A further reply to Jonathan Burke on the devil in the Gospel temptation stories

1. Idiosyncrasies and Scholarly Consensus
2. History of Religions
3. Genre and Form
4. Matthew's Mountain
5. Alleged Rabbinic Parallels
6. Conclusion

In this article I respond to Jonathan Burke's article in an ongoing online dialogue over the interpretation of ὁ διάβολος in the temptation stories (TS) in Matthew 4, Luke 4 and Mark 1. My original exegetical works, which interacted with Christadelphian interpretations of the TS, can be found here and here; a summary of the findings of both of these studies is here. Burke critiqued my studies here (part of a larger series responding to various other articles of mine). I responded to Burke's critique with a series of four blog posts (1, 2, 3, 4), and Burke's latest article (actually over a year old by now) is what I am addressing now.1

In the interest of conciseness and time management, I will not be referring to academic literature to support every claim - particularly claims for which I have done so in previous installments in this series. Any reader who shares Burke's conviction that I have a 'complete lack of familiarity with the relevant scholarship' is encouraged to consult the bibliography and supplementary online materials pertaining to my two co-authored studies on New Testament Satanology recently published in the Journal for the Study of the New Testament.2 Instead of replying point-by-point I will make observations on some of the key issues that have been raised. I will also try to follow my self-imposed rules of engagement for online theological discussions, including avoidance of a discussion about the discussion. I would ask Burke to do the same if he opts to carry the discussion further.

1. Idiosyncrasies and Scholarly Consensus

First, I want to comment on the issue of 'idiosyncratic' interpretations. Burke brought this term into the discussion, labelling my interpretation of the TS as idiosyncratic. Well, my interpretation of one small aspect of one of the TS may be idiosyncratic (namely, how Jesus was able to see all the kingdoms of the world from atop a very high mountain in Matthew's version). However, the main point of contention (the meaning of ὁ διάβολος/ὁ σατανᾶς in the TS) does not rest on the validity of my interpretation of this detail. This is obvious since, despite not sharing my 'idiosyncratic' interpretation of the mountaintop experience, the vast majority of scholars share my view on the meaning of ὁ διάβολος.

Thus, when it comes to weighing the scholarly support for Burke's position and mine, the most important point is that Burke's interpretation of the term ὁ διάβολος itself (namely, that it refers to a personification of the yetzer hara, the evil inclination in human nature, and not to a personal being) has very little support in either the academy or the church. Burke admits this: he says that 'the declared aim of my article is to challenge the existing consensus that the temptation account is mythological'. He also admits, 'my view that Jesus was tempted by his own desires is marginal within scholarship', but avers that it is 'certainly not idiosyncratic', citing two scholars in support of this interpretation. One is Kesich, who in a work on early church history devotes all of three sentences to the TS.3 While Kesich says that the TS have been 'translated into figurative language', he does not state that the devil in the TS is a figurative way of referring to Jesus' own desires.4 The other is Lachs, who agrees more closely with Burke's interpretation but still not exactly.5 It is worth noting the language with which Lachs introduces this interpretation, which indicates a speculative suggestion rather than a firm claim: 'the confrontation with Satan could be seen as Jesus’ struggle with himself and overcoming the yezer hara'. Moreover, the only scholarly review of Lachs' Rabbinic Commentary on the New Testament of which I am aware pans his work for its 'sloppy editing', 'errors of fact and interpretation', 'outdated scholarship' and 'uncritical methodology'.6

Leaving aside these three sources of questionable academic weight,7 one finds virtually no support for the ὁ διάβολος = yetzer hara interpretation of the TS in relevant monographs and exegetical journal articles on the TS or technical commentaries on the Synoptic Gospels. Thus, 'marginal' is an apt adjective to use for the status of Burke's interpretation (this seems to be one of the few areas of agreement between Burke and myself on the TS). Since Burke is challenging a near-unanimous scholarly consensus, my recommendation would be that he submit his exegesis of the TS to a peer-reviewed biblical studies journal so that it can be tried in the court of scholarly opinion, rather than limiting himself to blog posts. Furthermore, Burke needs to provide detailed commentary on the Matthaean and Lucan TS to explain what every clause would mean under his interpretation of ὁ διάβολος and show why this interpretation has, overall, more explanatory power than the standard view down to the level of syntax. To my knowledge, despite writing extensively on the TS he has yet to do this.

2. History of Religions

One of Burke's three arguments against the scholarly consensus that ὁ διάβολος in the TS refers to a cosmic foe is a history of religions argument:

the most common terms used in pre-Christian Second Temple literature for a supernatural evil being, are not used in the Synoptics. In fact most of them are not used in the New Testament at all (Beliar is used once). In contrast, the terms used in the Synoptic temptation accounts have almost no pre-Christian witness in Second Temple literature as a reference to a specific supernatural evil being.

More specifically, Burke claims:

In Second Temple Period literature the term ‘satan’ (whether in Hebrew or Greek), is predominantly used as a common noun rather than a personal name, the term ‘the devil’ (ὁ diaboloV), is rarely if ever used to refer to a supernatural evil being, and the terms ‘the tempter’ (ὁ peirazwn), and ‘the evil one’ (ὁ ponēroV), have no pre-Christian witness with such a meaning. The term satan, whether in Greek (satanaV), or Hebrew (śāṭān), is used rarely in pre-Christian literature and never as a proper name.

I have so far not responded in detail to these religion-historical claims, partly because I intend to make a study of the origins of early Christian Satanology against the background of Second Temple Judaism the topic of my Honours dissertation over the next year. So I hope in due course to address these issues in detail. For now, I will limit myself to five observations.

1) The term ὁ δίαβολος is used in Job 1-2 LXX and Zech. 3:1-2 LXX. In both cases, ὁ δίαβολος refers to a cosmic being, and in the first case, ὁ δίαβολος has a testing function: he solicits God to bring calamity into Job's life because he believes this will induce Job to sin ('stretch out your hand and touch all that he has, and he will curse you to your face' - Job 1:11; cf. 2:5).8 If I have understood Burke correctly, he regards these texts as irrelevant to the interpretation of the TS because ὁ δίαβολος may be an obedient servant of God rather than an enemy of God (as is widely believed for הַשָּׂטָן in Job and Zechariah MT). However, this feature does not derail the significance of the parallel, because the notion of Satan as God's servant (and prosecutor) continues even in the New Testament. After surveying the evidence, Page concludes:

The Bible portrays Satan as an implacable enemy of God, whose designs on humanity are malicious; however, it does not represent Satan as God’s equal or as one who acts independently of divine control. In the prologue of Job, the oldest text that speaks of a celestial Satan figure, he is clearly pictured as one who is subordinate to God and who operates only within the parameters that God sets for him. Although there is incontrovertible evidence of change and development in the concept of Satan in the biblical literature, this basic notion that Satan is under divine control appears repeatedly. This motif may stand in a certain degree of tension with the conception of Satan as a hostile force, but it is a persistent theme in the biblical record. Satan is an enemy of God, but he is also a servant of God.9

Moreover, we have pre-Christian evidence for the devil as an unambiguously evil being. There is a Jewish pseudepigraphon referred to by some as the Assumption of Moses, by others as the Testament of Moses, and by still others as the Moses fragment because of its uncertain identification with either of these two lost works. It is extant only in a Latin translation from Greek, which was probably the original language.10 It 'dates from the early years of the first century C.E.';11 Grierson estimates the terminus ad quem (latest possible date) at 30 C.E.12 The text contains the following apocalyptic prediction:

And then his (i.e. God's) kingdom will appear in his entire creation. And then the devil will come to an end, and sadness will be carried away together with him. (Testament of Moses 10.1)13

The Latin word translated 'devil' here is zabulus, which is an 'orthographic variant' of diabolus,14 the Latin transliteration of διάβολος. Because Latin lacks the definite article, we cannot be certain whether the Greek read ὁ διάβολος or just διάβολος. However, it seems clear that an individual being is in view, and that he is evil since he will come to an end.15 It is also worth noting that the ending of this document, which is lost but presumably contained an account of Moses' death, is widely regarded by scholars as the source of the allusion in Jude 9 to a quarrel between Michael and the devil over Moses' body.16

2) For similar reasons, הַשָּׂטָן in Job 1-2 and Zech. 3:1-2 is relevant to the interpretation of ὁ σατανᾶς in Mark 1:13 and Matt. 4:10. There are numerous other occurrences of the term 'satan' in Second Temple literature that support a cosmic referent in the TS. I will leave discussion of these for a later date once I've completed my dissertation, but for now I would draw attention to one obvious instance: Paul's assertion in 2 Cor. 11:14 that 'even Satan disguises himself as an angel of light'. Even apart from Paul's identity as a Christian, this is a relevant reference to ὁ σατανᾶς in a Second Temple Jewish text which predates the Gospels by only a few years (Mark) or decades (Matthew), and clearly places Satan in a cosmic, rather than anthropological, context. Earlier in the same epistle, Paul uses the term 'Beliar' (2 Cor. 6:15), which Burke recognizes (as Belial) among the 'established Second Temple terms for supernatural Satan figures' (my emphasis). Moreover, this text ticks all the boxes for the definition of cosmological dualism that Burke uses: 'the world is divided into two opposing forces of good and evil, darkness and light'. 2 Cor. 6:15 explicitly draws an antithesis between 'righteousness and lawlessness', 'darkness and light', and 'Christ' (supernatural, personal leader of the forces of light) and 'Beliar'.17 Burke does not mention any New Testament texts (even those that pre-date the Gospels) in his survey of the background to the Satanological terminology in the Synoptic Gospels.

Is 'satan' a proper name in Second Temple Jewish literature outside the New Testament? Burke says no:

The term satan, whether in Greek (satanaV), or Hebrew (śāṭān), is used rarely in pre-Christian literature and never as a proper name. Consequently, Laato notes that ‘we lack an established tradition whereby the name of the personal Evil or the leader of demons is Satan’.

