When Matthew and Luke share material that is lacking in Mark, the common assumption is that Matthew and Luke depend on a now-lost source that scholars call Q (see Stephen J. Patterson, “Yes, Virginia, There Is a Q,” BR 11:05). When, as in this case, the shared material is presented differently, the question that immediately arises is, Which gospel most accurately represents the hypothetical source Q?


The difference in duration is explained by the principle that a day in the life of an individual is equivalent to a year in the life of a people. This principle emerges from a number of texts, e.g., “According to the number of the days in which you [the spies] spied out the land, 40 days, for every day a year, you [the people] shall bear your iniquity 40 years” (Numbers 14:34); “I assign to you a number of days, 390 days, equal to the number of years of their punishment, and so you shall bear the punishment of the house of Israel. When you have completed these, you shall lie down a second time, but on your right side, and bear the punishment of the house of Judah: 40 days I assign you, one day for each year” (Ezekiel 4:5–6).


In Exodus 17:7 Moses also names the site Meribah (from the verb “to find fault”).


Where exactly did the second test take place? The Greek has to pterygion tou hierou (Matthew 4:5), which is variously translated as the “pinnacle of the temple,” the “parapet of the temple” or the “highest point of the temple.” This is the only appearance of the term pterygion (literally, “winglet,” or “little wing”) in the New Testament.

A fundamental theme of the Exodus is Yahweh’s protection of Israel. In the most dramatic and memorable of these texts, Yahweh is compared to an eagle protecting its nest: “As an eagle stirs up its nest, and hovers over its young, as it spreads its wings, takes them up, and bears them aloft on its pinions, the Lord alone guided them…he set them atop the heights of the land” (Deuteronomy 32:11–13). Apparently the term “winglet” was used in the second test because it evoked divine protection. For the rabbis, the Temple was the center of God’s protection because the divine presence was concentrated there.


The Mishnah is a collection of rabbinic oral teachings written down in about 200 A.D.


On the meaning of leb, see Robert North, “Did Ancient Israelites Have a Heart?” sidebar to “Ancient Medicine” BR 11:03.


It was the arrest of John the Baptist that brought Jesus back to Galilee. See my “Why Jesus Went Back to Galilee,” BR 12:01.



André Feuillet, “Le récit lucanien de la tentation,” Biblica 40 (1959), pp. 613–631. Similarly, Joseph A. Fitzmyer, The Gospel According to Luke I-IX, Anchor Bible 28 (Garden City, NY: Doubleday, 1985), pp. 507–508; W.D. Davies and Dale C. Allison, A Critical and Exegetical Commentary on the Gospel According to Saint Matthew (Edinburgh: T & T Clark, 1988), vol. 1, p. 364.


See in particular Jacques Dupont, “L’arrière-fond biblique du récit des tentations de Jesus,” New Testament Studies 3 (1956–57), pp. 287–292.


Birger Gerhardsson points out that the “the verb ain-nun-hey (pi’el) not only has the general meaning ‘to humble’ but also the special meaning ‘to humble (oneself) by fasting’” (The Testing of God’s Son (Matt 4:1–11 & Par) [Lund: Gleerup, 1966], pp. 41–42).


For a variety of opinions on where the second test took place, see Joachim Jeremias, “Die ‘Zinne’ des Temples (Mt 4.5; Lk 4.9),” Zeitschrift des deutschen Palästina-Vereins 59 (1936), pp. 195–208. The Greek to pterygion tou hierou is not an architectural term. Apparently “wing” was chosen precisely because it evoked divine protection (Gerhardsson, Testing of God’s Son, p. 59).


Dupont, “L’arrière-fond biblique,” p. 296.


The discovery of this principle must be credited to Gerhardsson, who also tracked down the interpretation in Mishnah tractate Berakoth (Testing of God’s Son, pp. 71–76). The skepticism shown by Davies and Allison (Matthew, p. 353) is unjustified. Doubts regarding other hypotheses of Gerhardsson regarding the use of the Shema as a structural principle elsewhere in Matthew’s gospel in no way impugn the accuracy of his insight here.


