By The Finger of God

A Literary Analysis of the Beelzebul Controversy

By Vlad



This paper, as part of a larger study of the Beelzebul controversy, will examine Matthew 12.25-28, Luke 11.17-20, and Mark 3.23-26 in order to distill from the synoptists the theology and self-understanding of Jesus. It is my contention that the most original dominical saying at the apogee of the encounter, in Matthew 12.28 // Luke 11.20, includes the phrase “the finger of God.” I will present an argument for this conclusion and discuss its significance.


I take for granted a literary relationship between the synoptic Gospels, though the same conclusions could be reached assuming only common oral sources. My allusions to the hypothetical document Q should be taken to refer broadly to a sayings source used by Matthew and Luke and not an acceptance of the entire reconstructed document by the International Q Project and others. I also maintain a high view of Scripture and find no tension in this vis-à-vis the two-source solution to the synoptic problem.


Establishing the Reliability of Luke

It should not be overlooked that the close agreement in Matthew and Luke starts well into the pericope with the beginning of Jesus’ dialogue. To facilitate a close scrutiny of the parallel passages they are included below, with the specific portions of the texts where they disagree and which will be addressed marked.


            Matthew 12.25,26
He knew what they were thinking and said to them, “Every kingdom divided against itself is laid    waste, and no city or house divided against itself will stand. If Satan casts out Satan, he is divided          against himself; how then will his kingdom stand? (NRSV)


            Luke 11.17,18
But he knew what they were thinking and said to them, “Every kingdom divided against itself        becomes a desert, and house falls on house. If Satan also is divided against himself, how will his        kingdom stand? —for you say that I cast out the demons by Beelzebul. (NRSV)


The strong editorial agreement in these two accounts contra Mark (3.23) in addition to the close agreement in Jesus’ words indicates that both are following the same source, one which is both more detailed and makes a stronger point than Mark. Mark gives no back-story and little of the drama preserved in the tradition underlying Matthew and Luke. Yet it is the differences between them that provide an interesting study and indeed provide the data crucial to working out Jesus’ wording in the climax of this story.


Though masked by the translation, Matthew and Luke use different words for “thinking.” I am not convinced that there is any significance to Matthew’s enthumeseis (ἐνθυμήσεις), that the word is “characterized by godless evil and foolishness” in the New Testament. There are only three other occurrences of the word (Matthew 9.4; Acts 17.29; Hebrews 4.12), and in each the stated connotation is not apparent unless you are looking for it. In looking at other relevant New Testament texts I find no reason to think it was any more negative than Luke’s dianoemata (διανοήματα). They appear entirely synonymous. If, however, enthumeseis is significant in this way it adds weighty evidence for Matthew’s theological bent, part of what I am trying to establish.


More important is that Luke 11.17 is the only occurrence of dianoema. Clearly, whether or not the word choice is significant, Matthew preferred to use the more common synonym, one that was already a part of his vocabulary and that he had used in the very same context of Jesus’ knowing the thoughts of his opposers (9.4). While this in itself is not decisive, it points us in the direction of Matthean change. Conversely, Luke has a word he uses nowhere else, that is not well-attested in the relevant literature, and for which there were more common synonyms. The preponderance of evidence shows that Luke preserves the more original word.


Another disagreement is in the idea of houses collapsing on one another in Luke’s “house falls on house” (οἶκος ἐπὶ οἶκον πίπτει) versus Matthew, perhaps following Mark, who writes of a house unable to stand. He also sweetens the aphorism by adding “city,” providing an even more dramatic backdrop for Jesus’ devastating logic. Matthew then extends the passage again in 12.26, “Satan casts out Satan.” We are seeing a clear tendency to conflate or add to the narrative in this portion of Matthew’s gospel. Luke, however, lacks this kind of flourish, displaying an earlier version of the saying and a fidelity to his source.


The causal clause “for you say that I cast out the demons by Beelzebul” also points to this conclusion. Though it could be explained as explanatory redaction, an effort at clarity by the narrator, it is hard to see why any of the evangelists would insert this comment, much less Luke. Satan/Beelzebul is well in view in the immediately surrounding text—why could clarification be needed? Luke would rather clean up bad writing than add to it, and this redundancy should have disappeared at his hand. Not only, then, is Luke here closer to the original than Matthew, but he is evincing a propensity to preserve his exemplar as is. Further, the linguistic form of the clause suggests that this exemplar is a faithful presentation of the underlying oral tradition; oral communication is naturally full of redundancy. It is therefore curious that Robinson’s (et al.) Q does not have this phrase. Nevertheless, the translation in Miller’s gospel compilation has it, as well as the translation of Q specialist Burton Mack.


