Skip to main content Accessibility help
Hostname: page-component-7f7b94f6bd-5bz6h Total loading time: 0.274 Render date: 2022-06-30T05:52:13.926Z Has data issue: true Feature Flags: { "shouldUseShareProductTool": true, "shouldUseHypothesis": true, "isUnsiloEnabled": true, "useRatesEcommerce": false, "useNewApi": true } hasContentIssue true

Ancient Binding Spells, Amulets and Matt 16.18–19: Revisiting August Dell's Proposal a Century Later*

Published online by Cambridge University Press:  27 May 2016

Seon Yong Kim*
The University of Chicago Divinity School, 1025 East 58th Street, Swift Hall, Chicago, IL 60637, USA. Email:


The goal of this essay is to reopen the unfortunately unrefined but still illuminating proposal of August Dell a century ago – a proposal to read the motif of ‘bind and unbind’ in Matt 16.19 against the backdrop of ancient magical conventions – and to corroborate this hypothesis with philological precision and contextualisation. The present study seeks to demonstrate how central the motif of ‘bind and unbind’ was to most binding spells and amulets, and how the verbs λύειν and δέειν and their cognates might have evoked the conventions of binding magic when heard in religious and ritual contexts. We see Matt 16.19 as a literary reappropriation of a typical binding spell, crafted to highlight Peter's authority to ‘bind’ and ‘break’ whatever entities he should choose. Jesus' guarantee that he/God will ‘unbind’ whatever target Peter ‘unties' is to be understood as a clear indication that Peter is endowed with an invincible potency over every other spiritual entity, symbolised by the power to break any ‘binding’ spell. The security of the church against malevolent and harmful spiritual entities is further guaranteed by Christ's promise, if read in ancient binding magical conventions: Christ will give Peter the key of the kingdom of heaven so that Peter as the key-bearer may become a figure who can channel the full potency of the keyed object.

Copyright © Cambridge University Press 2016 

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)



In memory of Sung-Hee Jung, who showed an unwavering faith in Christ's power to bind and loose.


1 Dell, A., ‘Matthäus 16, 17–19’, ZNW 15 (1914) 149CrossRefGoogle Scholar.

2 The traditional dichotomy between magic and religion as academic categories in the study of religion has received criticism, and virtually collapsed particularly in anthropology and history of religions. For a perceptive observation that shows the inappropriateness of this dichotomy regarding Greek magical papyri, see J. Z. Smith, ‘Trading Places’, Ancient Magic and Ritual Power (ed. M. Meyer and P. Mirecki; Leiden: Brill, 1995) 132–7. However, some scholars are not persuaded, and still defend the usefulness and relevance of the dichotomy in the study of ancient religion. See, for example, Versnel, H. S., ‘Some Reflections on the Relationship Magic–Religion’, Numen 38 (1991) 177–97CrossRefGoogle Scholar. For a more nuanced approach, see J. Braarvig, ‘Magic: Reconsidering the Grand Dichotomy’, The World of Ancient Magic (ed. D. R. Jordan, H. Montgomery, E. Thomassen; Bergen: Norwegian Institute at Athens, 1999) 21–54. Although I agree with the problematisation of the dichotomy, the present article does not engage with this huge problem, because it focuses on the complex appropriation of magical notions and terms that appear in Matt 16.18–19, without any value judgement on the practice of magic.

3 Dell's argument is one of the earliest yet most substantive attempts at a ‘magical’ reading. Earlier than Dell, F. C. Conybeare, under the influence of Frazer, briefly commented that the language of bind and loose was ‘directly borrowed from contemporary magic’ (Christian Demonology iii’, JQR 9 (1897) 468–70Google Scholar). Walter Köhler attempted to read the motif of ‘bind and loose’ in light of broader Greco-Roman religions including magic (Die Schlüssel des Petrus: Versuch einer religionsgeschichtlichen Erklärung von Matth. 16, 18.19’, ARW 8 (1905) 236–9Google Scholar).

