1. Introduction
Sometimes scholarly disapproval of a given view solidifies without any hard evidence. August Dell's proposal (1914)Footnote 1 to read the motif of ‘bind and unbind’ in Matt 16.19 against the backdrop of ancient magicalFootnote 2 conventions has long suffered such disapproval – and with no substantial philological argument against it. Rudolf Bultmann was one of the earliest exegetes to dismiss Dell's proposal that magical conventions underlie the language of ‘bind and loose’ in Matt 16.19.Footnote 3 In his famous book History of the Synoptic Tradition, Bultmann rejects Dell's thesis that λύειν and δέειν in Matt 16.19 are to be understood as technical terms of magic – terms denoting ‘to set the magic bound free’ and ‘to bind by magic’, respectively. For Bultmann, the verb δέειν and the immediate context of Matt 18.18 (a verbatim repetition of 16.19) pose a major difficulty for this kind of ‘magical’ reading: ‘For if we try to think of the authority of λύειν in terms of Mk 735; Lk 1316 as the ἐξουσία τῶν πνευμάτων τῶν ἀκαθάρτων (Mk. 67), how shall we interpret δέειν? The context in Matt 1818 shows quite plainly that Matthew did not so understand them [as magical incantations].’Footnote 4 Yet this is the very place where Bultmann's lack of the knowledge of ancient binding magic – an area of magic whose formulae are full of the verb δέειν and its cognates (notably καταδέειν) – becomes evident. Also problematic is Bultmann's appeal to the immediate context of Matt 18.18, which construes the language of ‘bind and untie’ as referring to ‘doctrinal’ or ‘disciplinary’ authority, as evidence for his preferred reading.Footnote 5 That interpretive move in fact undercuts Bultmann's own view that Matt 18.18 is a variation and a later formulation of Matt 16.19.Footnote 6
Unfortunately for Dell's thesis (and for our understanding of the passage itself), many have followed in Bultmann's footsteps, refuting the magical background for understanding Matt 16.19 without either substantially engaging with ancient magical sources or carefully measuring the likelihood of this reading in comparison with other interpretive options. The two most thorough and authoritative modern commentaries on Matthew are representative examples: Luz's Hermeneia commentary and Davies and Allison's ICC,Footnote 7 which provide virtually no discussion about why their ‘rabbinic’ reading is to be preferred to a ‘magical’ reading.Footnote 8
However, thanks to the growing scholarly interest in Greco-Roman magic, along with the ongoing publication of the collections of magical primary sources and translations,Footnote 9 we are now in a better position to test the validity of the so-far-neglected proposal of Matt 16.19's magical background. By demonstrating some remarkable linguistic and thematic affinities between this verse and the ancient binding spells, the present essay seeks to promote a ‘magical’ reading as one of the plausible interpretive options for Matt 16.19. Indeed, the magical context illuminates even Matt 16.18. In so demonstrating, the study will improve and rectify Dell's insightful but insufficiently supported proposal, which engages with unnecessarily broad comparanda (sweeping from Babylonian tablets to rabbinic literature) and unduly delimits the meaning of ‘bind and loose’ to an authority over evil spirits.Footnote 10 As we shall show below, in Greco-Roman binding magic, a ritual ‘binding’ can have almost any type of target, including a lover, horse or chariot, parts of human anatomy, demons and cities. The language of binding thus bears much broader implication than a mere authority over demons. Moreover, it becomes clear that the terms and themes in Matt 16.18–19 can be understood more potently and relevantly if the comparanda are narrowed down specifically to Greco-Roman binding magic and amulets.
A binding spell is a performative ritual utterance imposing some restriction on a victim or victims so that the target's willpower and the function of the target's faculties do not operate properly, for the benefit of the defigens (the one who issues the ‘binding’). Hence, it is understandable that binding magic was most often practised agonistically, owing to the intense competitiveness of Greco-Roman life.Footnote 11 This agonistic aspect of the ritual binding should be seriously considered in interpreting the Matthean passage.
A caveat is in order here: this thesis does not suggest that Matt 16.19 endorses magical practice or that the early Christianity reflected in this passage closely resembles ancient magic. Rather, the contention is that ancient binding and unbinding spells can shed light on some of the phrases and themes in Matt 16.18–19, in the sense that the passage judiciously and tactfully appropriates the imagery of the popular binding magic and imbues into it some radical twists and turns, so as to vividly represent the Christian community as a ‘safety zone’ on the battlefield of religious rivals, and as an entity that can overpower various threatening spiritual powers.
