^{page 317 note 1} The second theme is indicated rather imprecisely by Bultmann, R., Geschichte der synoptischen Tradition (Göttingen, 1958 ⁴), p. 327Google Scholar, ‘sein Weilen im Tempel, das seine religiöse Bestimmung kundtut’. B. van Iersel rightly pointed out this imprecision in ‘The Finding of Jesus in the Temple’, N.T. 4 (1960), esp. p. 168.Google Scholar

^{page 317 note 2} Schürer, E., Geschichte des jüdischen Volkes im Zeitalter Jesu Christi (Leipzig, 1901–1909 ^3–4) II, 496–7.Google Scholar

^{page 318 note 1} Strack, H. L.–Billerbeck, P., Kommentar zum N.T. aus Talmud und Midrach II (Munich, 1924, 1965⁴), 144–7.Google Scholar

^{page 318 note 2} Schürer, , Geschichte, 2, 497 n. 41.Google Scholar

^{page 318 note 3} Rengstorf, K. H., Das Evangelium nach Lukas (Göttingen, 1952 ⁶)Google Scholar(Das Neue Testament Deutsch), p. 51.Google Scholar

^{page 318 note 4} Schürmann, H., Das Lukasevangelium, I (Freiburg/Basel/Wien, 1969)Google Scholar (Herders theol. Komm. zum N.T.), 134.Google Scholar

^{page 319 note 1} Caird, G. B., The Gospel of St Luke (Harmondsworth, 1963)Google Scholar (The Pelican Gospel Commentaries), p. 66.Google Scholar

^{page 319 note 2} Cf. Rengstorf, in Th. Wb. zum N.T., 2, 322Google Scholar. By a ‘round’ number I mean not only that such numbers convey a particular impression of completeness, but also that because of their function they can be used to indicate quantities rather smaller or greater than the ‘round’ figure, i.e. as approximations.

^{page 319 note 3} Radermacher, L., ‘Christus unter den Schriftgelehrten’, Rhein. Mus. 73 (1920), 232–9Google Scholar. Independently of Radermacher, M. Dibelius saw the figure twelve as a round number, but also referred to the sacral character and biological significance of the age of twelve. This explains why the same age also plays a role in legends of the Buddha and Si Osire. Dibelius, M., Die Formgeschichte des Evangeliums (Tübingen, 1959 ³), p. 104.Google Scholar

^{page 319 note 4} See e.g. Kaser, M., Das römische Privatrecht, 11 (Munich, 1959), III and 78–9Google Scholar; Strack-Billerbeck, , 11, 10Google Scholar (ad Mark, v. 42); p. 374 (Google Scholarad John, ii. 1)Google Scholar. For Gortyn, see Delling, G., in RAC 4, 684Google Scholar. For Egypt, Ibid. p. 685. For Rome, Ibid. p. 687. The marriageable age for a man was fourteen according to Talmud, Babylonian, Kiddushin 29bGoogle Scholar. In ecclesiastical sources of the third and fourth centuries it is recommended that boys should be married as adolescentes (Rufinus' translation of νέοı) before their sexual desires reach a climax (Ps. Clem., Hom. 5, 25Google Scholar; id., Ep.ad, Jac. viiGoogle Scholar; Gonst. Ap.Iv, II)Google Scholar. Apparently, marriage between the ages of fourteen and eighteen was regarded as ideal for boys, while in Christian circles also girls were regarded as marriageable at the age of twelve. Indeed in the Protev. Jac. viiiGoogle Scholar, Mary is betrothed to Joseph at twelve, because of her puberty. True, Mary's marriage with Joseph did not lead to a matrimonium consummatum, but in c. x they were certainly regarded by the priest as married: see Strycker, E. de, La forme la plus ancienne du Protévangile de Jacques (Brussels, 1961), p. IIIGoogle Scholar. In Ev. Ps.-Mt. viiiGoogle Scholar, Mary was betrothed to Joseph at the age when other women ‘viros in coniugio suns adeptae’.

^{page 320 note 1} Schürmann, , p. 490.Google Scholar

^{page 320 note 2} Varro, according to Censorinus xiv. 2 and the scholiast on Lucian's Catapl. 1:Εφηβοı καλονταı οí απò ıέ (see Thes. 1. Graecae, s.v. έφηβος). Nilsson, Martin P., Die hellenistische Schule (Munich, 1955), p. 34Google Scholar, writes ‘die Pubertät trat nach der allgemeinen Meinung der Griechen in dem fünfzehnten Lebensjahr ein’, but does not cite any proofs.

^{page 320 note 3} Xenophon, , Cyropaedia, 1, iiGoogle Scholar. 4 and 8 says that the Persians place the boundary between παīδες and έφηβοı at the age of sixteen or seventeen. Cf. p. 322, n. 1.

^{page 320 note 4} Boll, F., ‘Die Lebensalter’, Neue Jahrbücher für das klassische Altertum 15 (1913), 89–145, esp. pp. 115–17.Google Scholar

^{page 320 note 5} Marrou, H. I., Histoire de l'eacute;ducation dans l'antiquité (Paris, 1965 ⁶), p. 161Google Scholar. Marrou mentions five classical authors who go back to Hippocrates.

