts. Third, it is an important source for historical 
interpretation. Fourth, it is more like English in certain respects.

The following will explain each of these four aspects, primarily for the 
Sahidic version of the Coptic.

1. Both the Latin and Syriac are representative of the Western textual family
(which is generally viewed as inferior), while the Sahidic ranks with the 
best papyri and the "B" Uncial (Vatican Library 1209) as representative of 
the Proto-Alexandrine textual family, which is generally recognized as the 
best and earliest group of manuscripts. The Westcott-Hort, UBS, and Nestle-
Aland master texts, and thus most modern translations, are based on this 
family of manuscripts.

2. The Old Latin and Syriac are represented by two extant manuscripts each,
which can be dated as being from before the fifth century (400 AD). All four
are of the Gospels only. On the other hand, Sahidic (and other Coptic) 
manuscripts dating from the same period are more numerous, more complete and
and represent many more books of the Bible.
  
Below is a comparison chart of the extant manuscripts for each language.

Language  Manuscript      Contents                Century   Condition
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Latin:   Vercellensis     Gospels                 IV        partial
         Bobiensis        Gospels                 IV/V      partial

Syriac:  Sinaitic         Gospels                 IV        fragmented
         Curetonian       Gospels                 IV/V      fragmented

Sahidic: Crosby           1Peter                  III       complete
         Brit. M. 7594    Acts                    III/IV    fragmented
         Mich. 3992       John 1Cor. Titus        III/IV    fragmented
         Berlin 408       Rev. 1John Philemon     IV        partial
         Kahle 22         Eph. 1Pet. 1John James  IV        fragmented
         Lectionary 1604  Matthew                 IV        fragmented
         Berlin 15926     Acts                    IV        fragmented
         Garrido          Matthew                 IV        fragmented
         Rainer V(p41)    Acts                    IV/V      partial
         Bodmer XIX       Matthew Romans          IV/V      fragmented
         Kahle 21         1Timothy Titus          IV/V      fragmented

Other:*  Achmimic         Matt(p62)               IV        fragmented
         Achmimic         John James(p6)          IV        fragmented
         Achmimic         Luke                    IV        fragmented
         Fayumic          John                    IV        partial
         Oxyrhynchite     Matthew                 IV        complete
         Oxyrhynchite     Acts                    IV        partial
         Proto-Bohairic   John                    IV        partial
         Sub-Achmimic     John(Q)                 IV        partial
         Achmimic         Galatians               IV/V      fragmented
         Fayumic          Acts                    IV/V      partial
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
* Other Coptic dialects.

3. On the subject of textual criticism, Kurt and Barbara Aland write (in 
"The Text of the New Testament"):

   Important as the Latin and Syriac versions may be, it is of far
   greater importance to know precisely how the text developed in Egypt.
  
On the Sahidic's importance in the field of historical interpretation, 
well known coptologist J. M. Plumley observed:

   By and large the Coptic version can be a valuable aid to the scholar
   engaged in textual criticism, and because in certain passages it
   preserves very ancient traditions of interpretation, it ought to be
   of considerable interest to the scholar working on the history and
   development of Christian doctrine. [Cf. Addendum 2 below]

4. While English is related to Latin, in some important respects it is quite
similar to the Sahidic. Where Greek has the definite article (the) but no
indefinite article (a, an) and Latin and Syriac have no articles at all, 
Sahidic has both the definite and indefinite articles. Moreover, Sahidic 
article usage is quite similar to English. Thus Thomas Lambdin wrote (in his 
"Introduction to Sahidic Coptic"): 

   Because the use of the Coptic articles, both definite and indefinite,
   correspond closely to the use of articles in English, only exceptions
   to this general correspondence will be noted.
  
In light of the above, the dialects of the Coptic language are a valuable, 
albeit greatly untapped, resource for New Testament researchers and 
translators. It is a pity that more research has not been done into these 
important version of the New Testament.

For these reasons a new version of the New Testament in Sahidic Coptic is
being produced.