Rather than dealing with the problem of the tradition-history of Matt 16.18–19,Footnote 12 we seek to account for an inner religious logic and some cultural assumptions that run through the verses, regardless of the answer to the question whether the composition of the piece was done by the First Evangelist or anonymous hand(s) before him. Several recognised but counter-interpretations of Matt 16.19 will not be surveyed or assessed, since that information can be easily obtained by consulting various fine commentaries and articles.Footnote 13 The goal is to reopen the unfortunately unrefined but still illuminating proposals of August Dell and his followers,Footnote 14 and to corroborate this hypothesis with philological precision and contextualisation.
The thesis is based on the following five observations: (1) there is some lexical and topical relationship between Matt 16.18 and 16.19, such as the motif of a ‘power struggle’ between the two realms (represented by the verb κατισχύειν), the paired imagery of key and door, and the spatial contrast between Hades and the kingdom of heaven, which make Matt 16.18–19 a single unit; (2) the ‘rabbinic’ interpretation fails to understand Matt 16.18; (3) binding magic in Greco-Roman society and the notion of the ritual ‘binding’ and ‘loosing’ associated with it were remarkably popular in Greco-Roman society; (4) a similar assumption underlies ancient magic and Matt 16.19 – the transference of power between two ‘ritually connected’ realms; (5) breaking the spell requires a stronger power than that used in ‘binding’. The first and second observations reveal important desiderata and limitations of the ‘rabbinic’ interpretation, and at the same time will pave the ground for the ensuing interpretive steps. Observations 3 and 4 demonstrate that the magical reading is a plausible interpretive candidate, and the fifth illuminates the meaning of Matt 16.19 in light of magical conventions.
The present thesis draws its validity from both internal and external factors: its explanatory power for many concepts, language and assumptions in Matt 16.18–19, for one, and also the remarkable popularity of binding magic during the compositional period of the material behind Matt 16.18–19 and its early reception. If we take into consideration the above five points all at once, we arrive at a fresh reading of 16.18–19: this Matthean text deliberately adopts the ideas and the language of binding magic and amulets to highlight ironically that Christ endows Peter with the supreme power as the key-bearer who breaks all the hostile, ‘binding’, anti-God powers and thereby represents the Christian assembly as a stable and secure site in the midst of intimidating entities. I will first show that the paired motif of ‘bind and loose’ was a stock phrase for the practice of binding and unbinding spells. Then the verb δέειν and its cognates will be discussed in light of binding magic, and this is followed by the exploration of some religious implications of ‘unbinding’. After establishing the magical context of Matt 16.19, I shall interpret several important expressions and ideas such as ‘to prevail’ and the ‘gates of Hades' in Matt 16.18 in light of the ancient context of binding magic.
2. Reading Matt 16.18–19 in Light of Ancient Binding Magic
Our task is to find an underlying religious logic behind Matt 16.18–19. For this we first need to show that the logia in 16.18 and 16.19 are closely related. These two verses have many points of contact, both lexically and thematically. First of all, the gate and the key constitute a well-known imagery pair. In addition, the antithetical spatial relationship between Hades and the kingdom of heaven corroborates more strongly the literary bond that conflates 16.18 with 16.19. Further, as Davies and Allison correctly observe, these two verses also have some structural resemblance, that is to say, both ‘contain three sentences, each of which consists of three parts – a statement of theme plus an antithetical couplet’.Footnote 15 Indeed, as will be shown below, the words clustered here – key, Hades, give, prevail and bind – are some of the significant components of ancient binding spells. Given that correlation, it is particularly noteworthy in our Matthean passage that the verb κατισχύειν (to conquer, to overcome, to prevail, to be superior, or to overpower)Footnote 16 appears in conjunction with Hades. This verb sets the tone of 16.18–19 as a ‘power struggle’ between the two realms or entities, namely, God (with his church) and the anti-God powers.
Considering the close connection between these two verses, the inadequacy of the ‘rabbinic’ interpretation for understanding 16.18 becomes clear. Davies and Allison, for example, see binding and loosing ‘as metaphors for halakhic decision’.Footnote 17 This interpretive choice leads them to opine that ‘just as the kingdom itself is taken from the Jewish leaders and given to the church (21.43), so are the keys of the kingdom taken from the scribes and Pharisees and given to Peter … Jesus replaces the Jewish academy with his own “chief rabbi”.’Footnote 18 However, their interpretation cannot address certain pressing questions: What does ‘the authority of the halakhic decision’ have to do with the gates of Hades? What does the verb prevail have to do with the halakhic authority?