^{page 320 note 6} Varro, apud Gellius III, x. 9Google Scholar; Augustus apud Gellius xv, vii. 2–3Google Scholar; Plin. Epp. 2, xx. 3–4Google Scholar; Tert. Idol. ix. 9Google Scholar and the commentary on the last passage by Nat, P. G. van der (ed.), Q. S. F. Tertulliani De Idolo-latria (Leiden, 1960), pp. 135–6.Google Scholar

^{page 320 note 7} Aëtius, , v, 23Google Scholar (Diels, Doxographi, p. 435)Google Scholar περ δέ τήν δευτέραν έβδομάδα ννοıα γíνεταı καλο τε καί κακο καί τ δıδασκαλίας ατν See, for a discussion of this passage, Bonhoeffer, A., Epictet and die Stoa (Stuttgart, 1890), pp. 204–7.Google Scholar

^{page 320 note 8} Opif. mundi 102–5.Google Scholar

^{page 320 note 9} Boll, F., ‘Lebensalter’, p. 121.Google Scholar

^{page 320 note 10} Marrou and Nilsson disagree on whether epheboi outside Egypt entered the gymnasium at 14/15 (Nilsson) or at 18 (Marrou). The Icarian epitaph mentioned below seems to support Nilsson. See Nilsson, , p. 36Google Scholar, and Marrou's review of Nilsson, in L'Antiquité Classique 25 (1956), 234–40.Google Scholar

^{page 321 note 1} Jitta, A. N. Zadoks Josephus, in W. J. Verdenius et al., Antieke Jeugd (s. 1., 1968), p. 18.Google Scholar

^{page 321 note 2} Peek, W., Griechische Vers-Inschriften, I (Berlin, 1955), 36Google Scholar, no. 119. The important place of έφηβ(ε);íα in the life of boys and their parents is the subject of Griessmair, E., Das Motiv derGoogle Scholar mors immatura in den griechischen metrischen Grabinschriften (Innsbruck, 1966)Google Scholar (Commentationes Aenipontanae XVII), pp. 55–60.Google Scholar

^{page 322 note 1} Cyropaedia 1, ii. 8Google Scholar. According to Xenophon's account, Persians were παīδες until their sixteenth or seventeenth year, έφηβοı until their twenty-sixth or twenty-seventh, τέλεıοı άνδρές until their fifty-first or fifty-second, and γεραíτεροı from fifty-two. Cf. p. 320, n. 3.

^{page 322 note 2} Diog. Laert., x, 2. Diogenes is probably citing Heraclides Lembus. Apollonius of Tyana is also reputed to have begun to enjoy the higher educational subjects of rhetoric and philosophy at the age of fourteen, that is the age at which other boys were only beginning their secondary education (Philostr. I, vii).

^{page 322 note 3} Marrou, Histoire, p. 310:Google Scholar ‘II (l'enseignement philosophique) suppose, au départ, un étudiant ayant achevé sa formation secondaire.’

^{page 322 note 4} Völter, D., Die evangelischen Erzählungen von der Geburt and Kindheit Jesu kritisch untersucht (Strassburg, 1911), p. 77.Google Scholar

^{page 322 note 5} Burrows, E., The Gospel of the Infancy (London, 1940), pp. 23 and 3.Google Scholar

^{page 322 note 6} Ant. ix. 6, 230Google Scholar, already cited by Wettstein, ad Luke ii. 47:Google Scholar σνεσıς δέ ο καάα τήν ήλıκίαν έφετο ατ, το δέ τατης μέτρου πολ κρεíττω καì πρεσβυτέραν δıεδεíκνυε τατης τήν περıουσíαν έν ταīς παıδεíαıς

^{page 322 note 7} I (= III) Kings ii. 12. The same tradition occurs in Eupolemus (second century B.C.), apud Eus., Praep. Ev. 9, 30Google Scholar, and in a number of patristic authors mentioned by Cotelier, J. B.ad Const. Apost. 11Google Scholar, I (in the Clericus edition of 1724, I, 216).

^{page 322 note 8} Ignatii et Polycarpi epistulae ed. Zahn, Th. (Lipsiae, 1876), p. 176Google Scholar. Cf. p. 323, n. 5.

^{page 323 note 1} Ignatius, ad Magnesios, long recension, iii. IGoogle Scholar; Ps.-Chrysostom, (PG 55, 567 and 56, 43)Google Scholar; Sulpicius Severus (PL 20, 128)Google Scholar; the Syro-hexaplaric translation of Susannah (in Walton's, Polyglot, London1657, vol. Iv)Google Scholar; this version of Susannah 1 (= Daniel xiii. t) runs: ‘Cum esset Daniel annorum duodecim, vir erat nomine Joacim. (2) Qui uxorem habebat Susannam…; The Thousand and One Nights, 394 (tr. Littmann, E., Leipzig, 1928, III, 528).Google Scholar

^{page 323 note 2} Ps.-Callisthenes, , Vita Alexandri, ed. Thiel, H. van (Darmstadt, 1974), 1, xiv(–xvi).Google Scholar

^{page 323 note 3} Midrash Shemoth Rabbah (eleventh/twelfth century?) v. 2Google Scholar (translated by Lehrman, S. M. in Midrash Rabbah, ed. Freedman, H. and Simon, M., 111, Exodus (London, 1951), p. 82).Google Scholar

^{page 323 note 4} Herodotus, III, iiiGoogle Scholar, varia lectio cited by Wettstein, ad Luke ii. 42.Google Scholar

^{page 323 note 5} Epistola Mariae Castabalensis, ed. Zahn, (cf. p. 322, n. 8), p. 176:Google Scholar Σολομν δέ ό σοφòς δυοκαíδεκα τυγχάνων έτν συνκε τò μέλα τς άγνωσíας τν γυναıκν έπì τοīς σφετέροıς τέκνοıς ήτημα ς πάντα τòνλαòν έκστναı έπì τ τοσατ το παıδòς σοφíα καì φοβηθναı οχ ς μεıράκıον, άλλ' ς τέλεıον ἃνδρα. Origen's remark quoted by Smith, H., Ante-Nicene Exegesis of the Gospels I (London, 1925), 273–4:Google Scholar ‘He (Jesus) did not display wisdom beyond His age, but at the time when even in us reason is wont to be completed through judgment, i.e. at the twelfth year’, seems to be wrong and conflicting with his statement in Hom. XIX in Lucam in Werke IX (Berlin, 1959)Google Scholar (GCS), p. 115Google Scholar; πρò γρ έτν δδεκα Ψıλòς ἃνθρωπος ο χωρεī πληρóτητα σοφíας

