¹ Among the most interesting and compelling studies are the following: Aland, Barbara, ‘Erwählungstheologie und Menschenklassenlehre: Die Theologie des Herakleon als Schlüssel zum Verständnis der christlichen Gnosis?’ in Gnosis and Gnosticism (ed. Martin, Krause; NHS 8; Leiden: E. J. Brill, 1977) 148–81Google Scholar; Mühlenberg, E., ‘Wieviel Erlösungen kennt der Gnostiker Herakleon’, ZNW 66 (1975) 170–93CrossRefGoogle Scholar; Pagels, Elaine, The Johannine Gospel in Gnostic Exegesis: Heracleon's Commentary on John (Nashville and New York: Abingdon, 1973);Google Scholar and Janssens, Yvonne, ‘Héracleon: Commentaire sur l'Évangile selon S. Jean’, Le Muséon 72 (1959) 101–51Google Scholar; 277–99. For a recent comparative analysis of Heracleon and Origen, see Poffet, Jean-Michel, La méthode exégétique d'Héracléon et d'Origène. Commentateurs de Jn 4: Jésus, la Samaritaine et les Samaritains (Fribourg: Éditions Universitaires, 1985)Google Scholar. A relatively full biographical sketch can be found in the older work of Brooke, A. E., The Fragments of Heracleon (TextS 1.4; Cambridge: University, 1891) 31–41Google Scholar; a more recent and briefer discussion can be found in Bammel, C., ‘Herakleon’, TRE 15 (1986) 54–7.Google Scholar

² I have addressed one aspect of this problem in extenso in my book, The Orthodox Corruption of Scripture: The Effect of Early Christological Controversies on the Text of the New Testament (New York: Oxford, 1993)Google Scholar. In addition, see my overview in ‘The Text as Window: NT MSS and the Social History of Early Christianity’, The Text of the New Testament in Contemporary Research: Essays on the Status Quaestionis (ed. Bart D. Ehrman and Michael W. Holmes; Studies and Documents; Grand Rapids: Eerdmans, forthcoming).

³ These are collected in Brooke, Fragments. I have given a liberal estimation of the number of lines; in fact, most of these fragments also include Origen's rebuttal of Heracleon's exposition. For an English translation of just the expositions themselves, see Foerster, Werner, Gnosis: A Selection of Gnostic Texts (tr. R, McL. Wilson; Oxford: Clarendon, 1972) 1.162–83.Google Scholar

⁴ I.e., not counting such scraps as P⁵². P⁶⁶ and P⁷⁵ are typically dated to around the year 200 CE, or perhaps somewhat later.

⁵ The Roman provenance becomes important in assessing the text-critical significance of Heracleon's citation (see below, pp. 178–9). For the arguments in its favour, see my article ‘Origen, Heracleon, and the Text of the Fourth Gospel’, VC 47 (1993) 105–18.Google Scholar

⁶ See, esp., the article by Gordon D. Fee on the use of Patristic evidence in ANRW 11.26.1246–65. These breakthroughs led Fee to launch a new monograph series, The New Testament in the Greek Fathers: Texts and Analyses. See further the discussion in my contribution to the series, Didymus the Blind and the Text of the Gospel (NTGF 1; Atlanta: Scholars, 1986) ch. 1.Google Scholar

⁷ Here it may be noted that Origen typically introduces his quotations of Heracleon with set phrases that serve almost as quotation marks; e.g. ὁ δὲ Ὴρακλέων τò μέσος ὐμν στήκει φησιν …(Io.Com 6.39.194) or ὁ μέντοι γε Ὴρακλέων τὁ μετὰ τοἀτο κατῦβη εῦς καφαρναοέμ αἰτός (Io.Com 10.11.48) or σφόδρα δὺ ὐπαρατηρήπως ὲ ἀΗρακλέων οὀεται τ· ἴζὸλος τοὀ οῆκου σου καταφάγεταί με (Io.Com 10.34.223).