If it is right that Matt 16.18–19 is a well-crafted, single unit and that the ‘rabbinic’ reading fails in its reading of this passage as such a unit, what then would be the proper background behind our passage? Ancient binding spells provide the requisite background, because (1) the remarkable popularity of binding magic might have created some expectation of magical connotation when the audience heard the concepts ‘bind and unbind’ in a religious context and (2) this proposal reveals significant explanatory power for construing many ideas and terms in the focal Matthean text.
To begin with, Pliny's remark with regard to the pervasive fear and the popularity of magic across the regions of the Roman Empire is revealing:
ita possessis hominum sensibus triplici vinculo in tantum fastigii adolevit, ut hodieque etiam in magna parte gentium praevaleat et in oriente regum regibus imperet.
Accordingly, holding men's emotions in a three-fold bond, [magic] rose to such a height that even today it has sway over a great part of mankind, and in the East commands the Kings of Kings. (HN 30.1 (LCL, Jones))Footnote 19
Although Pliny here refers to the conventions of magic in general, his point is also relevant to the binding spell in particular, since in an earlier section he reports that ‘there is indeed nobody who does not fear to be spell-bound by imprecations' (defigi quidem diris deprecationibus nemo non metuit, HN 28.4.19). Pliny's remark can be securely verified by the overwhelming numbers of extant artefacts discovered across the vast region and across time in various media, such as inscriptions, magical papyri, lead tablets and bowls.
Among the many media of binding spells, lead curse tablets – κατάδεσμοι/defixiones – are particularly notable and relevant for our study, given their pervasive popularity and thorough occupation with the language of ‘binding’. Indeed, ‘[t]he Greek term κατάδεσμος is derived from the verb καταδέω (which appears in Attic dialect in contracted form καταδῶ) “bind down” or “bind fast”’.Footnote 20 Therefore, in the Greco-Roman world the theme of binding must have evoked in the minds of many some notion of magical and ritual binding, particularly if mentioned in a religious context. Owing to the great popularity of binding magic, the pair ‘bind and unbind’ functioned as a shorthand for the whole scope of ancient conventions of binding spells and of breaking spells. For instance, in a curse tablet and in a magical papyrus we see these phrases:
I bind and shall not loose Antikles, son of Antiphanes and Antiphanes … I bind them all before Hermes of the underworld … I shall not loose [them] (καταδῶ καὶ οὐκ ἀναλύσω Ἀντικλέα Ἀντιφάνος και Ἀντιφανην … καταδῶ τούτος ἅπαντας πρὸς τὸν Ἑρμῆν τὸν [τὸν] χθόνιον … καὶ οὐκ ἀναλύσω).Footnote 21 With blood from the hand or foot of a pregnant woman, write the name given below on a clean piece of papyrus; then tie it about your left arm by a linen cord and wear it. Here is what is to be written: ‘SHTĒIT CHIEN TENHA, I am the one who binds and who releases (ⲁⲛⲟⲕ ⲉⲧⲥⲱⲛϩ ⲉⲧⲧⲓⲟⲩⲱ) …’ (PGM iv.79–82; Betz, 38, with modification)Footnote 22
As the first example shows, the most frequent word employed in the curse tablets is καταδέω. However, δέω and its equivalents are not uncommon (pace Bultmann). I cite a few instances:Footnote 23
‘Borphor’ syllables – babaie, mighty Betpyt, I hand over to you Eutychian (παραδίδωμι σοι Εὐτυχιανόν), whom Eutychia bore, that you may chill him and his purposes, and in your dark air also those with him. Bind (Δῇς) in the unilluminated aiôn of oblivion and chill and destroy also the wrestling that he is going to do in the … this coming Friday …Footnote 24 Eulamos, restrain … restrain … bind down … in order that you bind Domninos the son of Fortuna, especially, Aretmis the [son] … ([Εὐλάμων κατ]έχι … κατήσχ[ητε] … καταδήσητε … εἵνα … λήνην δέσθε Δομ[νῖνον τὸν υἱὸν] Φορτούνας, μάλειστ[α δὲ Ἀρτέ]μιν τὸν …). (DT 164) OREOBARZAGRA AKRAMMACHARI PHNOUKENTABAÔTH OBRABAU, you holy angels, ambush and restrain LULATAU AUDÔNISTA them. The spell – OIATITNOUNAMINTOU MASKELLI MASKELLÔ PHNOUKENTABAÔTH OREOBARZA, now attack, bind, overturn, cut up, chop into pieces (συνέρθατε, δήσατε, καταστρέψατε, λεπτοκοπήσατε, <σ>φάλετε) the horses and the charioteers of the Blue colors – Numphikos, Thalophoros … CHRAB, bind (δῆσον) and CHRAB, damage(?) the hands, sinews of the horses and charioteers of the Blue colours. (SGD 167 = SEG vii.123 = Gager, no. 5)Footnote 25
A variation of the verb δέω is also adopted: ‘I bind (δεσμεύω) Isias, before Hermes the Restrainer, the hands, the feet of Isias, the entire body’ (SGD 64). As we noted above, the word-group denoting binding (καταδέω, δέω) is the core of the binding spells in PGM and defixiones. The ritual binding intends to put the victim under the control of the invoked deity, aiming to invalidate some normal function (such as love, speech, wrestling or horse racing) against the will of the victim and for the benefit of the curse-maker.