Mishnah, Berakoth 9.5. Translation adapted from Herbert Danby, The Mishnah (Oxford: Oxford Univ. Press, 1933), p. 10.


For example, “I form light and create darkness. I make weal and create woe” (Isaiah 45:7) simply means that God is all powerful. See in particular Gustave Lambert, “‘Lier-délier’: L’expression de la totalité par l’opposition de deuz contraires,” Vivre et Penser 3 (1943–44) [=Revue biblique 52], pp. 91–103.


BT, Berakoth 61b, trans. Maurice Simon, in Babylonian Talmud, vol. 1, Seder Zerai’im (London: Soncino Press, 1948), p. 386.


Simon, BT, Berakoth 61b, p. 385.


This is a further argument for the originality of the structure of Matthew’s narrative.


This approach was first developed by Hans Holtzmann (Hand-Kommentar zum Neuen Testament, 3rd ed. [Tübingen, 1901], vol. 1, bk. 1, pp. 45–48), who found the reality behind the three tests (in the order of Matthew) in three episodes of chap. 8 of Mark’s gospel (vv.1–9, vv. 11–13, vv. 31–33). Some 40 years later, Herbert Preisker revived this line of thinking, but he identified the three tests with episodes in John’s gospel (John 4:31–34, 6:15, 7:2–6) (Preisker’s study was published in 1939 in an obscure festschrift that was not available to me; his conclusions were reported in Philippe Menoud, L’évangile de Jean d’aprés les recherches récentes [Neuchätel-Paris, 1947], p. 29). Raymond Brown (“Incidents that Are Units in the Synoptic Gospels but Dispersed in St. John,” Catholic Biblical Quarterly 23 [1961], pp. 152–155) preferred John’s version of the multiplication of the loaves (6:26–34) as the real parallel to the first test.

All these suggestions are far wide of the mark. The two points of contact with the multiplication of the loaves are (a) a miracle, and (b) bread. These are purely material. Jesus refuses a miracle in one case but performs one in the other, which involves the multiplication of bread not the transformation of a stone. Brown goes a step further by evoking the discussion that followed the miracle, in which a contrast is drawn between ordinary bread and God’s teaching as the bread of life (John 6:22–59). Jesus, Brown maintains, is virtually commenting on Deuteronomy 8:3. True, but this proves no more than a common use of Deuteronomy 8:3. Moreover, there is no hint in the multiplication of the loaves that Jesus is being tested. On the contrary, he exhibits sovereign power.


The apparent ambiguity of hoi par’ autou, literally, “those with him” (Mark 3:21), has been exploited by those who are unhappy with the negative picture of Jesus’ family painted here. (Presumably the same motive explains the omission of this incident by both Matthew and Luke. The fabrication of this event at a later stage, when the mother and brothers of Jesus were prominent disciples [Acts 1:14], is impossible.) In terms of Marcan narrative technique, however, it is certain that “those with him” are those later explicitly identified as “his mother and his brothers” (Mark 3:31). See the discussions in Vincent Taylor, The Gospel According to Mark (London, 1963), pp. 235–237; and in Brown et al., eds., Mary in the New Testament: A Collaborative Assessment by Protestant and Roman Catholic Scholars (Philadelphia, 1978), pp. 55–56.


Crucial to the understanding of this episode is Mark’s technique of structuring a narrative in an A-B-A’ pattern, in which two parts of a single story are separated in order to bracket an entirely different narrative (as in Mark 5:21–43, 6:7–30, 11:12–25, 14:1–11). In this case, the intrusive element (B) is made up of the two accusations of the scribes and Jesus’ responses to them in inverse order (Mark 3:22–30). If we leave these verses out of consideration, we find a coherent narrative.