The last print version of the entire Sahidic version came out in seven
volumes from 1911 to 1924. This edition is quite rare. The only available
copy found by a recent used book search (12/2003)* was partial and priced 
at $1300. Until 2000 the available only editions were electronic. One from 
the Packard Humanities Institute (PHI) in a obsolete database format. 
Another was put out by the Saint Shenouda the Archimandrite Coptic Society 
in 1998 in PDF format. Each of these volumes cost about $50 each. Neither 
edition is in a standardized format. 

Starting in 2000, using the PHI edition by permission, I began compiling 
and collating a new edition in a standardized format. The Sahidica edition
will be available online, free of charge.

The purpose of the edition is to make available a free, diglot edition of
the Sahidic Coptic version and along with original Greek version, presented 
in parallel, as an easy to use html version. The edition is designed to be
useful primarily to students and researchers, but also of a quality to be
of use to scholars and translators.

* A more recent search (11/2005) found this edition no longer available.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Note on Dialect Names

Sahidic is also called Thebaic
Bohairic is also called Memphitic
Fayumic is also called Bashmuric
Fayumic is also spelled Fayyumic and Fajumic
Oxyrhynchite is also called Middle-Egyptian
Proto-Bohairic is also called Old Bohairic
Sub-Achmimic is also called Lycopolitan

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Primary Sources:

Aland and Aland, "The Text of the New Testament." Michigan, 1989.

Metzger, "The Early Versions of the New Testament: Their Origin, Transmission, 
and Limitations." Oxford, 1978.

Metzger, "The Text of the New Testament. Its Transmission, Corruption, and
Restoration." Third Edition. Oxford, 1992.

Schmitz and Mink, "Liste Der Koptischen Handschriften Der Evangelien: Die 
Sahidischen Handschriften Der Evangelien." Berlin - New York, 1986.

Other sources include: Crum, Hintze and Schenke, Horner, Husselman, Kahle, 
Kasser, Kenyon and Adams, Lambdin, Plumley, Quecke, Schenke, Schussler, 
Thompson and Willis.

J. Warren Wells

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

ADDENDUM 1 (2004): Brief Overview of Greek-Coptic Diglots.

    "The Coptic version should be studied in conjunction with the original
    Greek; only in this way can the reader gain a clear understanding of 
    the translation techniques employed and of the influence the original 
    has had on the grammar, vocabulary, and style of the Coptic
    translation."  -- Thomas Lambdin

Although very rarely stated in either secular or scholarly literature, several
early Greek manuscripts (in fact about 7 percent of all papyri and uncials)
are Greek-Coptic diglots. These include 6 papyri dating from the fourth 
century to eighth century and 20 uncials dating from the fifth century to the 
tenth century. (This is according to Aland and Aland, ppg.96-128, which fails 
to note that p6 contains portions of John and James in Coptic.)

Included among the papyri are:

  p2,  John, VI
  p6,  John, James, IV
  p41, Acts, IV/V (VIII)
  p42, Luke, VII/VIII
  p62, Matt, IV
  p96, Matt, VI

Included among the uncials are:

  029,  e, V  (= 0113, 0125, 0139) 
  070,  e, VI (= 0110,  0124, 0178-0180, 0190, 0194, 0202)
  086,  e, VI
  0114, e, VIII
  0164, e, VI/VII
  0177, e, X
  0184, e, VI
  0200, e, VII
  0204, e, VII
  0205, p, VIII
  0236, a, V
  0237, e, VI
  0238, e, VIII
  0239, e, VII
  0275, e, VII
  0276, e, VIII
  0298, e, VIII/IX
  0299, e, X/XI

Plus two uncials now listed as lectionaries:

  Lectionary 963, e, VII (= 0100, 0195)
  Lectionary 1575, e, IX (= 0129, 0203)

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

ADDENDUM 2 (2005): Note on Christology in the Coptic Versions of John.

It was already noted that Coptic was the first language the New Testament
was translated into that has the indefinite article; and the only language 
with the indefinite article that was produced during the Koine Greek period. 

The is of interest because, in Coptic versions, John 1:1b is commonly 
translated "the word was with God and the word was a God" using the Coptic 
indefinite article; with some variation in word order.