On the other hand, the popularity and reputation of binding spells themselves are based on the firm belief in the difficulty of ‘unbinding’ the given curse: otherwise, there is no point at all for the burdensome preparation of any defixio.Footnote 26 A little stretch of the imagination is enough to suppose that the pervasiveness of binding curses engenders some urgent need for ‘protection’. It is no surprise that breaking the formidable binding spells was often called unbinding, or loosing. Breaking a binding spell was believed to require a much stronger counter-spell or the invocation of a deity more powerful than the binding deity. As Fritz Graf correctly observes, ‘Countermeasures seem to have been more complicated: the ritual binding asked for a ritual unbinding. This could be difficult; a spell from Republican Rome even asserts, “Nobody shall undo me if not he, who did this”’.Footnote 27
Indeed, there seem to have existed certain specialists and recipes for counteracting binding curses. In a fragment, the comic poet Magnes ‘combines interpreters of dreams and “untiers”’:Footnote 28 ὀνειροκρίταισιν ἀναλύταις (fr. 4).Footnote 29 Unbinding incantation often employs the verb λύειν or its compounds ἀναλύειν and ἀπολύειν, and was sometimes called λυσιφάρμακον (λύσις + φάρμακον). The following two magical texts are exemplary:Footnote 30
You will have honor, happiness; you will receive inheritances, have good fortune, be unaffected by potions and poison; you will break spells and conquer your enemies (ἔνδοξος, μακάριος ἔσει, κληρονομίας ἕξεις, εὐτυχήσεις, φάρμακα νικήσεις, καταδέσμους ἀναλύσεις καὶ ἐχθροὺς νικήσεις). (PGM iv.2175–8; Betz, 76) A charm to break spells (λυσιφάρμακον): Take lead and draw on it a unique figure holding a torch in its right hand, and in its left – and at the left – a knife, and on its head three falcons, and under its legs a scarab and under the scarab an ouroboros serpent. (PGM xxxvi.178–84; Betz, 273)
As can be clearly noted in the latter papyrus, the term λύσις in a magical context explicitly denotes a breakage of binding spells. We have another interesting example from an inscription that describes a removal of a curse by using the verb λύειν. Here a woman, Tatia, is suspected of setting up a sceptre and placing a curse (probably a defixio) against her son-in-law Ioukoundos. She and her offspring went through some misfortunes, and eventually her descendants and relatives decided to remove the sceptre and curses, and instead to set up a stele in order to propitiate and praise gods:
Therefore great are the gods of Axiottenos! They set about to have ‘loosened’ the sceptre and the curses that were in the temple, the ones the estate of Ioukoundos and Moschios had sought to undo … (TAM v.1 (1981) no. 318, lines 23–8 = Gager, no. 137)Footnote 31
So far we have seen 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. What insight, then, does this observation give into Matt 16.19? As we noted, breaking a binding curse, at least in theory, requires extremely strong power and authority. Correspondingly, 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. Our suggestion for the magical background of Matt 16.19 can be corroborated when we see how some phrases, such as ‘the gates of Hades’ and ‘giving the key of the kingdom of heaven’, resonate within the semantic world of ancient binding curses.
