¹ It may be questioned whether there was any Colossian heresy: perhaps the letter was written to immunise the congregation. So M. D. Hooker, ‘Were There False Teachers at Colossae?’, in Christ and Spirit (Fs. C. F. D. Moule., Cambridge: Cambridge University, 1973) 315–31, W.Schenk., ‘Der Kolosserbrief in der neueren Forschung, 1945–1985’, ANRW 2.25.4, 3327–64, 3350. But Hooker is forced to render τι δογματί**εσθε (2.20) by ‘Why submit?’ rather than ‘Why do you submit?’ (317), which would require τί θέλετ εδογματιεσθαιGoogle Scholar

² Weiss, H.-F., ‘Gnostische Motive und antignostische Polemik im Kolosser- und im Epheserbrief’, in Tröger, K.-W., ed., Gnosis und Neues Testament (Gütersloh: Mohn, 1973) 311–24; W. Schmithals, ‘Corpus Paulinum and Gnosis’, in A. H. B. Logan, and A. J. M. Wedderburn, ed., The New Testament and Gnosis (Fs. R. McL. Wilson; Edinburgh: Clark, 1983) 107–24.Google Scholar

³ Pearson, B. A., Gnosticism, Judaism and Egyptian Christianity (Studies in Antiquity and Christianity; Minneapolis: Fortress, 1990) argues from Philo that Gnosticism was a pre-Christian Jewish phenomenon; but see below, pp. 608–11.Google Scholar

⁴ So Lohse, E., Colossians and Philemon (Hermeneia; Philadelphia: Fortress, 1971 = KEK; Göttingen: Vandenhoeck & Ruprecht, 14th ed. 1968); E. Schweizer, The Letter to the Colossians (London: SPCK, 1982 = EKK; Zürich: Benziger/Neukirchen: Neukirchener, 1976). The Judaic element is minimised by Schweizer, 125–70. The ϕιλοσοϕία of 2.8 is strongly suggestive of a love of Jewish wisdom, i.e. halakha. ‘See that no one makes spoil of you through ϕιλοσοϕία… according to the tradition of men … Let no one therefore judge you about eating or drinking or over a festival or new moon or sabbath … according to the commandments and teachings of men’ (2.8–22). The appeal is to Isa 29.13, the text by which Mark rebutted the claims of ‘the tradition of the elders’ (Mark 7.6, cf. 1 Cor 2.13).Google Scholar

⁵ Francis, F. O. and Meeks, W. A., Conflict at Colossae (Missoula: Scholars, 1975), a collection of discussions of the ‘heresy’, includes an essay by Francis, ‘Humility and Angelic Worship in Col. 2:18’, pp. 163–95. His conclusions are not far from those of the present article, but are made unclear by adducing material from the Hermetic documents.Google Scholar

⁶ Cf.Pearson, , Gnosticism, 29–38.Google Scholar

⁷ Weiss, , ‘Motive’, gives a useful discussion of antignostic polemic in Col., with the suggestion that the author has taken on some gnostic motifs himself; but the definition of gnostic is too loose to be useful, and the argument lacks focus from the lack of reference to particular documents.Google Scholar

⁸ Giversen, S., Apocryphon Johannis (ATD 5; Copenhagen: Munksgaard, 1963) 45. Michel Tardieu dates the Berlin Papyrus (8502) version a little later, c. 400: Ecrits gnostiques: Codex de Berlin (Paris: Cerf, 1984) 45. The three Nag Hammadi versions were published by M. Krause and P. Labib in Die drei Versionen des Apokryphon Johannis im Koptischen Museum in Alt-Kairo (ADAIK Kopt. Reihe 1; Wiesbaden, 1965); and Krause compares them with Irenaeus' account in W. Foerster, ed., Gnosis (Oxford: Clarendon, 1972) l.lOOff.Google Scholar

⁹ Schenk, , ANRW (n. 1), 3327ff., treats the arguments of W. Bujard (Stilanalytische Untersuchungen zum Kolosserbrief[Göttingen: Vandenhoeck & Ruprecht, 1973]) as ‘bahnbrechend’ and as excluding Pauline authorship. Bujard appealed (with many tables) to the loosely attached style of the sentences, the absence of rhetorical questions or objections, the rare connecting conjunctions, etc. But there are passages in the agreed Paulines like this; and Colossians might adopt a different technique from Paul's customary argumentative style, taking over the opposition language as far as possible and glossing it. Cf. my ‘The Visionaries of Laodicea’, JSNT 43 (1991) 16–39. The issue of Pauline authorship does not greatly affect the date of Colossians: Petr Pokorný cites Ernst Käsemann's dictum, ‘If authentic, as late as possible on account of style; if not authentic, as early as possible’, Colossians (Peabody Mass.: Hendrikson, 1991 = ThHkNT 10/1; Berlin: Evangelische, 1991) 4. In either case the date will be in the 60s.Google Scholar