The link-word 'Consequently' leaves the reader with the impression that Laato agrees with the judgment that 'satan' is never used in pre-Christian literature as a proper name. In fact, he suggests in the same article that שָׂטָן may have transitioned to a proper name ‘already in the Old Testament’, that ‘there is no obstacle to regarding Satan in 1 Chronicles 21 as a proper name’, and that the word could, linguistically speaking, function as a proper name in Job 1-2 and Zechariah 3 despite the presence of the definite article.18

In fact, however, the question of whether 'satan' functions as a proper name in the Old Testament, Second Temple Jewish literature or the New Testament is not of decisive importance. Stokes, for instance, does not see evidence of early use of 'Satan' as a proper name. Nevertheless, he sees clear evidence for the emergence of an individualized Satan figure:

Two passages in the Hebrew Scriptures speak of a superhuman figure not simply as a satan, but as the Satan (Zech 3; Job 1-2). The definite article on the noun (הַשָּׂטָן) suggests the notion of a particular figure who holds the office of "Attacker/Executioner" in the divine court... The Hebrew Scriptures contain several different kinds of satans. They speak of humans in the capacity of attackers or executioners as satans. They speak of heavenly satans, serving the Deity as executioners of the wicked. They also speak of a particular satan, the Satan, who serves God as attacker or executioner of the wicked and, in the case of Job, as attacker of a righteous person. While it was this individual, the Satan, who especially piqued the imaginations of earlier interpreters and who would become the best known of the satans, early Jewish literature would continue to speak of other satans as well.19

After surveying Second Temple literature, he concludes:

In the Hebrew Scriptures and early Jewish writings, "satan" refers generically to an attacker or executioner. And many kinds of attackers/executioners could be called a satan. Some are human; others are superhuman. Some of them are something akin to evil spirits or demons. One of them is the Satan.20

Against this background, it does not really matter whether we consider ὁ σατανᾶς in the New Testament to be a 'personal name' or a 'definite title'.21 What matters is that it is not a common noun but the designation of a particular superhuman individual, which could be rendered in English either with 'Satan' (if understood as a personal name) or 'the Satan' (if understood as a definite title).

3) In his history-of-religions survey of Satanological terminology in the Synoptic Gospels, Burke restricts his attention to 'pre-Christian literature'. This is a serious methodological shortcoming for two reasons. (1) Christian literature that pre-dates the Synoptic Gospels (especially the Pauline epistles) is arguably the most relevant background data of all (see above). (2) Ancient texts which are likely to post-date the Synoptic Gospels are not irrelevant, especially if the temporal proximity is close. Although they obviously do not carry the same weight as sources that pre-date the Synoptic Gospels, they still help to fill in the religion-historical milieu in which the Synoptic evangelists wrote. Burke can hardly dispute this since he introduces much later rabbinic texts as the closest literary parallels to his interpretation of the devil in the TS.

4) In discussing the background to the term 'the evil one' Burke has inadequately represented his key source. Burke cites Black's discussion of Matt. 6:13b to support his claim that 'the evil one' (ὁ πονηρὸς) 'has no Second Temple pre-Christian witness as a reference to a supernatural evil being'. He summarizes Black's point thus:

Summarizing the lexicographical evidence, Black notes ‘this term or designation for Satan is, outside the New Testament and dependent patristic writings, nowhere attested in classical, Hellenistic, or Jewish Greek sources’, which he gives as the reason against reading it as ‘the evil one’ even in Matthew.

However, while Black was summarizing previous scholars' conclusions about the background to the term ὁ πονηρὸς, Black himself proceeds to dispute this summary, discussing two Jewish texts, one of which is a pre-Christian Second Temple text that, in his view, uses the term 'the evil one' (in Hebrew) for Satan. These texts are 4Q Amram^b, 4Q280, 286 (287) and Targum of Isa. 11.4. Black's own conclusion is as follows:

these two sets of texts...are the only passages in Jewish literature where the designations רשיעא/הרשע are used for Satan or a manifestation of Satan. The designation, however seems almost an inevitable one for the Prince of Darkness, so that it may well have been in more frequent use in Judaism than its extremely rare occurrence suggests? Was it perhaps dropped by the Synagogue when it was adopted by the early Church, in its almost literal Greek equivalent ὁ πονηρὸς? Such a term would no doubt commend itself widely as a general concept, immediately intelligible in the Hellenistic world, whereas the Hebrew/Aramaic terminology for Satan must have sounded strange and foreign in Greek ears.22

Hence, Black not only presents evidence (including a pre-Christian Second Temple text) for the use of 'the evil one' for a supernatural being, but also does not think that the paucity of evidence is particularly problematic. We are dealing with a low standard of research when one cites a source to support an argument from silence without reporting that the same source counters both the silence and the argument! Unfortunately, Burke repeats the same omission in his recent peer-reviewed publication when arguing that ὁ πονηρὸς does not refer to a supernatural evil being in Didache 8.2.23 Perhaps Burke would claim in his defense that he was interested only in the Greek term ὁ πονηρὸς and not its Hebrew equivalent; but this would be inexplicable since in the very same paragraph he shows concern for usage of 'satan' 'whether in Hebrew or Greek'.

Moreover, Burke's background survey inexplicably excludes the fairly widespread use of ὁ πονηρὸς for Satan in other early Christian texts, some of which may predate Matthew (Eph. 6:16; 2 Thess. 3:3; John 17:15; 1 John 2:13-14; 3:12; 5:18-19; Barnabas 2.10; 21.3; cf. 4.13). While Burke would no doubt dispute that 'the evil one' refers to a supernatural being in any of the New Testament texts, he concedes elsewhere that this is the case in the Epistle of Barnabas. Hence, despite the extremely rare use of 'the evil one' for a supernatural being in Second Temple literature, it is historically certain that this term was in Christian use as a designation for a supernatural Satan within a few decades of the destruction of the temple. The question is when, not whether, the designation entered the Christian vocabulary; and the burden of proof lies with Burke to show that Pseudo-Barnabas understands the term differently than the various NT writers.

5) I previously (following Kelly) identified four parallels to the TS which consisted of a righteous man being tempted by a supernatural being: Abraham being tempted by an unclean bird identified as Azazel in the Apocalypse of Abraham, Mastema's request to tempt Abraham in Jubilees, stories about the Angel of Death and Moses, and Beliar's three nets in the Damascus Document. Burke did not regard this evidence as a problem for his interpretation because:

Firstly, the tempter or challenger in each case is not a supernatural opponent of God, but an obedient servant, whether Azazel, Mastema, the Angel of Death or Beliar. Secondly, in none of these cases is a righteous man tempted by a supernatural being called Satan.

In all these texts, as in the standard interpretation of the TS, we have a supernatural tempter tempting a righteous man. This is a striking parallel, and it is not clear why Burke dismisses it on the grounds that these tempters have a better relationship with God than the devil in the TS (which is itself debatable.24) A possibly different moral disposition of the supernatural tempter does not imply that 'There are no Old Testament or Second Temple parallels to the temptation accounts'! Again, it is not clear how the parallel is invalidated simply because the supernatural tempter has a different name in each of these stories. Burke himself refers to Mastema, Belial and Azazel as 'supernatural Satan figures'.

3. Genre and Form

Burke accuses me of an internally inconsistent 'hermeneutic of convenience' in that I object to classifying the TS as haggadic midrash and yet appeal to midrashic parallels. Rather, my argument runs as follows:

(1) It is debatable whether the TS can be classified form-critically as haggadic midrash.
(2) Even if we classify the TS form-critically as haggadic midrash, it is debatable how helpful this category is for understanding the TS in their canonical form.
(3) Even if we classify the TS form-critically as haggadic midrash, this supports rather than undermines mythological interpretation of the tempter, since the closest midrashic parallels involve supernatural tempters.

Hence, my argument was that regardless of the position we take on the 'haggadic midrash' issue, we do not have grounds for rejecting a mythological interpretation of ὁ διάβολος. (Hence why a 'haggadic midrash' view of the TS has considerable support but the ὁ διάβολος = yetzer hara view does not.)

I reiterate that Burke does not adequately distinguish between the form of the TS tradition and the final literary products of Matthew and Luke. Even if the temptation dialogue in Q originated as an invented midrash based loosely on trials Jesus faced (a view I reject), it is obviously neither Matthew nor Luke who composed it. Hence, Burke needs to show not merely that the TS was understood as midrash rather than historical narrative at some pre-canonical stage of development; he needs to show that Matthew and Luke so understood it.25 This requires close attention to how Matthew and Luke incorporate the tradition into their wider narrative. This is where narrative criticism comes in - a methodology that Burke has heretofore ignored, despite his protest to the contrary.26 This is also why the syntactic features that I stressed in my original study remain decisive: they show that the events and dialogue are narrated in a way typical of the Gospels' stories. Although the differences between Matthew's and Luke's TS show that at least one of them has redacted rather freely, neither evangelist gives any indication of a shift in genre. It should also be noted that even if the pre-canonical TS underwent dramatic theological expansion, we can be quite certain that the term ὁ διάβολος is not dramatic theological expansion but was part of the story before this expansion occurred, since the synonymous term ὁ σατανᾶς is present in Mark's version of the TS, which is independent of Q and lacks its (alleged) 'dramatic expansion'.

I reiterate that Burke has not adequately interacted with the notion of myth as a literary genre or category in Gospel interpretation. His discussion of myth is limited to his claim that the Synoptic Gospels do not reflect cosmological dualism and that the Synoptic Satan/devil is not a cosmological figure. He describes the following as a 'line of evidence' against reading the TS mythologically:

Ethical dualism and psychological dualism are dominant in the Synoptics, rather than the cosmological dualism which would be expected if a supernatural evil being was present in the temptation accounts.

This, however, is simply an assertion of his conclusion; it is not evidence. And if we turn to Burke's article on dualism in the Synoptics where we might expect to find this point argued in detail, we simply find a definition of three kinds of dualism, and no evidence or argumentation that ethical and psychological dualism is dominant in the Synoptics while cosmological dualism is absent. One would certainly expect to find an extensive discussion of texts such as the Beelzebul Controversy, the parable of the strong man, the parable of the wheat and the weeds, Matt. 25:31, 41 and Luke 10:18-19 as part of any claim that cosmological dualism is lacking in the Synoptic Gospels.

4. Matthew's Mountain

The issue of the 'very high mountain' in Matt. 4:8 has long been the cornerstone of Christadelphian interpretation of the TS.27 There is no mountain on earth from which one can see all the kingdoms of the world, and this point is used to rule out a literal reading of the entire TS, justifying a figurative interpretation even of ὁ διάβολος. However, Matthew's mountain is not nearly the exegetical trump card that Christadelphians make it out to be.

Moreover, it is not true, as Burke claims, that I acknowledge 'that the description of the satan [sic] taking Jesus to a mountain high enough to see all the kingdoms of the world is impossible to read literally, and (resort) to describing this experience as a supernatural experience'. Rather, I argue that the description is possible to take literally precisely because it is a supernatural experience. My own suggestion is that Jesus was transported to the top of a literal mountain where he had a mystical experience. Burke describes this interpretation, with some justification, as idiosyncratic. However, for the purposes of this discussion, I do not need to prove conclusively that it is correct. Indeed, I do not even need it to be correct. To neutralize the 'mountain problem rules out historical narrative' argument, I only need to show that there are plausible ways to understand the mountain within an 'historical narrative' reading of the TS. And several such ways can be found in the scholarly literature. Accordingly I would make the following observations.

1) Classifying a pericope as historical narrative does not commit the exegete to a woodenly literal interpretation of every last detail. For instance, I do not, by virtue of my approach to the TS, feel compelled to infer from Matt. 4:4 that Matthew thought God to have a literal mouth (though he may have done). That said, I see no reason to doubt that Matthew had a literal mountain in mind (whether a known mountain in Israel, a cosmic mountain or a heavenly mountain).