^{page 323 note 6} To my knowledge, Erasmus, in his Paraphrases, is the only expositor to draw attention, at the words, ‘when he was twelve years old’, to the boundary between pueritia and ephebia though in a sense different from that given above: ‘Porro quum iam accedentibus annis firmior met pueritia, ad ephebiam accedens, annos nato duodecim…’

^{page 324 note 1} ‘Diem unum iter fecerant, altero remensi erant idem iter, tertio demum quaesitum inveniunt.’ Grotius was not the first to suggest this interpretation, which had already been given by Brugensis, Fr. Lucas (1549–1619) in his Commentaries in sanctum J.C. evangelium secundum Lucam (Antverpiae, 1712)Google Scholar, ad loc. Lucas Brugensis adds to his interpretation: ‘Sic recte Euthymius distinguit dies’. For Euthymius Zigabenus' commentary, see Migne, , PG 129, 898.Google Scholar

^{page 324 note 2} Holtzmann, H. J., Die Synoptiker (Tübingen/Leipzig, 1901 ³)Google Scholar (Hand-Commentar zum N.T., 1), p. 322.Google Scholar

^{page 324 note 3} Walker, N., ‘After three days’, N.T. Iv (1960), 261–2.Google Scholar

^{page 324 note 4} Radermacher, L., ‘Christus unter den Schriftgelehrten’, Rhein. Mus. 73 (1920), 232–9Google Scholar. Laurentin, R., Structure et théologie de Luc I–II (Paris, 1957), p. 212, no. 335Google Scholar, states wrongly that Radermacher's article appeared in Hermes. The same mistake occurs, oddly enough, in Neirynck, F., Evangelium Lucae (Louvain, 1966)Google Scholar (Studiorum N.T. auxilia 3), p. 106.Google Scholar

^{page 324 note 5} Usener, H., ‘Dreiheit’, Rhein. Mus. 58 (1903), I ff., 161 ff. 321 ff.Google Scholar

^{page 325 note 1} This last point is misunderstood by E. Klostermann, who in his commentary on Luke in the Handbuch zum N.T. objects to Radermacher, ‘niemand verkennt, daß dies an sich eine typische Zahl ist (gegen Radermacher)’. In many cases, according to Radermacher, three is not only intended as a ‘typical’ number but also not completely exact. Cf. Boll, F., ‘Lebensalter’ (see p. 320, n. 4), p. 99:Google Scholar ‘Daß…die Drei… einmal die unbestimmte Vielheit dargestellt hat, ist mir…sehr wohl glaublich.’

^{page 325 note 2} Bauer, J. B., ‘Drei Tage’, Biblica 39 (1958), 354–8Google Scholar. See also Jeremias, J., ‘Die Drei-Tage-Worte der Evangelien’, in: Tradition und Glaube. Festgabe für K. G. Kuhn (Göttingen, 1971), pp. 221–9Google Scholar, see p. 226: ‘… die Zeitangabe “am dritten Tag” (bezeichnet) keineswegs notwendig den dritten Kalendertag, sondern hat häufig, ja überwiegend, die vage Bedeutung “nach ein paar Tagen”.’

^{page 325 note 3} Th. Wb. z. N.T. 9 (1969), esp. p. 219, lines 7–16:Google Scholar ‘Es bestätigt sich auch durch andere Angaben in Ag, dass Lk in seiner Darstellung die Ereignisse absichtlich durch Angaben von Zeitabstnden miteinander verbindet. Mangels genuerer Daten sieht er sich dabei des Öfteren zur Einsetzung runder Zahlen genötigt, wie für die Monatsangaben ohne weiteres deutlich ist.’

^{page 326 note 1} The number three often has an analogous function in Latin literature. Tres is repeatedly used to denote a small number (see Lewis and Short) but it can also be a stereotyped figure denoting a relatively large number; see, e.g., Gellius v, xiv, 24Google Scholar, triennium totum, and 26Google Scholar, viam ferme tridui. The stereotyped use of ‘three’ in indications of time can also be well illustrated from Evangelium infantiae arabice, ed. Sike, H. (Trajecti ad Rhenum, 1697):Google Scholartriduum pp. 39, 91, 157Google Scholar; triennis, pp. 23, 125Google Scholar; triennium, p. 71Google Scholarbis.

^{page 326 note 2} Dupont, J., ‘Jésus à douze ans’, in Fête de la Sainte Famille (Bruges, 1961)Google Scholar (Assemblées du Seigneur 16), pp. 25–43, esp. p. 42.Google Scholar

^{page 326 note 3} Glombitza, O., ‘Der zwölfjährige Jesus’, N.T. 5 (1962), 1–4, esp. p. 2.Google Scholar

^{page 326 note 4} Laurencin, R., Jésus au temple (Paris, 1966), pp. 101–2.Google Scholar

^{page 326 note 5} Elliott, J. K., ‘Does Luke 2: 41–52 anticipate the resurrection?’ Exp. T. 83 (1971–1972), 87–9.Google Scholar