¹⁰ Barbelo occurs in The Apocryphon of John, The Three Steles of Seth, Zostrianos and ‘frequently in related literature’ (Giversen, 165). Yald(t)abaoth also comes in Apoc. Joh., The Origin of the World, The Trimorphic Protennoia and many other Gnostic documents.Google Scholar

¹¹ Gnosis: the Nature and History of an Ancient Religion (ET Edinburgh: Clark, 1983 = Die Gnosis, Leipzig: Koehler and Amelang, 1977, Göttingen: Vandenhoeck & Ruprecht, 3rd ed. 1990) 80.Google Scholar

¹² P. 73, citing Scholem, G..Google Scholar

¹³ Histoire critique du gnosticisme 1–3 (Strasbourg: Levrault, 1843–1844)2.201.Google Scholar

¹⁴ Apocryphon Johannis, 165–6.Google Scholar

¹⁵ ‘Gnosticism and the New Testament’, Vig. Chr. 19 (1965) 65–85.CrossRefGoogle Scholar

¹⁶ A Separate God (ET London: Darton, 1991 = Le Dieu Séparé, Paris: Cerf, 1984) 93–4.Google Scholar

¹⁷ In On the Origin of the World, II.100.13–14, the name Yaldabaoth is rendered ‘O youth, pass over here’.Google Scholar

¹⁸ Harvey, W. Wigan, ed., Irenaei libros quinque adversus haereses 1–2 (Cambridge: University, 1857) 1.221, n. 2. The head of the sect was Simon, according to Harvey, and as a Samaritan he would have spoken [a language akin to ?] Syriac.Google Scholar

¹⁹ Hauptprobleme der Gnosis (FRLANT 10; Göttingen: Vandenhoeck & Ruprecht, 1907) 14.Google Scholar

²⁰ Church and Gnosis (Cambridge: University, 1932) 54–5, 58ff. But the second century seems too early for Coptic: Demotic would be more likely for second-century Egyptian (Giversen, 166).Google Scholar

²¹ The Gnostic Scriptures (London: SCM, 1987) 15.Google Scholar

²² Écrits gnostiques, 259.Google Scholar

²³ Wisse's, F., translation from Robinson, James M., ed., The Nag Hammadi Library (San Francisco: Harper, 1977) 106–7, with my glosses in brackets.Google Scholar

²⁴ The Apoc. Joh. tradition is involved in some contradiction because Gen 1.26 has God say, ’Let us make man in our image and in our likeness';, and the image has to be the image of the ultimate God, the IVS. Hence II. 15.2–3, ‘Come, let us create a man according to the image of God and according to our likeness.’ Forced exegesis usually lands the expositor in this kind of trouble.Google Scholar

²⁵ Cf., for example, Philo Op. Mund. 72, ‘when he made the heaven and the earth and the sea’; Jub. 2.2, ‘He created the heavens which are above and the earth and the waters.’Google Scholar

²⁶ This section of A.H. is preserved in Latin; Theodoret gives a part of it in Greek with Βαρβηλώθ. Burkitt, Church and Gnosis, 60, says that the Latin is in general an accurate version; Theodoret is less accurate, and -oth is a standard magical termination, and valueless.Google Scholar

²⁷ ‘Jaldabaoth Reconsidered’, in Mélanges d'histoire des religions offerts à Henri-Charles Puech (Paris: Presses universitaires de France, 1974) 405–21. Scholem gives a list of earlier explanations.Google Scholar

²⁸ ibid., 416, ‘a magic word … obviously derived from an abridgement of Sabaoth’. Obviously?