2) Since Matthew and Luke differ on the mountain detail, it is uncertain whether the mountain was in Q and dropped by Luke, or whether the mountain was introduced by Matthew. Seemingly, the majority of scholars favour the first option,28 but there is also considerable support for the second.29 Let us consider the implications of both possibilities.

a. What if the mountain was present in Q, and omitted by Luke? This suggests that the mountain was part of the story very early. The words 'very' and/or 'high' may be redactional in Matthew even if 'mountain' is not.30 If the TS is regarded as authentic history, then one possibility is that the devil took Jesus to the top of an actual mountain in Israel. This finds support in the identification of Mount Tabor as the mountain of temptation in the Gospel of the Hebrews.31 (The identification of a specific, literal mountain suggests that the author of the Gospel of the Hebrews understood the account literally.) How could Jesus have seen 'all the kingdoms of the world' from the top of Mount Tabor or another mountain in Israel? He couldn't have, naturally - just as Moses couldn't have naturally seen all the places that Deut. 34:1-4 says he saw from the top of Mount Nebo. So, as noted previously, the TS, like Deuteronomy, implies some kind of supernatural, mystical experience atop the mountain. This is not a particularly radical suggestion given that Matthew records a supernatural, mystical experience in his only other reference to a high mountain (Matt. 17:1ff).

The second issue, if the mountain was present in Q, is why Luke omitted it. Luke simply has the devil 'take Jesus up'. This is significant because Luke's TS must be interpreted in its own right and not simply harmonized with Matthew's: it is possible that Luke and Matthew understood the TS differently. Thus, whatever the implications of the 'mountain problem', the exegete must still deal with Luke, where this problem is absent. Scholars have made several suggestions as to why Luke may have chosen to omit the mountain. One popular suggestion is that Luke is concerned to emphasise the temporal aspect of the panorama ('at a moment of time') than the spatial aspect ('to a very high  mountain').32 Others think Luke was embarrassed by the idea of Jesus seeing all the kingdoms of the world from a very high mountain because he knew no such mountain existed.33 Hence, his redaction serves to depict this temptation as 'visionary or imaginary',34 or alternatively as a heavenly journey,35 whether of the body36 or (less plausibly in my view) the soul.37 Whatever the case, Luke is explicitly concerned with ensuring his account is believable as historical narrative.38 If the whole episode were a dramatic theological expansion, we would not expect such historiographical concern.

b. What if the mountain was added by Matthew? (As noted above, a separate possibility is that the mountain was present in Q but 'very high' was added by Matthew.) Given the important theological role that mountains play in Matthew's narrative (cf. Matt. 5:1; 17:1; 28:16), this would presumably have been done to add a theological flourish to the story. If so, then if the TS basically recounts authentic historical events, the 'very high mountain' is not part of that authentic history and we can call back the Sherpas from the search for it. The actual historical events might then have included a heavenly journey (as Luke seems to imply), 'whether in the body or out of the body'. There remains the question of how Matthew understood the 'very high mountain' that he added to the story. This has been discussed in a previous article: possibly Matthew and his earliest readers understood it literally à la Donaldson's 'cosmic mountain' motif. They might well have believed that such a mountain existed. That it does not in fact exist means we must read the narrative critically, not figuratively. Alternatively, Matthew may have envisioned the mountain as a heavenly mountain as per Orlov's suggestion.

3) I reiterate that understanding the TS as a mystical experience is consistent with it being a historical narrative. Even if the TS were to be interpreted as an imaginary or dream-like experience, this would not imply that ὁ διάβολος lacked external existence any more than appearances of angels in dreams (Matt. 1:20; 2:13; 2:19) imply that those angels lacked external existence. However, while I am willing to allow for a mystical aspect to the temptations (i.e. a heavenly journey in Luke and a mountaintop visionary experience in Matthew), I think the TS clearly describe a physical bodily experience. 2 Cor. 12:2-4 shows that Second Temple Jews could conceive of mystical experiences as being either physical or non-physical. More importantly, all three temptations entail the involvement of Jesus' body (eating, leaping, prostrating). A temptation to throw oneself off a high building is meaningless if imaginary: it could not really be acted upon (and Jesus would be able to mount a robust defense against any claim that he was morally responsible for an act he performed within a dream). When actually standing at the top of a high building peering over the edge, it is a different matter.

4) I reiterate that, in a first century Jewish or Roman context, 'all the kingdoms of the world' does not mean 'every inch of the earth known since the Copernican revolution to be spherical'. It means all the kingdoms of 'a flat earth limited to the Mediterranean world'.39 We should not infer that, if the TS are historical narrative, Jesus must have gazed upon New Guinea, Cape Horn and the Kamchatka Peninsula. In insisting that 'the Roman Empire and its environs' is 'a poor temptation for a man who was promised rulership of the entire world by God', Burke is imposing a theological reading that is far removed from the first century context. In any case, I readily concede that it is not possible with natural human vision to see the whole Mediterranean world from any mountain in the Mediterranean world. However, Jesus' companion is explicitly capable of transporting him great distances to great elevations, and Luke's ἐν στιγμῇ χρόνου makes the supernatural character of the 'showing' unmistakable. Within such a narrative context, it makes little sense to limit the range of Jesus' eyesight on the mountaintop to what would be naturally possible.

As is evident from the above, there are several plausible ways of explaining Matthew's 'very high mountain' which do not conclude either (a) that he intended the entire TS to be read figuratively, including ὁ διάβολος; or (b) that there must be a stupendous physical mountain somewhere on earth from which the entire globe is visible.

5. Alleged Rabbinic Parallels

Burke introduces new evidence in the form of four rabbinic texts. The first two texts, in Burke's words, 'show the yetzer personified to the point that it is depicted as an independent being, even while being identified as an internal impulse to sin'.

Now in terms of the admissibility of this evidence, it is noteworthy Burke here uses rabbinic texts written centuries after the New Testament, and deems them relevant to the exegesis of the TS, whereas in his history-of-religions survey of Synoptic Satanological terminology, he is only prepared to place significant weight on pre-Christian Second Temple Jewish texts. For instance, concerning the Testament of Job, he judges that 'its very uncertain date precludes its use as a reliable source of contextual data for the New Testament.' However, the latest date in the range usually assigned to the Testament of Job is still earlier than the rabbinic literature cited by Burke, to which he does appeal as a reliable source of contextual data for the New Testament. This is a clear methodological inconsistency. Moreover, Burke certainly overstates things when he concludes based on these texts that a particular literary device 'was well established in Judaism at least as early as the Tannaitic period.' For one of the texts he cites (Sifre Numbers), the final redaction may be dated in the 'mid to late-third century', 'towards the end of the tannaitic period'.40 For the other texts, from the Babylonian Talmud, the final redaction may be dated to 'between the fifth and the seventh centuries'.41 One of them is paralleled in the Palestinian Talmud, the final redaction of which is dated by most scholars to the first half of the fifth century.42 Hence, Burke provides only one text that can be dated toward the end of the Tannaitic period - still about two centuries after the Gospels were written. If Burke wants to claim that the traditions within these rabbinic texts go back much earlier than the date of redaction, he will have to support this with historical-critical analysis, which he has not done. Neusner's famous dictum, 'What we cannot show, we do not know' is relevant at this point.43

Nevertheless, let us proceed with caution to consider the significance of the parallels. My overall assessment is that they are helpful in illustrating the contrast between 'a highly anthropomorphized yetzer ha ra' and the tempter in the Gospel TS. The key distinction is expressed by Burke himself: 'in these texts the yetzer is 'personified to the point that it is depicted as an independent being, even while being identified as an internal impulse to sin'. In both texts cited by Burke, the yetzer hara is explicitly identified as such. Indeed, while Burke states that Rosen-Zvi notes 'that this temptation is not depicted as a mere impulse, but as an independent person arguing cogently with Boaz', the sentences immediately prior to his supporting quotation show that for Rosen-Zvi, the identification of the yetzer hara signals the internal nature of the temptation:

The appearance of the yetzer moves the venue from the interpersonal sphere to the inner arena of the protagonist's desires and proclivities: the sexual drama here is one of thoughts and reflections no less than of actions. The dangers lurking for man are internal, not external.44

By contrast, in the TS the tempter is designated only with personal nouns and is not identified as the yetzer hara. We do not have 'the appearance of the yetzer' to move 'the venue from the interpersonal sphere to the inner arena of the protagonist's desires and proclivities'; the TS remain in the interpersonal sphere.

In the second text, the internal nature of the temptation receives further explicit emphasis: 'my evil impulse grew proud within me'. By contrast, in the TS the tempter is never described using internalizing language. Finally, nothing in these rabbinic 'parallels' approaches the self-awareness of the tempter in the TS. In the Boaz midrash, the yetzer only speaks to Boaz about Boaz and not about itself. In the Nazirite tale, the yetzer's words are not recorded but it merely beseeches the Nazirite. By contrast, in the Lucan TS the tempter makes elaborate claims about his prerogatives using a divine passive: 'it has been delivered to me, and I give it to whom I will' (Luke 4:6). In both Matthew and Luke the tempter demands a physical act of worship from Jesus, which is simply impossible if the tempter is internal to him. This would make the very temptation itself - which is no incidental detail but a key part of the story - nonsensical. Burke has never yet offered a satisfactory explanation for these features of the TS.

The other two rabbinic texts cited by Burke are cases where he says the Rabbis 'identified the yetzer in Old Testament passages which were speaking plainly of human beings', showing that they 'had anthropomorphized the yetzer to the extent that it was now natural to depict it as an individual separate from humans'. Again, these rabbinic texts involve explicit assertions made about the yetzer, whereas the Gospel TS do not mention the yetzer. Moreover, in the case of Micah 7:5, Rosen-Zvi notes a specific feature of the Hebrew text that made the yetzer interpretation possible: 'The homilist (in a typically midrashic move) reads רֵע (friend) as רַע (evil)'.45 Without the vowel points (which were added to the text by the Masoretes, probably after this midrash was developed),46 the words 'friend' and 'evil' are identical in Hebrew. The second text into which the rabbis read the yetzer is 2 Sam. 12:4 (incorrectly cited by Rosen-Zvi and Burke as 2 Sam. 2:12). Here, the yetzer is identified with the traveler who came to the rich man in an allegorical parable told by Nathan to David. It is surely the allegorical nature of the passage which prompted this identification. Nathan tells David that he is the rich man in the story. The poor man is obviously Uriah the Hittite and the ewe lamb Bathsheba, but what is the referent of the traveler? One can hardly imagine the rabbinic imagination concluding that it lacks an allegorical referent. Thus, as in Micah 7:5, the biblical text itself supplies the basis for the midrashic elaboration. That the rabbis chose to interpret this human character as the yetzer remains impressive, but as Rosen-Zvi states concerning the whole passage:

Taken together, these three homilies present a yetzer more developed than anything we find in the earlier Tannaitic literature: a sophisticated, dynamic, and demonic enemy.47

Thus, in Rosen-Zvi's judgment, far from demonstrating that such a highly anthropomorphized yetzer 'was well established...at least as early as the Tannaitic period', this text represents a development beyond anything found in the earlier Tannaitic literature. Thus, any attempt to use this text as contextual data for interpreting the Gospel TS is historically suspect.