4 R. Bultmann, The History of the Synoptic Tradition (New York: Harper & Row, 19682) 138.

5 Bultmann's conjecture (History, 138–9) has been further refined by later generations of exegetes who interpret ‘bind and loose’ as the ‘authority of halakhic decision’ and the ‘power to discern valid prohibitions from invalid ones’. So hereafter the present study calls this line of reading the ‘rabbinic’ reading. A succinct expression of a ‘rabbinic’ reading can be found in Marcus, J., ‘The Gate of Hades and the Keys of the Kingdom (Matt 16.18–19)’, CBQ 50 (1988) 452Google Scholar: ‘it seems best to side with those exegetes who interpret “binding and loosing” in Matt 16:19 as declaring forbidden or permitted, i.e., promulgation of authoritative halakah. Our passage speaks of the revelation to Peter in the earthly sphere of the interpretation of the law that has been decided in heaven. He is given total power on earth to distinguish valid from invalid prohibitions, “binding” upon human beings the observance of certain of them – even some not explicit in the Mosaic torah – and “loosing” them from the observance of others of them – even some enjoined by Moses.’ One of the earliest advocates of this view in modern scholarship is Georg Eduard Steitz, who insists, ‘Die mit ihm in dem engsten Zusammenhange stehende Redensart: binden und lösen Matth 16,19 und 18,18 ist ferner eine konstante Formel des rabbinischen Sprachgebrauchs und wurzelt gleichfalls in althebräischen Vorstellungen und Anschauungen’ (‘Der neutestamentliche Begriff der Schlüsselgewalt’, Theologische Studien und Kritiken: Eine Zeitschrift für das Gesamte Gebiet der Theologie (Gotha: F. A. Perthes, 1866) 436).

6 Bultmann, History, 141.

7 U. Luz, Matthew 8–20: A Commentary (Hermeneia; Minneapolis: Augsburg Fortress, 2001); W. D. Davies and D. C. Allison, A Critical and Exegetical Commentary on the Gospel according to Saint Matthew (3 vols.; ICC; Edinburgh: T&T Clark, 1991).

8 Davies and Allison insist that the magical reading ‘is possible only for some pre-Matthean stage. Matthew's animus towards magical practices is established’ (Gospel according to Saint Matthew, ii.637). However, it is necessary to note that Matthew's disapproval of magical practice does not necessarily exclude his creative appropriation of the language and imagery of magic aiming at emphasis in his own case (here the church's superiority and power vis-à-vis anti-God powers). Luz does not even try to discuss the unlikelihood of a magical reading; he simply discards it (Matthew 8–20, 365).

9 A. Audollent, Defixionum tabellae (Paris: Alberti Fontemoing, 1904) (= DT); H. D. Betz, ed., The Greek Magical Papyri in Translation: Including the Demotic Spells (Chicago: University of Chicago Press, 19922) (= Betz); R. Wünsch, ed., Inscriptiones Atticae aetatis Romanae, defixionum tabellae, inscriptiones Graecae iii.3, Appendix (Berlin: Reimer, 1897) (= DTA); Papyri Graecae Magicae: Die Griechischen Zauberpapyri (2 vols.; ed. K. Preisendanz and A. Henrichs; Stuttgart: Teubner, 1973–42)’ (= PGM); Jordan, D. R., ‘A Survey of Greek Defixiones Not Included in the Special Corpora’, Greek, Roman, and Byzantine Studies 26 (1985) 151–97Google Scholar (= SGD); J. Gager, Curse Tablets and Binding Spells From the Ancient World (New York; Oxford: Oxford University Press, 1992) (= Gager); R. Kotansky, Greek Magical Amulets: The Inscribed Gold, Silver, Copper and Bronze Lamellae, Part i: Published Texts of Known Provenance (Opladen: Westdeutscher, 1994); A. Kropp, Defixiones: Ein aktuelles corpus Lateinischer Fluchtafeln. Dfx. (Speyer: Kartoffeldruck-Verlag Kai Brodersen, 2008); R. W. Daniel and F. Maltomini, eds., Supplementum Magicum, vols. i–ii (Opladen: Westdeutscher, 1990–2).