In the proto-Bohairic version (Papyrus Bodmer III, the text of which was 
partially reconstructed by Rodolphe Kasser) the first occurrence of "God"
in John 1:1 is in the Nomina Sacra form, whereas the second occurrence is 
spelled out. In John 1:18 the word "God" (which no one has seen) is in 
the Nomina Sacra form, while the word "God" (only-begotten) is spelled out.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

ADDENDUM 3 (2006): A Tabulation of Kasser's Timeline for Coptic Dialects.

 Date CE    Sahidic           Bohairic
------------------------------------------------
  150       Preliminary       ...
  
  200       Pre-Classical     ...
  
  250       Classical         ...
  
  300       ...               Pre-Classical
  
  500       ...               Classical
  
  650       Final             ...
  
 1000       ...               Final
------------------------------------------------
 The source is Kasser's essay as reproduced in 
 Metzgers "Early Versions of the New Testament
 Their Origin, Transmission, and Limitations." 
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
First edition 1989
Eighth revision 2008 
Copyright 1989-2008 by J. Warren Wells.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Why a New Edition

Since Horner's edition was published, a large number of Sahidic manuscripts have come to light. Some have been collated and published. The only new edition of the entire Sahidic New Testament to appear since Horner is a computer database version from the Packard Humanities Institute (PHI).[3]

The PHI version updated Horner's work with more recent collations. However, while the PHI version is a major improvement on Horner, the PHI version has its own idiosyncrasies.

A new text of the Sahidic New Testament has been developed. It is called the Sahidica text and is based on the Coptic New Testament developed by the Packard Humanities Institute (PHI). In turn, the PHI version was based on the edition by George Horner.[4]

PHI Version Basis

The following newer editions were used to update Horner.[6]

• For Matthew: El Evangelio de San Mateo Copto Sahidico, edited by Gonzalo Aranda Perez. Madrid, 1984. (Based on M 569 in the Pierpont Morgan Library.
• For Mark: Das Markusevangelium Saidisch: Text der Handschrift Ppalau Inv. Nr. 182 mit den Varianten der Handschrift M 569, edited by Hans Quecke. Barcelona, 1972.
• For Luke: Das Lukasevangelium Saidisch: Text der Handschrift Ppalau Inv. Nr. 181 mit den Varianten der Handschrift M 569, edited by Hans Quecke. Barcelona, 1977.
• For John: Das Johannesevangelium Saidisch: Text der Handschrift Ppalau Inv. Nr. 183 mit den Varianten der Handschrift 813 und 814 der Chester Beatty Library und der Handschrift M 569, edited by Hans Quecke. Barcelona, 1984.
• For Acts and the Pauline corpus: The Coptic Version of the Acts of the Apostles and the Pauline Epistles in the Sahidic Dialect, edited by Sir Herbert Thompson. Cambridge 1932.

In many ways, the PHI version improved on Horner's text. This can be illustrated by considering research involving the textual history of the shorter and longer endings of Mark in the Sahidic version. When Horner prepared his edition, his text for the ending of Mark was based on two fragmentary manuscripts; Horner's 50 (ninth century) and 108 (eleventh century). By 1951, research into newer discoveries allowed P. E. Kahle to present a clearer history of the text based on four additional manuscripts.

This history is presented in Kahle's article "The End of St. Mark's Gospel. The Witness of the Coptic Versions."[7] Based on the text in Horner's two manuscripts and the four additional manuscripts,[8] Kahle pointed out that only one manuscript[9] (the latest) "regards [Mark] 16.9-20 as part of the original text and indicates no alternative." The other five manuscripts all contain both the shorter and longer endings of Mark but "indicate by short notes that these [two alternative endings are] found in some manuscripts." On this basis, he goes on to say, "the other [earlier] Sahidic manuscripts . . . all contain evidence that some (older) manuscripts ended at [Mark] 16.8." Kahle's observation was later proven correct.

The Barcelona manuscript, edited by Quecke[10] and published in 1972, dates from the first half of the fifth century.[11] It is older that those compared by Kahle and it ends at Mark 16:8. It contains neither the short nor the long ending of Mark; thus substantiating Kahle's conclusions.

In the PHI version, the text from this earlier manuscript replaces Horner's; which was based on later manuscripts. In this way, the PHI version improved on Horner by incorporating a later discovery that probably reflects an earlier stage of the Sahidic New Testament. However, while this simple, direct method of updating Horner has definite benefits, it also has drawbacks.