3. Binding Magic, Hades, Gates and Key
That magical background can also illuminate the contextual meaning of ‘the gates of Hades’ in Matt 16.18, which has puzzled generations of scholars. One of the strengths of the present proposal lies in its power to solve the problem of actively ‘attacking’ gates: ‘Gates do not overcome, but resist.’Footnote 32 There have been several suggestions to deal with this problem, and the most convincing of them is to see ‘gates of Hades’ as a synecdoche for the entire city of the dead, or even for the ‘ruler’ of the underworld.Footnote 33 However, this solution has an important limitation: as Joel Marcus's survey shows, there is no extant Jewish text that describes ‘the gate(s) of Hades (or Sheol)’ as a synecdoche for the attacking demonic rulers.Footnote 34 A ‘magical’ reading can address this problem.Footnote 35
In magical sources, we have several examples that explicitly invoke ‘the ruler(s)’ of the gates of Hades, thereby providing context for the imagery of ‘attacking’ gates of Hades. The following two texts, among others, that can illustrate how ‘the gate of Hades’ is related to the images of attacking. In a curse tablet found at Cyprus that aims at ‘binding’ the tongues of the opponents in a legal procedure (DT 22 = SGD 193 = Gager, no. 45), we encounter the phrase ‘the ruler of the gates of Hades’, who is to ‘attack’ the victim:Footnote 36
Daimones under the earth and daimones whoever you may be; … take over the passion of Ariston which he has toward me, Sotêrianos also called Limbaros, and his anger; and take away from him his strength and power and make him cold and speechless and breathless, cold toward me … and hand him over to the doorkeeper in Hades (παράδοτε τῶ κατ’ Ἅδη θυρουρῶ), MATHUREUPHRAMENOS, and (to/of?) the one who is appointed over the gate to Hades and the door bolts of heaven (τὸν ἐπὶ τοῦ πυλῶνος τοῦ Ἅ[δους κ]ὲ τῶν κλήθρων τοῦ οὐρανοῦ), STERXERX ÊRÊR[XA] earthshaker ARDAMACHTHOUR PRISGEU LAMPADEU. And bury in this mournful (grave) the one whose name is written on this curse tablet which brings about silence … Hear the great name, for the great SISOCHOR rules over you, the ruler of the gates of Hades (ὁ ἐξάγων τοῦ Ἅδους τὰς πύλας). Of my opponent Ariston bind (κατα[δ]ήσατε) and put to sleep the tongue, the passion and the anger he holds toward me … I invoke you by the gods in Hades, OUCHITOU, the dispenser of tombs, AÔTH IÔMOS TIÔIE IÔEGOÔEOIPHIRI, who in heaven rule the upper kingdom … I invoke you gods who were exposed by Kronos … and you, earthshaker, who holds the keys of Hades. Carry out for me, you … Provide … (emphasis added)
This magical text has a number of points of contact with our Matthean text: the verb ‘bind’, the figure who is appointed over the door bolts of heaven, the heaven as the upper kingdom, the key, and the gates of Hades. The one who is in charge of the gate of Hades and the door bolts of heaven is addressed to ‘bind’ the victim, and the curser's main concern is to prevail in a lawsuit. Another similarly informative text may be added. Pluto, a god equivalent to Hades, was often called upon in curse tablets.Footnote 37 In the following Latin curse tablet, we find a cluster of words that indicate the prevailing force of the gates of Hades in its binding power, which reminds us of our Matthean passage: Pluto ( = Hades), the verb ‘overcome’, the reference to the gate of Hades (here symbolised by Cerberus), an indication of ‘binding force’, and escaping a curse:
Good and beautiful Proserpina or Salvia, if you prefer that I call you so, wife of Pluto, snatch away the health, the body, the complexion, the strength, and the faculties of Plotius. Hand him over to Pluto, your husband (Tradas [Plutoni] viro tuo). May he not be able to escape this (curse) by his wits (Ni possit cogitationibus sueis hoc vita) … Let them overcome him to the point where they snatch away his soul (ev[in]cant, [vincant,] usq[ue dum animam eiu]s eripia[nt) … Summon for me the triple-headed hound (M[e mittas a]recessitum canem tricepitem) to snatch away the heart of Plotius … These I will offer you, Proserpina Salvia, when you complete this in an orderly fashion … (Gager, no. 134; Latin tablet from Rome, 1st century bce)
These two magical texts describe Hades or the doorkeeper/key-bearer of Hades (SISOCHOR, Cerberus) as a major agent for carrying out some ‘attack’ on the victim through ‘binding’. Also noteworthy is the employment of the words denoting ‘overcome’ or ‘victory’ or ‘destroy’ in rendering the effect of the binding curse. This observation, in my view, can shed light on the verb κατισχύειν in Matt 16.18. The verbs that connote ‘overcome’ or ‘prevail’ are often adopted in PGM and in defixiones, precisely because the concerns of the clients are situated in agonistic contexts, either in the battlefield of competing deities or in the realm of victory or defeat between the curse-maker and the victim, such as a lawsuit or race.