^{page 326 note 6} It must be admitted that in both cases, striking ‘minor agreements’ between Matthew and Luke are to be seen; see the synopsis on Mark, viii. 31 and x. 34Google Scholar. Yet for three reasons it is not absolutely necessary to assume that Luke and Matthew in these passages are preserving the text of a proto-Mark, from which Mark himself has deviated. Firstly, in the case of Mark, x. 34Google Scholar, Luke and Matthew differ in the order in which they put ήμέρῷ and τρíτᾳ. The order used in Luke, xviii. 33Google Scholar occurs frequently in Luke, never in Matthew; cf. Acts xiii. 33 έν τῷ Ψαλμῷ τῷ δευτέρῷ Luke, i. 26Google Scholar τῷ μενì τῷ ἓκτῳ, xxiii. 44 ⋯ρα ἓκτη, Acts x. 9 ώραν ἓκτην, Luke, i. 59Google Scholar τ ήμέρα τ όγδóŋ (cf. Acts vii. 8 = Gen. xxi. 4), Acts xix. 9 ἓρας πέμπτης, Luke, iiiGoogle Scholar. έν ἓτεı δέ πεντεκıδεκάτῳ. The order in Luke, xviii. 33Google Scholar, therefore, looks redactional. Secondly, Luke, xviii. 33Google Scholar has άναστήσεταı as in Mark, x. 34Google Scholar and only Matthew has έλερθήσεταı this situation is most easily explained by assuming that Matthew made an alteration in Mark, and not vice versa. Thirdly, Jesus' resurrection on the third day was a central theme in early Christian theology, for which a fairly definite terminology existed. It is not impossible that Matthew and Luke, independently of each other, adapted the phraseology of Mark, viii. 31Google Scholar to the expression of I Cor. xv. 4 έλήλερταı τ ήμέρα τ τρíτŋ while retaining the Marcan word order (numeral-noun-verb). In the episode of the last supper, too, Luke, (xxii. 19–20)Google Scholar shows a strikingly close relationship to I Cor. Xi. 24–5.

^{page 327 note 1} The reason why, according to Luke, the resurrection did not take place μεα τρεīς ήμέρας, but τ τρíτŋ ήμέρᾳ is, presumably, that in Lucan idiom, τρεīς could refer to an indefinite number, while τρίτος simply referred to the exact number three. With reference to the resurrection, which according to the ancient system of inclusive reckoning took place after three days precisely (Th. Wb. z. N.T. 7, 29, n. 226)Google Scholar, the ambivalent expression was less suitable, and the more exact one with τρíτος was preferred by Luke.

^{page 327 note 2} Cadbury, H. J., The Style and Literary Method of Luke (Cambridge, 1920), II, 153:Google Scholar ‘In the position of the numeral adjective [by which Cadbury means the nomen numerale cardinale ] Luke's changes tend towards the normal order, [i.e. postposition of the cardinal].’ Luke, ix. 13Google Scholar // Mark, vi. 38Google Scholar; Luke, ix. 14Google Scholar // Mark, vi. 44Google Scholar; Luke, ix. 17Google Scholar // Mark, vi. 43Google Scholar; Luke, ix. 33Google Scholar // Mark, ix. 5Google Scholar; Luke, xi. 26Google Scholar // Matthew, xii. 45Google Scholar; in Acts Tpsis precedes the noun on three occasions and follows it on nine, while in one case the order is uncertain (xxviii. 7).

^{page 327 note 3} Schürmann also rightly states this, but without any argument: Das Lukaseoangelium, p. 134Google Scholar, ‘An die drei Tage bis zur Auferstehung soll man hier schwerlich denken.’

^{page 327 note 4} Cloppenburg, J., Deliciae Biblicae Brielenses, sise collationes…cum L. de DieuGoogle Scholar, most easily accessible in Critici Sacri, IX (London, 1660¹), cols. 3967–4004Google Scholar, see 4000–1, sub ‘17 Julii/Septemb./Octobris 1632’; in the Amsterdam edition of 16983 this is vol. wit, cols. 1457–8.

^{page 328 note 1} Lightfoot, J., Horae hebraicae, in Opera Omnia, 11 (Roterodami, 1686), 500Google Scholar; 11, 500 in the Franeker edition of 1699. Oddly enough, Lightfoot located Luke, ii. 46 ffGoogle Scholar., in his Chronica (Opera Omnia, 11, 5–6)Google Scholar not in a synagogue, but in an assembly room of one of the three synedria which he distinguished.

^{page 328 note 2} Antiquitates sacrae veterum Hebraeorum (Trajecti ad Rhenum, 1741 ⁴), p. 44Google Scholar, I owe these references to Relandus, Vitringa and Carpzov to Edersheim (see p. 329, n. 1).

^{page 328 note 3} Vitringa, C., De Synagoga vetere libri iii (Franequerae, 1696) (and Leucopetrae, 1726), pp. 38–9Google Scholar. Vitringa ascribes the ‘gloss’ quoted by Lightfoot (‘synagoga proxima erat Atrio in monte domus’) to Jarchi, but ‘Jarchi’ is the name often given, by Jews and Christians, to Rashi.

^{page 328 note 4} Carpzov, J. G., Apparatus historico-criticus (Francoforti/Lipsiae, 1748), p. 136.Google Scholar

^{page 328 note 5} Strack-Billerbeck, , 11, 150.Google Scholar

^{page 328 note 6} Th. Wb. z. N. T. 3, 235Google Scholar. Schrenk cites Dalman, G., Orte and Wege Jesu (1924 ²), p. 317Google Scholar, but in the first and second editions which were accessible to me ( 1919 ¹, p. 277Google Scholar; 1921 ², p. 242)Google Scholar Dalman locates Luke, ii. 46Google Scholar ‘wahrscheinlich in den Hallen dieses Hofes (= des Frauenhofes!)’.