6. Conclusion

I will conclude by listing features within the TS that support a literal interpretation of ὁ διάβολος / ὁ σατανᾶς as a supernatural personal being.

• Only personal nouns are used for the tempter, and there is no explicit identification as yetzer or explicit internalization of the tempter (as seen in the alleged rabbinic parallels)
• The tempter 'comes' at the beginning of the story, and 'leaves' at the end, just as angels 'come' at the end (in Matthew)
• The tempter demands a physical act of worship from Jesus: 'fall down and worship me' (Matt. 4:9); 'worship before me' (Luke 4:7), which makes sense only if the tempter is an external person
• The tempter demonstrates strong self-awareness independent of Jesus (in Luke): 'To you I will give all this authority and their glory, for it has been delivered to me, and I give it to whom I will.' (Luke 4:6)
• The statement, 'I give it to whom I will' simply does not make sense if it is Jesus' yetzer speaking. To whom, other than Jesus, might Jesus' yetzer possibly envision giving the kingdoms of the world?
• Understood in the context of Roman law, showing Jesus the kingdoms of the world represents a proposed property transaction, which implies the presence of a distinct seller and buyer. This feature of the narrative is inexplicable if the seller and buyer are the same person, since then no transaction would be necessary
• Whatever the tradition-history of the Q TS, both Matthew and Luke weave it seamlessly into their narrative, offering no indication that it should be read less literally than other events in the life of Jesus
• As narrative critics argue, Satan/the devil features as a distinct character in the plot of each of the Synoptic Gospels
• The identification of the tempter as Satan/the devil is common to both independently transmitted versions of the TS, those of Mark and Q. Consequently, the externality of the tempter cannot be attributed to a progressive theological dramatization of the Q TS known from Matthew and Luke
• If the Q TS is identified as haggadic midrash, we should make recourse to the closest religion-historical parallels from this genre, which all feature the testing of a righteous man by a supernatural being (in fact, the contours of this motif are visible already in Job). The testing of a righteous man by his yetzer, depicted as an external person capable of speech, is first attested in the mid to late third century C.E. (according to evidence presented by Burke)
• There are plausible explanations of Matthew's 'very high mountain' within a broadly literal reading of his TS, so this detail is no justification for proposing a figurative reading of the entire pericope. Moreover, the 'mountain' detail is found only in Matthew and so is irrelevant to the interpretation of the Marcan TS. It affects interpretation of the Lucan TS only in terms of the need to explain Luke's omission of the mountain (if it was present in the Q TS).
• New research concerning the devil's use of Psalm 91 (an apotropaic48 psalm) in the TS has helped to 'situate Satan's invocation of the psalm within the larger context of early Jewish demonological tradition'.49

For these reasons, I believe the scholarly and traditional consensus that ὁ διάβολος / ὁ σατανᾶς in the Gospel TS refers to a supernatural being remains well-founded and secure despite the arguments raised in Burke's online writings. And I reiterate my observation that Burke has yet to offer a blow-by-blow commentary on the TS.

