Grace Theological Journal 6.1 (1985) 49-66

                Copyright © 1985 by Grace Theological Seminary.  Cited with permission.

 

                     THE TEXT OF JOHN 3:13

 

                                   DAVID ALAN BLACK

 

      Examination of the external and internal evidence for the reading

of John 3:13 indicates that the longer reading (which includes the

clause o[  w}n  e]n  t&?  ou]ran&?) should be regarded as authentic. This

longer reading has extensive external attestation. Furthermore, tran-

scriptional probabilities and John's style and theology lend strong

internal support for this reading. Therefore, John 3:13 is a proof of

the omnipresence of the earthly Jesus.

 

*  *  *

 

INTRODUCTION

 

TEXT-CRITICAL studies on the Gospel of John have concentrated

mainly on the pericope of the adulterous woman, which is placed

in modern editions of the Greek NT between 7:52 and 8:12 (some-

times relegated to the critical apparatus). There is, however, at least

one other major textual problem in John which calls for special atten-

tion.1 The present article examines the text of John 3:13 in which the

final clause, "who is in heaven" o[  w}n  e]n  t&?  ou]ran&?, is lacking in

important Greek witnesses to the text of John. It is argued on the

basis of both external and internal considerations that the words were

original and later were deleted to avoid saying that Jesus was simul-

taneously present in heaven. Hence, the disputed reading in John 3:13

should be allowed to stand as an explicit statement of the omni-

presence of the Son of Man, even as he walked on the earth.

 

1 An exhaustive list of the more problematic textual variants in John is given by

R. Schnackenburg, The Gospel According to St John. trans. K. Smyth (New York:

Herder, 1968) I. 182-87. The author specifies some 53 examples of textual variation "to

give an impression of the need for textual criticism on John" (p. 182). The editorial

committee of the UBS Greek NT has considered 207 places of variation in John, set-

ting forth the reasons for including certain variants in the text and for relegating others

to the apparatus. See B. M. Metzger. A Textual Commentary on the Greek New

Testament (New York: United Bible Societies, 1971) 195-258. I am grateful to my

col1eague Harry Sturz for bringing this variant to my attention. I also acknowledge a

special debt to past teachers Bo Reicke and Markus Barth for encouraging me to delve

into the textual history of John.

 



50                                GRACE THEOLOGICAL JOURNAL

 

EXTERNAL CRITERIA

 

The text of John 3: 13 circulated in the early church in two basic

yet quite distinct forms, one which included the words o[  w}n  e]n  t&? 

ou]ran&?, and another which lacked them. The former, which has been

traditionally regarded as authentic, is represented by a diversified

array of witnesses, primarily non-Alexandrine in character. The other

form is attested chiefly by the Alexandrian group of manuscripts, in

particular the uncials x and B, and by early papyrus codices of the

Bodmer collection. This section examines in greater detail the external

textual evidence for and against the reading o[  w}n  e]n  t&?  ou]ran&?.

            With the UBS Greek New Testament2 and Tischendorf's 8th Edi-

tion3 serving as sources, evidence from the manuscripts, versions and

Fathers has been accumulated and segregated under the leading text-

types or groups of witnesses (Table 1).

In assessing the evidence, the following observations can be made.

First, considerations of external evidence clearly demonstrate that

readings (3) and (4) are secondary. The former has only versional evi-

dence in support, while the latter is supported only by two Greek

manuscripts and the Sinaitic Syriac. Each of these readings is an

apparent attempt, each in its own way, to alter reading (I) to avoid

suggesting that Jesus was at once on earth and in heaven.

Variant reading (2) is also supported by a relatively small number

of witnesses. This minority, however, comprises those manuscripts

considered to be of the highest quality (as noted by Westcott4). The

Bodmer papyri p66, 75 attest the shorter reading, as do the fourth cen-

tury uncials Sinaiticus (x) and Vaticanus (B) which are the earliest

and best uncial representatives in John of the Alexandrian text-type.

The testimony of the Coptic and Ethiopic translations, as well as that

of Origen, add further early versional and patristic support to this

important array of Greek manuscripts. Thus, if the traditional read-

ing be accepted as original, some attempt must be made to explain

how the words were omitted in such early and noteworthy witnesses

to the text of the NT.