²⁹ ‘Notes on Gnosticism’, Vig. Chr. 11 (1957) 148–9; on which cf. Scholem, art. cit., 407.Google Scholar

³⁰ A Separate God, 43ff.Google Scholar

³¹ A counter-suggestion has been made of an Aramaic derivation: either the Begetter of Fatherhood, ‘abahuta’ (A. Adam, ‘Ist die Gnosis in aramäischen Weisheitsschulen entstanden?’, in Bianchi, U., ed., The Origins of Gnosticism [SHR 12; Leiden: Brill, 1967], 291–301); or the Begetter of Abomination, bahuth, like the Hebrew bosheth (M. Black, ‘An Aramaic Etymology for Jaldabaoth?’, in A. H. B. Logan and A. J. M. Wedderburn, ed., The New Testament and Gnosis [Fs. R. McL. Wilson; Edinburgh: Clark, 1983], 67–72). Black's criticisms of Scholem are bypassed by the suggestions of Grant and Pétrement (and myself) above.Google Scholar

³² Tobin, Thomas H., The Creation of Man (CBQ MS 14; Washington DC, 1983) 113.Google Scholar

³³ Tobin, , Creation, 112.Google Scholar

³⁴ Xenophon reports Socrates as speaking of the Demiurge in Mem. 1.4.9; Plato expounds the idea of a creator-god in Tim. 29d–30c without using the term, which however was applied to it in Middle Platonism. Cf. Tobin, 44–55; F.M. Cornford, Plato's Cosmology (London and New York: Kegan Paul, 1937) 34–6.Google Scholar

³⁵ Quaest.Gen. 1.8a, ‘Why does he place the moulded man in Paradise, but not the man who was made in his Image? Some, believing Paradise to be a garden, have said that since the moulded man is αίσθητός…’; cf. Tobin, 102, n. 2.Google Scholar

³⁶ So Conzelmann, H., 1 Corinthians (Hermeneia; Philadelphia: Fortress, (1975) = 1st German ed. 1969) 286.Google Scholar

³⁷ For Jewish speculations on the name of God, cf. Fossum, J., The Name of God and the angel of the Lord (WUNT 36; Tübingen: Mohr, 1985). Fossum argues for a Samaritan/Jewish origin for Gnosticism in an entirely convincing way; though the point at which ‘Gnosticism’ emerges may be a matter of definition.Google Scholar

³⁸ Tobin, , Creation, 108ff.Google Scholar

³⁹ The most general view, represented by Rudolph, Gnosis, 275–308, sees Gnosticism as a pre-Christian religious movement, rooted in Jewish apocalyptic and Wisdom circles, but influenced by Iranian and various Greek thought. Rudolph does not explain how monotheistic Jews came to adopt a radically dualist theology; the steps towards this are set out by Fossum and Pearson.Google Scholar

⁴⁰ This is well argued by Simone Pétrement, A Separate God, esp. 46–7; the book is wayward, and full of improbable suggestions, but the central thesis, that Gnosticism was from the beginning a Christian movement, is convincing.Google Scholar

⁴¹ Logan, A. H. B., ‘John and the Gnostics: the Significance of the Apocryphon of John for the Debate about the Origins of the Johannine Literature’, JSNT 43 (1991) 41–69.Google Scholar

⁴² Hegermann, H., Die Vorstellung vom Schöpfungsmittler im hellenistischen Judentum und Urchristentum (TU 82; Berlin: Akademie, 1961), derives the Colossian ‘hymn’ from Hellenistic synagogues, whose use of the language of the mysteries can be found in Philo; but Apoc. Joh. offers closer parallels, and to more of the letter. J. Fossum, in an interesting article, ‘Colossians 1.15–18a in the Light of Jewish Mysticism and Gnosticism’, NTS 35 (1989) 183–201, adduces other Jewish/Gnostic parallels to a conclusion similar to mine.Google Scholar