Footnotes

• 1 I already addressed it briefly in some comments at the bottom of my last blog post; this is a more detailed response.
• 2 Farrar, Thomas J. & Williams, Guy J. (2016). Diabolical Data: A Critical Inventory of New Testament Satanology. Journal for the Study of the New Testament, 39(1), 40-71; Farrar, Thomas J. & Williams, Guy J. (2016). Talk of the Devil: Unpacking the Language of New Testament Satanology. Journal for the Study of the New Testament, 39(1), 72-96.
• 3 Kesich, Veselin (2007). Formation and Struggles: The Church, AD 33-450 (Vol. 1). Crestwood: St. Vladimir's Seminary Press, p. 12.
• 4 The comment that Burke cites specifically concerns the 'historicity of the temptations'. Kesich suggests that 'Jesus himself was most likely the source of his trials in the wilderness, translated into figurative language'. What Kesich means by Jesus being the source of his trials is not that they were strictly internal but that Jesus was the historical source for the TS - since he was the only human being present, only he could have transmitted the story to his disciples (if in fact it is rooted in historical events). Burke infers from 'translated into figurative language' that Kesich interprets ὁ διάβολος as referring to Jesus' own desires, but Kesich himself does not say so.
• 5 When Lachs says that the yetzer hara 'is externalized in the literature by the figure of Satan', it is unclear whether he thinks that Satan is an external figure in the literature or is just an apparently external figure in the literature. In rabbinic literature, Satan is an external figure.
• 6 'To sum up, Lachs's Rabbinic Commentary is marred by sloppy editing, errors of fact and interpretation, and outdated scholarship. It is fatally flawed, however, by its uncritical methodology coupled with a lack of serious reflection on the implications of such a commentary for understanding New Testament literary history.' (Visotzky, Burton L. (1988). Review: Lachs' "Rabbinic Commentary on the New Testament". The Jewish Quarterly Review, 78(3/4), 340-343, here p. 343.)
• 7 The third, as both Burke and I have noted previously, being Phipps, William E. (1993). The Wisdom and Wit of Rabbi Jesus. Louisville: Westminster John Knox Press, p. 38.
• 8 The verb in both the MT and LXX literally means 'bless', but this should be understood as a euphemism, similar to the famous tannaitic birkat ha-minim ('blessing on heretics'). See Mangan, Celine (2002). Blessing and Cursing in the Prologue of Targum Job. In Paul V.M. Flesher (Ed.), Targum and Scripture: Studies in Aramaic Translations and Interpretation in Memory of Ernest G. Clarke (pp. 225-230). Leiden: Brill, p. 226.
• 9 Page, Sydney H.T. (2007). Satan: God's Servant. Journal of the Evangelical Theological Society, 50(3), 449-465; here p. 465.
• 10 'It is undisputed that the Latin of the Moses fragment was translated from Greek...as there is no strong evidence for the Moses fragment having existed in a Semitic language, there is no need to look beyond a Greek original for a Hebrew text' (Grierson, Fiona (2008). The Testament of Moses. Journal for the Study of the Pseudepigrapha, 17(4), 265-280; here p. 275.)
• 11 Collins, John J. (2016). The Apocalyptic Imagination: An Introduction to Jewish Apocalyptic Literature. Grand Rapids: Eerdmans, p. 160.
• 12 Grierson, op. cit., p. 276.
• 13 Translation: Tromp, Johannes (1993). The Assumption of Moses: A Critical Edition with Commentary. Leiden: Brill, p. 19.
• 14 Tromp, op. cit., p. 229. Tromp suggests the text refers to the devil being barred from the heavenly council or, more likely, is a less concrete reference to the demise of satanic forces similar to what we find in other apocalyptic literature of the period.
• 15 Tromp, op. cit., p. 229, notes the similarity to Jesus' statement in Mark 3:26 that Satan 'has an end'.
• 16 See discussion in my article on this subject; also Tromp, op. cit., pp. 271f. Grierson takes a more conservative approach. She thinks either the Assumption of Moses or the Testament of Moses was the source of Jude's allusion, and 'cautiously assert[s] that it was the ToM'. However, because she finds that 'there is not enough evidence in the text itself or in external sources to identify the Moses fragment with either ToM or AoM, she concludes 'that no direct connection can be made between the Moses fragment and Jude 9' (Grierson, op. cit., p. 279).
• 17 Cf. 2 Cor. 4:4, where 'the light of the gospel of the glory of Christ, who is the image of God' is contrasted with 'the god of this age' who 'has blinded the minds of the unbelievers'.
• 18 Laato, Antti (2013). The Devil in the Old Testament. In I. Fröhlich & E. Koskenniemi (Eds.), Evil and the Devil (pp. 1-22). London: Bloomsbury T&T Clark, pp. 4-5, 20-21.
• 19 Stokes, Ryan E. (2016). What is a Demon, What is an Evil Spirit, and What is a Satan? In J. Dochhorn, S. Rudnig-Zelt & B. Wold (Eds.), Das Böse, der Teufel und Dämonen (pp. 259-272). Tübingen: Mohr Siebeck, pp. 268-69.
• 20 Stokes, op. cit., p. 271.
• 21 These are the two possibilities raised concerning ὁ σατανᾶς in the Pauline epistles by Williams, Guy (2009). The Spirit World in the Letters of Paul the Apostle: A Critical Examination of the Role of Spiritual Beings in the Authentic Pauline Epistles. Göttingen: Vandenhoeck & Ruprecht, p. 88.
• 22 Black, Matthew (1990). The Doxology to the Pater Noster with a Note on Matthew 6.13b. In Davies, P.R. & White, R.T. (Eds.), A Tribute to Geza Vermes: Essays on Jewish and Christian Literature and History. (pp. 327-338). Sheffield: Sheffield Academic Press, p. 336.
• 23 Burke, Jonathan (2016). Satan and Demons in the Apostolic Fathers: A Minority Report. Svensk Exegetisk Årsbok, 81, 127-168. On p. 137, n. 56, Burke cites Black as support for his claim that the use of ὁ πονηρὸς for a supernatural being 'has no pre-Christian witness'.
• 24 Matt. 4:1 says that Jesus was led by the Spirit into the wilderness to be tempted by the devil, which implies divine complicity in allowing the devil to tempt Jesus.
• 25 For Christian readers, it is the final, canonical form of the TS that is authoritative for constructing theology.
• 26 In his latest article, Burke writes, 'Farrar wrongly claims I ignore narrative criticism'. However, not only does he then neglect to provide any evidence of his previous interaction with narrative criticism; he does not rectify the omission by interacting with any of the narrative critics I had cited.
• 27 See Christadelphian writings cited in my The Devil in the Wilderness, p. 9.
• 28 See the various opinions summarized in Carruth, Shawn & Robinson, James M. (1996). Q 4:1-13, 16: The Temptations of Jesus - Nazara. Reconstructions of Q Through Two Centuries of Gospel Research: Excerpted, Sorted and Evaluated. Leuven: Peeters, pp. 264-272. After surveying previous research, four evaluators (Carruth, Hartin, Chang and Robinson) pass judgment and all four agree that ὄρος was in the Q source. Similarly, 'Most scholars believe that Matthew took εἰς ὄρος ὑψηλὸν from Q; only rarely does one encounter the opposite opinion. The fact that specific settings (desert, temple) are given for the other two temptations, points in the former direction, for it leads one to expect that a setting for the third temptation would be specified as well. Moreover, Luke's version of this temptation shows signs of extensive editorial revision. In particular, the verb ἀνάγω, which appears in Lk. 4.5 in place of the mountain reference, is a characteristically Lukan word, and the omission of a mountain setting along with the addition of ἐν στιγμῇ χρόνου can be seen as the result of Luke's desire to clarify the purely visionary nature of this temptation. Thus both the form of the narrative and certain features characteristic of Luke's redactional style favour the view that the mountain reference was originally part of the Q account.' (Donaldson, Terence (1987). Jesus on the Mountain: A Study in Matthew. London: Bloomsbury, pp. 87-88).
• 29 'The third temptation, according to the redaction of 4:8, occurs on "a very high mountain." Luke's version (4:5) simply reads, "And leading him up..."' (Waetjen, Herman C. (1976). The Origin and Destiny of Humanness: An Interpretation of the Gospel According to Matthew. San Rafel: Crystal Press, p. 76); 'This is probably redactional. There is no parallel in Luke. Matthew will add a mountain in 5.1; 8.1; 15.29; and 28.16; and he likes λίαν. For ὑψηλός (Mt: 2; Mk: 1; Lk: 1) with ὄρος (Mt 16; Mk: 11; Lk: 12) see 17.1 (from Mark; cf. Gen 7.19-20; Deut 12.2; Jth 7.4; Isa 2.14; 14.13; Jer 3.6; Ezek 40.2).' (Davies, W.D. & Allison, Dale C., Jr. (1988). A Critical and Exegetical Commentary on The Gospel according to Saint Matthew (Vol. 1). London: T&T Clark, p. 369); 'To be the object of εἰς, the evangelist adds ὄρος ὑψηλὸν λίαν, "an exceedingly high mountain," i.e., higher yet, since according to Matthew's version Jesus has already been led up (v 1). ὑψηλὸν anticipates the description of the Mount of Transfiguration (17:1), and ὄρος and λίαν belong to Matthew's favorite vocabulary (6,1; 3,1). Above all, the added phrase carries forward the parallel between Jesus and Moses: Jesus views all the kingdoms of the world from a mountain just as Moses viewed not only all the land of Canaan (Deut 34:1-4), but also "the west and north and south and east" (Deut 3:27), from Mount Pisgah, or Nebo.' (Gundry, Robert Horton. (1994). Matthew: A Commentary on His Handbook for a Mixed Church Under Persecution (2nd edn). Grand Rapids: Eerdmans, p. 57.); 'Matthew may have added ὄρος ὑψηλὸν λιάν to his mountain motif. Though the absence in Luke of a location is irregular, perhaps "mountain" is to be understood, or perhaps the journey was not of an earthly, but rather cosmic, even apocalyptic sort. This is an attractive idea, since it would resolve the logistics problem of viewing all the kingdoms from a mountain in the first place. Wherever this "mountain" is, it too is a mythic reference, and woulud be a logical addition by Matthew. Vaage's retention of ὄρος is logical, still allowing ὑψηλὸν λίαν to be a Matthean embellishment. This would give Matthew some credit for mountain-top elaboration, and at the same time preserve (or add) some of the parallelism and balance to the scenes. It doesn't explain Luke's ommission however. I would retain ἀνάγειν and assume the most logical location for such a viewing: the sky, adding an element of an apocalyptic sort, but it seems inevitable. "In a moment of time" is the mythological duration of the panoramic viewing; appropriate if the location is the sky, inappropriate on a mountain.' (Robbins, C. Michael (2007). The Testing of Jesus in Q. New York: Peter  Lang, p. 149); At Matt 4:8 the expression "to a very high mountain" (ὄρος ὑψηλὸν λίαν) may be redactional, since it is not in the corresponding passage at Luke 4:5; on the other hand, Luke may have chosen to omit it. In any case, it seems entirely likely that Matthew intends us to see the mountain at 4:8 as corresponding to the mountain at 28:16' (Bryan, Christopher. (2011). The Resurrection of the Messiah. Oxford: Oxford University Press, p. 297 n. 51.); 'Matthew has followed his source, which has Jerusalem (= Matthew's "holy city") and the temple in the source of this narrative, sandwiched by desert on the one side, and Matthew's redactional "an exceedingly high mountain," on the other side... Matthew has modified Q's third temptation - "Then the devil led him up and showed him in an instant all the kingdoms of the world" - by situating it on a mountain.' (Cohen, Akiva. (2016). Matthew and the Mishnah. Tübingen: Mohr Siebeck, p. 227).
• 30 See discussion in Carruth & Robinson pp. 274-79.
• 31 As noted by Edwards, Origen and Jerome quote from the Gospel of the Hebrews in a way that appears to indicate its identification of Mount Tabor as the mount of temptation (Edwards, James R. (2015). The Gospel According to Luke. Grand Rapids: Eerdmans, p. 128 n. 101). For example, 'But if someone accepts the Gospel according to the Hebrews, where the Savior himself says, "My mother, the Holy Spirit, took me just now by one of my hairs and carried me off to the great mountain Thabor"...' (Origen, Commentary on John 2.12.87, in Heine, Ronald E., trans. (1989). Origen, Commentary on the Gospel of John, Books 1-10. Washington: Catholic University of America Press, p. 116).
• 32 '[Luke's] shift from a local to a temporal scene can account for his omission of the mountain in the source (Loisy, Vogels, Creed, Vosté, Morgenthaler, Dupont, Kruse, Fitzmyer, Sevenich-Bax).' (Carruth, in Carruth & Robinson, op. cit., p. 271).
• 33 'That no such mountain exists from which one can view the whole earth provides the key to explaining Luke's omission of this phrase.' (Hartin, in Carruth & Robinson, op. cit., p. 271); 'It is unclear to what extent Luke would have been influenced by pragmatic considerations: Even with a flat earth limited to the Mediterranean world, no mountain is high enough. (The limitation is actually that of eyesight, not of altitude.)' (Robinson, in Carruth & Robinson, op. cit., pp. 271-72).
• 34 'The omission of the setting on "a very high mountain" (see Matt 4:8), and especially the reference to "in an instant" (Gk. lit. "in a moment of time"), implies something visionary or imaginary. It is thus more spiritual than the physical version of the temptation in Matt 4:8-9.' (Edwards, op. cit., p. 128).
• 35 See Robbins, op. cit., p. 149 (quoted above).
• 36 'There may also be a heightened mythological component: "Leading him up" may imply an ascent through the air, e.g. somewhat like what may be meant in Acts 8:39: πνεῦμα κυρίου ἥρτασεν τὸν Φίλιππον. Luke also refers to visible descents from above: ἐθεώρουν τὸν σατανᾶν ὡς ἀστραπὴν ἐκ τοῦ οὐρανοῦ πεσόντα (Luke 10:18, a saying only in Luke); καὶ καταβῆναι τὸ πνεῦμα τὸ ἅγιον σωματικῷ εἴδει ὡς περιστερὰν ἐπ’ αὐτόν (Luke 3:22-the bodiliness is only in Luke). Jesus' own final ascension would then be a kind of inclusio (note the recurring prefix ἀνα-): Acts 1:2: ἀνελήμφθη; Luke 24:51: διέσθη ἀπ’ αὐτῶν καὶ ἀναφέρετο εἰς τὸν οὐρανόν, anticipated already in Luke 9:51: ἀνάλημψις.' (Robinson, in Carruth & Robinson, op. cit., pp. 271-72).
• 37 'The temptation story also gives a description of a soul flight, where Jesus is transported from one place to another, from the wilderness to the pinnacle of the temple in Jerusalem. Then, in another soul journey, the devil takes Jesus to a high mountain, where he is shown all the kingdoms of the earth.' (Gagné, André. (2016). Narrative Depictions of Altered States of Consciousness in 1 Enoch and the Synoptic Tradition. In Loren T. Stuckenbruck & Gabriele Boccaccini (Eds.), Enoch and the Synoptic Gospels: Reminiscences, Allusions, Intertextuality (pp. 19-30). Atlanta: SBL Press, p. 27).
• 38 'As Fitzmyer says (1980, 507) Luke makes "minor modifications in the story (which) reveal a Lucan concern to present the temptations in a plausible form"' (Hartin, in Carruth & Robinson, op. cit., p. 271); 'How can we explain Luke's redaction?... Luke substitutes a temporal expression ἐν στιγμῇ χρόνου for Q's spatial expression εἰς ὄρος ὑψηλὸν out of a concern for verisimilitude. He realizes that no mountain, however high it might reach, could furnish a vantage point from which to survey all the kingdoms of the world. A similar concern for verisimilitude surfaced in the first temptation where Luke switched to a single stone to avoid having the landscape strewn with loaves of bread. Instead of the spatial image, Luke substitutes a temporal expression which refers to the instantaneous vision that unfolds before Jesus. Luke thus frames the temptation as no ordinary occurrence but rather as a visionary experience.' (Fleddermann, H.T. (2005). Q: A Reconstruction and Commentary. Leuven: Peeters, p. 249.)
• 39 (Robinson, in Carruth & Robinson, op. cit., p. 271).
• 40 Gruschcow, Lisa (2006). Writing the Wayward Wife: Rabbinic Interpretations of Sotah. Leiden: Brill, p. 5.
• 41 Zellentin, Holger M. (2011). Rabbinic Parodies of Jewish and Christian Literature. Tübingen: Mohr Siebeck, p. 10 n. 33.
• 42 Stemberger, Günter (1996). Introduction to the Talmud and Midrash (2nd edn). Edinburgh: T&T Clark, p. 171: 'Most scholars...would accept a later date for the last named teachers in PT, thereby inferring a date of redaction in the first half of the fifth century.'
• 43 Neusner argues at length that historical criticism must be used in the study of rabbinic literature, just as in the study of the Gospels. The burden of proof is on the one who claims a particular saying is early enough to be relevant to New Testament research. If this cannot be shown, it is not known. See Neusner, Jacob (1994). Rabbinic Literature and the New Testament: What We Cannot Show, We Do Not Know. Eugene: Wipf & Stock, esp. pp. 1-17.
• 44 Rosen-Zvi, Ishay (2009). Refuting the Yetzer: The Evil Inclination and the Limits of Rabbinic Discourse. Journal of Jewish Thought and Philosophy, 17(2), 117-141. Here p. 121.
• 45 Rosen-Zvi, op. cit., p. 131 n. 39.
• 46 'From about A.D. 500 to 800 the Masoretes added vowel points, accents and the Masorahs (to help safeguard the text from error) as well as many scribal corrections.' (Wegner, Paul D. (2006). A Student's Guide to Textual Criticism of the Bible: Its History, Methods and Results. Downers Grove: InterVarsity Press, p. 77.)
• 47 Rosen-Zvi, Ishay (2009). Sexualising the Evil Inclination: Rabbinic 'Yetzer' and Modern Scholarship. Journal of Jewish Studies, 60(2), 264-281. Here p. 272; emphasis added.
• 48 'Apotropaic' activity refers to 'preventative measures...taken, via petition or incantation, to ensure safety from future demonic harm', in contrast to exorcistic activity which cures people from current demonic affliction (Morris, Michael (2016). Apotropaic Inversion in the Temptation and at Qumran. In J. Dochhorn, S. Rudnig-Zelt & B. Wold (Eds.), Das Böse, der Teufel und Dämonen (pp. 93-100). Tübingen: Mohr Siebeck, p. 93.
• 49 Morris, op. cit., p. 93. Morris concludes that the TS portrays 'an aggressive manipulation in which the Devil mocks the apotropaic efficacy of Psalm 91 in order to intimidate Jesus' (p. 99). This, of course, makes little narrative sense if there is no demonic element to the temptation.