10 Dell, ‘Matthäus’, 45–6.

11 C. Faraone, ‘The Agonistic Context of Early Greek Binding Spells’, Magika Hiera: Ancient Greek Magic and Religion (ed. C. A. Faraone and D. Obbink; New York: Oxford University Press, 1991) 3–32.

12 Davies and Allison argue that ‘vv. 17–19 came to Matthew as a piece’ (Gospel according to Saint Matthew, ii.602), pace Luz (Matthew 8–20, 356–7), who insists that ‘v. 19b, c was a traditional logion that was originally independent of v. 19a’ (p. 356).

13 For this, see R. H. Hiers’ succinct lists (“Binding” and “Loosing”: The Matthean Authorization’, JBL 104 (1985) 233–5Google Scholar).

14 Hiers, ‘“Binding” and “Loosing”’, 233–50.

15 Davies and Allison, Gospel according to Saint Matthew, ii.602. Although I agree that Matt 16.17 is likely to be a part of the single unit, I have decided to focus on Matt 16.18–19 in order to offer a clear representation of the magical theme in the passage.

16 LSJ s.v. ‘κατισχύω’; W. Grundmann, ‘ἰσχύω, κτλ.’, TDNT iii.398.

17 Davies and Allison, Gospel according to Saint Matthew, ii.635.

18 Davies and Allison, Gospel according to Saint Matthew, ii.639.

19 In most cases, I follow the translations provided by the literature cited in n. 9 and those of LCL, with some modification where necessary (mostly for more literal rendering). I provide translation of my own when there is no other translation available.

20 Faraone, ‘Agonistic’, 21.

21 Wilhelm, A., ‘Über die Zeit einiger attischer Fluchtafeln’, JOAI 7 (1904) 121–2Google Scholar (= SGD no. 18); my translation.  The writer of the spell made the mistake of writing -ος instead of -ους. Ἀντιφάνος is a mispelling of Ἀντιφάνους (genitive of Ἀντιφάνης). τούτος is simply a misspelling of τούτοuς. I am indebted to David Martinez of the University of Chicago for figuring out these grammatical errors of this spell.

22 A recipe for the phylactery for the foregoing after initiation. The underlined text is in Coptic.

23 In addition to the several examples quoted, see also DTA 88, 108; DT 164, 163, 187, 249, 252. I cite an instance here from DT 81.a3 (= Wünsch, DTA, praef., p. viii): Ἑ[ρμ]ῆν [κι]κλήσκω χθόνιαν καταδίδημι Δ[ι]ονυσίαν καὶ Φερσεφόνηαν δῆσαι Διονυσίας γλῶς[σ]αν . . ατε το . .

24 Inv. no. IL 950: 205–55, at 214–15 (Jordan, D. R., ‘Defixiones from a Well near the Southwest Corner of the Athenian Agora’, Hesperia 54 (1985) 214–15CrossRefGoogle Scholar).

25 On CHRAB, see Gager, 55: ‘The same four Greek letters precede the two verbs of binding, though they do not make Greek words. Mouterde … suggests, appropriately in a region where Hebrew and Aramaic would have been familiar, that behind these letters lies the Hebrew or Aramaic verb ḥrb, “to destroy”.’

26 Pace Faraone, who insists that this formula (see n. 21) ‘seems to imply that a binding curse could have a limited duration or be loosed at a later date’ (‘Agonistic’, 26).

27 F. Graf, Magic in the Ancient World (Cambridge, MA: Harvard University Press, 1997) 168. DT 37: ne quis solvat nisi nos qui fecimus. A similar curse is found in a lead attraction-curse tablet (found in a cemetery at Macedonia, dated to the late fourth century bce or later; Inv. I.160.79/1987): ‘Of Melissa of Apollonia. Pausanias puts a binding spell (καταδεῖ) on Ainis. May s/he not be able to touch a victim nor be able to get possessed of any other good, before Ainis is gracious to Pausanias. And may no one other than Pausanias undo these things (ταῦτα δε[ὶ] μηδεὶς ἀναλύσαι ἀλλ’ ἢ Παυσανίας).’ (Greek transcript and translation from D. Jordan, ‘Three Curse Tablets’, World of Ancient Magic, 122).