^{page 329 note 1} Edersheim, A., The Life and Times of Jesus the Messiah, 1–11Google Scholar. I used the New York edition of 1931 (II 742–3: ‘Appendix X, On the supposed Temple Synagogue’). Earlier editions appeared from 1883¹ (London, 1886³)

^{page 329 note 2} Hoenig, S. B., ‘The Suppositious Temple-Synagogue’, Jewish Quarterly Review 54 (1963), 115–31CrossRefGoogle Scholar, reprinted in Gutmann, J., ed., The Synagogue, Studies in Origins, Archaeology and Architecture (New York, 1975), pp. 55–71.Google Scholar

^{page 329 note 3} p. 129 according to the pagination in J.Q.R.; p. 69Google Scholar in Gutmann, , The Synagogue.Google Scholar

^{page 329 note 4} Haenchen, E., Die Apostelgeschichte (Güttingen, 1968 ²)Google Scholar (Meyers Reihe), p. 164Google Scholar. Luke, in fact, had no accurate ideas of the geography of Palestine as a whole and his account of Jesus' last journey to Jerusalem (ix. 51–xviii. 14) cannot be reconstructed on a map of Palestine. Luke had no such map. See Conzelmann, H., Die Mitte der, Zeit (Tübingen, 1954, 1957²), p. 61 n. 6.Google Scholar

^{page 329 note 5} Nilsson, M. P., Die hellenistische Schule, pp. 30–3:Google Scholar ‘Die Gebäude’. Nilsson, refers to Vitruvius V. IIGoogle Scholar, where it is said that the palaestrae of gymnasia were surrounded on four sides by colonnades. Characteristic examples of gymnasia with colonnades are those of Priene and Pergamum.

^{page 330 note 1} Dupont, pp. 31–2: ‘le voici donc dans la capitale, devant les maîtres les plus illustres de sa nation; c'est dans cette Sorbonne du judaisme que son intelligence force l'attention, mérite l'admiration, provoque la stupeur’. Schürmann, , p. 134Google Scholar, gives a similar interpretation.

^{page 330 note 2} Cf. the way in which Luke uses the census in fi. I to manoeuvre Mary and Joseph to Bethlehem, and the presentation in the temple (ii. 22) to set the scene for the song of Simeon.

^{page 330 note 3} Augustine, , Serino 51Google Scholar, cap. xi, no. 18, PL, 38, 343Google Scholar, cited by Laurencin, , Jésus, pp. 217–18.Google Scholar

^{page 331 note 1} Klostermann, E., Lukaseuangelium, p. 47:Google Scholar ‘Der Gegensatz von το πατρóς μου zu v. 48 steht in zweiter Linie (anders 8:21).’

^{page 331 note 2} Laurentin, R., Jésus au temple (Paris, 1966), pp. 38–72.Google Scholar

^{page 331 note 3} Easton, B. S., The Gospel according to St Luke (Edinburgh, 1928), p. 32.Google Scholar

^{page 331 note 4} Scott, J. A., Classical Weekly XL. (1947), 70, in a review of the Revised Standard Version (New York, 1946).Google Scholar

^{page 331 note 5} Laurentin, , Jésus, p. 56Google Scholar. The possibility which Laurentin wrongly dismisses here is used in the French translation of Louis Segond, and the New Translation of the Dutch Bible Society (1951).

^{page 332 note 1} Job xviii. 19; Tobit vi. 11; Esther, vii. 9.Google Scholar

^{page 332 note 2} Moulton, J. H., Einleitung in die Sprache des N.T. (Heidelberg, 1911), p. 167.Google Scholar

^{page 332 note 3} Laurentin, , Jésus, pp. 58–60.Google Scholar

^{page 332 note 4} Ibid. pp. 60–1.

^{page 332 note 5} Guillaumont, A. et al. , eds., The Gospel according to Thomas (Leiden/London, 1959), pp. 34–5Google Scholar. ‘To give from’: Matthew, xxv. 8Google Scholar, Rev. iii. 9: έξ; Luke, xx. 10:Google Scholar άπó

^{page 333 note 1} Acts iii. 21. Luke's view that the task which Jesus had to fulfil continued after his stay in the temple cannot be derived from the fact that δεī is in the present tense. Even if Jesus’ task had been at an end in ii. 49, Luke could have written δεī. Cf. xxiv. 21 έλπíομεν ότı ατóς έστıν. Blass-Debrunner, , Grammatik, § 324Google Scholar; Robertson, A. T., A Grammar of the Greek N.T. in the light of Historical Research (Nashville, Tenn., 1934, reprint of the 4th edition), pp. 1029–30.Google Scholar

^{page 333 note 2} Luke, iv. 43Google Scholar, ix. 22, xiii. 33, xvii. 25, xxii. 37, xxiv. 7, xxiv. 26, 44–6; Acts iii. 21, xvii. 3.

^{page 334 note 1} Laurentin, , Jésus, p. 71Google Scholar, states that this conflict has been observed by various authors, and recognizes that there is a ‘tension paradoxale’ (p. 71) and an ‘apparente contradiction’ (p. 130) between 49 and 50. But in Laurentin's opinion this conflict is only illusory, because Jesus' words ‘I must be with the Father’ form a prophetic allusion to his resurrection and ascension. As this would not take place for another eighteen years, Jesus was not in conflict with his words in 49 when he returned to Nazareth: his hour had not yet come (John ii. 4). After all, Laurentin too attaches two meanings to the words ν τοι το πατρóς μου: (i) ‘with my Father in the temple’; (2) ‘with my Father in heaven after the resurrection and ascension’. In my opinion, the text of Luke, ii. 41–51Google Scholar offers no support for this second interpretation.

^{page 334 note 2} Dupont, p. 33, speaks of ‘une invitation à nous efforcer de comprendre’.

