J. Gresham Machen
The Virgin Birth of Christ
Excerpt on the Genealogies of Matthew and Luke


This paper presents J. Gresham Machen's comments on the genealogies of Jesus in Matthew and Luke. It represents a verbatim excerpt taken from his book, The Virgin Birth of Christ (New York: Harper & Brothers Publishers, 1930). Machen's footnotes are not included.

pdf Printable Version of This Page


The Genealogies of Matthew and Luke

One of the most serious questions of harmony between the representations of the birth of Jesus in Matthew and in Luke has not been discussed so far, because it belongs in a somewhat different category from those questions with which we have just dealt. The difference is that this question, unlike those others, concerns not the relation between the two infancy sections, but the relation between one infancy section and a detail contained in a later chapter of the Gospel that contains the other. We refer to the matter of the two genealogies in Mt. 1:1-17, and Lk. 3:23-38.

At first sight, the genealogies may seem to be in hopeless contradiction. From Abraham to David they run alike; but the son of David through whom the line is traced is, according to Matthew, Solomon, and according to Luke, Nathan, and from that point on two divergent lists of names are given down to the father of Joseph, the husband of Mary. How can the divergence of the two lines be explained? How, in particular, could the father of Joseph be Jacob (as Matthew says) and also Heli (as he is said to be by Luke)? The two genealogies seem to be directly contradictory even with regard to a point so near at hand as that.

According to one solution of the difficulty, the genealogy in Luke is not really a genealogy of Joseph at all, but a genealogy of Mary. In that case, the difficulty regarding the divergence at the end of the genealogies disappears; Jacob was simply the father of Joseph and Heli the father of Mary.

But is this solution exegetically possible? The verses at the beginning of the Lucan genealogy read, according to the usual interpretation, when literally translated, as follows: And Jesus Himself was, when He began, about thirty years old, being the son, as was supposed, of Joseph, who was the son of Heli, who was the son of Matthat, etc.

According to this rendering, Heli was the father of Joseph. But some scholars, who find in this genealogy a genealogy of Mary, remove the comma between "as was supposed" and "of Joseph," and translate somewhat as follows: "being the son (of Joseph as was supposed) of Heli, of Matthat, etc." That is, Jesus was supposed to be the son of Joseph, but was really the son of Heli, etc. Heli would then be the father of Mary, and the word "son" would be taken in the wider sense of "descendant," the name of the mother of Jesus being omitted because it was not customary for women to be included in a genealogy.

Undoubtedly this interpretation would remove a difficulty in the comparison between the two genealogies; and it has won the support of noteworthy scholars, including, for example, Bernhard Weiss. But on the whole it seems rather unnatural. The strictly parenthetical interpretation of "as was supposed of Joseph" would hardly occur to a reader who had not the advantage of modern marks of punctuation. And instead of taking every name in the genealogy as depending directly upon the initial word "son" (in the sense of "descendant"), it surely seems more natural to take every name as depending upon the immediately preceding name.

A more natural way of interpreting the genealogy as a genealogy of Mary is to say that Joseph is here represented as having become "son" or heir of Heli by his marriage to Mary. If Mary was Heli's daughter, and if she had no brothers, then she would become an heiress in accordance with the provisions of Num. 27:1-11; 36:1-12. In the former passage, it seems to be provided that the "name" of a man who had daughters but no sons should be preserved. This could be accomplished if the husband of one of the daughters should become identified with the family of his wife. Conceivably, therefore, Joseph may be designated in the Lucan genealogy as the son of Mary's father, Heli. This interpretation would at least have the advantage of avoiding the linguistically unnatural treatment of the words, "as was supposed of Joseph," which is involved in the view just mentioned; every link in the genealogy would now be joined naturally to that which precedes; Jesus would be represented as the "son" of Joseph, Joseph as the "son" of Heli, Heli as the "son" of Matthat.

On the whole, however, it seems better to follow the usual view, in accordance with which the genealogy is the genealogy of Joseph. This view is in accordance with what we saw to be the more natural interpretation of Lk. 1:27. If in that verse the author of the Third Gospel calls attention to the Davidic descent of Joseph, and does not call attention to the Davidic descent of Mary, it is natural to find that the genealogy which he inserts at a later point in his book is a genealogy of Joseph and not of Mary.

But if the Lucan genealogy is a genealogy of Joseph, how shall we explain the apparent discrepancy with the genealogy in Matthew? How shall we explain the fact that according to Luke Joseph's father is Heli, and according to Matthew, Jacob?

The most probable answer is that Matthew gives the legal descendants of David--the men who would have been legally the heir to the Davidic throne if that throne had been continued--while Luke gives the descendants of David in that particular line to which, finally, Joseph, the husband of Mary, belonged. There is nothing at all inherently improbable in such a solution. When a kingly line becomes extinct, the living member of a collateral line inherits the throne. So it may well have been in the present case.