In addition to the text cited above, a few more examples will suffice as evidence for this agonistic aspect of binding magic:
By your name I make this amulet so that it may fortify this person and the threshold of the house and any possessions he may have. I bind the rocks of the earth and tie down the mysteries of heaven. I overcome/suppress/subdue them (כבשתינון) … I rope, I tie, I overcome/suppress/subdue all demons and harmful spirits in the world … In the name of Gabriel, the mighty hero, who kills all heroes who are victorious in battle … In the name of Yah, Yah, Yah, Sabaoth. Amen, Amen, Selah. (Gager, no. 123, with a slight modification)Footnote 38 I consign and entrust Karpime Babia, the weaver of garlands, to the Fates who exact justice so that they may expose her acts of insolence, and to Hermes of the Underworld, to Earth, to the children of Earth, so that they may overcome (κατεργάσων<τ>αι) and completely destroy her … (Gager, 37 n. 92)Footnote 39
The following example is illustrative in that it explicitly relates ‘breaking the spell’ to ‘victory’ over the enemy:
You will have honor, happiness; you will receive inheritances, have good fortune, overcome potions and poison [or, magic]; you will break spells [lit. loosen the binding curse] and conquer your enemies (ἔνδοξος, μακάριος ἔσει, κληρονομίας ἕξεις, εὐτυχήσεις, φάρμακα νικήσεις, καταδέσμους ἀναλύσεις καὶ ἐχθροὺς νικήσεις). (PGM iv.2175–8; Betz, 76, with my modification)
How can these observations provide insight into Matt 16.18b–c? Here the dominical saying puts stress on a particular aspect of the church, namely the church as the invulnerable and invincible site built on ‘this rock’. If ‘the gates of Hades’ metonymically represents some ‘attacking’ demonic ruler(s) of the underworld (evidenced especially in binding magic), the message of our passage would be that the church should not be afraid of such an evil force, however intimidating it appears to be. The ‘gates of Hades’ or any other threatening binding force associated with it cannot prevail or do harm to the church built on ‘the rock’. The security of the church against malevolent and harmful spiritual entities is further guaranteed by Christ's twofold promises: (1) 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; (2) Christ promises that whatever objective the church ‘binds’ and whatever spells it ‘unbinds’ will be bound or loosed exactly in the same ways in heaven. Let me unpack these two points in detail.
4. ‘Endowment of a Key’ in the Context of Magic
Given the imagery of the conflicting force fields in v. 18 that is conjoined with the theme of ‘binding and loosing’, the common verb δίδωμι in v. 16.19a attains a technical connotation of ancient binding spells. In many binding curses, the verb δίδωμι and its cognates (such as παραδίδωμι, and the verb do in Latin tablets) are employed almost as a synonym for δέω, a referent to the ritual ‘binding’ or ‘entrusting’ the target under the control of an invoked deity. For instance, in the curse tablet targeting a wrestler and discovered in a well near the Athenian Agora quoted earlier,Footnote 40 ‘binding’ and ‘handing over’ are manifestly related and even used somewhat interchangeably: ‘Bind (Δῇς) in the unilluminated aiôn of oblivion … I hand over (παραδίδω[μι) to you Eutychian whom Eutychia bore …’
In another curse tablet found in the south-west corner of the Athenian Agora, δίδωμι and παραδίδωμι are adopted as synonyms:
Mighty Typhon, I hand over to you (παραδίδωμί σοι) Tyche, whom Sophia bore, that you may do her harm … <Bind into> the unilluminated aiôn of oblivion … I give you (δίδωμί σοι) Tyche, whom Sophia bore, the aforenamed, whose hairs these are, rolled up … Bind (δ̣ῆσον), twist … that she may no longer rise up, walk around …Footnote 41
These curse formulae indicate that the ritual act of ‘giving’ intends to put the victim within the controlling power of the addressed deity. The following curse makes this point explicit:
Proserpina Salvia, I give over (dono) to you the nostrils, lips, ears, nose, tongue, and teeth of Plotius, so that he may not be able to say what is causing him pain … No matter what he may have written, great or small, just as he has written a proper spell and commissioned it (against me), so I hand over (trado) and consign (mando) Plotius to you, so that you may take care of him by the month of February. Let him perish miserably …Footnote 42
Now compare these binding curses and Matt 16.19a:
δώσω σοι τὰς κλεῖδας τῆς βασιλείας τῶν οὐρανῶν (Matt 16.19a)
I give you (δίδωμί σοι) Tyche … (Inv. no. IL 1737 (Jordan, ‘Defixiones from a Well’, 252–3))
Proserpina Salvia, I give over (dono) to you the head of Plotius … (Gager, no. 134)
We note that in Matt, while the language of binding magic is employed, both the power hierarchy between the giver and the receiver and the characteristics of the direct object to be ‘bound’ are reversed. In curse tablets, the ritual ‘giver’ is a human, and the ‘receiver’ is a superhuman being. But in Matthew, the giver is Christ; the receiver, a human being. While the direct object in these curse formulae is the victim of the binding spell, Matt 16.19 has ‘the key of the kingdom of heaven’ as its direct object. Nevertheless, both in the defixiones and the Matthean text, the receivers are the agents who are to perform ‘binding’. For this reason, we see Matt 16.19 as a literary reworking of a typical binding spell, crafted to highlight Peter's authority to ‘bind’ and ‘break’ whatever entities he should choose; Jesus giving the key of the kingdom of heaven to Peter can be understood almost as Jesus ritually ‘binding’ or entrusting the key of his kingdom under the ‘control’ of Peter, quite a remarkable gift indeed!