PHI Version Methodology

In a similar instance the Barcelona manuscript and the PHI version omit the phrase "son of God" from Mark 1:1. By comparison Pierpont Morgan MS XI (M 569) does contain the phrase. In fact, the Nestle-Aland Greek text indicates that the Barcelona manuscript stands alone in omitting the phrase.[12] Other Sahidic texts include it. So, in this case, the PHI version does not represent the Coptic as a whole.

Moreover, in some cases, other Sahidic manuscripts represent the Greek more faithfully than the Barcelona and PHI versions. For instance, both omit the phrase epecht eqalassa, found in Mark 5:13. In addition, both omit the entire verse at Mark 15:14.[13]

In all fairness, however, it must be pointed out that the PHI version was never meant to be a critical text. The goal was more pragmatic. According to David Brakke, the editor of the PHI version, the goal was to make an up-to-date electronic version available for people to use.[14] This goal, which necessarily precludes any idea of a critical apparatus, was achieved by the PHI version.

In addition, any apparent inconsistencies in the PHI version are, to some degree, defensible. This is because the original Sahidic texts themselves are far from being internally consistent in content and usage. This fact is pointed out by Rodolphe Kasser in describing the emergence of the official, classical Sahidic version during the late third century.

"Even though this first [official] Coptic version rested upon the Greek text, it was far from being a robe without seams. On the contrary it was a heterogenous work composed of disparate pieces, as one can prove by analyzing its vocabulary. It was not the work of a single translator or of a group of translators working together. It reveals, rather, the labours of extremely diverse authors, separated by time (or space), whose productions were put together finally because they found favour in the eyes of a commission of experts who made a more or less thorough revision."[15]

In a sense, the PHI text began an effort not unlike the one just described. Like those who put together the classical Sahidic version, the PHI editors took the first step of putting together what were considered the best texts. But unlike their third century predecessors, the editors of the PHI text did not attempt to take the second step; to create a revision.

Analysis of the Resultant Text

The above example of the shorter and longer endings of Mark illustrates part of this problem. For a clearer view, this example must be put in perspective. Bear in mind that several individual editions were used to replace readings in Horner's single edition. Each replacement text represented one manuscript, and one modern editor, in matters of usage and style. Thus a patchwork of textual and editorial idiosyncrasies were introduced. This is especially true in the use of supralinear marks and punctuation, as well as nomina sacra usage. This approach resulted in a disjointed and inconsistent text. While it is indeed an improvement, the PHI version is actually less consistent than Horner.

Two aspects of this problem introduced into the PHI text involve the use of the supralinear stroke and punctuation. Each replacement work could reshape a single book by following the usage in a single manuscript, by editorial preferences, or by both. As a result, usage of the supralinear stroke and punctuation in the PHI version is inconsistent to the extreme.

Consider the analytical statistics. In the PHI version there is an average of 5.34 supralinear strokes per verse overall. Yet, while Peter's epistles average over 9 per verse, Revelation averages less than 1 per 3 verses (0.32). Consider also that Mark, Luke, and John (all from Quecke) have about 5.34 supralinear strokes per verse, while Matthew (from Aranda Perez) has a total of only 3 supralinear strokes in the entire book.

As for punctuation, the standard Greek text (UBS4) has an average of 2.49 punctuation marks per verse (mostly commas, with only 0.71 periods per verse). The PHI text has an average of 2.07 punctuation marks per verse. Yet where the standard Greek text is editorially consistent in the matter of punctuation, the PHI version is most definitely not.

At one end of the PHI spectrum, 1 John has an average of 3.27 periods per verse.[17] At the other end of that spectrum we find Revelation with a total of 4 punctuation marks in the entire book. In addition, of all the punctuation marks used, there are only two commas and two semicolons in the entire PHI New Testament.

Another problem involves consistency in spelling.[18] Users of standard Greek texts have come to expect a high degree of consistency in spelling, even though the ancient Greek papyri and uncials are by no means consistent in this regard.