Close analysis has revealed both the earlier, underlying form of the story, and the proclivities of the individual evangelists. In this portion of the narrative Luke follows Q very closely and has resisted making changes to the tradition even when his literary sensibilities would seem to make him likely to do so. Luke’s hapax legomenon and redundant clause indicate not only that he is following his source scrupulously, but that we have at least some reason to think that he is very close to Jesus’ original words. Conversely, Matthew leans toward editing his source for literary or theological reasons.


Close Wording and Textual Variants


To proceed, we need to look more closely at the next couple of verses where we find even closer verbal agreement, the significance of which will be discussed following.


Matthew 12.27,28

kai ei ego en Beelzeboul ekballw ta daimonai, hoi huioi humon en tini ekballousin; dia touto autoi kritai esontai humon. ei de en pneumati theou ego ekballo ta daimonia, ara ephthasen eph humas he basileia tou theou.


            Luke 11.19,20
εἰ δὲ ἐγὼ ἐν ‹Βεελζεβοὺλ ἐκβάλλω τὰ δαιμόνια, οἱ υἱοὶ ὑμῶν ἐν τίνι ἐκβάλλουσιν; διὰ τοῦτο             ˜αὐτοὶ ὑμῶν κριταὶ ἔσονται˜. εἰ δὲ ἐν δακτύλῳ θεοῦ °[ἐγὼ] ἐκβάλλω τὰ δαιμόνια, ἄρα ἔφθασεν             ἐφ’ ὑμᾶς ἡ βασιλεία τοῦ θεοῦ.


The agreement here is striking in its exactness even in word order, with relatively few textual variants. The single, pivotal exception will be considered shortly. It is true that corruption due to scribal harmonization of the two accounts could produce this indistinguishable similarity, but that kind of unwarranted skepticism would throw out the synoptic gospels altogether.  What militates against this possibility, besides general text-critical principles, is that this is not the kind of devotional or confessional passage that is likely to be so memorable as to be conflated.


The minor differences in how they begin (If/Now If, in the New Revised Standard Version) should be ignored, as the Greek is essentially equivalent. We could speculate that Luke (δὲ) saw more contrast than did Matthew (καὶ), or Matthew saw δὲ in Q a simple conjunction. The variants on Beelzebul are inconsequential spelling differences, indicative perhaps of the provenance of the archetypal manuscripts. The variance within the phrases rendered ‘they will be your judges’ (αὐτοὶ κριταὶ ἔσονται ὑμῶν, Matthew) is to be expected given the flexibility the language. The words themselves are the same and so the New Revised Standard Version translators took Matthew and Luke to be identical, the difference in their final forms in the current critical editions of the Greek New Testament being a function of their individual transmission.


The explanation for the omission of Luke’s ego (ἐγὼ) in some manuscripts is self-evident: since the subject of the verb is clear from its inflection the first-person pronoun is unnecessary, and was therefore easily overlooked by scribes at various points in the transmission process. For this reason the weight of the manuscripts that omit it have no bearing. Consider that while ego serves a purpose in the first verse (of both accounts), stressing Jesus’ exorcism over others’, it does not do so in the second verse making accidental omission more likely.


My purpose in dragging this out is to underscore that textual corruption has not obscured what the evangelists wrote and that Matthew and Luke must both be following the tradition perfectly to get this kind of agreement. This identicalness heightens the importance of their divergence. Matthew has “the Spirit of God” as the means of exorcism, Luke has “the finger of God.” Why do they differ? What did Jesus actually say?


Ipsissima Verba Jesu


In determining the more primitive version, Matthew’s narrative leading into the pericope and his conclusion to it weigh heavily. In 12.15-18 Matthew has Jesus fulfilling the words of Isaiah 42, the servant upon whom is active God’s spirit. Concomitant, and even more telling, is Matthew’s arranging of dominical sayings to condemn the Pharisees. He has inserted two paragraphs of denunciation—to him part of the refutation of the Pharisees’ sacrilege—between the pericope under consideration and the ‘wandering spirit’ (Matthew 12.43-45 // Luke 11.24-26), with which Luke follows immediately. Matthew ends the Beelzebul encounter, not with the rather soft Lukan “Whoever is not with me is against me…” (11.26), but with the severest of strictures: “Whoever speaks against the Holy Spirit will not be forgiven, either in this age or in the age to come” (Matthew 12.32, NRSV).