The first objection to this view which might occur to a modern reader is found in the use of the word "begat" by Matthew. How could that word be used if merely legal heirship and not physical descent were intended? But an examination of Semitic usage soon shows that this objection is entirely without force. Indeed, it is clear in the course of the genealogy itself that the word "begat" is used in a very broad sense. Thus any reader of the Old Testament would know that in the strict sense Joram did not "beget" Uzziah, but that three generations are here omitted between these two kings. As Burkitt has pointed out, it is probable that the author of the genealogy knew his Old Testament as well as we do. Evidently, therefore, he is using the word "begat" in a broader sense than that in which we employ the English word.

Lord Hervey, who adopts this general solution of the problem of the harmony between the genealogies, cites a number of instances of double genealogies in the Old Testament--that is, a number of cases where a man was reckoned with the family of one who was not in a physical sense his father. This incorporation into another family was practised, he thinks, when a man came into possession of property which belonged to some one other than his own father.

It is not necessarily an objection to this view of the relation between the genealogies that they coincide, in the middle of the divergent sections, in two names, Shealtiel (Salathiel) and Zerubbabel; for we should only have to suppose either that Jeconiah, who is said to have "begotten" Shealtiel, had no son, or else that his son, because of the curse recorded in Jeremiah 22:30, could not be the heir to the Davidic throne, so that Shealtiel (the living representative of the collateral line recorded in Luke) had to be inserted next by Matthew in the kingly line. Then, at the death of Zerubbabel, or of Abiud (if he is the same person as the Joda of Luke), the line of descent of Joseph's ancestors began to diverge again from the line of legal heirs to the throne, because the (potentially) reigning line came to an end with Jacob, who is said to have "begotten" Joseph, so that Joseph, who was the son, not of Jacob but of Heli, became legally Jacob's heir.

Of course, we cannot say, on this view, how many times in the genealogy of Matthew, between Zerubbabel and Joseph, the line of descent was broken; for all that we can tell, there may be several places in Mt. 1:13-16 where a family came to an end and thus had to take a descendant of a collateral line into itself as its heir. But it would only be at the end that a representative of that particular line to which Joseph, the foster-father of Jesus, belonged became heir to the Davidic throne.

A difficulty, indeed, does arise at this point, when we examine the Old Testament records. In 1 Chron. 3:19, according to the Hebrew text, Zerubbabel is represented as the son of Pedaiah and as the nephew, not the son, of Shealtiel, so that when he is elsewhere in the Old Testament called Shealtiel's "son," the word "son" designates an adoptive, not a physical, relationship. How then, if Luke in his genealogy is giving the line of physical, rather than merely legal, descent of Joseph, could he have designated Zerubbabel as the son of Salathiel, as he does in Lk. 3:27?

Three answers are possible. In the first place, one may hold that the "Zorobabel" and the "Salathiel" of Lk. 3:27 are different persons from the "Zorobabel" and the "Salathiel" of Mt. 1:12 and from the "Zerubbabel" and the "Shealtiel" of the Old Testament. That opinion has actually been held, but in view of the juxtaposition of the two names it seems perhaps to be unlikely. In the second place, one may follow certain manuscripts of the Septuagint at 1 Chron. 3:18 f., instead of following the Hebrew text. In that case, Pedaiah drops out as the father of Zerubbabel, and Zerubbabel may be regarded as the actual son of Shealtiel. But it is certainly far more likely that the Hebrew text is correct. Probably, therefore, the third of our three alternatives is to be chosen. By that third alternative, we shall simply have to modify our view of the Lucan genealogy. We shall have to say that there is at least one link in that genealogy in which something other than actual physical paternity is designated. This admission would not involve the total abandonment of our hypothesis; it would not involve the relinquishment of our distinction between the Lucan and the Matthæan genealogy. We should still be able to say that, while the Matthæan genealogy traces the successive heirs to the throne of David from David to Joseph, the Lucan genealogy traces the ancestors of Joseph back to David. Suppose, as is very probable, that Pedaiah "raised up seed" to his brother, Salathiel, in accordance with the legal provisions about Levirate marriage, which we shall speak of in a moment. The Lucan genealogy could then designate Pedaiah's son, Zerubbabel, as being the son of Salathiel, without at all becoming confused with the Matthæan genealogy, supposing that that genealogy involved breaks where the scion of a more or less widely separated collateral line had to be taken into the succession of the heirs to the throne. Luke's genealogy would not, indeed, in its successive links, always indicate actual physical paternity, but it would mean that every successive link did involve at least a very close adoptive relationship between the two persons named.

