Sacred Texts  Christianity  Early Church Fathers  Index  Previous  Next 

Chapter VII.—The Alleged Discrepancy in the Gospels in regard to the Genealogy of Christ.

1. Matthew and Luke in their gospels have given us the genealogy of Christ differently, and many suppose that they are at variance with one another. Since as a consequence every believer, in ignorance of the truth, has been zealous to invent some explanation which shall harmonize the two passages, permit us to subjoin the account of the matter which has come down to us, 107 and which is given by Africanus, who was mentioned by us just above, in his epistle to Aristides, 108 where he discusses the harmony of the gospel genealogies. After refuting the opinions of others as forced and deceptive, he give the account which he had received from tradition 109 in these words:

2. “For whereas the names of the generations were reckoned in Israel either according to nature or according to law;—according to nature by the succession of legitimate offspring, and according to law whenever another raised up a child to the name of a brother dying childless; 110 for because a clear hope of resurrection was not yet given they had a representation of the future promise by a kind of mortal resurrection, in order that the name of the one deceased might be perpetuated;—

3. whereas then some of those who are inserted in this genealogical table succeeded by natural descent, the son to the father, while others, though born of one father, were ascribed by name to another, mention was made of both of those who were progenitors in fact and of those who were so only in name.

4. Thus neither of the gospels is in error, for one reckons by nature, the other by law. For the line of descent from Solomon and that from Nathan 111 were so involved, the one with the other, by the raising up of children to the childless and by second marriages, that the same persons are justly considered to belong at one time to one, at another time to another; that is, at one time to the reputed fathers, at another to the actual fathers. So that both these accounts are strictly true and come down to Joseph with considerable intricacy indeed, yet quite accurately.

5. But in order that what I have said may be made clear I shall explain the interchange of the generations. If we reckon the generations from David through Solomon, the third from the end is found to be Matthan, who begat Jacob the father of Joseph. But if, with Luke, we reckon them from Nathan the son of David, in like manner the third from the end is Melchi, 112 whose son Eli was the father of Joseph. For Joseph was the son of Eli, the son of Melchi.

6. Joseph therefore being the object proposed to us, it must be shown how it is that each is recorded to be his father, both Jacob, who derived his descent from Solomon, and Eli, who derived his from Nathan; first how it is that these two, Jacob and Eli, were brothers, and then how it is that their fathers, Matthan and Melchi, although of different families, are declared to be grandfathers of Joseph.

7. Matthan and Melchi having married in succession the same woman, begat children who were uterine brothers, for the law did not prohibit a widow, whether such by divorce or by the death of her husband, from marrying another.

8. By Estha 113 then (for this was the woman’s name according to tradition) Matthan, a descendant of Solomon, first begat Jacob. p. 92 And when Matthan was dead, Melchi, who traced his descent back to Nathan, being of the same tribe 114 but of another family, 115 married her as before said, and begat a son Eli.

9. Thus we shall find the two, Jacob and Eli, although belonging to different families, yet brethren by the same mother. Of these the one, Jacob, when his brother Eli had died childless, took the latter’s wife and begat by her a son 116 Joseph, his own son by nature 117 and in accordance with reason. Wherefore also it is written: ‘Jacob begat Joseph.’ 118 But according to law 119 he was the son of Eli, for Jacob, being the brother of the latter, raised up seed to him.

10. Hence the genealogy traced through him will not be rendered void, which the evangelist Matthew in his enumeration gives thus: ‘Jacob begat Joseph.’ But Luke, on the other hand, says: ‘Who was the son, as was supposed’ 120 (for this he also adds), ‘of Joseph, the son of Eli, the son of Melchi’; for he could not more clearly express the generation according to law. And the expression ‘he begat’ he has omitted in his genealogical table up to the end, tracing the genealogy back to Adam the son of God. This interpretation is neither incapable of proof nor is it an idle conjecture. 121

11. For the relatives of our Lord according to the flesh, whether with the desire of boasting or simply wishing to state the fact, in either case truly, have handed down the following account: 122 Some Idumean robbers, 123 having attacked Ascalon, a city of Palestine, carried away from a temple of Apollo which stood near the walls, in addition to other booty, Antipater, son of a certain temple slave named Herod. And since the priest 124 was not able to pay the ransom for his son, Antipater was brought up in the customs of the Idumeans, and afterward was befriended by Hyrcanus, the high priest of the Jews.