Inconsistencies in the PHI text often involve variant spelling. For example, words with double letters and the ei combination are often inconsistent. Thus we have jekaas and jekas, naHrn and nnaHrn, joeis and jois. In addition, there are variations in imported Greek words like farrisaios and farisaios, and porneia and pornia. We also find variants like eeie, eie, and ei. Moreover, due to the similarity of consonantal sounds, we find errors like nefarmagos in Re.21:8, and the correct form nefarmikos in Re.22:15. Other changes involve irregular usage of the aspirate H. We find both abraHam and abraam, parHousia and parousia, etc. Following the example of standard Greek editions by standardizing spelling in the Sahidic version would undoubtedly be beneficial; especially in comparing the Sahidic and Greek.

Yet another aspect of the problem is found in nomina sacra usage. Today's standard Greek texts do not use nomina sacra forms, even though these are common in the papyri and uncials.

While some Greek words represented by nomina sacra forms are found in the PHI text; they are inconsistent. The word Jerusalem, for example, has four nomina sacra forms,[19] as does Israel. Stauros, with its related verb, has six forms.

In addition, some words are represented in multiple forms within individual books of the PHI version. For example, the nomina sacra form of joeis (Lord)s[20] occurs only in the book of Revelation in the PHI version. Yet even there it occurs inconsistently; in Re.15:3 the word appears in the phrase pjoeis pnoute ppantwkratwr and then in Re.16:7 as pjs noute ppantwkratwrs[21] with the nomina sacra form.

On the other hand, the nomina sacra form of pneuma (spirit) is consistently represented as pma throughout the PHI text. But, unlike Greek texts, is applied globally. That is, the PHI text represents God's spirit, wicked spirits, and spirit in general with the "sacred" form.

Another facet of this consistency problem with some nomina sacra forma is the fact that the electronically encoded format used for marking them in the PHI database is itself inconsistent. For instance, the name Jesus is represented as <IS>, [IS], \I\S, and I\S. These variant methods also apply to other words; e.g. "Christ."

Obviously, all such variation in the usage and format of the nomina sacra forms makes reading, translation, and especially automated analysis difficult. As mentioned above, following the example of Greek editions by replacing all nomina sacra forms with the normal words would facilitate reading and analysis. Again, this would be helpful in comparing the Sahidic and Greek.

One more item must be considered. This is the issue of scribal error; in this case, by modern scribes. There are cases where the PHI version differs from the source texts due to errors in database input. One such instance is found in Mark 1:5 where the word ebol was missed during input. Another error is found at Mark 2:24, where neqarissaios[22] was input into the PHI database, quite obviously by mistake.

This is not to say that the PHI edition was shoddily done. To the contrary, there is evidence that it was quite carefully done. For example, Quecke himself erred when transcribing Mark 16:7, which error has been corrected in the PHI version. If one compares the last leaf of Mark's gospel in P. Palau Rib. 182 with Quecke's edition, one will see where the last line of the left column in the manuscripts[23] says tn etgali yet Quecke has tn etgagi.[24] However, when examining the copy of Quecke actually used by PHI editor David Brakke, one can clearly see where the spurious g has been crossed out and the correct l has been written right below it. Thus the PHI editor spotted and corrected the error.

The Sahidica Text and eMss

The development of this new version has been based, to a large extent, on statistical analysis of various source documents. To facilitate this effort several Sahidic texts have been put in an electronic format to allow automated verification and analysis.

These electronic manuscripts (eMss) are the actual basis for the Sahidica text. An example text can be found on the Sahidica Project website.[25] This example text has the entire text of first Peter from the Nestle-Aland Greek and three Sahidic manuscripts,[26] along with the Sahidica text.

The development of the Sahidica text has, from its inception, been based on automated systems. More specifically, it has been developed using proprietary analysis, formatting, and pattern-recognition software created by the editor. The Sahidica text was based on The Coptic New Testament, an electronic publication, which was edited by David Brakke (Cop0001.Txt on PHI-CD 5.3 dated 7/2/91. Los Altos. Packard Humanities Institute. 1991).

However, the Sahidica text is very different from the PHI version. Literally thousands of changes have been made to the text. These changes involve primarily specific standardization and a partial level of clitic disintegration (i.e. the separation of certain proclitic words). Such changes were made to enhance readability and simplify statistical analysis of the text. Changes to the actual content of the text are fewer in number and have been made only where the PHI version has errors or idiosyncratic readings that fail to represent the overall Sahidic textual tradition.