28 Graf, Magic, 168.

29 R. Kassel and C. Austin, eds., Poetae comici Graeci, vol. v (Berlin/New York: de Gruyter, 1986) 630; Graf, Magic, 168 and 286 n. 155. Also see Gager (223): ‘a text of Magnes, a comic dramatist, indicates that there existed specialists (analutai) who offered to dissolve (analuein) spells directed against the clients’.

30 For an additional example, see PGM xxxvi.256–64:

Take a triangular ostracon from the intersection of a road – pick it up with your left hand – write on it with myrrh-ink and then hide it – “ASSTRAÊLOS CHRAÊLOS, destroy every spell (λύσατε πᾶν φάρμακον) prepared against me (so-and-so), for I invoke you according to the great and frightful names which the winds fear and which make the rocks split apart at its sound” (Gager, no. 130)

31 A confessional inscription Asia Minor, Lydia, 156 ce. H. S. Versnel cites two terms from two inscriptions that relate the notion of ‘impossibility to unbind’ to a threatening, unbreakable curse: the gods who are called ἀνεπιλύ[τους] (Eugene Lane, CMRDM, no. 155), and the ‘sceptres’ that are ἄλυτα (τὰ ἄλυτα σκῆπτρα) (Petzl, G., ‘Inschriften aus der Umgebung von Saittai (i): Encekler, Hamidiye, Ayazviran’, ZPE 30 (1978) 249–76Google Scholar, at 260 (for a discussion, see Versnel, H. S., ‘‘May he not be able to sacrifice’: Concerning a Curious Formula in Greek and Latin Curses’, ZPE 58 (1985) 247–69Google Scholar, at 261–3). In this regard, see an interesting remark from Philo's On the Confusion of Tongues (LCL) 167, where he describes the Jewish God as the possessor of unbreakable binding power: ‘by God, who has fastened chains which can never be broken round the universe, namely, his own powers, with which he binds everything, willing that it shall never more be released’ (ὑπὸ θεοῦ, ὃς τοῖς ὅλοις δεσμοὺς τὰς ἑαυτοῦ δυνάμεις περιῆψεν ἀρρήκτους, αἷς τὰ πάντα σφίγξας ἄλυτα εἶναι βεβούληται).

32 J. Maldonatus, A Commentary on the Holy Gospels (2 vols.; London: Hodges, 1888; orig. 1596–7) i. 41. Re-quoted from Marcus, ‘The Gate of Hades’, 444. Also see J. Jeremias, ‘πύλη, πυλῶν’, TDNT vi.926.

33 Cf. Davies and Allison, Gospel according to Saint Matthew, ii.632. ‘πύλαι ᾅδου is a pars-pro-toto term for the ungodly powers of the underworld which assail the rock’ (Jeremias, ‘πύλη, πυλῶν’, TDNT vi.927) For a general discussion of the various meanings of ‘key’ in Greco-Roman and Jewish literature, see J. Jeremias, ‘κλείς’, TDNT iii.745–53.

34 Marcus, ‘The Gate of Hades’, 446: ‘In none of these [Jewish] examples, however, is the word “gates” used through synecdoche for the demonic rulers. Rather, the gates are literal gates, through which the demonic powers pass; nor are the gates portrayed as attacking.’

35 I do not insist that the term the ‘gates of Hades’ is always to be understood in a magical context. My point is that the appearance of Hades with the theme of binding and loosing, accompanied by the verb denoting ‘overcome’, strongly indicates a magical context.

36 This defixio comes from around the mid-third century ce.

37 For Pluto, see also Dell, ‘Matthäus’, 28.

38 Aramaic incantation bowl 5. For the text and discussion, see J. Naveh and S. Shaked, eds., Amulets and Magic Bowls: Aramaic Incantation of Late Antiquity (Leiden: Brill, 1985) 158–63.

39 =R. S. Stroud, Corinth: Results of Excavations Conducted by the American School of Classical Studies at Athens, vol. xviii.6:The Sanctuary of Demeter and Kore: The Inscriptions (Princeton: American School of Classical Studies at Athens, 2013) 104–5.  A curse tablet ‘found rolled and pierced by two iron nails’ in the temple of Demeter at Acrocorinth (Stroud, Corinth, 104).