An assessment of the accommodation theory of demon possession and exorcism in the Synoptic Gospels

I haven't blogged in a few weeks, partly because it seemed right to take some time off over the holidays, and partly because I've been busy with some larger projects. One of those projects is now available online:

‘When an unclean spirit goes out of a person’: An Assessment of the Accommodation Theory of Demon Possession and Exorcism in the Synoptic Gospels

This is a thorough study of the references to demons, unclean spirits, possession and exorcism in the Synoptic Gospels (Matthew, Mark, and Luke) with a view to evaluating the accommodation theory. Because it is a lengthy paper, I'm going to offer an overview here for those who may not be inclined to read 50 pages on the subject.

A surface reading of the Synoptic Gospels gives the impression that Jesus, the disciples and the narrator all share the belief of their contemporaries in the reality of demons. There are, broadly speaking, three theories offered to explain this impression. The first two theories, which I'll refer to as the 'reality theory' and the 'error theory', agree that all is as it seems: Jesus, the disciples and the narrators (the Synoptic writers) did actually believe in demons. Proponents of the reality theory claim that they were correct in this belief, while proponents of the error theory claim that they were mistaken. The choice between these two theories will depend on one's view of Scripture, and also to some extent on empirical evidence for demon possession and exorcism today. The third theory, the accommodation theory, claims that all is not as it seems: Jesus, the disciples and the Synoptic writers did not believe in demons. They used the language and terminology of demon possession and exorcism, either because there were no other words available to describe these phenomena, or because it wasn't a priority for them to correct these misconceptions, or they used it ironically with the intention of correcting their contemporaries' beliefs about demons. As you can see, the accommodation theory can take different forms, which I classify broadly into two categories: 'benign accommodation', in which Jesus and his followers were content to leave existing ideas about demons uncorrected, and 'subversive accommodation', in which Jesus and his followers intended through their apparent accommodation to subvert popular beliefs and construct an alternative, non-supernatural demonology.

Now that we understand the options on the table, it's time for a quick history lesson. Throughout most of Christian history, from the Patristic era down to the Reformation, the reality theory has been the unquestioned position of the church. After the Reformation, and especially as the Age of Reason and subsequently the Enlightenment got underway, people started to challenge the existence of demons, and they brought their skepticism to the biblical text - which was, however, still regarded by most as inerrant. It seemed impossible that Jesus and the New Testament writers could have shared a belief that was now regarded as obsolete and even irrational, and so the accommodation theory was born. This view came to prominence in England in 1737 with the anonymous publication of a tract entitled, An Enquiry Into the Meaning of Demoniacks in the New Testament (the author was later revealed to be A.A. Sykes). This set off a heated debate between traditionalist and rationalist students of Scripture that lasted for many decades.

As the 19th century developed, the rise of biblical criticism brought a new challenge, not only to the traditionalists but also to the accommodationists. D.F. Strauss, a pioneer of 'historical Jesus' studies, published a lengthy historical-critical study of the Gospels in 1835-36 entitled The Life of Jesus which included a devastating critique of the accommodation theory. Strauss himself did not believe in demons but was an early proponent of the error theory. Throughout the 19th and early 20th centuries the debate continued, but the accommodation theory gradually waned in influence. This could be seen clearly in the controversy that arose in the Church of England in the 1970s over the existence of demons and propriety of exorcism, in which both sides presupposed that Jesus himself had believed in demons and been an exorcist. Today, the majority of New Testament scholars agree that Jesus, his earliest followers and the Synoptic Evangelists all shared a real belief in demons and exorcism (see the plethora of quotations provided in the article, pp. 12-16). The accommodation theory has been pushed to the periphery in academic circles.

It is only among those who disbelieve in demons but have a very high view of Scripture that the accommodation theory remains popular. The Christadelphians are an example of a group with such beliefs. Christadelphians have always subscribed to some kind of accommodation theory or other. However, over the past half century some Christadelphian writers have proposed a more radical form of the theory, which I've referred to above as subversive accommodation. They hold that the references to demons in the Synoptic Gospels are ironic, intending to teach something completely different to the literal meaning of the words. To my knowledge, this idea is unknown outside of the Christadelphian community.

So, do theories of accommodation satisfactorily explain the evidence of the Synoptic Gospels?

Virtually everyone, accommodationists included, are in agreement that belief in demons was prevalent among first century Palestine, though there is scholarly debate on just how prevalent it was, and in particular, how prevalent belief in demon possession was. Nevertheless, there is little doubt as to the literal meaning of phrases like 'cast out demons' or 'come out of him, you unclean spirit' in such a setting. So, is there any reason to think that when Jesus, his disciples and the Synoptic writers used such language, they merely did so neutrally without espousing these beliefs? Or that they did so ironically, intending an altogether different meaning?

Accommodationists have offered several arguments in favour of their theory. For the most part, these are arguments from silence, which are intrinsically weak. Firstly, it is argued that the Synoptic Gospels contain no systematic demonology. They show no interest for instance in the origin or precise nature of demons. This, it is said, shows that they were not drawing on existing Jewish beliefs about demons. Now, there is a strong consensus among scholars that the Synoptic Gospels do draw on existing Jewish beliefs about demons, but we will come to that. For now, suffice it to say that the Gospels are about the life, ministry and person of Jesus. They refer frequently to angels without providing a systematic angelology, because they are not about angels. Angels appear in the text incidentally, and so do demons. We need to let the Gospels be Gospels and not make them out to be demonological treatises.

The second accommodationist argument is that the Synoptic Gospels do not always distinguish the demon and its actions from the possessed individual, which suggests that the 'demon' can be regarded as an attribute of the sufferer and not a separate being. This argument ignores the fact that the Synoptic Gospels often do distinguish the demon from the possessed person. Moreover, the limitations of language in describing a phenomenon such as demon possession, in which one 'person' takes over the body of another, make it virtually necessary to vacillate between ascribing actions to the demon and to the demon-possessed person. One New Testament scholar states that this state of affairs had given rise to a figure of speech whereby "A person 'is' the spirit which dwells in the person concerned" (see p. 20 of paper).

The third accommodationist argument is another argument from silence. It is claimed that Jesus' exorcism ministry was restricted to the northern part of Palestine (mainly Galilee) and that the Synoptic Gospels never have him performing exorcisms in Judea. This is said to show that belief in demons was minimal in Judea. Consequently, he accommodated belief in demons in Galilee, but did not do so in Judea because there was no need. Several objections can be raised against this line of argument. Firstly, the Synoptic Gospels mention only two distinct healing miracles of Jesus in Judea, plus one summary statement about an unspecified number of healings in the temple (the evidence of John's Gospel will be addressed separately below). By contrast, they mention at least eighteen distinct healing and exorcism miracles of Jesus in the north, plus numerous summary statements describing 'many' additional healings and exorcisms. Thus it is apparent that the vast majority of Jesus' healing ministry in general took place in Galilee and surrounding areas. The sample size from Judea is far too small for an argument from silence to carry any weight. Moreover, there is no evidence that afflictions which were characteristically regarded as demonic in the Galilean context were characteristically regarded as natural in the Judean context. Furthermore, all three Synoptic Gospels have a summary statement about Jesus' healings and exorcisms in Galilee which reports that a great multitude of people from all Judea and Jerusalem were following him (Luke 6:17-19 and parallels). There is every reason to think the exorcisms mentioned here included Judeans and Jerusalemites. We also know for certain that the apostles engaged in a major exorcism ministry in Jerusalem for the benefit of Judeans (Acts 5:16). We also know that the de facto leader of the church at Jerusalem, James, alluded to exorcism in his epistle (James 2:19). Finally, we know that, apart from the Sadducees, the Jewish religious leaders from Jerusalem believed in demons (Mark 3:22; John 7:20; 8:48; 8:52; 10:20; cf. Acts 19:12-18). The notion that Judeans and Jerusalemites, as a rule, did not believe in demons and would not have required accommodation is demonstrably false.

The fourth and final major line of argument raised by accommodationists is yet another argument from silence. In this case it is argued that the omission of exorcism stories in the Gospel of John shows the real state of the case. In particular, it is claimed that the Synoptic Gospels were written for the uninformed and spiritually immature, who needed accommodation, while the Gospel of John were written for mature Christians who no longer believed in demons and thus needed no accommodation. In the first place, as discussed in the paper, this sharp distinction in audience and purpose between the Synoptic Gospels and John is reductionist and has virtually no scholarly support. There is good reason to think that Matthew, Mark and perhaps even Luke were written primarily for use by the church, including but not limited to evangelistic purposes. Moreover, there is good reason to think that John was written at least partly for evangelistic purposes (John 20:31). So the clear distinction in audience and purpose that is supposed to explain the absence of exorcism in John's Gospel does not actually exist.

Secondly, the Gospel of John contains a grand total of four healing miracles, and mentions only four distinct types of health problems, whereas the Synoptic Gospels contain over 20 healing miracles, plus summary statements, and mention more than a dozen different types of afflictions. Thus, while it is striking that John contains no exorcisms, we should be cautious about making an argument from silence on this basis. We cannot be certain of why John chose not to include any exorcism stories, because he doesn't tell us. It is possible that he did not believe in demons, but this is only a conjecture. Scholars have offered a number of other possible explanations (see paper, pp. 29-31) which are equally if not more plausible. Moreover, even those scholars who take John's silence to mean that he did not believe in demons do not regard this as overriding the evidence of the Synoptic Gospels; they instead argue that John had a different view than Jesus and the Synoptic writers.

In summary, there is very little substance to the arguments that are raised in support of accommodation. What about the arguments against accommodation? There are seven mentioned in the paper. The first is an argument from silence, but a weighty one. Subversive accommodationists claim that the references to demons in the Synoptic Gospels are ironic. Now irony is a literary technique in which one makes explicit one attitude or evaluation but implies a different attitude evaluation that is often the opposite of what is expressed. In short, one doesn't mean what one says. Verbal irony is subtle, particularly in written language, and the writer must provide the reader with clues enabling him or her to detect the irony. In my paper I cite a couple of examples of verbal irony in the New Testament and show how the writer has made the irony easily detectable. The question for accommodationists is, where are the clues informing the reader that the references to demon possession and exorcism are ironic? The answer is simple: there are none. Moreover, the vast majority of readers of the Synoptic Gospels down through history have failed to detect any irony in these statements. These are already compelling reasons to reject the subversive accommodation theory.