12. And having been sent by Hyrcanus on an embassy to Pompey, and having restored to p. 93 him the kingdom which had been invaded by his brother Aristobulus, he had the good fortune to be named procurator of Palestine. 125 But Antipater having been slain by those who were envious of his great good fortune 126 was succeeded by his son Herod, who was afterward, by a decree of the senate, made King of the Jews 127 under Antony and Augustus. His sons were Herod and the other tetrarchs. 128 These accounts agree also with those of the Greeks. 129

13. But as there had been kept in the archives 130 up to that time the genealogies of the Hebrews as well as of those who traced their lineage back to proselytes, 131 such as Achior 132 the Ammonite and Ruth the Moabitess, and to those who were mingled with the Israelites and came out of Egypt with them, Herod, inasmuch as the lineage of the Israelites contributed nothing to his advantage, and since he was goaded with the consciousness of his own ignoble extraction, burned all the genealogical records, 133 thinking that he might appear of noble origin if no one else were able, from the public registers, to trace back his lineage to the patriarchs or proselytes and to those mingled with them, who were called Georae. 134

14. A few of the careful, however, having obtained private records of their own, either by remembering the names or by getting them in some other way from the registers, pride themselves on preserving the memory of their noble extraction. Among these are those already mentioned, called Desposyni, 135 on account of their connection with the family of the Saviour. Coming from Nazara and Cochaba, 136 villages of Judea, 137 into other parts of the world, they drew the aforesaid genealogy from memory 138 and from the book of daily records 139 as faithfully as possible.

15. Whether then the case stand thus or not no one could find a clearer explanation, according to my own opinion and that of every candid person. And let this suffice us, p. 94 for, although we can urge no testimony in its support, 140 we have nothing better or truer to offer. In any case the Gospel states the truth.” And at the end of the same epistle he adds these words: “Matthan, who was descended from Solomon, begat Jacob. And when Matthan was dead, Melchi, who was descended from Nathan begat Eli by the same woman. Eli and Jacob were thus uterine brothers. Eli having died childless, Jacob raised up seed to him, begetting Joseph, his own son by nature, but by law the son of Eli. Thus Joseph was the son of both.”

17. Thus far Africanus. And the lineage of Joseph being thus traced, Mary also is virtually shown to be of the same tribe with him, since, according to the law of Moses, intermarriages between different tribes were not permitted. 141 For the command is to marry one of the same family 142 and lineage, 143 so that the inheritance may not pass from tribe to tribe. This may suffice here.



“Over against the various opinions of uninstructed apologists for the Gospel history, Eusebius introduces this account of Africanus with the words, τὴν περὶ τούτων κατελθούσαν εὶς ἡμᾶς ἱστορίαν.” (Spitta.)


On Africanus, see Bk. VI. chap. 31. Of this Aristides to whom the epistle is addressed we know nothing. He must not be confounded with the apologist Aristides, who lived in the reign of Trajan (see below, Bk. IV. c. 3). Photius (Bibl. 34) mentions this epistle, but tells us nothing about Aristides himself. The epistle exists in numerous fragments, from which Spitta (Der Brief des Julius Africanus an Aristides kritisch untersucht und hergestellt, Halle, 1877) attempts to reconstruct the original epistle. His work is the best and most complete upon the subject. Compare Routh, Rel. Sacræ, II. pp. 228–237 and pp. 329–356, where two fragments are given and discussed at length. The epistle (as given by Mai) is translated in the Ante-Nicene Fathers, Am. ed. VI. p. 125 ff.

The attempt of Africanus is, so far as we know, the first critical attempt to harmonize the two genealogies of Christ. The question had been the subject merely of guesses and suppositions until his time. He approaches the matter in a free critical spirit (such as seems always to have characterized him), and his investigations therefore deserve attention. He holds that both genealogies are those of Joseph, and this was the unanimous opinion of antiquity, though, as he says, the discrepancies were reconciled in various ways. Africanus himself, as will be seen, explains by the law of Levirate marriages, and his view is advocated by Mill (On the Mythical Interpretation of the Gospel, p. 201 sq.); but of this interpretation Rev. John Lightfoot justly says, “There is neither reason for it, nor, indeed, any foundation at all.”