On the other hand, it is also evident that the shorter reading is

supported by a single text-type. In the Greek manuscript evidence,

the omission is found only in the Alexandrian text-type. However,

other Alexandrian witnesses, most notably several manuscripts of the

 

2Kurt Aland and others, eds., The Greek New Testament (3d ed.; New York:

United Bible Societies, 1975) 329.

            3C. Tischendorf, Novum Testamentum Graece (8th ed.; Graz, Austria: Akademische

Druck. und Verlagsanstalt, 1965) I. 765.

            4B. F. Westcott, The Gospel According to St John (reprint; Grand Rapids:

Eerdmans, 1975) 65.



TABLE  1

                 Witnesses to the Text of John 3:13 (Black:  John 3:13)

 

Byzantine                                                   Alexandrian                      Western                      Caesarean

(1)  a]nqrw<pou  o[  w]n  e]n t&?  ou]ran&?     892  copbo mss  Origenlat      it a.aur,b,c,f,ff-2,j,l,q,r2 vg                      Q f1 f13 28 565 arm geo

A*vid (omit w}n)  A cEFGHKMSU          Dionysius                      syrh Diatessarona             Cyril

VGLPY 050 063 (qeou? for                                                        Hippolytus Novatian 

a]nqrw<pou) 1195 1344 1646 Byz                                               Lucifer

Lect Eustathius Jacob-Nisibis

Aphraates Epiphanius Basil

Amphilochius Didymus

Chrysostom Nonnus Thedoret

 

(2)  omit                                                        p66, 75 xBLTbWsupp    Diatessaron carm, v           Apollinaris Cyril

                                                                  33 1241 copsa.bomss,

                                                                        eth  Origenlat  Didymus

 

(3)  o{j  h#n e}n t&? ou]ran&?                                                               ite syr c

 

(4) o[  w{n e}k  tou?  ou]ranou?                                                              0141 80 syrs



52                                GRACE THEOLOGICAL JOURNAL

 

Bohairic dialect, indicate that the o[  w}n  e]n  t&?  ou]ran&? were also

known early in Egypt. Moreover, concerning the patristic evidence,

the testimony of Origen, an Alexandrian Father, indicates only that

he was acquainted with the local text as preserved in Greek witnesses

and versions. Otherwise, ecclesiastical tradition points to the general

acceptance of the phrase as original. Summarizing, then, it appears

that the strongest evidence in favor of the shorter reading is the fact

that the words  o[  w}n  e]n  t&?  ou]ran&? are lacking in the early Alexandrian

manuscripts p66, 75, x and B.

The evidence for the inclusion of the words (reading 1) is as

follows. The phrase is found in nearly all the uncial and minuscule

manuscripts of the NT as well as in nearly every ancient version,

including the Bohairic of lower Egypt. Support for the longer reading

is also found in the great majority of the earliest patristic witnesses,

including Origen5 himself, whose testimony at this point is divided

equally between readings (I) and (2). Moreover, this reading is not

limited to manuscripts of only one geographical area, as is its omis-

sion. The reading was accepted as genuine over a wide geographic

area, encompassing most of the then civilized ancient world: Rome

and the West, Greece, Syria and Palestine, and even Alexandria, the

literary capital of Egypt.

These considerations are significant according to generally ac-

cepted canons of textual criticism which apply to the external evi-

dence of readings. Greenlee, for example, states that any reading

supported by one text-type exclusively is suspect since "no ms. or

text-type is perfectly trustworthy.’”6 Conversely, "a reading which is

supported by good representatives of two or more text-types is gener-

ally preferable to a reading supported by one text-type exclusively.7

This line of thinking favors the longer reading. The external evidence

shows almost the entire ancient tradition supporting the disputed

phrase {including the Old Latin [Itala], which establishes the date of

the longer reading as at least the last quarter of the second century).8

Also significant is the geographical distribution of the witnesses in

support of the longer reading. Being from such a wide geographical

 

5 "Non dixit qui fuit, sed qui est in caelo" (cited in Tischendorf, Novum Testa-

mentum, I. 765).

6 J. H. Greenlee, Introduction to New Testament Textual Criticism (Grand Rapids:

Eerdmans, 1972) 119.

7 Ibid.

8 Greenlee (Introduction, 46) dates the origin of the Itala "before the second cen-

tury had passed," while B. M. Metzger (The Text of the New Testament [Oxford:

Oxford University, 1968] 72) places the earliest Latin translations in North Africa

within "the last quarter of the second century," and adds that "not long afterward

translations were also made in Italy, Gaul, and elsewhere."



