Today we will explore a subject that has a bearing on the following twelve questions:
● Matthew 27:17: did Pilate describe Barabbas as if Barabbas was also called Jesus?
● Mark 1:1: did Mark describe Jesus as the Son of God, or not?
● Luke 23:42: Did the dying repentant thief call Jesus “Lord,” or not?
● John 1:18: does the author refer to the only-begotten Son, or to the only-begotten God?
● John 9:35: did Jesus ask the no-longer-blind man if he believed on the Son of God, or on the Son of Man?
● Acts 20:28: did Paul refer to the
, or to the church of the Lord, or to the
church of the Lord, even God? church of God
● Romans 14:10: Did Paul mention the judgment seat of Christ, or the judgment seat of God?
● First Corinthians 10:9: Did Paul say that the Israelites in the days of Moses tempted Christ, or that they tempted the Lord, or that they tempted God?
● Philippians 4:13: Did Paul specify, “I can do all things through Christ who strengthens me,” or should the final phrase be, “through the one who strengthens me”?
● First Timothy 3:16: did the original text refer to “God manifest in the flesh,” or not?
● First Peter 3:15: who should we sanctify as Lord in our hearts: Christ, or God?
● Jude 1:5: did Jude state that the Lord saved a people out of
, or that Jesus had done so? Egypt
It is not my intention to resolve any of these textual contests today. Instead I want to explore a scribal mechanism which may have been involved in their creation: the reverential use of contractions to write sacred names. These contractions, which appear in almost all New Testament Greek manuscripts, are called nomina sacra (singular: nomen sacrum), sacred names.
Other contractions can be found in Greek manuscripts. The words και (and) and περι (about) are often abbreviated; in chapter-lists which precede each Gospel in many manuscripts, περι is typically shortened to its first two letters, written vertically, as the first word of chapter-titles. And very frequently, when the letter νυ is at the end of a line of text, it is represented by an overline instead of being written. The nomina sacra, however, form a special class of contractions; they were not made in order to conserve time or materials, but as expressions of reverence.
Even before any New Testament books were composed, copyists of books of the Old Testament already treated the name of God with special reverence. In Hebrew, this name consists of four Hebrew letters, and for this reason is known as the Sacred Tetragrammaton. In some of the Dead Sea Scrolls, when the Sacred Tetragrammaton appears, it is written in its own distinct script; the copyists used ordinary lettering elsewhere but for the name of God, paleo-Hebrew letters were used. 11Q Psalms, a scroll of Psalms, has many examples of this.
|The Sacred Tetragrammaton|
In intertestamental times, the Old Testament was translated into Greek; the most popular Greek translation was known as the Septuagint. In the first half of the 200’s, the patristic writer Origen, in his Homily on Psalm 2, as he commented about the second verse, made this observation about some copies of the Septuagint: “In the most accurate manuscripts, the name [i.e., the name of God] occurs in Hebrew characters – not in modern-day Hebrew, but in the very ancient lettering.”
|A replica of the Tetragrammaton in one of the |
Dead Sea Scrolls. See the photo by Shai Halevi
at the Leon Levy Dead Sea Scrolls
Digital Library - Image B-371124.
The Habbakuk Commentary
has other examples.
Jerome, in about 390, as he wrote his preface to the books of Kings – the Prologus Galeatus, or “Helmeted Prologue” – said something similar: “And the name of the Lord, the Tetragrammaton, we find in certain Greek volumes written, up to now, in ancient lettering.”
In Greek manuscripts of Old Testament books produced by Christian copyists, the name of God was replaced with the Greek word Κυριος, or with the contraction ΚΣ, accompanied by a line above the letters. Other copyists, however, continued to give the name of God special treatment. For example, in P. Oxy 3522 (a fragment with text from Job 42:11-12, produced in the first century A.D.) and Papyrus Fouad 266 (a fragment with text from portions of Genesis and Deuteronomy, produced in the first century B.C.), the name of God is repeatedly written in paleo-Hebrew letters, even though the main text is Greek. The Nahal Hever Scroll of the Minor Prophets also clearly displays the use of paleo-Hebrew (or paleo-Aramaic) letters reserved for the Tetragrammaton.
In a Greek fragment of Leviticus among the Dead Sea Scrolls (4Q120), the Sacred Name is treated differently: the copyist wrote neither the Hebrew Tetragrammaton nor the contraction for Κυριος; instead he wrote ΙΑΩ, a series of Greek vowels which could be used to vocalize the Tetragrammaton; these letters served as a proxy for the Tetragrammaton itself. (The IAW-vocalization is also found on some charm-talismans.)
|P. Oxy 3522 (See the detail |
at the Oxyrhynchus Online website)
Even later, in the 400’s-500’s, some copyists of Greek manuscripts still gave the Tetragrammaton special treatment. This is seen in a fragmentary palimpsest of portions of Aquila's Translation of First & Second Kings found at the Cairo Genizah in the 1800’s and published by F. C. Burkitt in 1897. (It was found with some other palimpsests.) In the lower writing, which consists of text from First & Second Kings (or, Third & Fourth Kings, using the Septuagint’s titles), the Tetragrammaton, written in Hebrew letters, is repeatedly embedded in the Greek text.
|Enough of this page of Aquila's translation has |
survived to show that when it was intact, it
featured the Tetragrammaton in Second Kings 23:16.