Upon the supposition that both genealogies relate to Joseph the best explanation is that Matthew’s table represents the royal line of legal successors to the throne of David, while Luke’s gives the line of actual descent. This view is ably advocated by Hervey in Smith’s Bible Dictionary (article Genealogy of Jesus). Another opinion which has prevailed widely since the Reformation is that Luke gives the genealogy of Mary. The view is defended very ingeniously by Weiss (Leben Jesu, I. 205, 2d edition). For further particulars see, besides the works already mentioned, the various commentaries upon Matthew and Luke and the various lives of Christ, especially Andrews’, p. 55 sq.


Eusebius makes a mistake in saying that Africanus had received the explanation which follows from tradition. For Africanus himself says expressly (§15, below) that his interpretation is not supported by testimony. Eusebius’ error has been repeated by most writers upon the subject, but is exposed by Spitta, ibid. p. 63.


The law is stated in Deut. xxv. 5 sq.


Nathan was a son of David and Bathsheba, and therefore own brother of Solomon.


Melchi, who is here given as the third from the end, is in our present texts of Luke the fifth (Luke iii. 24), Matthat and Levi standing between Melchi and Eli. It is highly probable that the text which Africanus followed omitted the two names Matthat and Levi (see Westcott and Hort’s Greek Testament, Appendix, p. 57). It is impossible to suppose that Africanus in such an investigation as this could have overlooked two names by mistake if they had stood in his text of the Gospels.


We know nothing more of Estha. Africanus probably refers to the tradition handed down by the relatives of Christ, who had, as he says, preserved genealogies which agreed with those of the Gospels. He distinguishes here what he gives on tradition from his own interpretation of the Gospel discrepancy upon which he is engaged.




γένος. “In this place γένος is used to denote family. Matthan and Melchi were of different families, but both belonged to the same Davidic race which was divided into two families, that of Solomon and that of Nathan” (Valesius).


All the mss., and editions of Eusebius read τρίτον instead of ιόν here. But it is very difficult to make any sense out of the word τρίτον in this connection. We therefore prefer to follow Spitta (see ibid. pp. 87 sqq.) in reading ιόν instead of τρίτον, an emendation which he has ventured to make upon the authority of Rufinus, who translates “genuit Joseph filium suum,” showing no trace of a τρίτον. The word τρίτον is wanting also in three late Catenæ which contain the fragments of Africanus’ Epistle (compare Spitta, ibid. p. 117, note 12).


κατὰ λόγον. These words have caused translators and commentators great difficulty, and most of them seem to have missed their significance entirely. Spitta proposes to alter by reading κατ€λογον, but the emendation is unnecessary. The remarks which he makes (p. 89 sqq.) upon the relation between this sentence and the next are, however, excellent. It was necessary to Africanus’ theory that Joseph should be allowed to trace his lineage through Jacob, his father “by nature,” as well as through Eli, his father “by law,” and hence the words κατὰ λόγον are added and emphasized. He was his son by nature and therefore “rightfully to be reckoned as his son.” This explains the Biblical quotation which follows: “Wherefore”—because he was Jacob’s son by nature and could rightfully be reckoned in his line, and not only in the line of Eli—“it is written,” &c.


Matt. i. 6.


See Rev. John Lightfoot’s remarks on Luke iii. 23, in his Hebrew and Talmudical Exercitations on St. Luke.


This passage has caused much trouble. Valesius remarks, “Africanus wishes to refer the words ς ἐνομίζετο (‘as was supposed’) not only to the words ιὸς ᾽Ιωσήφ, but also to the words τοῦ ῾Ηλὶ, which follow, which although it is acute is nevertheless improper and foolish; for if Luke indicates that legal generation or adoption by the words ς ἐνομίζετο, as Africanus claims, it would follow that Christ was the son of Joseph by legal adoption in the same way that Joseph was the son of Eli. And thus it would be said that Mary, after the death of Joseph, married his brother, and that Christ was begotten by him, which is impious and absurd. And besides, if these words, ς ἐνομίζετο, are extended to the words τοῦ ῾Ηλὶ, in the same way they can be extended to all which follow. For there is no reason why they should be supplied in the second grade and not in the others.”