For this condemnation to stick, it must be clear to Matthew’s audience that the Pharisees have indeed spoken against this pneuma. “Finger of God” is a reference to God’s pneuma, but it is oblique and lacks the directness of Matthew’s pneumati theou (πνεύματι θεοῦ), Jewish audience or not. As I will argue, the finger of God is a much more nuanced allusion to God’s involvement. So, for Matthew, God’s finger would not do as the means of exorcism.  He must have the pneuma there and must therefore emend the tradition if it is not so. This, together with Matthew’s redactional activity thus far, makes his text suspect.

Luke might well have an agenda also. Since he has just concluded relating Jesus’ teaching on prayer with “how much more will the heavenly Father give the Holy Spirit to those who ask him” in 11.8—immediately preceding the Beelzebul encounter—he would be predisposed to have Jesus’ exorcism accomplished by means of this spirit. So it is exceptionally hard to explain why Luke would write this odd anthropomorphic saying if he had a much easier reading in front of him, one that so suited him. Attempts to make sense of this fail utterly. Turner, after discounting one unsatisfactory hypothesis, says:


            The shift in terminology (a clear reference to Ex 8:19 [LXX Ex 8:15]) is probably in the interest     of Luke’s prophet like Moses christology but still refers to the Spirit; cf. the parallel term “the           hand of the Lord,” which was interpreted to refer to the Spirit.


This does not satisfy. Luke, despite his facility with the Scriptures, would not have changed the text to such an archaic phrase; certainly not for Theophilus. It is Matthew who makes the change avoiding a difficult reading, just as he did with “—for you say that I cast out the demons by Beelzebul.” Luke, then, has the earlier expression.


In Luke 11.14-21 the evangelist has demonstrated great fidelity in his work. His text is in close agreement with the independent Matthean tradition. Additionally, he has preserved an obscure word, a redundant clause, and a difficult Hebraism, another indicator that we are close to the very words of Jesus.


The Finger of God Saying


We must now turn our attention to the climax of Jesus’ response to the Beelzebul accusation. Turner, as quoted above, is right to connect finger of God to hand of God. This phrase, in turn, is part of the arm of God trope, which stands for God’s power.


The symbol of the arm outstretched, or made bare (much the same idea in view of E dress), is used especially of the Lord to portray his mighty acts, referring often to the deliverance of Israel from Egypt (Ex. 6:6, etc.), also to other acts of judgment or salvation evidenced or sought (Is. 51:9; Ezk. 20:33).[…] The powerful arm of the Lord is contrasted with the puny arm of man, ‘an arm of flesh’ (2 Ch. 32:8).


But the idea of strength does not exhaust the metaphor of the hand of God which can have the sense of ‘manipulation,’ of activity and work. For example, Abel’s blood, or life, was ‘taken’ by the hand of Cain (Genesis 4.11). The finger of God goes further in this direction. It is not simply metonymic for hand but synecdochic, and therefore does something more. With this more specific metonym we are at an expression once removed from hand as a figure of speech related to God’s power. We can say this because of the nature of this kind of idiomatic expression. Wheels works better for car than most other parts of an automobile because the wheels are critical for what it does (compared to the doors at least). Wheels do not indicate the heaviness of the machine or its size, but its locomotion, its ability to transport. A finger’s association to the hand is not in the strength of the hand, but rather its manipulation, its discrete action. The finger is an instrument of creative work, of a more specialized use than an arm or a hand; an action on a narrower, more specific scale.


Returning, then, to the primary texts alluded to in the climax of the Beelzebul controversy, we can see the nuance of finger of God. The Jews would not have missed Jesus’ clear allusion to the Torah. At Exodus 8.19 the Egyptians confess that the gnats are brought on by God’s finger. To be sure, this does point to God’s power, the strength in God’s hand to act. There is a connection to Exodus 7.5 where Yahweh promises this kind of recognition from the Egyptians. He would use his hand to bring the Israelites out of Egypt. This, however, was a general statement and was clearly not meant to predict the response of specific Egyptians, much less the magicians. The finger of God in 8.19, then, does not stand for the hand of God of 7.5. Further, the Egyptian magicians were not awed merely by the power of the act, but by their inability to reproduce it. This was beyond their magic and revealed that a true Deity was at work.