⁸ See my discussion of these instances in ‘Heracleon, Origen, and the Text of the Fourth Gospel’.

⁹ I have cited all such instances in the apparatus.

¹⁰ I have drawn these data from the volume I have co-edited with Gordon Fee, D. and Holmes, Michael W., The Text of the Fourth Gospel in the Writings of Origen (NTGF 3.1; Atlanta: Scholars, 1992).Google Scholar

¹¹ See note 7 above.

¹² C. Blanc, Origène: Commentaire sur Saint Jean (SC 120, 157, 222, 290; Paris: Cerf, 1975–82).

¹³ Even, e.g., Origen's claim that ἔοικε δἴ βασιλικἔν ὲ Ήρακλέων λέγειν τὸν Δημιουργὀν which might be taken to suggest that H. read βασιλικός rather than βασιλισκός in 4.46 – is so allusive and so clearly put in Origen's own words, that there is simply no way to know.

¹⁴ I have not marked the Latin witnesses as ‘not applicable’ in cases of singular readings, since in these instances none of the witnesses can be used to establish textual affinities.

¹⁵ E.g., the variation involving διψήσει/διψήσηin 4.14.

¹⁶ E.g., the absence of γάρ in 4.37 or δέ in 4.39. Patristic citations tend to conform such connectives to the context, making their testimony unusable in all but the rarest incidences. See the discussion in Ehrman, Fee, and Holmes, The Text of the Fourth Gospel in the Writings of Origen, 26.

¹⁷ E.g. in John 1.27, where ὸρχόμενος is followed by őς ὸμπροσθέν μου γέγονεν in many but not all witnesses. There is simply no way to divine Heracleon's text in such instances.

¹⁸ Origen notes that Heracleon appeals to this verse in explicit support of his exegesis of 4.24. When he earlier discusses Heracleon's interpretation of 1.3, he quotes it in the form found otherwise throughout his own writings (οἐδἔ ὐν). As I argue in ‘Heracleon, Origen, and the Text of the Fourth Gospel’, he appears to do so in order to make Heracleon's construal look particularly forced: ‘Apart from him was made not even one thing’ … except for some things (the Pleroma and what was in it)!

¹⁹ This reconstruction is based on Origen's discussion of Heracleon's exegesis, which focuses on the meaning of ὲν αἕτἐ rather than the significance of the verb tense. What is striking is that Heracleon understands the verse to refer to the pneumatics, who because of the revelatory work of the Logos have come to their true being ‘in him’, and in that sense are, in Origen's words, ‘the same as the Logos’. The present tense is clearly amenable to this kind of exegesis. Combined with this is the circumstance that ἕστιν is in fact preserved in the two witnesses that Heracleon otherwise supports in unusual readings (K D; see below). Here we are dealing with a verse that was so familiar to Origen – he cites this portion of it ten times in his extant writings, always in exactly the same form – that even when discussing Heracleon's exegesis he appears to have quoted it the way he had learned it. See further my discussion in ‘Heracleon, Origen, and the Text of John’.

²⁰ Origen notes that Heracleon attributes these words to John the Baptist.

²¹ Origen notes that Heracleon attributes these words to John the Baptist.

²² The reference is probably too allusive to draw any conclusions concerning the word order of Heracleon's text or the presence of the article (οὐδεὶς τὸν θέον for θεὸν οὐδείς).

²³ This quotation comes from Origen, but Heracleon's text may have influenced its form: this is the only clear citation of the passage in which Origen omits the article before μονογενής. The rest of the citation, however, is Origen's: Heracleon apparently did not read ὁ ὤν (see the following reference).

²⁴ C. Blanc (SC 157, p. 140) follows Preuschen in supplying the clause (ó ν), so as to bring the passage into closer conformity to the prevailing text of the MS tradition of the Fourth Gospel (ἐξηγήσατο (ó ν) εἰς τὸν κόλπον τοὐ πατρός). As I have argued in ‘Heracleon, Origen, and the Text of the Fourth Gospel’, however, this overlooks both the exegetical possibilities of the shorter text and its preservation in a stream of the NT textual tradition.