Most of these changes have been made by pattern-recognition and analysis software, developed for this purpose by the editor. In turn, the resultant text has been further analyzed, validated, and formatted using automated systems.

At this juncture, it is important to make a clarification. As in the case of the master text for first Peter (mentioned above), the resultant Sahidica text was not produce by the automated systems. Rather, it was produced by hand using the information supplied by the automated systems. Therefore, while automation does much of the collation and analysis, editorial decisions are still left to the editor.

Sahidica Version Basis

Section	Primary Basis	Collated Primarily Against
Matthew	Pierpont Morgan Library M-569	Bodmer XIX
Mark	PPalau Rib. Inv.Nr.182	Pierpont Morgan Library M-569
Luke	PPalau Rib. Inv.Nr.181	Pierpont Morgan Library M-569
John	PPalau Rib. Inv.Nr.183	Pierpont Morgan Library M-569
Acts	Chester Beatty B (Copt.Ms.814)	Berlin P.15926 + British Museum MS Or.7594
Pauline	Chester Beatty A (Copt.Ms.813)	Pierpont Morgan Library M-570 + M-571
James-Jude*	Pierpont Morgan Library M-572	Pierpont Morgan Library M-573 + M-601
Revelation	Berlin MS Or.408 = British Museum Ms Or.3518	British Museum MS Or.6803

*Note: First Peter was collated primarilly against the Crosby-Schøyen Ms.193.

Sahidica Version Goal: Usability

The Sahidica text is intended to be a coherent, consistent, and easy-to-use edition. Moreover, the text is meant to facilitate analysis (both visual and algorithmic). In addition, it was felt by the editor that a new edition of the Sahidic should take definite steps toward the kind of uniformity in spelling, punctuation and other usage as is found in standard Greek editions. Thus the overall goal is to create a standardized version for people to use.[27] In turn, the methodology to reach this goal requires high levels of uniformity, and low levels of complexity; thus making standardization and simplicity the primary guidelines.

It should be clear that the Sahidica text is intended primarily for use by at students and instructors, and to a much lesser degree as a reference tool for translators and textual critics. Part of the goal is to make the Sahidica text freely available and easily accessible in electronic formats. This includes plain text and a electronic "print" version with parallel Sahidic and Greek[28] texts, which uses a single standard font in order to facilitate comparison of the two versions.

With this limited goal in mind, we can now describe the methodology used to develop the Sahidica text.

The Sahidica Methodology

One change involves a move toward standardized spelling. Users of standard Greek texts have come to expect a high degree of consistency in spelling, even though the ancient papyri and uncials are by no means consistent in this regard. It is obviously desirable to attain such consistency in the Sahidic, is it not?

In light of the inconsistencies in the PHI text (as mentioned above), the spellings of many words in the Sahidica text have been standardized to facilitate reading, translation and analysis.

Another change involves usage of the Nomina Sacra. Standard Greek texts do not use the nomina sacra, even though these are common in the papyri and uncials. In Horner, PHI, and other collations they do appear.

All such variation makes reading, translation, and especially automated analysis difficult. It was therefore decided that all nomina sacra forms would be replaced by the normal word in the Sahidica text. Again facilitating comparison with the Greek text.

Supralinear strokes and punctuation changes have been made. Standard rules for supralinear marking have been applied globally to the Sahidica text. So where Matthew in the PHI version has only 3 supralinear strokes, Matthew in the Sahidica text has about 4000. Additional direct editing of the text will undoubtedly produce even more changes.

Please note however that all supralinear strokes have been omitted in the current version.

Punctuation has been added (and in some instances removed) to correspond more closely to the standard Greek texts - again to facilitate reading and comparison.

Clitic disintegration is another change. To some degree, proclitic elements have been broken out, thus becoming stand-alone words. This also was done in order to enhance the readability and analysis of the text. Where practical, other cumbersome proclitic structures have been disintegrated.