⁴³ Cf. Käsemann's comment cited above, n. 9.Google Scholar

⁴⁴ The Coptic is a form of the Greek άόρατος.Google Scholar

⁴⁵ The language of Col 1.15–18 is for the main part clearly Pauline. The whole phrase őςέστιν είκών το θεο recurs at 2 Cor 4.4; πρωτότοκος comes at Rom 8.29 and πᾱσα (ή) κτίσις at Rom 8.22 (κτίσις 9x in agreed Paulines); έν αύτῷ 9x; κτίζειν 2X; τά πάντα 16x; έν (τοỖς) ούρατοςῷ… έπί (τς) γς comes in 1 Cor 8.5, as does τά πάντα δί αύτο/είς αύτόν; άόρατος comes in Rom 1.20, where it is also contrasted with the visible creation; εῖτε is a Pauline favourite, occurring in 35 passages, often multiple as here (+ 5x in the doubtful Paulines and 2x in 1 Peter in the NT); άρχαί and έξουσίαι come together in 1 Cor 15.24, and other categories of powers are named in 1 Cor 3.22 and Rom 8.38–9; in 1 Cor 15.20 Christ is the firstfruits (άπαπχή) νεκρῷν. ∑υνιστάνειν is virtually a Pauline word (12x), though the intransitive use here is unique. Of course it is possible to ascribe all these Pauline uses to a clever imitator; cf. R. F. Collins, Letters That Paul Did Not Write (Wilmington, 1988) 184–5. It is difficult to see how discussion can advance against such hard-nosed scepsis; but the examples of the Pastorals, the Lucan sermons in Acts, and Hebrews as attempted imitations of Paul do not favour the sceptics. It is often alleged that it is un-Pauline to think of Christ as the head of the Church; but Paul is elastic in his use of imagery (both Church and Christian may be temples of the Holy Spirit), and there is a reason in Colossians and Ephesians for this changed emphasis—the Gnostic pleroma which reduced Christ to one of many aeons. It seems curious that the Colossian church should (ex hypothesi) know by heart a hymn also known to Paul but not cited elsewhere. The counter-arguments in Lohse, 41–61 (two or three hapax legomena, άρχή a concept not found elsewhere in Paul) seem weak; we cannot argue convincingly that the apostle never used new concepts, or lapsed into rhythmical prose when moved.Google Scholar

⁴⁶ Docetic should be understood in the sense of a fully human Jesus temporarily possessed by an angelic Christ, who left Jesus in his Passion and only seemed to suffer — Cerinthus' teaching (Irenaeus A.H. 26.1); for the early emergence of this doctrine see my ‘The Pre-Marcan Gospel’ forthcoming in SJT (1995).Google Scholar

⁴⁷ The Coptic uses the Greek words τέλειος and γνώσις, where I have rendered with Gnosis.Google Scholar

⁴⁸ Unnik, W.C.van, ‘Die jüdische Komponente in der Entstehung der Gnosis’, Vig. Chr. 15 (1961) 65–82, concludes correctly that Apoc. Joh. comes from a community which did not know the OT in Hebrew, knew OT names from hearsay and probably derived them from the Greek text (p. 80). I am arguing that Philo is dependent on a Hebrew exegesis of Gen 1 originating in Palestine, and that the first part of the Apoc. Joh. myth, that concerned with Barbelo and the creation of the aeons, and not that covering the fall of Sophia, the generation of Yaldabaoth and the Gnostic plight of mankind, was developed there before the Christian era. It is an important element in this argument, though, that even the name Yaldabaoth has a Hebrew origin.CrossRefGoogle Scholar

⁴⁹ Giversen, , 176–7, suspects some muddle behind Irenaeus' list, and conjectures the ten names in the ‘original decad’ himself.Google Scholar

⁵⁰ There looks like some overlap here with Aletheia in the Decad; but such duplication is common in lists of Gnostic names.Google Scholar

⁵¹ Blinzler, J., ’Lexikalisches zu dem Terminus στοι*khgr;εῖα το κόσμου bei Paulus', Ana. Bibl. 17/18.2 (1963) 429–43, and D. Rusam, ‘Neue Belege zu den στοιχεῖα κόσμου (Gal.4,3.9; Kol.2,8.20)’, ZNW 83 (1992) 119–25, list instances in which the phrase is used, virtually exclusively, for the four (five) Greek elements of which the world consists, earth, air, fire and water. But it is very difficult to see what sense this gives in the Pauline contexts. F. F. Bruce, The Epistle to the Galatians (NIGNT; Exeter: Paternoster, 1982) 193–4, takes the alphabetic meaning to be primary, and cites 2 Pet 3.10, 12, where στοιχεῖα are stellar powers, and Pseudo-Callisthenes, Alexander Romance 1.1, where the same is true of τά κοσμικά στοιχεῖα.Google Scholar

⁵² The word emanations is theologically pejorative: Gnostics speculate about emanations. But Paul never explains how he conceived of Christ as being on an equality with God, his Image, etc. Perhaps we should allow that he accepted contemporary ideas (speculations); or perhaps he knew about it by revelation.Google Scholar

⁵³ I am grateful to Prof. R. McL. Wilson for his generous help in suggesting criticisms, improvements and documentation for this article.Google Scholar