^{page 334 note 3} It is true that Luke, ii. 41–52Google Scholar shows a number of striking similarities to I Sam, . i–iiiGoogle Scholar; this has been pointed out in particular by Völter, D., Die evangelischen Erzählungen, pp. 76–9Google Scholar. But there is a significant difference. Samuel, after he had been brought to the house of the Lord (είς οīκον κυρίου, i. 24), remained in the temple, attached to the temple-worship; Jesus, on the other hand, returned to Nazareth after the presentation (Luke, ii. 39)Google Scholar and continued to live with his parents. For that reason it could not be expected that they would know that he must be in the temple.

^{page 335 note 1} Dupont, p. 34: ‘la parole énigmatique’; Laurentin, , Structure, p. 143:Google Scholar ‘phrase énigmatique’; Jésus, p. 72:Google Scholar ‘in the house of my Father’ is an expression of ‘caractère volontairement énigmatique’.

^{page 335 note 2} This appears to be a conclusion which can be justified from the Greek fathers (see Laurentin, , Jésus, pp. 59–61)Google Scholar, pace Moule, C. F. D.'s observation, ‘a priori the Authorised Version about my Father's business seems the more natural’, An Idiom book of N.T. Greek (Cambridge, 1968 ²), p. 75.Google Scholar

^{page 335 note 3} The word δεī is ambivalent too, as it can represent a present tense as well as an imperfect in oratio recta, cf. p. 333 n. 1.Google Scholar

^{page 335 note 4} Laurentin, , Jésus, p. 54Google Scholar. For the meaning ‘to be engaged in’, see Soph. Oed. Tyr. 562Google Scholar έν τ τέχνŋ; Plat. Phaedo 59AGoogle Scholar ἓν φıλοσοφíῷ (Liddell, and Scott, , p. 488)Google Scholar, Prot. 317CGoogle Scholar and Meno 91 EGoogle Scholar έν τ τέχνŋ; Thuc. VIII, 14 έν τεıχıσμῷ καì παρασκευ Xen. Hell. 4, viii. 7Google Scholar έν τοıοτοıς; Plut. Mor. 2, 342BGoogle Scholar έν τοīς κυρıωτάτοıς τς ήγεμονíας; Aelian, V.H. 1, 31 έν γεωρλíᾳ. Cf. Field, F., Notes on the Translation of the N.T. (Cambridge, 1899), p. 52.Google Scholar

^{page 335 note 5} The Latin translations are also unsatisfactory, either because literalness makes them obscure (ff² l q r¹: ‘in patris mei’), or because they give only one of the two meanings (β: ‘ in patris meidomum’; aur c df vg: ‘in his quae patris mei sunt’), or neither of them (b: ‘in propria patris mei’).

^{page 335 note 6} E.g. A Harmony of the Evangelists in English, with … notes for the we of the Unlearned, by Priestley, J. (London, 1780), p. 16Google Scholar; the Dutch Willibrord version, the Gospel of Luke in contemporary Dutch, Vrij! (Amsterdam, 1970)Google Scholar, the Bible de Jérusalem, and the Version synodale de la société biblique de France.

^{page 336 note 1} Laurencin, , Jésus, pp. 95–109.Google Scholar

^{page 337 note 1} I feel obliged to correct two mistakes of another sort in Laurentin's work. (i) On p. 51 he cites from a nineteenth-century commentary the name of a scholar called Valcken, who is said to have interpreted τ το πατρóς as ‘affairs’. Laurentin adds that he had not been able to see Valcken's work himself. He apparently failed to realize that Valcken was an abbreviation of L. C. Valckenaer, professor of Greek at Leiden from 1765 to 1785, who was a good scholar in the fields of textual criticism, language and interpretation of the New Testament. His notes on Luke, ii. 41–51Google Scholar occur in his Selecta e Scholis (ed. Wassenbergh, E.), I. (Amstelodami, 1815). (ii) On p. 156Google Scholar Laurencin includes, among the authors who have compared Luke, ii. 41–51Google Scholar with a legend of the Buddha, two scholars, Van den Bergh and Van Ensinga (sic), their names being separated by a comma, and listed separately in the index. In reality, these scholars were one and the same person, i.e. G. A. van den Bergh van Eysinga (sic), professor of New Testament at the University of Amsterdam from 1936 until 1944. He wrote, as well as hundreds of reviews, about 200 books, brochures and articles including several in French. With Couchoud, P. L. of Paris, he also edited the Annales d'histoire du Christianisme.Google Scholar

^{page 338 note 1} Neirynck, F., ‘Maria bewaarde al de woorden in haar hart’, Collationes Brugenses et Gandavenses 5 (1959), 433–66, see p. 463.Google Scholar

^{page 338 note 2} Iersel, B. van, ‘The Finding of Jesus in the Temple’, N.T. 4 (1960), 161–73, see p. 162.Google Scholar

^{page 338 note 3} Meyer, B. F., ‘But Mary kept all these things, Lc 2, 19, 51’, C.B.Q. 24 (1964), 31–490Google Scholar

^{page 338 note 4} The exegetes who see 51b as the conclusion of 41 ff. include Unnik, W. C. van, ‘Die rechte Bedeutung des Wortes treffen, Lukas 2, 19’, in Verbum, Essays… dedicated to Dr H. W. Obbink (Utrecht, 1964), pp. 129–47, see p. 131, n. 3:Google Scholar ‘Dieser Passus steht als Abschluss der Geschichte vom zwölfjährigen Jesus im Tempel.’

^{page 338 note 5} The pericope contains 170 words. The word μἓσω in 46 is the 85th word and the phrase έν μέσῳ τν δıδασκάλων therefore forms the mathematical centre of the pericope. In the scheme of the structure of the pericope, which we give in the text, another mathematical balance is concealed, to which Professor Smit Sibinga drew my attention: A+B+B′+A′ = C+C′+X = 85 words.