Secondly, the Synoptic writers carefully distinguish between demonic and non-demonic cases, which shows that they were not accommodating a simplistic view which held all illness to be demonic and treatable with exorcism. The Synoptic writers show sophistication in their use of demon possession and medical terminology and expect sophistication of their readers in appreciating this discernment.

Thirdly, scholars who have compared the references to demons and exorcism in the Synoptic Gospels to the beliefs current in the Judaism of Jesus' day have found that the continuity is far more striking than the differences. Points of continuity include: (i) belief in an indefinitely large number of evil spirits; (ii) association of demons with definite localities, such as deserts, tombs and other desolate places; (iii) special reference to groups of seven evil spirits; (iv) the destiny of demons being imprisonment and final fiery destruction; (v) the belief that demon possession could give people superhuman strength.

Besides this, scholars who have closely studied Jesus' exorcism techniques find significant points of continuity with exorcism techniques known from other ancient literature. In particular, the words of Jesus to the demons parallel incantations known from other literature: 'Be quiet'; 'Come out of him'; 'What is your name?' and 'No longer enter into him'. The use of objects to which to transfer demons (as in the pigs episode) also has ancient parallels. Now it should be noted that there are also points of difference between Jesus and other exorcists of his day. He does not seem to have used any mechanical devices in his exorcisms, he did not use 'proofs' to indicate the success of his exorcisms, he did not pray during his exorcisms, he did not invoke a power authority in his exorcisms, and he seems not to have used the formula 'I bind you'. Jesus' uniqueness as an exorcist lies particularly in commanding the demons on his own authority instead of invoking a higher authority. This tells us something about the person of Christ. 

On the whole it is evident that neither Jesus (as an exorcist) or the Evangelists (as writers) were trying to distance themselves from contemporary conceptions of demons and exorcism. Jesus behaved like an exorcist, and the Evangelists wrote as though they shared the belief in demons common to their contemporaries.

A fourth argument against the accommodation theory lies in the great theological significance that Jesus attached to his exorcisms. For him they were not merely incidental or just one of his many types of healing miracles; they heralded the breaking in of God's rule into the world and the defeat of Satan.

Fifthly, there are at least three supernatural elements in the exorcism accounts which cannot be explained in rationalist terms. The first is the superhuman strength of the Gerasene demoniac which enabled him to tear chains apart and break shackles in pieces. The second is the demons' request to be transferred to the pigs, to which Jesus gave them permission. The third and most striking is the demons' supernatural knowledge about Jesus' identity. The demons consistently know and declare that Jesus is the Son of God, long before the people in the narrative make such a confession. All of this strongly suggests that the Synoptic writers regarded demon possession as a supernatural phenomenon.

Sixthly, there is a general saying about the operations of unclean spirits in Matthew 12:43-45 and Luke 11:24-26 that presupposes the existence of such beings. While these texts raise some interesting hermeneutical issues (e.g. is it a parable or a straightforward saying?), in Luke's case at least, the positioning of the saying immediately after an exorcism and the ensuing Beelzeboul controversy makes it obvious that he regarded the saying as concerning actual spirits. This is true even if there is a higher symbolic meaning.

Lastly, Jesus was not content merely to behave like an exorcist himself. He solemnly conferred on his disciples the authority to cast out demons and sent them out twice on missions in which exorcism played a central role. This enthusiasm is difficult to explain if he was merely humouring or containing existing beliefs. He celebrated when his disciples were successful at exorcism (Luke 10:17-20) and offered them coaching tips when they were unsuccessful (Mark 9:28-29). Moreover, Jesus endorsed the exorcisms of a person who was not his disciple, and therefore had not received any secret teaching that Jesus may have given his disciples about the true nature of demons (Mark 9:38-40). This is virtually impossible to explain if Jesus sought to subvert existing beliefs and practices regarding demons.

In conclusion, it is with good reason that the accommodation theory has fallen out of favour among biblical scholars. The evidence supporting it is virtually non-existent, while the evidence against it is voluminous. Jesus, his disciples, and the Synoptic writers believed in evil spirits, demon possession, and exorcism. And this means that the question which logically follows from this conclusion cannot be papered over: should we?

The reader is encouraged to access the full paper for a more thorough treatment of the subject, with references.

Who tempted Jesus in the wilderness? Ten points to ponder

Accounts of Jesus' temptations in the wilderness are found in Mark 1:13, Matthew 4:1-11, and Luke 4:1-13. In all three accounts an agent of temptation is identified. But who was 'he'? Traditional Christian teaching has identified this 'devil' as a supernatural personal being. Today, however, many Christians regard such a doctrine as an embarrassing relic of pre-modern thought. Some claim that while Jesus and the early church no doubt believed in such a devil, we cannot, and so we have a warrant to reinterpret, 'demythologize' or 'psychologize' this tempter. Others go further and claim that these texts were never intended to convey the idea of a personal devil in the first place. On the vanguard of this school of thought have been the Christadelphians, a millenarian group which formed in the mid-19th century. 

The founder of the movement, a British medical doctor named John Thomas, understood the biblical devil to be a figurative depiction of sin. However, in the case of the wilderness temptations he taught that the tempter was an unspecified human being; thus, still an external tempter. He was followed in this interpretation by his successor as the de facto head of the movement, Robert Roberts. However, it was eventually supplanted by a different view (familiar to but rejected by Roberts) which regarded the devil as a personification of Jesus' own internal desires and hence interpreted the whole account figuratively. In other words, it is held that the Gospel writers (and ultimately Jesus himself) internalized and psychologized the devil. Thus the 'modern' view of the devil is not a modern innovation; it was there in the text all along, just waiting to be discovered! This latter view dominates Christadelphian teaching today.

The Christadelphian interpretations allow Christians to circumvent having to come to terms with a Jesus who believed in a personal devil. However, as attractive as this feature might be to the modern mind, the main question that must be asked is how the original readers of the Gospels are likely to have understood the temptation narratives. What follows is a ten-point summary of more detailed analyses of the temptation narratives (here and here), intended as talking points to assist Christadelphians and others in arriving at a sound biblical answer to the question posed in the title.

1. The genre of the Gospels is narrative.

There can be no doubt that, broadly speaking, the genre of the Gospels is narrative; that is, they report a series of related events. The writers expected their readers to understand these documents as narrating factual, literal events in the life of the historical person Jesus of Nazareth. Thus, when we encounter the temptation accounts, which certainly sound like narrative, our first impression is surely to read them as factual, literal events in the life of Jesus.

To say that these accounts are not straightforward, literal narrative but rather figurative prose is to claim that the writers are using a very subtle and sophisticated literary technique here. Besides being subtle and sophisticated, this technique is completely without parallel in the Gospels, which otherwise stick rigidly to the narrative genre.

True, there is material within the Gospel narratives that falls under other genres, such as parables and discourses. Such material is more likely to contain literary devices such as personification. However, this material is easily distinguished from the narrative itself inasmuch as it is invariably spoken by one of the characters in the story - usually Jesus himself. By contrast, the temptation accounts are not spoken by Jesus but are presented as a story involving Jesus. And there are simply no other stories involving Jesus in the Synoptic Gospels which could be construed as anything other than straightforward narrative.

This is why appeals to personification in other parts of Scripture, such as Lady Wisdom in Proverbs 8, fail to establish a precedent for a figurative reading of the temptation accounts: they fail to take into account the vast difference in genre between poetic wisdom literature and Gospel narrative.

Before even looking at the text in detail, on the grounds of genre alone, a straightforward, literal reading of the temptation narratives is necessarily the default interpretation and a figurative reading must be judged an unlikely possibility which carries a very heavy burden of proof.

2. The importance of Mark's account

Mark's very brief temptation account is sometimes passed over as though it has nothing to tell us about the tempter that is not found in Matthew or Luke. However, Mark's version is actually very significant for two reasons.

Mark is generally agreed to have been written before Matthew and Luke, the latter two having used Mark as a source. However, Matthew and Luke evidently had a different (and probably common) source for the temptation narrative. This hypothetical source is referred to by scholars as Q. It is clear that the agent of temptation in the Q source was ho diabolos (the devil). Matthew and Luke reflect a separate, more detailed temptation tradition which does not seem to be dependent on Mark. Neither Matthew nor Luke include's Mark's detail that "he was with the wild animals", and both Matthew and Luke use ho diabolos as the main designation for the tempter as opposed to Mark's ho satanas. Of course, ho diabolos and ho satanas are equivalent and interchangeable terms, as is clear from Matthew 4:10 and other passages such as Job 1:6 (Hebrew and LXX), Mark 4:15/Luke 8:12, and Revelation 12:9.

Thus, Mark shows that there are two independent and early strands of tradition which attribute Jesus' temptations to Satan/the devil. It is thus very likely that this attribution goes back to the teachings of Jesus himself rather than being the literary stroke of a later writer. (Of course, Jesus must have recounted the wilderness temptation to his disciples, or otherwise we cannot explain how the tradition came about, since no eyewitnesses appear to have been present.)

The other significant feature of Mark's version of the temptations is its very brevity. It is evident that Mark regarded the sentence, "And he was in the wilderness forty days, being tempted by Satan" as self-explanatory. Mark was able to assume that his readers would understand what he meant without providing any further clues as to the identity of this 'Satan'. This is remarkable since satanas is not a Greek word but a transliterated Semitic term. Mark often uses Semitic terms in his Gospel but usually provides a translation for his predominantly Gentile readers. In this case, he provides no translation or explanation. This suggests that Mark regarded ho satanas as a proper name or, at the least, as a specific theological term well known to his readers.

Thus, we need to ask ourselves, at the time Mark wrote his Gospel (c. 50s or 60s AD), what concept of ho satanas could have been well established in the church? The most likely answer is a concept of 'satan' found in the Old Testament and/or intertestamental Judaism. The problem is, while 'satan' is used for personal beings (and in some cases, arguably a specific personal being) in the Old Testament and intertestamental Judaism, there seems to be no evidence for a well-developed figurative concept of 'satan' at this time. Mark thus presents the Christadelphian view with a real historical difficulty.

3. Not a devil but 'the' devil; not a satan but 'the' satan

All three Gospels use the definite article when introducing the devil for the first time. Mark has ho satanas, literally 'the satan' (or 'Satan' if it is taken to be a proper name). Matthew and Luke have ho diabolos, 'the devil', whom Matthew also identifies as ho peirasmos ('the tempter'). The significance of the definite article here cannot be discounted. It is further evidence that the writers expected their readers to know who or what they meant by these terms. It was not merely a satan, or a slanderer, or a tempter, but THE satan, THE slanderer, THE tempter par excellence!