Here are a few examples of clitic disintegration:

• Common prepositions such as Hn, ejn, Hi, Hijn, etc.
• Common prepositional phrases like ebol Hn.
• Common phrases like ete pai pe.

Paragraphs and section breaks have been added. These are based primarily on the UBS4 Greek text. This also was done to enhance readability and to simplify comparison with the Greek text.

Notes on Textual Changes

• Where the PHI edition fails to represent the overall Sahidic tradition.
• Where the PHI editors omitted text or misspelled words.
• Where the PHI text reflects a manuscript-specific, idiosyncratic reading.
• Where the edition underlying a portion of the PHI text is in error.
• Where brackets are required to better represent the overall Sahidic tradition.
• Where brackets were required to better understand the underlying Greek text.

Yet, such cases are not common. So for the most part the Sahidica text differs from the PHI edition, not so much textually as technically; in enhancements meant to facilitate reading, comparison, and analysis (both visual and algorithmic). Again, the goal is simply to create a coherent, consistent, and easy-to-use edition; a text meant primarily for students and instructors of the Sahidic New Testament.

Notes on the Greek Version

The purpose is to provide a parallel Greek text for students and researchers who work with the Sahidic version. At the same time fabricating a hypothetical version of the Greek to correspond to the Sahidic has been avoided. To accomplish, this the CCAT lemmatized version of the UBS text corrected by James Tauber, was used as a basis. That Greek text has been altered to correspond to the Sahidic only where there is support in the UBS apparatus. This is a simplified text. The following marks are omitted: Accent, breathing, dieresis, subscript and elision. Punctuation marks have been removed except for periods and Greek question marks (";") which have been retained. This allows the student or researcher to compare differences between the Sahidic and the standard Greek texts. As a result this Greek text is encoded to be viewable with the Symbol font available in Microsoft Windows(R).

To better serve the student or researcher, the following steps have been taken in modifying the Greek text.

Brackets in the UBS/NA have been removed. Double brackets have been left. Commas have also been removed. Single brackets mark where the Sahidic witnesses read differently than the standard Greek text. Single brackets with hyphens within [--] indicates the Sahidic omits one or more words found in the Greek text. Where single brackets with hyphens within stand alone by a verse number, the Sahidic omits the entire verse. Double brackets with hyphens mark where the Greek text is not represented in the Sahidic witnesses. Double brackets with hyphens within [[--]] indicates the Sahidic omits a major section found in the UBS Greek.

Note on Interpolations

The shorter and longer endings of Mark appear in most Sahidic mss. Herein the longer is given first. The pericope of the adultress (John 7:53 - 8:11) does not appear in any Sahidic mss. Herein it is omitted.

Limitations of the Sahidica Text

In addition, the development and use of the analysis software is itself a learning process and will enable further refinements the text. Many of the changes discussed in this paper have been made by pattern-recognition and analysis software. These applications have been programmed by the editor.

Algorithms, however, cannot safely solve the countless problems involving something as mercurial as human language - even one that's frozen in the past. While the editor has been programming highly specialized pattern-recognition and analysis software, those years of experience clearly show that unforeseen errors are introduced into the text by the formatting programs. Therefore, any results produced by an automated approach must be, to some degree, suspect. For this reason, it must be assumed that much traditional editorial work is required.

Additional Collation

The CD is (at the time of this writing) available from the Saint Shenouda the Archimandrite Coptic Society in Los Angeles, California. The price of the CD when I purchased it was $50 US.[29]

The fact that the Sahidic NT on the CD was in PDF format allowed me to double check my work at that time. After temporarily reducing the PDF text to a binary form, I was able to compare the entire Sahidica, PHI, and Coptic Society texts in parallel and letter by letter. This allowed me to isolate all the differences in the three texts and correct mistakes where they could be identified as compared with the Coptic printed editions used in the collation.