^{page 339 note 1} For the term ‘concentric symmetry’, see Vanhoye, A., La structure littéraire de l'Épître aux Hébreux (Paris/Bruges 1963), p. 62.Google Scholar

^{page 339 note 2} Older literature is listed by Vanhoye, , La structure, pp. 60–3Google Scholar. It is not possible to give even a survey of the voluminous recent literature. We may content ourselves with a reference to Sibinga, J. Smit, ‘Melito of Sardis. The Artist and his Text’, Vig. chr. 24 (1970), 81–104, see p. 104, n. 59CrossRefGoogle Scholar; Léon-Dufour, X., ‘Trois chiasmes johanniques’, N.T.S. 7 (1960– 1961), 249–55Google Scholar; and Marco, A. di, ‘Der Chiasmas in der Bibel, 3. Teil’, Linguistica biblica 39 (December 1976), 37–85Google Scholar (especially on the gospels).

^{page 339 note 3} Vanhoye, , La structure, p. 60Google Scholar; Bailey, K. A., ‘Recovering the Poetic Structure of I Cor. i 17–ii 12’, N.T. 16 (1975), 265–96, see p. 270:Google Scholar ‘the climax of a poem of this [concentro-symmetric] type is always the centre’.

^{page 339 note 4} Jansen, J. F., ‘Luke ii, 41–52’, Interpretation, a, Journal of Bible and Theology 30 (1976), 400–4, see p. 401.Google Scholar

^{page 340 note 1} Bultmann, R., Geschichte der synoptischen Tradition (Göttingen, 1958 ⁴), p. 327.Google Scholar

^{page 340 note 2} Ibid. p. 334: ‘… das Interesse an seinem βíος…’.

^{page 340 note 3} Laurentin, R., Jésus, p. 147–51Google Scholar. Laurentin substitutes Plutarch's Oratio de Alexandri Magni for-tuna et uirtute, Mor. II p. 342BGoogle Scholar, for Plut. Alex. v.Google Scholar

^{page 341 note 1} Ps.-Herodotus, , Vita Homeri (ed. Wilamowitz-Moellendorf, )Google Scholar (Kleine Texte 137), iv–v.Google Scholar

^{page 341 note 2} Vita Aeschinis, in Franke, F. (ed.), Aeschinis Orationes (Lipsiae, 1851) (Teubner), pp. 1–2.Google Scholar

^{page 341 note 3} Otto, E., Die biographischen Inschnften der Ägyptischen Spätzeit (Leiden, 1954)Google Scholar. On the theme of intelligentia precox in Jewish sources, see Perrot, Ch., ‘Les récits d'enfance dans la haggada antérieure au IIe siécle de notre ére’, Rech. de sc. rel. 55 (1967), 481–518; see p. 4.84Google Scholar (on Noah), p. 486 (on Abraham), p. 503 (on Moses).

^{page 341 note 4} See also Vita Alexandri I, xvi (ed. Thiel, H. van)Google Scholar; Bieler, L., Theios Aner (Darmstadt, 1967, reprint of the Wien edition, 1935, vol. 1), p. 34.Google Scholar

^{page 342 note 1} Unnik, W. C. van, ‘Eléments artistiques dans 1'Evangile de Luc’, in Neirynck, F. (ed.), L'Evangile de Luc. Problémes littéraires et théologiques (Gembloux, 1973)Google Scholar (Biblioth. Ephem. Theol. Lov. 32), pp. 129–40, see pp. 136–7.Google Scholar

^{page 342 note 2} Ienel, B. van, ‘The Finding of Jesus in the Temple’, N.T. IV (1960), 161–73.Google Scholar

^{page 343 note 1} Schürmann, H., Das Lukasevangelium, I (Freiburg/Basel/Wien, 1969), 135.Google Scholar

^{page 343 note 2} Such a story is told, for example, of Rabbi Eliezer ben Hyrcanos, who ran away from his father to devote himself to the study of the Law, and was found by his father, who had come to Jerusalem, in a ‘house of study’. See Bultmann, R., Geschichte, p. 328.Google Scholar

^{page 343 note 3} The verb ⋯ξíσlταμαı characterizes astonishment at supernatural happenings such as the raising of Jairus' daughter (Luke, viii. 56)Google Scholar, the resurrection of Jesus (Luke, xxiv. 22)Google Scholar, the magical arts of Simon the Magician (Acts viii. 9, II), Philip's miracles (13), the miracle of the tongues at Pentecost (Acts ii. 7, 12).

^{page 343 note 4} Hawkins, J. C., Horae Synopticae (Oxford, 1909 ², reprinted 1968), p. 23.Google Scholar

^{page 344 note 1} Conzelmann, H., Die Apostelgeschichte (Tübingen, 1963)Google Scholar(Handbuch zum N.T.), p. 59Google Scholar. Cf. also Acts ii. 12.

^{page 344 note 2} Blass-Debrunner, , Grammatik §442, 16.Google Scholar

^{page 344 note 3} Brugensis, Fr. Lucas (cf. p. 324, n. I)Google Scholar paraphrased os et sapientiam as sermonem sapientem. Modern translations, such as The Gospel of Luke in Contemporary Dutch, Vrij! (Amsterdam, 1970)Google Scholar or the New Testament in Today's Dutch, Groot Nieuws voor U (Amsterdam/Boxtel, 1972)Google Scholar have ‘wijze woorden’ (wise words). The first translates ‘I shall have you say such wise words that…’ and the second, ‘I shall put such wise words into your mouth that…’.

^{page 345 note 1} Hawkins, , Horae Synopticae, pp. 28–9Google Scholar; cf. Neirynck, F., ‘Hawkins's Additional Notes to his ‘Horae Synopticae’’, Eph. Theol. Lov. 61 (1970), 78–111, see p. 86Google Scholar where the expression is not mentioned.