The fact that Mark inserted the accusations of the scribes—“He is possessed by Beelzebul!” and “It is by the prince of demons that he casts out demons” (Mark 3:22)—at precisely this point clearly indicates how the evangelist understood “he is out of his mind.” Further, krateô means “arrest” in Mark 6:17, 12:12, 14:1, 44, 46, 49, 51. These two points effectively refute attempts to water down the obvious meaning of the accusation, such as that by Marie-Joseph Lagrange, who wrote that Jesus was in such an exalted state that he forgot to eat his meals (Évangile selon saint Marc [Paris, 1911], p. 64), or Henry Wansbrough, who argued that it was the crowd that was out of control with enthusiasm (“Mark iii.21: Was Jesus Out of His Mind?” New Testament Studies (1971–72), pp. 233–235.


Like Mark 3:20–21, this episode is not a late redactional creation. It antedates the composition of Luke 1–2 because (a) it does not know of the virgin birth—note “his parents” (v. 43) and “your father and I” (v. 48; contrast “the supposed son of Joseph” in Luke 3:23); and (b) the parents have not profited (v. 50) by the revelations of Luke 1:32–35, 2:11, 17, 19 regarding the identity of Jesus. See in particular Baas Van Iersel, “The Finding of Jesus in the Temple: Some Observations on the Original Form of Luke ii 41–51a, ” Novum Testamentum 4 (1960), pp. 161–173. This is not to say, however, that the editor of the gospel did not make additions.

Commentators are extremely coy about the historicity of this event. Fitzmyer is typical in concentrating exclusively on the role of the episode in Luke’s framework (Luke I-IX, pp. 434–39). As far as I am concerned, there is nothing implausible in the basic elements of the story, particularly if it is conceded that the source may have heightened the dramatic effect of certain aspects, e.g., lost for three days rather than a couple of hours.


On the question of authenticity, see Norman Perrin, Rediscovering the Teaching of Jesus (New York, 1976), p. 144.


As regards the reality behind the second test of Jesus (Matthew 4:5–7), Holtzmann suggests Mark 8:11–12, while Preisker and Brown propose John 7:1–4. Neither hypothesis has any real merit.

In Mark 8:11–12 the Pharisees demand a sign from heaven, but the only point in common with the second test of Jesus is the fact that they were “testing” him. The episode takes place in Galilee, and no one would have ever thought of it in terms of the second test were it not the immediate sequel of the multiplication of the loaves, with which Holzmann, as we have seen, mistakenly associates the first temptation.

In John 7:1–4 Jesus’ skeptical brothers suggest that he should cease embarrassing the family in Galilee and go and perform his miracles in Jerusalem. The mention of the Holy City does furnish a link with the second test, but that is all. There is a radical difference between Jesus working a miracle for another’s benefit and God working a miracle in favor of Jesus by saving his life.


On the historicity of this episode, see in particular John B. Tyson, “Jesus and Herod Antipas,” Journal of Biblical Literature 79 (1960), pp. 239–246.


Fitzmyer, Luke I-IX, p. 779.


The references are assembled in Fitzmyer, The Gospel According to Luke X-XXIV, Anchor Bible 28A (Garden City, NY: Doubleday, 1985), p. 230.


The full references are Mark 13:9–13 = Matthew 24:9–14 = Luke 21:12–19.


Holtzmann’s suggestion that it corresponds to Peter’s reaction to the first prediction of the passion (Mark 8:31–33) is based purely on the material correspondence of the phrases “Begone, Satan!” and “Get behind me, Satan!” Apart from the name Satan, the two narratives have nothing in common.


This passage was first proposed by Raymond Brown.


This conclusion is made all the more probable by the fact that Jewish theologians created the same sort of story. According to the Babylonian Talmud, God tested Abraham with many trials. In one such trial, Satan and the patriarch trade texts from Scripture, exactly as Satan and Jesus do in the gospels (BT, Sanhedrin 89b, which is quoted in Davies and Allison, Matthew, p. 352).