BLACK: JOHN 3:13                                                53

 

area, it is highly improbable that there is any genealogical relation-

ship between them. The testimony of the Greek manuscripts, ancient

versions, and Church Fathers thus forms, as it were, a strong three-

cord strand which is not easily broken, If, therefore, the reading

which is both early and supported by independent witnesses from a

wide geographical area is more likely to be original, as Greenlee sug-

gests, then clearly reading (1) should be preferred.

The retreat at this point by many scholars, such as Morris,9 to

the early uncials Sinaiticus and Vaticanus is understandable. The

reading of x and B where they agree, and of B alone where they

disagree, has long been accepted as original in places of variation.

However, despite the acknowledged antiquity and worth of these great

uncials, it has become increasingly common since the days of Westcott

and Hort to question the reading of these witnesses when they stand

alone. Greenlee writes: "The agreement of B N remains one of the

most highly regarded witnesses to the New Testament text, but it is

generally doubted that the text is as pure as W-H believed it to be."10

Metzger concurs:

 

A s a rule of thumb, the beginner may ordinarily follow the Alexandrian

text except in the case of readings contrary to the criteria which are

responsible for its being given preference in ?eneral. Such a pocedure,

however, must not be allowed to degenerate into merely looking for the

reading which is supported by Band N (or even B alone, as Hort was

accused of doing); in every instance a full and careful evaluation is to

be made of all the variant readings in the light of both transcriptional

and intrinsic probabilities.11

 

All of this does not mean, of course, that the Alexandrian witnesses

have become less important in the actual practice of textual criticism.

It does mean, however that the readings of x and B: even when sup-

ported by early papyri cannot be accepted prima facie, for the Idea of

Hort's "neutral text" is untenable and no longer should be accepted.

Critics of the text are thus in general agreement that, in the present

state of research, no single group of manuscripts can be given an

absolute preference.12

 

9"The words 'who is in heaven' are absent from some of the most reliable manu-

scripts and they should probably be omitted" (L. Morris, The Gospel According to

John [NICNT: Grand Rapids: Eerdmans, 1971] 224). For a similar judgment see E. F.

Harrison (John. The Gospel of Faith [Chicago: Moody, 1962] 26), who writes, "The

last clause of verse 13 lacks sufficient manuscript authority to be accepted as part of

the text."

10 Greenlee, Introduction,86

11 Metzger, Text of the NT, 218

12However, though it claims to be eclectic, there is evidence that the UBS Greek

NT is a text dominated by N and B,as is argued by J. K. Elliott in "The United Bible



54                                GRACE THEOLOGICAL JOURNAL

 

Undoubtedly one's idea of the history of the text and one's

principles of textual criticism will influence his decision in the present

case. My own view, simply stated, is that an early reading supported

by representatives from two or more text-types is preferable to a

reading supported by witnesses of a single text-type, even a text-type

regarded (properly or not) as the best ancient recension. It seems

highly unlikely that such a localized reading could have a better claim

to originality than a reading which is both early, widespread, and

heavily attested. On the basis of external criteria, it therefore appears

that the disputed words are original.

Greenlee's summary of the Alexandrian text is worth quoting for

those who still may have qualms about rejecting the reading of p66, 75, x

and B: “As such, it is probably the best single text of the local texts; but

like the others its readings cannot be accepted uncritically but must be

submitted to the principles of criticism.”13 An important factor militat-

ing against an uncritical acceptance of the early Alexandrian manu-

scripts is that they show a capacity to support readings which-even

in the eyes of the editors of the UBS Greek NT -are likely to be

wrong. For example, in I Cor 2: 10 the reading given in the text of the

UBS Greek NT3 is de< but p46 BI739 Clement read in its place ga<r, a

conjunction which Metzger says "has the appearance of being an

improvement introduced by the copyists."14 Another and more signifi-

cant example is the reading Xristou? in 1 Cor 1:8, which is omitted in

p46 B, as Metzger says, "either accidentally in copying. ..or perhaps

deliberately for aesthetic reasons."15 The short reading of B 1216 in

Matt 13:44 leads Metzger to speak of "the Alexandrian penchant for

pruning unnecessary words.”16 Even in the Gospel of John itself there

are readings in Alexandrian manuscripts which the UBS Greek NT

editors have attributed to scribal error. The omission of  ]Ihsou?j by

p75 x B W in 5:17, of qeou? by p66, 75 B W in 5:44, and of ei] o[ qeo>j

e]doca<sqh e]n  au]t&?  by p66  x* B C* DLW in 13:22, as well as the

substitution of e@ligma for mi<gma by x* B W in 19:39 are but four

__________________________________________________________________

Societies' Textual Commentary," NovT 17 (1975) 131ff. Elsewhere Elliott speaks of "the