²⁵ Origen is paraphrasing Heracleon's text explicitly to show that he failed to account for the article before προφητής.

²⁶ This is the one instance in Book 6 in which Origen clearly cites Heracleon's form of the text. In all of his other quotations of the passage in Book 6, and earlier in Book 2, Origen cites the verb in the perfect (see the following note). This suggests that in the quotation in 6.39.197, Origen has given the clause in his own customary form, even though it might appear at first to be from Heracleon. It is to be noted that the quotation of 1.27 that immediately follows appears also to be Origen's, in that Heracleon's exegesis appears to presuppose a different wording.

²⁷ With the exception of Io.Com 6.39.194 – which happens to preserve Heracleon's text -the alternation of ἕστηκεν/στήκει in Origen's citations follows a regular pattern: he uses the perfect tense early in his career (John Commentary, Books 1–6), the present tense late (Book 32, and the Contra Celsum). This appears then to be an instance in which Origen continued using an Alexandrian MS during his early residence in Caesarea, before changing MSS later. See further ‘Heracleon, Origen, and the Text of the Fourth Gospel’, 108–9.

²⁸ Origen appears to have cited Heracleon's text precisely except in his substitution of the ἄξιος for Heracleon's ἰκανός. In fact, Heracleon's own exposition of the verse in 6.39.198 explicitly uses the latter term. This is another verse that was so intimately familiar to Origen (he cites it six times in his extant writings, in every case with ἄξιος), that he apparently made an inadvertent modification of Heracleon's text when citing it. Cf. 1.4 above.

²⁹ The first citation is too terse to be taken as evidence that Heracleon's text lacked the article before ὀπίσω, or even the introductory clause αὐτῷς ὲστίν. I have therefore not cited these variants in the apparatus. The case is otherwise with Origen himself. See Ehrman, Holmes, and Fee, The Text of the Fourth Gospel in the Writings of Origen.

³⁰ The witnesses locate ἐγώ in a variety of places in the clause.

³¹ I take this to be the text of Origen's own MSS, even though he himself emended the text to Βηθαβαρᾴ. This emendation itself proved influential, as later scribes incorporated it into their texts. Origen admits, however, that his preference is based not on MS evidence, but on having realized that his MSS could not be correct: Bethany is not near the River Jordan. His statement that ‘nearly’ all MSS read Βηθαβαρᾴ (6.40.204) should probably be taken, then, as hyperbole; so far as we can tell, none of his MSS read otherwise.

³² Origen is referring to Heracleon's exegesis of the text.

³³ Again, Origen is referring to Heracleon's exegesis.

³⁴ Origen is referring here to Heracleon's interpretation of the verse.

³⁵ It has sometimes been argued that, along with C*, the exemplar Origen used for his commentary on the Fourth Gospel had inadvertently omitted the phrase οὐ μή…αὐτ ῷ from 4.14 (because of homoeoteleuton; thus, e.g., Fee, ‘The Text of John’, 377). The lemma of Io.Com 13.1.2 omits the phrase, as does the full quotation of Io.Com 13.3.14 and the adaptation of 13.4.20. It is nonetheless to be noted that Origen explicitly quotes the first portion of the phrase in question in the nearby context (Io.Com 13.10.60), countering Heracleon's interpretation of it with his own. It would appear then, that Origen's earlier citation and allusion have been shortened for contextual reasons (Origen is interested in commenting on what the final clause might mean) and that the lemma has in turn been influenced by his quotations ad loc.

³⁶ It is impossible to determine whether Heracleon's text included the name ‘Ιησοῦς, with or without the article. I have therefore not cited the variants in the apparatus.

³⁷ Both Heracleon and Origen attest the variant word orders.