But against Valesius, Stroth says that Africanus seeks nothing in the words ς ἐνομίζετο, but in the fact that Luke says “he was the son of,” while Matthew says “he begat.” Stroth’s interpretation is followed by Closs, Heinichen, and others, but Routh follows Valesius. Spitta discusses the matter carefully (p. 91 sq.), agreeing with Valesius that Africanus lays the emphasis upon the words ς ἐνομίζετο, but by an emendation (introducing a second ς ἐνομίζετο, and reading “who was the son, as was supposed, of Joseph, the son of Jacob, who was himself also the son, as was supposed,—for this he also adds,—of Eli, the son of Melchi”) he applies the ς ἐνομίζετο only to the first and second members, and takes it in a more general sense to cover both cases, thus escaping Valesius’ conclusions expressed above. The conjecture is ingenious, but is unwarranted and unnecessary. The words which occur in the next sentence, “and the expression, ‘he begat’ he has omitted,” show that Africanus, as Stroth contends, lays the emphasis upon the difference of form in the two genealogies, “Son of” and “he begat.” The best explanation seems to me to be that Africanus supposes Luke to have implied the legal generation in the words “the Son of,” used in distinction from the definite expression “he begat,” and that the words ς ἐνομίζετο, which “he also adds,” simply emphasize this difference of expression by introducing a still greater ambiguity into Luke’s mode of statement. He not only uses the words, the “Son of,” which have a wide latitude, admitting any kind of sonship, but “he also adds,” “as was supposed,” showing, in Africanus’ opinion, still more clearly that the list which follows is far from being a closely defined table of descent by “natural generation.”


This seems the best possible rendering of the Greek, which reads τὴν ἀναφορὰν ποιησ€μενος ἑ& 240·ς τοῦ ᾽Αδὰμ, τοῦ θεοῦ κατ᾽ ἀν€λυσιν. οὐδὲ μὴν ἀναπόδεικτον κ.τ.λ., which is very dark, punctuated thus, and it is difficult to understand what is meant by κατ᾽ ἀν€λυσιν in connection with the preceding words. (Crusè translates, “having traced it back as far as Adam, ‘who was the son of God,’ he resolves the whole series by referring back to God. Neither is this incapable of proof, nor is it an idle conjecture.”) The objections which Spitta brings against the sentence in this form are well founded. He contends (p. 63 sqq.), and that rightly, that Africanus could not have written the sentence thus. In restoring the original epistle of Africanus, therefore, he throws the words κατ᾽ ἀν€λυσιν into the next sentence, which disposes of the difficulty, and makes good sense. We should then read, “having traced it back as far as Adam, the Son of God. This interpretation (more literally, ‘as an interpretation,’ or ‘by way of interpretation’) is neither incapable of proof, nor is it an idle conjecture.” That Africanus wrote thus I am convinced. But as Spitta shows, Eusebius must have divided the sentences as they now stand, for, according to his idea, that Africanus’ account was one which he had received by tradition, the other mode of reading would be incomprehensible, though he probably did not understand much better the meaning of κατ᾽ ἀν€λυσιν as he placed it. In translating Africanus’ epistle here, I have felt justified in rendering it as Africanus probably wrote it, instead of following Eusebius’ incorrect reproduction of it.


The Greek reads: παρέδοσαν καὶ τοῦτο, “have handed down also.” The καὶ occurs in all the mss. and versions of Eusebius, and was undoubtedly written by him, but Spitta supposes it an addition of Eusebius, caused, like the change in the previous sentence, by his erroneous conception of the nature of Africanus’ interpretation. The καὶ is certainly troublesome if we suppose that all that precedes is Africanus’ own interpretation of the Biblical lists, and not a traditional account handed down by the “relatives of our Lord”; and this, in spite of Eusebius’ belief, we must certainly insist upon. We may therefore assume with Spitta that the καὶ did not stand in the original epistle as Africanus wrote it. The question arises, if what precedes is not given upon the authority of the “relatives of our Lord,” why then is this account introduced upon their testimony, as if confirming the preceding? We may simply refer again to Africanus’ words at the end of the extract (§15 below) to prove that his interpretation did not rest upon testimony, and then we may answer with Spitta that their testimony, which is appealed to in §14 below, was to the genealogies themselves, and in this Africanus wishes it to be known that they confirmed the Gospel lists.