Thus the difficulty about Shealtiel and Zerubbabel requires a modification, rather than an abandonment, of our hypothesis. We shall still be able to say that the difference between the two genealogies, taken broadly, is due to the fact that for the most part--perhaps even in every link except one--the Lucan genealogy traces the actual physical ancestors of Joseph back to David, while the Matthæan genealogy enumerates the successive heirs to the Davidic throne. The Lucan genealogy, in other words, starts with the question, "Who was Joseph's 'father'?" The answer to that question is, "Heli." Then, in the course of the genealogy, we come to the question, "Who was Zerubbabel's 'father'?" The answer is, "Salatheil," even though the relationship of Salathiel to Zerubbabel was not that of physical paternity. And so on up to David. In the Matthæan genealogy, on the other hand, we start with the question, "Who was the heir to David's throne?" The answer is, "Solomon," and so on down to Joseph. When we consider the matter in this way, it becomes evident that our distinction between the two genealogies does not depend upon the assumption that actual physical descent is designated in the Lucan genealogy in every link, though no doubt it is designated in the vast majority of the links. There may well have been two perfectly valid ways of exhibiting Joseph's Davidic descent, even though the general principle of one at least of these two ways--supposing that general principle was the exhibition of actual physical paternity--was not followed with complete uniformity throughout.

The correctness of this view of the purpose and meaning of each genealogy is confirmed by the fact that the genealogy in Luke begins at the end and works backward, whereas the genealogy in Matthew begins at the beginning. Where the point was to trace the actual decent of Joseph back to David, that could be done by recording the tradition of the family as to his actual father, Heli, and then the actual father of Heli, and so on up to Nathan the son of David. But where the point was to mention the successive heirs of the Davidic throne, it was natural to begin with David and work down.

The view which we have set forth above as to the latest links in the two genealogies is based upon the assumption that the Matthan of Mt. 1:15 is not the same person as the Matthat of Lk. 3:24. If these two names do refer to the same person, then a difficulty seems to arise. For if Jacob and Heli were both sons of the same person, why should not the elder of them have been the heir? And if Jacob was the elder, how could Joseph, the son of the younger, Heli, have been in the line of legal heirship?

This difficulty, however, is quite readily removed. We should need only to suppose that Jacob died without issue, so that his nephew, the son of his brother Heli, would become his heir. One could also think, at this point, of the institution of Levirate marriage, in accordance with which when a man died without issue his brother married the widow and "raised up seed" to the deceased. The question which the Sadducees addressed to Jesus about this matter shows that the custom was not forgotten, whether or not it was frequently practised, in the time of Christ. Possibly, therefore, Heli married the widow [of] Jacob, so that the children, while physically his own, belonged legally to his dead brother.

If Matthan and Matthat are not the same person, then it is less natural to appeal to Levirate marriage; for if Heli was the brother of Jacob, how could he have had a different father (Matthat instead of Matthan)? The only answer, apparently, would be the ancient one that Jacob and Heli were half-brothers--that is, that they had the same mother but not the same father. But then the question might be raised whether Levirate marriage was practised in the case of half-brothers, and in general the hypothesis would seem to be overloaded.

If Matthat and Matthan are the same person, then the question how different persons could each be the father of the same man, which on the other view arises in connection with the father of Joseph, arises in the case of the father of Matthat (Matthan). And here again it would be unnatural to appeal to Levirate marriage; because that would assume that Levi, the father of Matthat (Matthan) according to Luke, and Eleazar, the father of the same man according to Matthew, were brothers--in which case their father ought to be the same. If, therefore, Matthat and Matthan are the same person, it seems best to explain the divergence regarding their father, not by Levirate marriage, but by the fact that the kingly line became extinct with Eleazar, who is said by Matthew to have "begotten" Matthan (Matthat), so that a scion of a widely divergent collateral line became his heir. Matthat (Matthan) would thus be the legal heir of Eleazar, but the actual son of Levi, who appears in the genealogy in Luke.

Thus on the view that Matthat and Matthan are the same person, the custom of Levirate marriage may plausibly be cited to explain the divergence as to the father of Joseph (Heli in one genealogy, Jacob in the other), but not to explain the divergence as to the father of Matthat (Matthan). On the view that Matthat and Matthan are not the same person, Levirate marriage is probably not to be appealed to at all.

We are not endeavoring to discuss the intricate question of the genealogies with even the slightest approach to completeness. But enough, we think, has been said to show that the differences between the two genealogies are not irreconcilable. Reconciliation might conceivably be effected in a number of different ways. But on the whole we are inclined to think that the true key to a solution of the problem (however the solution may run in detail) is to be found in the fact that Matthew, in an intentionally incomplete way, gives a list of incumbents (actual or potential) of the kingly Davidic throne, while Luke traces the descent of Joseph back through Nathan to David. Thus the genealogies cannot properly be used to exhibit contradiction between the Matthæan and the Lucan accounts of the birth and infancy of our Lord. Here, as in the other features of the two accounts, there is complete independence, but no contradiction.

Return to Genealogies Page

Those interested in more information on the difficulties surrounding the genealogies will enjoy reading Ken's paper, "Write This Man Childless! The Crisis in the Davidic Dynasty and the Genealogies of Jesus."


© Copyright 2017 Rediscovering the Bible. All Rights Reserved. | Contact Us | Email Webmaster