Where the world comes to study the Bible

The Problem of Luke 2:2 "This was the first census taken when Quirinius was governor of Syria"

Related Media

One of the greatest difficulties in the Bible, in terms of its accuracy, is the census mentioned in Luke 2:2—a census that purportedly led Joseph and Mary to Bethlehem, where Jesus would be born. The Greek text reads as follows: αὕτη ἀπογραφὴ πρώτη ἐγένετο ἡγεμονεύοντος τῆς Συρίας Κυρηνίου. This text casts serious doubts on Luke's accuracy for two reasons: (1) The earliest known Roman census in Palestine was taken in AD 6-7, and (2) there is little, if any, evidence that Quirinius was governor of Syria before Herod's death in 4 BC. In light of this, many scholars believe that Luke was thinking about the census in AD 6-7, when Quirinius was governor of Syria. At the same time, Luke demonstrates remarkable historical accuracy overall, and even shows both an awareness of this later census (cf. Acts 5:37) and an understanding that Jesus was not born this late (cf. Luke 1:5).

This issue cannot be resolved with certainty, though a couple of views are unlikely.  First, it is rather doubtful that πρώτη here is used superlatively: “first of at least three.”  Not only is the usage of πρῶτος for a comparative well established in the NT (cf., e.g., Matt 21:28 [“a man had two sons; he came to the first. . .”]; John 20:4 [“the other disciple came first to the tomb”]), but it is unnecessary to compound the historical difficulty this text presents.  A second census is hard enough to find!1

Second, it has sometimes been suggested that the text should be translated, “this census was before the census which Quirinius, governor of Syria, made.”2 It is argued that other comparative expressions sometimes have elided words (as in John 5:36 and 1 Cor 1:25) and, therefore, such is possible here. In spite of the ingenuity of this translation, the basis for it is insufficient, for the following reasons: (a) In both John 5:36 and 1 Cor 1:25, the genitive immediately follows the comparative adjective, making the comparison explicit, while in this text Κυρηνίου is far removed from πρώτη and, in fact, is genitive because it is part of a genitive absolute construction.3  Thus, what must necessarily be supplied in those texts is neither necessary nor natural in this one.4  (b) This view presupposes that αὕτη modifies ἀπογραφή.  But since the construction is anarthrous, such a view is almost impossible (because when a demonstrative functions attributively to a noun the noun is almost always articular);5 a far more natural translation would be “This is the first census . . .” rather than “this census is . . .”

Third, πρώτη is sometimes regarded as adverbial: “this census took place before Quirinius was governor of Syria.”6  The advantage of this approach is that it eludes the historical problem of Quirinius’ governorship overlapping the reign of Herod.  However, like the previous view, it erroneously presupposes that αὕτη modifies ἀπογραφή.  Further, it ignores the concord between πρώτη and ἀπογραφή, making the adjective most likely to function adjectivally, rather than adverbially. Actually, the adjective functions similarly to John 1:15, 30, but in both places a genitive immediately follows. Also, if this governed the participial phrase, as Hoehner believes, a number of other constructions would be far more natural (and we might justifiably expect Luke's grammar to be somewhat “natural,” especially in his editorial sections [since such sections are not from other sources, but are in Luke’s own words]).

In conclusion, facile solutions do not come naturally to Luke 2:2. This does not, of course, mean that Luke erred. In agreement with Schürmann, Marshall “warns against too easy acceptance of the conclusion that Luke has gone astray here; only the discovery of new historical evidence can lead to a solution of the problem.”7


Evangelicals often have a tendency to find implausible solutions to difficulties in the Bible and to be satisfied that they have once again vindicated the Word of God. On the other hand, critical scholars tend to find errors in the Bible where none exist. At bottom, our belief in the infallibility and authority of scripture is a faith-stance, just as our belief in the Deity of Christ is a faith-stance. This does not mean that we have no basis! Nor does it mean that we are obligated to solve all problems to our satisfaction before we can believe. As B. B. Warfield argued long ago, we believe in the accuracy of the Bible, first of all, because the biblical writers themselves both held and taught this view. And if we consider the biblical writers to be trustworthy as doctrinal guides, then their doctrine of the Bible must also be trustworthy. Certainly we need to make many adjustments in how we define that accuracy (allowing the biblical writers themselves to shape our understanding8); but if we were to deny their accuracy at one point, then we must either (a) deny that they held and taught such a view of the Bible, or (b) assume that they might not be trustworthy in other doctrinal areas as well. There is much to be done in this aspect of bibliology, not just in terms of vindication, but also in understanding.9 Responses that are implausible on their face certainly do not help the evangelical faith in the long run.

1Remarkably, Robertson in his massive grammar suggests that in Luke 2:2 the census is the “first in a series of enrollments as we now know” (669)!  He is basing his view on the work of W. M. Ramsay, though this particular point has long since been rejected.

2See Turner, Insights, 23-24, for a defense of this view. The view was found as early as the seventeenth century by Herwartus, and maintained by Huschke, Tholuck, Lagrange, Heichelheim, Bruce, Turner, et al.

3H. Hoehner, Chronological Aspects of the Life of Christ (Grand Rapids: Zondervan, 1977) 21, argues similarly: “one notable difference between Luke 2:2 and the other passages cited is that Luke 2:2 has the participial phrase, ‘when Quirinius was governor of Syria,’ which is cumbersome, namely, ‘This census was earlier than [the census] when Quirinius was governor of Syria.’”

4Winer-Moulton, 306, rightly calls this view “awkward, if not ungrammatical.”

5BAGD point out that “When the art. is lacking there is no real connection betw. the demonstrative and the noun, but the one or the other belongs to the predicate . . .” (οὗτος, 2.c. [597]).  They list but two exceptions, calling them “more difficult” (Acts 1:5) and “most difficult of all” (Acts 24:21). BAGD suggest that Luke 2:2 should be translated “this was the first census.”  Incidentally, M. Palmer lists three such exceptions to this general rule in his Levels of Constituent Structure in New Testament Greek, 118, n. 33.  In addition to Acts 1:5 and 24:21, he gives Luke 24:21, but this is not the only interpretation.

6Cf. A. J. B. Higgins, “Sidelights on Christian Beginnings in the Graeco-Roman World,” EvanQ 41 (1969) 200-1.

7I. H. Marshall, Luke: Historian and Theologian (Grand Rapids: Zondervan, 1971) 69, n. 5, enlisting H. Schürmann, Das Lukasevangelium (Freiburg, 1969) 1.98-101, on his side.

For an excellent treatment of the problem overall, especially from a historical perspective, see Hoehner, Chronological Aspects, 13-23.

8 For example, the biblical writers did not feel any compulsion to quote exactly from their sources, even from Jesus. This is evidenced by the synoptic parallels and by the use of the Old Testament in the New. We must not, therefore, impose a definition of accuracy on these writers that comes from our twentieth century perspective.

9 There is also much to be done in terms of method. Evangelicals tend to allow their doctrinal convictions to guide their research. It is better to not the left hand know what the right hand is doing: methodologically, investigate with as objective a mind as possible, allowing the evidence to lead where it will. At all times, pursue truth! Then, when possible conclusions have been met, come back to the presuppositions and wrestle with how both relate to one another. Most of us recognize that we must do this in textual criticism; why not in historical criticism, too?

Related Topics: Incarnation, Textual Criticism