^{page 345 note 2} Cadbury, H. J., The Style and Literary Method of Luke (2 vols. Cambridge, Mass., 1919–1920), II, 115–16.Google Scholar

^{page 345 note 3} Cadbury, , pp. 142–5Google Scholar; Iersel, Van, p. 171.Google Scholar

^{page 345 note 4} Huet, G., ‘Daniel et Susanne’ Revue de l'histoire des religions, 33 me ann´e, 65 (1912), 277–84.Google Scholar

^{page 346 note 1} Cadbury, , p. 110Google Scholar; cf. pp. 173 and 177.

^{page 346 note 2} Ibid. p. 117.

^{page 346 note 3} Hawkins, , p. 19Google Scholar. The word has an asterisk which marks it as a most distinctive and important Lucanism. For ‘l'importance accordée à Jérusalem, thème favori de Luc’, see Laurentin, , Structure, p. 103Google Scholar, and Jŕsus, pp. 95–9.Google Scholar

^{page 346 note 4} Cadbur, , pp. 154 ff.Google Scholar

^{page 346 note 5} Ibid. pp. 153–4.

^{page 346 note 6} Hawkins, , p. 18Google Scholar (the expression is asterisked) and p. 40. Cadbury, , p. 132.Google Scholar

^{page 346 note 7} Hawkins, , p. 23Google Scholar (the word is asterisked); Cadbury, , p. 172.Google Scholar

^{page 346 note 8} Hawkins, , pp. 22 and 28.Google Scholar

^{page 346 note 9} Hawkins, , p. 17Google Scholar (asterisked).

^{page 346 note 10} Cadbury, , p. 153, and see p. 327, n. 2 above.Google Scholar

^{page 346 note 11} Cadbury, , p. 201.Google Scholar

^{page 346 note 12} Ibid. p. 167.

^{page 347 note 1} Ibid. pp. 202–3; Hawkins, , p. 21.Google Scholar

^{page 347 note 2} Hawkins, , p. 24Google Scholar; Laurencin, , jésus, p. 36.Google Scholar

^{page 347 note 3} Hawkins, , p. 24.Google Scholar

^{page 347 note 4} τí ότı also occurs in the LXX and as a variant reading in Mark, ii. 16Google Scholar. Blass-Debrunner, , § 299.Google Scholar

^{page 347 note 5} Iersel, Van, p. 170 n. 2.Google Scholar

^{page 347 note 6} Cadbury, , p. 107Google Scholar; Laurentin, , jésus, p. 12.Google Scholar

^{page 347 note 7} Hawkins, , pp. 19 and 41Google Scholar; Cadbury, , p. 193Google Scholar; Michaelis, W., ‘Das unbetonte Kal airrb¸ bei Lukas’, Studia theologica 4 (1950), 86–93.CrossRefGoogle Scholar

^{page 347 note 8} Hawkins, , p. 21.Google Scholar

^{page 347 note 9} Blass-Debrunner, , § 353:Google Scholar ‘die meisten nt. Beispiele entfallen auf Lukas (Ev. and den ersten Teil der Acta)…’.

^{page 348 note 1} Th. Wb. zum. N.T. 7, 515, line 4.Google Scholar

^{page 348 note 2} Völter, D., Die evangelischen Erzählungen, pp. 80–1.Google Scholar

^{page 348 note 3} Th. Wb. zum N. T. 6, 713Google Scholar, line 5. (aopla, 40; avVE61S, 47).

^{page 349 note 1} Berger, K., ‘“Gnade” im frühen Christentum’, Ned. theol. tijdschrift 27 (1973), 1–25Google Scholar, tried to show that the wisdom books of the LXX assume that χάρıς, in the form of wisdom or understanding, is given by God to his elect. This idea is indeed present in Wisdom iii. 9, but not in any of the other passages cited by Berger, (p. 3Google Scholar, Prov. viii. 17; Sir. vi, 18, xxxvii. 21). It cannot therefore be accepted as a current and traditional opinion, which throws light on Luke ii. 40. More often wisdom or know-ledge is, in apocalyptic contexts, a part of eschatological salvation, sometimes simply called Xäpls (Berger, , p. 4:Google Scholar I Enoch v. 8; Test. Levi 18. 9Google Scholar; Luke, i. 17)Google Scholar. But this wisdom is to be shared by the many who will share in the time of salvation. Luke, ii. 40Google Scholar, in which wisdom and understanding are ascribed to Jesus alone, cannot be directly related to this tradition.

^{page 350 note 1} Conzelmann, H., Die Mitte der Zeit (Tübingen, 1954), p. 62Google Scholar, dealing with the triumphal entry into Jerusalem.

^{page 350 note 2} Conzelmann, , Mitte, p. 18.Google Scholar

^{page 350 note 3} Ibid. p. 64, ‘Im Blick auf die Kirche ist der Anspruch der Juden widerlegt. Ihre Berufung auf den Tempel, die Tradition, besteht zu Unrecht.’

^{page 351 note 1} Fascher, E., ‘Theologische Beobachtungen zu δεī’, in Neutestamentliche Studien für Rudolf Bultmann (Berlin, 1954)Google Scholar (BZNW 21), pp. 228–54, esp. 245–7.Google Scholar

^{page 352 note 1} Hahn, Ferdinand, Christologische Hoheitstitel (Göttingen, 1966 ³), pp. 319–33Google Scholar. Voss, P. G., Die Christologie der lukanischen Schriften in Grundzügen (Paris/Brügge, 1965), p. 120Google Scholar, wrongly dismisses the distinction between the traditions of the ‘Son of God’ and ‘the Son and the Father’.

^{page 352 note 2} Völter, D., Die evangelischen Erzählungen, pp. 78–9.Google Scholar