How then are we to understand the meaning of this ritual ‘binding’ of Peter to the kingdom of heaven in more specific terms? For this, we argue that the motif of Peter as the key-bearer provides some important clues. Traditionally, the key has been interpreted particularly with regard to its function of opening and closing. However, we should note that the ‘key-bearer’ is often portrayed as a forceful figure in many myths, and even specifically as an active performer of binding in magical sources. Anubis and Aeacus, for example, were sometimes invoked specifically as ‘key-holders’:Footnote 43
… Both Acheron and Aiakos, gatekeeper of the eternal bars, now open quickly, O thou Key-holder, guardian, Anubis (Ἀχέρων τε καὶ Αἴακε, πυλωρὲ κλείθρων τῶν ἀϊδίων, θᾶττον ἄνοιξον, κλειδοῦχέ τε Ἄνουβι φύλαξ). Send up to me the phantoms of the dead forthwith for service in this very hour. (PGM iv.1464–9; Betz, 66) And take a lead tablet and write the same spell and recite it … The spell to be written and recited is: ‘I entrust this binding spell to you, chthonic gods … and to mighty ANUBIS PSIRINTH, who holds the keys to Hades (τῷ τὰς κλεῖδας ἔχοντι τῶν καθ’ Ἅιδου) …’ (PGM iv.329–42; Betz, 44) I know all that and that you [Aeacus] are the gate-keeper (Οἶδα ταῦτα καὶ σέ, ὅτι πυλωρεῖς). (Lucian, Dialogues of the Dead 6 (20) (LCL, Macleod)) Even after his death Aeacus is honoured in the abode of Pluto, and keeps the keys of Hades (τιμᾶται δὲ καὶ παρὰ Πλούτωνι τελευτήσας Αἰακός, καὶ τὰς κλεῖς τοῦ Ἅιδου φυλάττει). (Apollodorus 3.12.6 (LCL, Frazer))
In another magical papyrus, a deity who is to fulfil an act of binding is addressed as a key-bearer of the ‘kingdom’:Footnote 44
Do the NN deed for me; you are the image, the whole of the universe, [you] who, after being selected, guarded the holy place of the great king. Do the NN deed for me, the one who keeps the keys of the triangular paradise of earth, which is the kingdom (τῷ κλειδοφύλακι τοῦ τρ[ι]γώνου παραδείσου τῆς γῆς, τῆς βασιλείας). (PGM iii.536–41; Betz, 32)
The invocation of a key-bearing supernatural entity in binding magic relies on a belief that key-holding signifies a magnificent power. In the texts above, we note that possession of a key represents the key-bearer as a commanding figure who can exert unlimited influence over the victim, with the full force of the objective for which the key is made. Since the possession of the key is believed to be particularly associated with the power of a given deity, it is not surprising to see the following formula, which relates ‘possessing the key of a deity’ to control over that deity:
For I have hidden this magic symbol of yours, your sandal, and possess your key. I opened the bars of Keberos, the guard of Tartaros … (PGM iv.2290–5; Betz, 79)
With this awareness of this element of control, we can read Matt 16.18 anew. As key-bearer, Peter is a corresponding or mirroring figure to the deities invoked in magic who, possessing the key, are to carry out binding. Peter the key-holder of the kingdom is entitled to exert controlling power over the kingdom of heaven (a fact that is confirmed by our interpretation of the verb δίδωμι in its magical context) and embodies the full potency of the kingdom, since Christ has entrusted his kingdom to him and promised that the gates of Hades cannot overcome the church based on ‘the rock’.Footnote 45
5. Heaven, Earth and the Underworld
One last point is in order. Matt 16.19 clearly shares one of the most common underlying assumptions behind binding magic: when the curse-maker issues a binding spell, he or she presumes that the ritual binding issued ‘on earth’ becomes effective through the binding in the underworld or the upperworld, as the following curse shows:
(Side A) I will bind … Aristias and Euaristes and Kaliiades …
(Side B) and Menon the son of Aristokles, both him and the actions of Menon and his tongue and words and actions; and that he may prove useless to the authorities … The god who restrains holds the advocates with Nikos and Hedule the daughter of Timokrates. (DTA 95 = Gager, no. 39; Greece, Attica, 4th century bce)