The question is, in a mid first century context what did ho satanas or ho diabolos (without qualification) refer to? Ultimately we should make recourse to Job 1-2; the writings of Second Temple Judaism also provide useful background. However, right in the Gospels we have an account of a dispute between Jesus and his contemporaries about a personal ruler of demons whom Jesus refers to as ho satanas. Now, it has been claimed that Jesus merely assumed this view of Satan for the sake of argument (though I've argued elsewhere why this interpretation doesn't stand up to scrutiny). However, what is more immediately relevant is that the dispute establishes that 'the satan' was in contemporary Jewish usage the title or name of a specific personal being. This also forms an important part of the larger Gospel context against which 'THE satan' and 'THE devil' must be understood.

Thus the use of the definite article in the temptation narratives shows that a particular being or figure is in view. This rules out any possibility of interpreting the tempter as an unspecified human opponent. Moreover, in light of the Beelzebul controversy we have grounds for claiming that Jesus' contemporaries would have understood ho satanas to be the designation of a specific supernatural being. The Christadelphian view faces a serious obstacle in the lack of evidence for a figurative concept of 'satan' that had become so entrenched in the church by the time the Gospels were written that the writers perceived no risk of misunderstanding in describing Jesus' tempter as the devil and the satan without qualification.

4. The devil came and said...

It was already mentioned that the reader's first impression upon reading the temptation accounts is that of a straightforward, literal narrative. This reading is borne out by a closer inspection of the text. Matthew tells us, "And the tempter came and said to him..." The verb translated 'came' here is proserchomai. This verb occurs 87 other times in the New Testament (50 of them in Matthew!) and in every one of them it takes a literal meaning. Among these 87 occurrences is Matthew 4:13, "angels came and were ministering to him." Thus in the same immediate context we have the verb being used literally of personal beings coming to Jesus.

Proserchomai can take a figurative meaning, similar to how we might say in English, 'I don't know what came over me.' However, this meaning is very rare - rare enough, as we have seen, not to be attested in the other 87 uses of this word in the New Testament. 

Furthermore, the fact that proserchomai is used together with another verb, epo, militates against taking it figuratively. The combination of proserchomai with another verb is a common feature of Matthew's style, and in most cases he uses it to introduce interpersonal encounters - particularly dialogues (Matthew 8:2; 8:5; 8:19; 8:25; 9:14; 9:28; 13:10; 13:27; 13:36; 14:15; 15:1; 15:12; 15:23; 16:1; 17:7; 17:14; 17:19; 17:24; 18:1; 18:21; 19:3; 19:16; 20:20; 21:23; 21:28; 21:30; 22:23; 25:20; 25:22; 25:24; 26:17; 26:49; 26:69; 26:73; 28:18). It would be distinctly odd for Matthew to use his stylistic idiosyncrasy here with a completely different meaning.

5. The devil left and angels came

As already mentioned, the verb proserchomai is used literally of angels coming to Jesus in Matthew 4:13. What is even more striking about v. 13 is that the coming of the angels is contrasted with the devil's leaving. The devil left and the angels came. One is very obviously a literal statement about personal beings; on what basis could we insist on taking the other as a figurative statement about a personification?

Furthermore, it should be noted that the statement that angels came to Jesus (corroborated in Mark 1:13) establishes beyond any doubt that Jesus did interact with supernatural personal beings while in the wilderness.

6. Dialogue between a person and a personification?

The main focus of the temptation accounts in Matthew and Luke is the dialogue that takes place between the devil and Jesus. If we are to interpret the accounts figuratively, then obviously no actual dialogue took place; instead this is a dramatic depiction of an internal struggle in Jesus' mind. There are two significant difficulties here. The first is a very simple matter. If the struggle is between aspects of Jesus' thought process, within Jesus, then why is "Jesus" one of the interlocutors, as opposed to, say, 'the servant' or some other figurative representation of the obedient aspect of Jesus' will? The fact that it is "Jesus" who is in dialogue with "the devil" makes it quite clear that "the devil" is entirely distinct from Jesus and not a part of Jesus.

Second, the dialogue stretches the limits of figurative language to the breaking point. In some cases in the Bible, impersonal entities are described as speaking or singing (but with no actual content of their speech or song specified). For instance, in Genesis 4:10 God tells Cain that Abel's blood cries to him from the ground (note that, while this figure of speech occurs within a narrative, it part of a statement by God and not an event described by the narrator). In Psalm 98:8 "the hills sing for joy". Rarely, in obviously figurative contexts, personified figures speak with the content of the monologue actually spelled out (e.g. Lady Wisdom crying in the streets in Proverbs 8, or the foot hypothetically talking to the hand in 1 Corinthians 12:15).

However, what we do not find are dialogues between personified figures in which one speaks and the other answers, back and forth. And we certainly do not find such dialogues between a literary device and a literal person! To claim that this is what is happening in Matthew 4 and Luke 4 is to require the writers to have invented a brand new genre -- and camouflaged it within a genre which is normally read in a straightforward, literal manner!

7. A physical act of worship

Both Matthew's and Luke's accounts of the temptation are explicit that what the devil demanded from Jesus in the third temptation (second in Luke's order) was a physical act of obeisance. In Matthew's case this is expressed by combining the verbs proskuneo (to worship) and pipto (to fall down). Proskuneo on its own almost never takes on an abstract or reflexive meaning; certainly it never does elsewhere in Scripture. Any lingering doubt about whether it is literal in Matthew 4:9 is removed by the addition of pipto. Jesus was to fall down and worship the devil. These two verbs are combined in two other passages in Matthew (2:11; 18:26) and in both cases the sense is a physical act of obeisance.

Luke is less explicit but his language still implies a physical act of worship since he modifies proskuneo with enopion, a word meaning 'in the presence of' or 'before'. So according to Luke, Jesus was to worship before the devil. The language in both Matthew and Luke is perfectly clear: what the devil demanded of Jesus was not merely an internal shift in allegiance but a physical act.

Now if the devil in fact represented an abstract concept, a component of Jesus' mind, a physical act of obeisance before the devil is meaningless: there is no physical object of worship! One is left to try to force 'fall down and worship me' to represent an internal decision, contrary to the plain meaning of the words. One cannot allow an interpretation which implies that what the text says Jesus was tempted to do is not what he was actually tempted to do!

8. A property transaction

There is a Roman legal term called traditio longa manu which sheds additional light on this third temptation. Traditio ('delivery') referred to the process of transferring ownership of property from one person to another. For movable assets this normally consisted of a physical handing over. In the case of immovable assets such as land, however, the law provided for the 'delivery' to take place by the seller bringing the buyer to the spot and pointing it out to him. As long as the seller and buyer both had the intent to exchange the land, the transaction was considered to have been effected. With this background in mind, the devil's move in showing Jesus all the kingdoms of the world and his offer to give them to Jesus can be understood as an offer to transfer this property to him. This shows that the temptation was transactional in nature; a transaction requires two distinct parties.

As in the case of the worship language, this legal background to the temptation is rendered meaningless if only one person was involved.

9. The devil's pitch

In Luke's version of the temptation account, the devil makes an extended pitch not found in Matthew. He says, "To you I will give all this authority and their glory, for it has been delivered to me, and I give it to whom I will." This pitch makes perfect sense if the devil is a personal being trying to persuade Jesus that he is capable of delivering on his offer. However, if this 'dialogue' is actually an internal struggle in Jesus' mind, this is a very odd line. If the authority has already been delivered to 'me', and 'me' and 'you' are not actually distinct persons but aspects of one person, what does it mean for 'me' to give it all to 'you'? Furthermore, if the whole dialogue is about Jesus and no one else, what possible meaning does the hypothetical 'whom' have in, "I give it to whom I will"? To whom else might the Son of God contemplate giving all authority and glory?

10. What about the very high mountain?

The only positive exegetical argument that Christadelphians typically raise against taking the temptation narratives literally involves Matthew's reference to a very high mountain from which the devil showed Jesus all the kingdoms of the world and their glory. Since there is no mountain on earth from which all the kingdoms of the world can be viewed, it is argued that the whole account must be taken figuratively.

However, this issue can be satisfactorily resolved without resorting to a figurative interpretation fraught with much more serious difficulties. First of all, in a first century context, "all the kingdoms of the world" does not refer to the entire globe but rather to the then known world, namely the Roman Empire and its environs. Secondly, being shown "all the kingdoms of the world" can be understood hyperbolically (Jesus didn't literally see the whole world, but a vast expanse of land which he and the devil understood to represent the whole world). 

Another alternative, more likely in my view, is that Jesus did see all the kingdoms of the world, but that this required a supernatural experience, such as supernaturally enhanced vision or being taken up from the mountain into the heavens. Several points can be raised in support of this interpretation. Firstly, Luke does not mention a mountain but says that "the devil took him up and showed him all the kingdoms of the world in a moment of time". The 'taking up' and 'moment of time' both emphasize the supernatural character of the experience. 

Secondly, the whole temptation narrative depends heavily on Deuteronomy (the forty days = forty years in the wilderness, the testing of Israel in the wilderness, and the fact that Jesus' responses to the devil all quote from Deuteronomy). In light of this, it is very likely that Matthew intends this mountaintop temptation to be read in light of Deuteronomy 34, in which Moses ascended Mount Nebo and God showed him all the land that the Israelites were about to receive. Geographically inclined commentators advise us that some of the places mentioned in Deuteronomy 34:1-3, such as Zoar, are not actually visible from Mount Nebo. They conclude, therefore, that this account must be understood as involving some kind of supernatural visionary experience on the part of Moses. Of course, this does not negate the literal nature of the narrative in Deuteronomy 34, nor even the literal nature of the mountain. A similar mountaintop visionary experience is described in 2 Baruch 76:3-4 (a Jewish text roughly contemporary with Matthew). Thus it may be reasonably supposed that first century readers of Matthew would have understood Matthew 4:8 in terms of a supernatural - but still objective - visionary experience atop a literal mountain.

Thus the reference to the very high mountain does not provide us with an escape hatch by which to justify interpreting interpreting the entire temptation account figuratively.

Conclusion

It should be apparent to the reader that, whatever difficulties it might present to the modern scientific mind, the only plausible interpretation of the wilderness temptation accounts is that this was a literal encounter between Jesus and a personal being known as Satan or the devil. This being had sufficient notoriety to be referred to as THE tempter par excellence. He was in a position to make a credible claim to be able to hand over the kingdoms of the world to whomever he would. He possessed the supernatural power necessary to set Jesus on the pinnacle of the temple, and take him up to show him all the kingdoms of the world in a moment of time.

There can be only one conclusion. The devil that tempted Jesus - and thus the devil of Scripture - is a supernatural personal being.