(Transcription and Plate from F. C. Burkitt.)
Unfortunately, the lines containing Second Kings 23:24 are not readable in the photograph provided in Burkitt’s book. Assuming that Burkitt’s transcript is correct, and noticing that the lines of Greek in the manuscript are not justified on the right side, this raises a new question: why wasn’t there room for the Tetragrammaton to be added? Or to put it another way: why didn’t the copyist simply start the Tetragrammaton on the next line?
|P. Berlin 17213, at the Papyrus Databank|
of the Egyptian Museum in Berlin
Before tackling that question, let’s look at a feature in Papyrus
This fragment, which contains text from Genesis 19:11-13 and -19, has been assigned to the 200’s. In verse 18 (on the recto), after the word δεομαι, there is
a small blank space. Normally there
would be a nomen sacrum here. In Scribal Practices and Approaches Reflected in the Texts Found in the Judean Desert
(published by E. J. Brill), Emmanuel Tov proposed that this blank space
“denotes a closed paragraph after Gen. 19:18,” but he grants (in a table on
page 266) that it is “possibly space left for divine name,” that is, we may see
here a blank space reserved for the nomen
sacrum ΚΕ (or the Tetragrammaton) that was never filled in. Berlin
Similarly, as Tov reported on page 226, in the Dead Sea Scroll fragment 4QpIsae, “A space was left open in 6 4 where MT (32:6) has a Tetragrammaton. This space may have been left for a Tetragrammaton, to be filled in possibly by a different scribe (or was the Tetragrammaton omitted intentionally, indicated by a space in the middle of the line?).”
Pages 226-229 of the
Kacmarcik Codex, produced in 1344,
contain a guide to writing
the nomina sacra.
Tov also observed that in the Dead Sea Scroll 11QPsa, “It appears that the scribe left irregular spaces, and that at a later stage someone, possibly the original scribe himself, penned in the Tetragrammata, sometimes squeezing them in between the surrounding words.” Tov also noticed the extra space surrounding the occurrences of the Tetragrammaton in P. Fouad 266b. He concluded: “The above evidence shows that at least in some
Qumran texts, the Tetragrammaton was filled in
after the writing of the main text, and this was also the case in one
manuscript of the LXX.” [Bold print added.] (This explains
why the copyist of Papyrus 17213 did not simply move along to the
next line: the next line had already
been written.) Berlin
The practice of leaving blank spaces, to be filled in later with the Tetragrammaton, developed as an improvement of the scribal custom of reserving one pen exclusively for writing the Tetragrammaton. As an alternative to putting down the primary pen, picking up the reserved pen, testing it, and writing the Tetragrammaton with it, putting it down, and picking up the primary pen again, over and over, efficiency was greatly increased by leaving blank spaces, and then, when the text was proof-read, inserting all the occurrences of the Tetragrammaton with the specially reserved pen.
Did this phenomenon have something to do with the origin of nomina sacra in New Testament manuscripts? I believe that it did, to a limited extent. As I describe the effects of this phenomenon, keep in mind that we can’t interview the ancient copyists; some things have been deduced and some of the deductions are calculated guesses.
Very early in the transmission-history of the Gospels, Christian copyists developed four names to be treated in a manner congruent to the Judaic treatment of the Tetragrammaton; the first four Christian nomina sacra probably appeared as a group: ΚΣ , ΘΣ, ΙΣ, and ΧΣ. It was at this early stage that some Christian copyists (probably copyists who were already engaged in the production of copies of the Septuagint), modifying the practice of leaving space for the Tetragrammaton and adding it in a second copying-stage, left overlined space for these four words in the first copying-stage, and they were added during the proof-reading stage. Occasionally the proof-reader worked from memory and interchanged the names, or failed to insert a contraction in the space reserved for it.
The next words to become nomina sacra were Πατηρ, Υιος, and Πνευμα. Following this, the group of contracted words was expanded to include words that were components of titles of Christ – the Son of Man, the Son of David – or which were paralleled in the Gospels by a sacred name (as, frequently, Matthew refers to the kingdom of heaven where Mark refers to the kingdom of God). The contraction of Σωτηρ probably began at the same time, in the same way; it was considered a title of God and/or Christ.