See above, chap. VI. notes 5 and 6.


We should expect the word “temple-servant” again instead of “priest”; but, as Valesius remarks, “It was possible for the same person to be both priest and servant, if for instance it was a condition of priesthood that only captives should be made priests.” And this was really the case in many places.


Appointed by Julius Cæsar in 47 b.c. (see chap. VI. note 1, above).


He was poisoned by Malichus in 42 b.c. (see Josephus, Ant. XIV. 11. 4).


Appointed king in 40 b.c. (see chap. VI. note 1, above).


The ethnarch Archelaus (see chap. VI. note 18) and the tetrarchs Herod Antipas and Herod Philip II.


Cf. Dion Cassius, XXXVII. 15 sqq. and Strabo, XVI. 2. 46.


It was the custom of the Jews, to whom tribal and family descent meant so much, to keep copies of the genealogical records of the people in the public archives. Cf. e.g. Josephus, De Vita, §1, where he draws his own lineage from the public archives; and cf. Contra Apion. I. 7.


χρι προσηλύτων. Heinichen and Burton read ρχιπροσηλύτων, “ancient proselytes.” The two readings are about equally supported by ms. authority, but the same persons are meant here as at the end of the paragraph, where προσηλύτους, not ρχιπροσηλύτους, occurs (cf. Spitta, pp. 97 sq., and Routh’s Reliquiæ Sacræ II. p. 347 sq., 2d ed.).


Achior was a general of the Ammonites in the army of Holofernes, who, according to the Book of Judith, was a general of Nebuchadnezzar, king of the Assyrians, and was slain by the Jewish heroine, Judith. Achior is reported to have become afterward a Jewish proselyte.


The Greek reads νέπρησεν αὐτῶν τὰς ἀναγραφὰς των γενων, but, with Spitta, I venture, against all the Greek mss. to insert π€σας before τὰς ἀναγραφὰς upon the authority of Rufinus and the author of the Syriac version, both of whom reproduce the word (cf. Spitta, p. 99 sq.). Africanus certainly supposed that Herod destroyed all the genealogical records, and not simply those of the true Jews.

This account of the burning of the records given by Africanus is contradicted by history, for we learn from Josephus, De Vita, §1, that he drew his own lineage from the public records, which were therefore still in existence more than half a century after the time at which Herod is said to have utterly destroyed them. It is significant that Rufinus translates omnes Hebræorum generationes descriptæ in Archivis templi secretioribus habebantur.

How old this tradition was we do not know; Africanus is the sole extant witness of it.


τοὺς τε καλουμένους γειώρας. The word γειώρας occurs in the LXX. of Ex. xii. 19, where it translates the Hebrew גֵּר The A.V. reads stranger, the R.V., sojourner, and Liddell and Scott give the latter meaning for the Greek word. See Valesius’ note in loco, and Routh (II. p. 349 sq.), who makes some strictures upon Valesius’ note. Africanus refers here to all those that came out from Egypt with the Israelites, whether native Egyptians, or foreigners resident in Egypt. Ex. xii. 38 tells us that a “mixed multitude” went out with the children of Israel (ἐπίμικτος πόλυς), and Africanus just above speaks of them in the same way (πιμίκτων).


δεσπόσυνοι: the persons called above (§11) the relatives of the Saviour according to the flesh (οἱ κατὰ σ€ρκα συγγενεις). The Greek word signifies “belonging to a master.”


Cochaba, according to Epiphanius (Hær. XXX. 2 and 16), was a village in Basanitide near Decapolis. It is noticeable that this region was the seat of Ebionism. There may therefore be significance in the care with which these Desposyni preserved the genealogy of Joseph, for the Ebionites believed that Christ was the real son of Joseph, and therefore Joseph’s lineage was his.


“Judea” is here used in the wider sense of Palestine as a whole, including the country both east and west of the Jordan. The word is occasionally used in this sense in Josephus; and so in Matt. 19:1, Mark 10:1Mark x. 1, we read of “the coasts of Judea beyond Jordan.” Ptolemy, Dion Cassius, and Strabo habitually employ the word in the wide sense.