References

1. Rodolphe Kasser. Quoted in The Early Versions of the New Testament: Their Origin, Transmission, and Limitations, by Bruce Metzger. pg. 129.
2. For example, William Willis states that the Crosby-Schøyen manuscript was "at least a copy of a copy, the original translation on which it is based must be pushed back to A.D. 200, perhaps even earlier." The Crosby-Schøyen Codex MS 193, pp. 137, 138.
3. This was released by the Packard Humanities Institute as a database on their PHI-CD 5.3. The database used was dated 2 July 1991.
4. The Coptic Version of the New Testament in the Southern Dialect otherwise called the Sahidic and Thebaic, by George Horner was released as 7 volumes from 1911 to 1924.
5. Cf. Kurt and Barbara Aland's The Text of the New Testament, pg. 201; Bruce Metzger's The Text of the New Testament, pg. 79, and The Early Versions of the New Testament: Their Origin, Transmission, and Limitations, pg. 109; and F. G. Kenyon's Text of the Greek Bible, pg. 137.
6. According to a document entitled "Text Editions Used in the Preparation of the Coptic New Testament" provided by the Packard Humanities Institute. The document was prepared by David Brakke (the editor of the PHI version) and dated February 1990.
7. Journal of Theological Studies, n.s. II (1951), pp. 49-57.
8. These four mss are: Pierpont Morgan MS XI (M 615); Pierpont Morgan MS IV (M 569); Vienna K. 9075, 9076; and British Museum Or. 7029. The respective dates of these mss are: seventh to eighth century; eighth to ninth century; eighth century; and late tenth century.
9. British Museum Or. 7029.
10. Located at the Seminario de Papirologia de la Facultad Teologica in Barcelona. PPalau Inv. Nr. 182.
11. The Early Versions of the New Testament: Their Origin, Transmission, and Limitations, by Bruce Metzger. pg. 113.
12. The critical apparatus in Novum Testamentum Graece (27th edition ) indicates that a single manuscript (ms) omits the phrase and rest of the Coptic texts (Co) include it.
13. Mark 15:14 in M 569 reads, peje pilatos de nau je ou pe ppeqooy ntavaav. Ntoou de nHouo auji Skak ebol eujw mmos je staurou mmov.
14. From David Brakke via personal email.
15. From Kasser's essay "Les dialectes coptes et les versions coptes bibliques" as quoted in Metzger's Early Versions of the New Testament: Their Origin, Transmission, and Limitations. pp. 130-131.
16. By relating consistency to usability I mean a level of standardization that enhances the readability and statistical analysis of the text. This level of usability is evident in the standard Greek texts, which are consistent in spelling, punctuation, accenting, etc. even though the Greek mss themselves are by no means consistent.
17. An extreme case is 1 John 2:14, where the PHI text has 11 periods in the verse. UBS4 has three commas and three periods.
18. It is noteworthy that the original Sahidic texts were quite divergent in matters of spelling. Though presented here as part of a "problem" with the PHI text, no blame in insinuated; either toward the PHI editors or the editors of the texts. The problem is one facing anyone wishing to create a consistent, yet truly representative, edition of the Sahidic version.
19. qilhm, ielhm, ilhm, and ihm. Jerusalem is also spelled out at least four different ways in the PHI text. Yet notably, even in modern Greek texts Jerusalem appears in two forms; the neutral plural ierosoluma and the feminine singular ierousalhm.
20. The form is js, which appears 16 times in Revelation.
21. Interestingly we also find pantwkratwr (almighty) spelled "correctly" (i.e. as the Greek) pantokratwr only at 2Co.16:7 and Re.1:8 in the PHI text.
22. I.e. "the Pharisees" misspelled "the Tharisees."
23. It just so happens that most anyone can look at this. The last page of the manuscript appears in plate 5 9 on page 202 of Aland and Aland's Text of the New Testament. There, the correct spelling is clearly visible.
24. Das Markusevangelium Saidisch: Text der Handschrift Ppalau Inv. Nr. 182 mit den Varianten der Handschrift M 569. Page 179.
25. You may view the text on the Sahidica.org website
26. These are: the PHI text (=Horner), Pierpont Morgan Library M 572, and Crosby-Schoyen Codex MS 193.
27. N.B. The goal is "a" standardized version, not "the" standardized version. However, it is hoped that the Sahidica text will be of sufficient quality to serve as a basis for future efforts; and perhaps, even for a critical edition.
28. Sahidica: The Sahidic New Testament with Parallel Greek.
29. Saint Shenouda the Archimandrite Coptic Society. Coptic CD Volume I. April 1998.