This might be overlooked were it not for the other occurrences of this idiom. In Exodus 31.18 the mere power of inscribing the tablets of the Decalogue was not at issue. What made them special was the very touch of God upon them. Deuteronomy 9.9-10 makes the point even more starkly:


            When I went up the mountain to receive the stone tablets, the tablets of the covenant that the Lord            made with you, I remained on the mountain for forty days and forty nights; I neither ate bread nor      drank water. And the Lord gave me the two stone tablets written with the finger of God; on them        were all the words that the Lord had spoken to you at the mountain out of the fire on the day of      the assembly. (NRSV)


The covenant had been made by Yahweh, and so had the tablets. The emphasis is on God’s own action and interaction with his people. The ancientness of the trope, intimated above, should also be noted at least in brief. Even under the documentary hypothesis it must be recognized as early. That it is both ancient and not otherwise appropriated in the literature makes its usage in our text important.


Conclusions About Jesus


We are now in a position to synthesize the preceding. The finger of God saying is so infrequently attested and so otherwise inexplicable that it must be original. Combining this with the setting of Jesus’ saying and the well-known authenticity of his Kingdom of God theme, we have a sure data point from which to sketch out some implications for the character and activity of Jesus of Nazareth. When Jesus said that he expelled the demons, not by Beelzebul, but by the finger of God, he was saying that God’s direct touch was involved. Yahweh, the redeemer of Israel, was working through Jesus in an intimate and immanent way. The Kingdom of God was truly upon them. It is clear why the Matthew and Mark follow with the culpability of Jesus’ opposers in failing to recognize the source of his actions. The pagan magicians recognized God’s ‘handiwork’ when they saw it, and so should have the scribes and Pharisees. To attribute such an expression of Yahweh’s own touch to a demon is a most severe blasphemy.


Two fundamental things can be said about the historical Jesus. First, his locus was in Judaism and the Tanakh. He was not a moral philosopher, but a Jewish prophet who operated within that ambit. Second, Jesus saw himself as inaugurating God’s kingdom in a real and physical way. It was no abstraction to him. This kingdom was not simply working in the hearts and minds of men—it was working in the world around them.


This is not an equivocation on Markan priority or the two-source hypothesis (or other permutations), which still seem to me to have the greatest explanatory power. However, I think the attempt to finely delimitate and reconstruct Q, much less to divine its community, is highly speculative.

New Revised Standard Version Bible, copyright 1989, Division of Christian Education of the National Council of the Churches of Christ in the United States of America. Used by permission. All rights reserved.

While it is not necessary to posit that Matthew and Luke knew Mark, it is inescapable that Matthew and Luke worked from a common source of some kind. Since Luke seems to have the more primitive reading, as I will attempt to show, I find it very unlikely that Luke used Matthew. This particular Matthew-Luke agreement in the triple tradition seems to me a very strong argument against the Griesbach hypothesis.

Exegetical Dictionary of the New Testament (Grand Rapids, MI: Eerdmans 1990), vol. 1, 454.

In his parallel, Luke uses the far more common ἐπίγνωσις (Luke 5.22).

Precise demonology is irrelevant here, though it looks as if Satan and Beelzebul are one and the same to Jesus.

James M. Robinson, Paul Hoffman, John S. Kloppenborg, The Sayings Gospel Q in Greek and English (Minneapolis, MN: Fortress 2002) 104.

Robert Miller, ed., The Complete Gospels: Annotated Schola’s Version (New York: HarperCollins 1993), 272.

Burton Mack, The Lost Gospel: The Book of Q and Christian Origins (New York: HarperCollins 1994), 90.

I have attempted to reproduce the sigla from the Nestle-Aland 26th in a way that I hope will be obvious.

By this I mean that, while both were looking at the same text, each got a different ‘feel’ from the language. This subjective feel manifested itself in their word choice. This is speculative, but I wish to show that such a minor difference is not only ignorable, but also easily explicable with a number of solutions.

Apparently no difference was found even in emphasis. I am making more of this than I should, especially given the equivocation in the manuscripts, but note that they did find a difference in emphasis in the καὶ and δὲ, so they were not insensitive to it.

Though I am persuaded that the relationship between Luke and Matthew is literary, an ‘oral Q’ could also account for this. It would necessitate a very early, widespread, and exact Greek oral tradition.

Isaiah 42.1 LXX, …τὸ πνεῦμά μου ἐπ’ αὐτόν.

Matthew 12.32, κατὰ τοῦ πνεύματος τοῦ ἁγίου.

M.M.B. Turner, IVP Dictionary of the New Testament (Downers Grove, IL: InterVarsity 2004), 485. I will deal with the background of the expression in the final part of this essay. For now it suffices to say that these connections cannot so easily be made, especially on the basis of a Moses christology.

New Bible Dictionary, Second Edition (Wheaton, IL: Tyndale House 1982), 83.