Here, the defigens ‘binds’ the victims ‘on earth’, and the restraining god ‘holds’ those victims in the underworld. This spell presumes a principle of ‘sympathetic transference of the power and representation’ based on a presupposition that ‘everything in the universe is interconnected’ in certain ways.Footnote 46 A detailed explanation of this presupposition of magic is found in Plotinus (Enn. 4.4.40, 42 (LCL, Amstrong)): ‘But how do magic spells work? By sympathy and by the fact that there is a natural concord of things that are alike and opposition of things that are different … [B]y the arts of physicians and magicians one thing is compelled to give something of its power to another.’Footnote 47
The principle of power translation and interaction across the two realms is certainly evident in Matt 16.19: ὃ ἐὰν δήσῃς ἐπὶ τῆς γῆς, ἔσται δεδεμένον ἐν τοῖς οὐρανοῖς· καὶ ὃ ἐὰν λύσῃς ἐπὶ τῆς γῆς, ἔσται λελυμένον ἐν τοῖς οὐρανοῖς. Here the passive must be understood as the ‘divine passive’, with Christ's promise presupposing both the sympathetic relationship between the earth and the heavens and God's willingness to address the wish of Peter. Because Christ and his Father in heaven are believed to possess the utmost power, Christ's guarantee that Peter's act of binding and unbinding will be fulfilled in the realm of his heavenly Father is particularly reliable. Compared to binding magic, this dominical promise can be read as a magnified (in terms of scope and non-conditionality) and inverted version of a binding spell: contrary to the magical conventions which often try to compel the invoked deity to fulfil the wish of clients, the Jesus in Matt 16.19 voluntarily and graciously makes a promise to Peter regarding the certainty of Peter's binding and unbinding effects. This reading supports the suggestion that the literary form of this Matthean passage is a type of reappropriation.
6. Concluding Remarks
Interestingly, the result of our reading is not so different from that proposed by those who advocate the ‘rabbinic’ interpretations (including Bultmann), when it comes to the emphasis on the imagery of a ‘power struggle’. For instance:
The church will rather be the site of the battle between the powers of Hades and the power of heaven. In the age inaugurated by Jesus's death and resurrection, the gates of the underworld will swing open and the horrors of the pit will erupt onto the earth with a roar, attacking everything on it – including the church – with unbridled fury. In the midst of this peril, however, Peter will be given the keys that unlock the gates of heaven …Footnote 48
One should probably think of the end-time scenario, when the powers of the underworld will be unleashed from below, from the abyss, and rage against the saints (cf. 1 En. 56.8; Rev 6.8, 11.7, 17.8). The promise is that even the full fury of the underworld's demonic forces will not overcome the church.Footnote 49
[I]n the end, when the powers of the underworld overcome mankind, the Church will be saved.Footnote 50
However, I find that all these conclusions evince some important inconsistency in their authors’ exegeses because their ‘rabbinic’ reading cannot generate this line of the ‘eschatological spiritual battle’ image. Moreover, this kind of eschatological interpretation is too restrictive, considering that the threat of the underworld is ubiquitous and atemporal, as many ancient curse texts show (cf. Pliny's remarks: ‘there is indeed nobody who does not fear to be spell-bound by imprecations’; ‘magic rose to such a height that even today it has sway over a great part of mankind’). Peter's authority to bind and unbind is strongest and is supported by Christ's guarantee that there should be no limit in its objective and that the Father in heaven will carry out the ritual action of Peter. We conclude that Matt 16.18–19 is a literary ‘reworking’ adopted to depict, via the popular image of the binding spell, the powerfulness and the security of the church vis-à-vis its surrounding religiosity.
In addition, although this is far from conclusive, if Matt 16.18–19 is originally a post-Easter logion, it is possible that Jesus’ power to ‘bind and untie’, which is entrusted to Peter, must have been perceived as particularly powerful because Jesus had died a violent and untimely death, yet was resurrected, possibly indicating that Jesus’ unbinding power should be far superior to other binding magic, which often appeals to the souls of the untimely and violently dead.Footnote 51
Finally, Richard Hays's call for overcoming ‘exegetical amnesia’Footnote 52 by studying the reception history applies here. By revisiting Dell's proposal I have shown how important it is to ‘recover what has been lost / And found and lost again and again’.Footnote 53