κ μνήμης. These words are not found in any extant mss., but I have followed Stroth and others in supplying them for the following reasons. The Greek, as we have it, runs: καὶ τὴν προκειμένην γενεαλογίαν žκ τε τῆς βίβλου τῶν ἡμερῶν κ.τ.λ. The particle τε indicates plainly that some phrase has fallen out. Rufinus translates ordinem supra dictæ generationis partim memoriter partim etiam ex dierum libris in quantum erat perdocebant. The words partim memoriter find no equivalent in the Greek as we have it, but the particle τε, which still remains, shows that words which Rufinus translated thus must have stood originally in the Greek. The Syriac version also confirms the conclusion that something stood in the original which has since disappeared, though the rendering which it gives rests evidently upon a corrupt text (cf. Spitta, p. 101). Valesius suggests the insertion of πὸ μνήμης, though he does not place the phrase in his text. Heinichen supplies μνημονεύσαντες, and is followed by Closs in his translation. Stroth, Migne, Routh, and Spitta read κ μνήμης. The sense is essentially the same in each case.


It has been the custom since Valesius, to consider this “Book of daily records” (βίβλος τῶν ἡμερῶν) the same as the “private records” (διωτικὰς ἀπογραφ€ς) mentioned just above. But this opinion has been combated by Spitta, and that with perfect right. The sentence is, in fact, an exact parallel to the sentence just above, where it is said that a few of the careful, either by means of their memory or by means of copies, were able to have “private records of their own.” In the present sentence it is said that “they drew the aforesaid genealogy (viz., ‘the private records of their own’) from memory, or from the Book of daily records” (which corresponds to the copies referred to above). This book of daily records is clearly, therefore, something other than the διωτικὰς ἀπογραφὰς, but exactly what we are to understand by it is not so easy to say. It cannot denote the regular public records (called the archives above), for these were completed, and would not need to be supplemented by memory; and apparently, according to Africanus’ opinion, these private records were made after the destruction of the regular public ones. The “Book of daily records” referred to must have been at any rate an incomplete genealogical source needing to be supplemented by the memory. Private family record books, if such existed previous to the supposed destruction of the public records, of which we have no evidence, would in all probability have been complete for each family. Spitta maintains (p. 101 sq.) that the Book of Chronicles is meant: the Hebrew דִּבְרֵי הַיָּמִים , words or records of the days. This is a very attractive suggestion, as the book exactly corresponds to the book described: the genealogies which it gives are incomplete and require supplementing, and it is a book which was accessible to all; public, therefore, and yet not involved in the supposed destruction. The difficulty lies in the name given. It is true that Jerome calls the Books of Chronicles Verba Dierum and Hilary Sermones Dierum, &c.; but we should expect Africanus to use here the technical LXX. designation, Παραλειπομένων. But whatever this “Book of daily records” was, it cannot have been the “private records” which were formed “from memory and from copies,” but was one of the sources from which those “private records” were drawn.


Compare note 3, above. Africanus’ direct statement shows clearly enough that he does not rest his interpretation of the genealogies (an interpretation which is purely a result of Biblical study) upon the testimony of the relatives of the Saviour. Their testimony is invoked with quite a different purpose, namely, in confirmation of the genealogies themselves, and the long story (upon the supposition that their testimony is invoked in support of Africanus’ interpretation, introduced absolutely without sense and reason) thus has its proper place, in showing how the “relatives of the Saviour” were in a position to be competent witnesses upon this question of fact (not interpretation), in spite of the burning of the public records by Herod.


The law to which Eusebius refers is recorded in Num. 36:6, 7. But the prohibition given there was not an absolute and universal one, but a prohibition which concerned only heiresses, who were not to marry out of their own tribe upon penalty of forfeiting their inheritance (cf. Josephus, Ant. IV. 7. 5). It is an instance of the limited nature of the law that Mary and Elizabeth were relatives, although Joseph and Mary belonged to the tribe of Judah, and Zacharias, at least, was a Levite. This example lay so near at hand that Eusebius should not have overlooked it in making his assertion. His argument, therefore in proof of the fact that Mary belonged to the tribe of Judah has no force, but the fact itself is abundantly established both by the unanimous tradition of antiquity (independent of Luke’s genealogy, which was universally supposed to be that of Joseph), and by such passages as Ps. cxxxii. 11, Acts ii. 30, xiii. 23, Rom. i. 3.





Next: Chapter VIII