The Controversy over Beelzebul (Mt 12:24–32; Mk 3:22–30; Lk 11:15–23 + 12:10 )

One of the most critical passages in the Synoptic Gospels is the Beelzebul controversy. The prominence in the gospel tradition of the accusation that Jesus casts out evil spirits by being in league with Beelzebul, the prince of evil spirits, suggests that it was one of the major ways of discounting Jesus' miracles used by his opponents. Moreover there are also parallels in John to the synoptic tradition, since there also Jesus is accused of having an evil spirit (Jn 7:20 ); on another occasion Jesus cites his power to work miracles in reply to such an accusation (Jn 10:20–1 ). The parallels are, however, sufficiently loose to be explained as dependent on oral rather than written tradition; the common point may be merely the memory that Jesus was accused of having an evil spirit. In John the accusation is made twice that Jesus ‘has an evil spirit’, and on the second occasion this is backed up with the question, ‘Can an evil spirit open the eyes of the blind?’ The circumstances of the accusation in the Synoptics and in John are entirely different. In the Synoptics the starting-point of the discussion is expulsion of evil spirits; it is more specific than in John: ‘They said that he has Beelzebul and casts out evil spirits through the leader of evil spirits.’ It then leads on to a full-blown controversy.

The use of this particular tradition is different in each of the gospels. In Mark it is the centrepiece of a typically Markan ‘sandwich’, showing how Jesus was misunderstood by different groups of people. This then leads on to the recourse to parables in Mk 4 . It is, then, part of Mark's demonstration of Jesus turning away from the crowds to instruct his special disciples, an important hinge in the structure of the first part of Mark's gospel. In Matthew the passage provides a commentary on the important quotation in Mt 12:18–21 of Isa 42:1–4 , including, ‘I shall place my Spirit upon him’; its message is, therefore, the contrast between the Spirit of Jesus and the spirit of Beelzebul. Not dissimilarly, in Luke the main part of the passage comes in the section on discipleship after the Lord's prayer and the promise ( 11:13 ) that the heavenly Father will give the Holy Spirit to those who ask him. It serves to contrast the spirit of the disciples with that of Jesus' opponents. Finally in John, the passages are part of the confrontation between Jesus and the temple authorities in Jerusalem, as part of the judgement theme that is so important in John.

From the point of view of the synoptic problem this passage has an importance all its own. Any solution to the problem must, of course, be shown to be valid for all the pericopes of the synoptic tradition. Nevertheless each theory has its special pericopes for which the proponents of the theory claim that their solution is obviously the best, while there are other pericopes where this solution is less obviously the best, and prima facie another solution would fit the facts equally well or perhaps even more easily. In the case of this pericope, however, it is claimed as primary evidence for their own theory by proponents of each of the principal solutions to the synoptic problem. The relationships between the synoptic passages have been claimed as evidence by proponents of the Griesbach theory, as evidence of Mark-Q overlap (in which Luke is closer than Matthew to Q) by proponents of the Two-Source theory, and by single-source theorists as evidence of editing of Mark by Matthew and Luke successively.

The basic relationship between the three synoptic texts is shown in Table 1 .

One of the chief arguments of proponents of the Griesbach theory is the claim that Mark combines Matthew and Luke by zigzagging between them: when Mark departs from the order they share, Mark follows first one and then the other (see C.1). This is claimed to be exemplified here. So, it is dubiously claimed, Mk 3:22b agrees with the order of words in Luke against Matthew. Then Mk 3:25 agrees with Matthew (there is nothing corresponding in Luke). Mk 3:26 agrees with both. Still, after a gap, Mk 3:27–8 agrees with Matthew. Finally, Mk 3:29 corresponds to Lk 12:10b (the aorist of the verb blasphēmeō, eis to pneuma to agion). The zigzag is, however, in this case hard to sustain. In fact Mark shares overwhelmingly with Matthew, never in this passage with Luke, though there are

Table 1. Relationship between synoptic texts

occasional elements in the triple-tradition verses where Mark is closer to Luke than to Matthew. In Mk 3:22b the phrases are indeed in the Lukan order (Beelzebul first, not second as in Matthew), but the relationship between the verses is more easily explained as independent improvement by Matthew and Luke of Mark's clumsy double-phrase. In Mk 3:29 there are equally strong correspondences with Matthew. The argument is perhaps plausible, but by no means compelling.