The contractions of “
” and “ Jerusalem ” may have originated as ordinary
abbreviations of names which obtained the same format as the nomina sacra to keep format-variations
to a minimum. (The presence of the
letters ι and η in these two nomina sacra
may have had something to do with their adoption, too.) The contraction of μητηρ was a result of
increased devotion to the Virgin Mary.
This leaves the origin of one nomen
sacrum unaccounted for: why would
σταυρος be considered a word worth venerating? Israel
The answer might have something to do with Greek numerals. The Greek numerals from 1-999 were written as combinations of the 24 letters of the Greek alphabet, combined with three obsolete letters to make the system work. Thus, as shown in the accompanying chart, each letter had a numerical value. This was no secret code; it was the normal way to write numerals.
In the second-century composition The Epistle of Barnabas, 9:7, the author mentions an allegorical, Christ-centered interpretation of Genesis 14:14; the author interpreted the unusual term “dedicated” (or “trained”) to mean that they were circumcised. The number 318, he says, illustrates a pattern of salvation; the idea is that just as Abraham’s servants rescued Lot from captivity, Jesus on the cross rescues souls held captive by sin, and this is shown by considering the component-parts of the quanitity of 318: “300” is the cross-shaped letter tau (300) and the letters iota and eta are the first two letters in the name Ιησους (that is, “Jesus” in Greek).
In addition to the idea that Jesus and the cross, together, had a spiritual significance, the numerical value of the word σταυρος may have had something to do with its adoption as a nomen sacrum. Using the normal list of numerical values, the total value of the letters in σταυρος is 1,270. But if the first two letters are combined as the obsolete letter stau, or stigma, then the numerical value of ϛ+α+υ+ρ+ο+ς = 6+1+400+100+70+200 = 777, which, to copyists in the 100’s, might be seen as a sufficient reason to place it alongside Jesus’ name (the numeric value of which is 888) among the nomina sacra.
Four Peripheral Subjects:
(1) Three-letter Forms
The main nomina sacra (ΚΣ , ΘΣ, ΙΣ, ΧΣ) do not always appear as two overlined letters. The copyist of Papyrus 45 switched between two-letter and three-letter forms. In Papyrus 66, two-letter forms are used. In Papyrus 46, three-letter forms (such as ΚΡΣ , ΙΗΣ, ΧΡΣ usually consisting of the first two letters plus its last letter) are used. This might be explained in a number of ways: (1) The three-letter forms might be the original forms of the nomina sacra, or (2) Copyists slightly expanded some nomina sacra to make them a little easier to read, or (3) In a location where Greek and Latin were both spoken and written developed, some of the Latin nomina sacra were expanded to lower the risk that one would be confused with another, and a sense of tidiness motivated scribes to similarly expand their Greek counterparts.
(2) The Copyist of Codex Vaticanus and the Nomen Sacrum for Πνευμα
|The copyist of Papyrus 66 probably did not read|
the contraction for PNEUMA in his exemplar.
We don’t know. Perhaps the copyist held what Larry Hurtado has called a “binitarian” approach to devotional expression, and was not quite willing to express the same level of devotion to the Holy Spirit that he gave to the Father and the Son. Or, perhaps the copyist felt that the status of Πνευμα as a nomen sacrum caused too many interpretive difficulties, because the word often refers to unclean spirits and to the human spirit (but he could have chosen to not contract the word in those passages, like some other copyists). Likewise he may have felt that references to ordinary men, fathers, and sons were best left uncontracted, and to avoid making case-by-case decisions according to context, he decided not to contract any of them.
Or, possibly, the copyist of Vaticanus carefully followed exemplars which had been made in an era when only ΚΣ , ΘΣ, ΙΣ, and ΧΣ were contracted – exemplars so early that they sometimes exhibited signs of the very first stage of the usage of nomina sacra in copies of New Testament books, when copyists left blank spaces for the nomina sacra, to be added during proof-reading. Several variants in Vaticanus may support this idea; its reading in James 5:14 (where Vaticanus does not have the words του Κυ) is particularly interesting.
(3) Unusual Treatments of Nomina Sacra
Ludwig Traube, on page 22 of his groundbreaking 1907 book Nomina Sacra (written in German), mentioned the unusual treatment given to some of the nomina sacra in Codex N (022), a deluxe Gospels-manuscript probably produced in the mid-500’s. Its text, written on purple-dyed parchment, is mostly written in silver ink, but the nomina sacra Κς, Θς, Ις, Χς, Πηρ, Υς, and Πνα are written in gold. The production of such manuscripts, in such a format, must have involved two stages of production – the first using silver, and the second using gold. (The use of gold and silver ink, besides being a display of imperial wealth, may have been inspired by a statement in Josephus’ Antiquities of the Jews, Book 12, chapter 2 (which in turn seems to be dependent upon The Letter of Aristeas, part 176), stating that when a group of men from Eleazar the high priest visited Ptolemy in Egypt, they brought with them parchment copies of the Law, written in golden letters.)