40 P. 385 with n. 24.

41 Ibid., 252–3. I first encountered these ‘Defixiones from a Well’ and a Latin tablet cited on the current page in the work of D. R. Smith, ‘Hand This Man Over to Satan’: Curse, Exclusion and Salvation in 1 Corinthians 5 (LNTS 386; London: T&T Clark, 2008) 84, 87–8, 98.

42 Gager, no. 134.

43 Cf. Dell, ‘Matthäus’, 28; D. E. Aune, Revelation 1–5 (WBC; Dallas: Word Books, 1997) 104–5.

44 See also DT 22 = SGD 193 = Gager, no. 45, quoted above.

45 So Dell, ‘Matthäus’, 29, 38. A given image or theme can be utilised and appreciated variously, even in opposite ways. Some stock phrases and technical terms in binding magic are adopted for various literary purposes, as this investigation shows. For instance, while the focal Matthean passage appeals to one of the characteristics of ‘rock’ – namely, its hard-to-break quality – the following magic formula appeals to a power of a deity who actually ‘breaks apart rocks’:

Lord, you who have a great name, you whom we all have each in our own heart; your name is BARPHANNETH RALPHAI NINTHER CHOUCHAI. You who break apart rocks and change the names of gods, enter in, appear to me, lord, you who have in fire your power and your strength, SESENGENBARPHARAGGES. (PGM iv.1020–5; Betz, 58)

46 H.-J. Klauck, The Religious Context of Early Christianity: A Guide to Graeco-Roman Religions (Minneapolis: Fortress, 2003) 214.

47 Also cited by Klauck, Religious Context, 215. For a more vivid presentation of the principle of ‘sympathetic’ magic, see this magical papyrus: ‘[H]aving tied a cord [to the plate] throw it into the stream – or into the sea – [and let it] be carried along. Use the cord so that, when you wish, you can undo [the spell]. Then should you wish to break [the spell], untie the plate’ (PGM vii.435–40; Betz, 129).

48 Marcus, ‘The Gate of Hades’, 455.

49 Davies and Allison, Gospel according to Saint Matthew, ii.633.

50 Bultmann, History, 139.

51 ‘Burial sites of those who died young or by violent means were the preferred choices because it was believed that their souls remained in a restless conditions near the graves until their normal life-span had been reached’ (Gager, 19). For instance, see PGM iv.335–43 (Betz, 44): ‘I entrust this binding spell to you, chthonic gods … to infernal gods and daimons, to men and women who have died untimely deaths (ἀώροις τε καὶ ἀώραις) …’

52 R. Hays, The Faith of Jesus Christ: The Narrative Substructure of Galatians 3:1–4:11 (Grand Rapids: Eerdmans, 20022) lii. Hays owes this expression to Dale Allison.

53 Ibid.

Save article to Kindle

To save this article to your Kindle, first ensure is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about saving to your Kindle.

Note you can select to save to either the or variations. ‘’ emails are free but can only be saved to your device when it is connected to wi-fi. ‘’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.

Find out more about the Kindle Personal Document Service.

Ancient Binding Spells, Amulets and Matt 16.18–19: Revisiting August Dell's Proposal a Century Later*
Available formats

Save article to Dropbox

To save this article to your Dropbox account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you used this feature, you will be asked to authorise Cambridge Core to connect with your Dropbox account. Find out more about saving content to Dropbox.

Ancient Binding Spells, Amulets and Matt 16.18–19: Revisiting August Dell's Proposal a Century Later*
Available formats

Save article to Google Drive

To save this article to your Google Drive account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you used this feature, you will be asked to authorise Cambridge Core to connect with your Google Drive account. Find out more about saving content to Google Drive.

Ancient Binding Spells, Amulets and Matt 16.18–19: Revisiting August Dell's Proposal a Century Later*
Available formats

Reply to: Submit a response

Please enter your response.

Your details

Please enter a valid email address.

Conflicting interests

Do you have any conflicting interests? *