³⁸ Origen appears to give Heracleon's precise text, even though it varies from all other witnesses in reading ἀληθὲς εἴρηκας (for καλῶς εἴπας). Either Heracleon or his MS of John has been influenced here by the phrasing at the end of v. 18. This in itself, however, indicates the wording of that passage (which is otherwise disputed) in Heracleon's text.

³⁹ Heracleon, or his MS of John, derived these words from 4.18; see note 38 above.

⁴⁰ This represents an actual quotation from Heracleon's text, contra Janssens, ‘Héracleon’, 135, n. 43. See my discussion in ‘Heracleon, Origen, and the Text of the Fourth Gospel’, 112–13.

⁴¹ It is impossible to judge whether Heracleon's text actually included the prepositional phrase πρὸς αὐτόν, or whether this is simply a loose citation adapted to the context. The phrase is otherwise unattested.

⁴² Origen actually cites the text with λαλίαν rather than μαρτυρίαν. But Heracleon's explanation of the verse shows that he read μαρτυρίαν: ούτοι οὐκέτι διὰ μόνην ἀνθρωπίνην μαρτυρίαν, ἀλλὰ δι᾽ αὐτὴν ἀλήθειαν πισεύουσιν. I have assumed, for the purposes of the reconstruction, that here again Origen reproduced Heracleon's text, but inadvertently substituted the key word as he himself remembered it.

⁴³ The MSS that attest the adverb locate it in a variety of places in the clause.

⁴⁴ The context of the discussion indicates that the reference is to 4.50 rather than 4.53.

⁴⁵ It is difficult to judge from this brief reference whether Heracleon stands against virtually the entire MS tradition of John in reading the verb in the aorist rather than the perfect.

⁴⁶ See n. 5 above.

⁴⁷ For the development of this method and its various refinements, see my article, ‘Methodological Developments in the Analysis and Classification of New Testament Documentary Evidence’, NovT 29 (1987) 22–45.Google Scholar

⁴⁸ I have included a proportionally greater number of Alexandrian witnesses and a number of ‘Caesarean’ witnesses because Origen, our source for Heracleon's text, has traditionally been used to establish the contours of these traditions.

⁴⁹ See notes 19,28, and 42 above.

⁵⁰ On the significance of group profiles for a full analysis of textual alignments, see my article ‘The Use of Group Profiles for the Classification of New Testament Documentary Evidence’, JBL 106 (1987) 465–86.Google Scholar

⁵¹ See the apparatus at 1.4,18 (fourth variant), 21; 4.17 (final variant), 24 (first variant), 37 (second variant), and 42 (third variant).

⁵² The exceptions occur in 1.27 (second variant) and 4.42 (ninth variant). It should be noted here that Heracleon agrees with 56% of the ‘Caesarean’ uniform readings (14/25) and with only 43.33% of the Byzantine (17/30).

⁵³ Assuming, that is, the validity of my reconstruction of 1.4.

⁵⁴ 1.27 (second variant); 4.15 (second variant), 30, 42 (second variant).

⁵⁵ 1.18 (fourth variant), 21; 2.14 (second variant); 4.24 (first and second variants), 27, 37 (second variant), 42 (third and fifth variants).

⁵⁶ He agrees in 1.18 (third variant), but not in 4.15 (second variant) or 4.51 (second variant).

⁵⁷ Agreeing in the first two variants of 1.27, but not in those of 2.15 (first variant), 4.15 (second variant), and 4.51 (second variant).

⁵⁸ He would not be expected to agree as frequently with these witnesses in such readings, of course, since here they tend to side with Origen.

⁵⁹ 1.4,18 (fourth variant), 21; 4.17 (fourth variant), 24 (first and second variants), and 42 (third variant).

⁶⁰ This is the burden of my article, ‘Origen, Heracleon, and the Text of the Fourth Gospel’. See further my full-length treatment, The Orthodox Corruption of Scripture.