On the Two-Source theory it is considered a passage of Mark-Q overlap. It is one of the five principal passages accepted as such by Streeter (along with the preaching of John the Baptist, the temptations, the mustard seed, and the commissioning of the disciples, see C.2). Matthew and Luke share 6½ verses absent from Mark, and in the triple-tradition verses there is persistent minor agreement between them against Mark. Some explanation must be given of these agreements, and if the Mark-Q overlap theory makes sense at all, it is a possible candidate as the explanation. Therefore a three-stage process is postulated: first comes Mark, then Q develops this tradition, then Matthew and Luke independently combine this Q tradition with their version of Mark.

In order to show, however, that at least in this case Mark-Q overlap is the most economical explanation it is necessary to show that Luke's version is the more primitive, and Mark has subsequently been edited by Matthew. So advocates of the Mark-Q overlap claim that Matthew has taken verses from various places in Q (the elements occur in three different sections of Luke) to make a skilfully unified composition, but that the elements of this composition are still visible in their original form in Luke. Advocates of this theory are posed the formidable task of showing that underlying Luke and/or Matthew is a unified theology or style that is distinct from that of the final authors, and can be considered characteristic of Q. So Kloppenborg (1987: 121–7) argues vigorously that Luke is the more primitive version, more coherent than Matthew's form. Luke's parable of the stronger man in 11:21–2 evokes warfare, which better fits the mention of ‘kingdom’ in the previous verses than does Matthew's household burglary. Matthew would then have adopted the earlier verses from Q, but reverted slavishly to Mark for the burglary. After the little Q-saying of Luke 11:23 , Luke would have added another passage (originally separate in Q, and used by Matthew at 12:43–5 ) to stress that mere expulsion of the evil spirit is not enough without a further positive response to the kingdom. For Kloppenborg both Mark and Q versions have the same origin: ‘the starting-point for this complex of Q-sayings is the traditional Beelzebul accusation and its refutation in Mark 3:20–6 ’ (ibid. 127). But Q has enlarged the scene in two ways, first by attributing the accusation not (as does Mark) to the scribes from Jerusalem but to ‘your sons’ in general, and secondly by applying Jesus' threat not only (as Mark) to those who accuse Jesus of complicity with Beelzebul, but to all who oppose Jesus (Lk 11:23–6 ).

Opponents of the Mark-Q overlap must show that the Matthean passage is so typical of Matthew that there is no trace of any written source other than Mark. So Goulder (1974: 332) maintains that the changes are best explained as introduced first by Matthew. He points out that in Mt 12:25 the balance of two similarly shaped phrases is a typically elegant Matthean improvement on Mark's rough phrase. Goulder then argues phrase by phrase that the expansions of Mark are so characteristic of Matthew that it would be a mistake to postulate any Q. Particularly the rhythm of vv. 31, 33, 35 is typical of Matthean formations, and such antitheses as ‘gather/scatter’, ‘good/bad’. It is then necessary to argue that Luke can best be explained as derived from Matthew. To begin with, it is pointed out that Luke often breaks up longer Matthean sections, and that the method of so doing is in this case typical of Luke (see C.4).

On the other side it is argued that Luke, with his stress on the Spirit, would never have substituted ‘finger of God’ (Lk 11:20 ) for Matthew's ‘Spirit of God’ if he had been following Matthew. This is taken as an indication that ‘finger of God’ must have been the original form in Q (e.g. Stanton 1992a : 177 n.3); to which Goulder (1989: 504) replies that this allusion to Moses' miracles in Ex 8:15 is typical of Luke, and that ‘Spirit’ occurs only twice in Luke's accounts of Jesus' teaching.

In this particular case it is unlikely that either side will finally convince the other. The particular question must be judged in function of the more general question whether a Mark-Q overlap makes sense, and particularly whether this overlapping Q is so close to Mark that some literary dependence of Mark on Q would need to be postulated. This in turn would raise the question of why Mark omitted so much of Q.