Relatively few such ornate copies of the Scriptures in Greek have survived to the present day. Purple Codices (which, besides the Greek manuscripts N, O, Σ, and Φ, also include the Gothic Codex Argenteus, Old Latin i, Codex Vindobonensis 1235, and others) were being made not only in the mid-500’s but also earlier, in Jerome’s lifetime, for he complains about them in his Preface to Job, with words to this effect: “Whoever wants to keep the old books, let him keep them, whether written on purple skins with gold and silver, or in uncial letters, as they are called, with plenty of lettering but not a lot of letters. Let them leave to me and mine our modest little pages which are not so beautiful, but are carefully proof-read.” Perhaps Jerome realized that the copyists of the deluxe purple codices who added the nomina sacra (in Old Testament volumes, such as the Psalterium Turicense, as well as in New Testament books) were thus distracted from the task of proof-reading.
(4) Were the nomina sacra in P46 added during the proof-reading stage?
|In P46, the nomina sacra frequently are|
accompanied by extra space, which may indicate
that they were added in a secondary copying-stage.
Although the custom of skipping sacred names in the initial copying-stage and adding them in the proof-reading stage was probably limited to only a few generations in a few locations, it may have endured long enough to be maintained by the copyist of Papyrus 46 (sometime in the late 100’s), one of the most important early manuscripts of the Epistles of Paul.
Repeatedly, unnecessarily large space surrounds a nomen sacra in P46, and this is accounted for by the theory that the copyist initially left an overlined blank space at this point, and filled in the blank spaces in a secondary copying-stage. Some examples, from pages of P46 at the
, are shown here. This phenomenon also explains the occasional
insertion of the wrong nomina sacra,
as seen in Hebrews 12:24. University of Michigan
The nomina sacra are a significant part of a manuscript’s meta-text. In descriptions of a manuscript’s format and secondary features (such as chapter-titles, section-numbers, lectionary apparatus, decorations, etc.), its copyists’ treatment of the nomina sacra should also be noted. Unusual treatments of the nomina sacra shared by manuscripts may indicate a link between them.
|Nomina sacra were used not only in Scriptures,|
but also in inscriptions such as this one, from
a mosaic in Megiddo which was part of
the floor of a building used for
Christian gatherings in the late 200's:
“Akeptous, she who loves God,
has offered the table to God Jesus Christ
as a memorial.”
(1) Even a text that is otherwise excellent may not be reliable where nomina sacra are involved; during the proof-reading stage of an ancestor-manuscript, when the nomina sacra were added, the proof-reader might have relied on his memory to a greater degree than the copyist relied on his exemplar. (For example, Codices B and À, in the hands of their initial copyists, both read Ις Χς in Matthew 16:21.)
(2) Evidence of secondary-stage insertion of nomina sacra, either in the production of an extant manuscript or in the production of a manuscript’s ancestor, might be detected via the detection of nomina sacra which are uniquely out-of-place. (For example, ΚΩ where the text should be ΚΥ in Matthew in À, and ΠΝΚΟΣ in First Corinthians and ΧΡΥ in Ephesians in P46.)
(3) Evidence of secondary-stage insertion of nomina sacra, either in the production of an extant manuscript or in the production of a manuscript’s ancestor, might also be detected via the detection of the loss of otherwise secure nomina sacra; the explanation being that a copyist, in the secondary copying-stage, simply failed to notice the overlined blank space. (See, for example, the loss and subsequent insertion of Κε in À in John 13:6 and 13:9.)
(4) The text, as read with nomina sacra, must be considered when evaluating rival variants. Sometimes a nomen sacrum could elicit a parableptic error which would not be elicited without the contraction. For this reason, publishers of Greek texts for textual critics ought to consider printing the nomina sacra in the text and in the apparatus.
(5) In a close contest between ancient rival variants which both (or all) consist of nomina sacra, when a form of Κυ is one of the readings, it should be preferred, on the grounds that it is the less specific reading.
(6) Inconsistencies in a copyist’s contraction or non-contraction of nomina sacra are sometimes opaquely arbitrary; something, though, they may indicate how the copyist, or the copyist of his exemplar or ancestor-copy, interpreted the text. Similarly, anomalous treatments of nomina sacra may alert researchers to other anomalies. (The non-contraction of Ιησουν in Mark 16:6 in À, for example, is part of the copyist’s attempt to stretch the text into the following column.)
(7) In passages where several nomina sacra occur in close proximity, a contraction could be lost if the overlines were not neatly separated, as the proof-reader, encountering what appeared to be one overline, casually assumed that one overline implied that one name should be inserted.
(8) If an overline and blank space were longer than necessary, the proof-reader might assume that two nomina sacra were called for. This, rather than a natural tendency for embellishment, may have contributed to expansions from one name to two names.