reluctance of the editors to deviate too far from these hypnotic mss. x B" ("A Second

Look at the United Bible Societies' Greek New Testament," BT 27 [1975] 328). If

Elliott's conclusions are correct, one could almost speak of the text of x and B as a

modern "Textus Receptus," the overthrow of which is as difficult today as it was

during the period of the struggle for a critical text.

13Greenlee, Introduction, 86-87.

14 Metzger, Textual Commentary, 546.

15Ibid., 544.

16Ibid., 34.



BLACK: JOHN 3:13                                                55

 

examples.17 The concurrence of these early witnesses behind doubtful

readings raises questions about their integrity as witnesses to the

original text.

Summarizing, then, in this case it appears that, according to

accepted canons of text-criticism, the reading most likely to be orig-

inal on the basis of external criteria is the one which includes the

words o[  w}n  e]n  t&?  ou]ran&?. The omission, though early and supported

by the chief representatives of the Alexandrian text, is less likely to be

original due to the scarcity and geographical limitation of manuscript

support (as well as the possibility that this text-type may not be as

inherently pure as it was once thought to be). Therefore, there appears

to be no conclusive reason based on external criteria for rejecting the

strong textual and historical testimony in favor of the longer reading.

 

INTERNAL CRITERIA

 

In assessing the text of John 3: 13, one must also take into con-

sideration internal evidence. This involves two kinds of criteria: tran-

scriptional and intrinsic probabilities. The former involves evaluating

the kinds of mistakes or alterations a scribe may make as he copied a

text while the latter considers what the author was more likely to

have written. Under transcriptional probabilities, four canons are

generally accepted: (1) the more difficult reading is to be preferred;

(2) the shorter reading is to be preferred except where parablepsis

may have occurred or where a "scribe may have omitted material

which he deemed to be (i) superfluous, (ii) harsh, or (iii) contrary to

pious belief, liturgical usage, or ascetical practice”;18 (3) the reading

which is verbally dissident is to be preferred to one which is verbally

concordant with a parallel passage; and (4) the reading which best

accounts for the other variants is to be preferred:

 

Prefer the More Difficult Reading

 

Preference for the longer reading established on the basis of

external evidence finds strong internal support in the first of these

 

17Cf. also the text of 8 L Origen in John 1:26 (sth<kei for e!sthken), the text of p45

x* 8 in 10:18 (h#ren for ai@rei), and the text of p66 x B Q in 20:31 (pisteu<hte for

pisteu<shte). The evidence for 3:13 is much like that for 10:18, where Metzger (Textual

Commentary, 231) writes, "a majority of the Committee judged that its external attesta-

tion was too limited in extent, representing, as it does, only a single textual type (the

Egyptian)." Elsewhere. I have argued along similar lines in relation to the text of

Eph I: I ("The Pecularities of Ephesians and the Ephesian Address," GTJ 2 [1981]

59-73).

            18Metzger, Textual Commentary, xxvii.



56                                GRACE THEOLOGICAL JOURNAL

 

canons, since it obviously is the more difficult reading. Assuming that

John 3:13 belongs to Jesus' narrative with Nicodemus,19 the longer

reading has Christ saying that he was at that moment present both in

heaven and on earth. The awkwardness of this saying would explain

the origin of readings (3) and (4), which undoubtedly were produced

to make the longer reading less objectionable (it is much more diffi-

cult to assert that Jesus "is in heaven" while speaking to Nicodemus

than to say that he "was in heaven" or that he "is from heaven").

Thus Metzger, writing on behalf of the minority of the UBS Greek

NT editorial committee, remarks, "If the shorter reading. ..were

original, there is no discernible motive which would have prompted

copyists to add the words o[  w}n  e]n  t&am