As I happen to know my Beza, I am almost sure he would never propose such a thing. And indeed, in his annotated New Testament editions, he discusses some variant readings on Luke 2:14, but there is no trace of a conjecture, in none of the five editions. So there must be an error somewhere, something “lost in transmission”.
Perhaps a small warning to the benevolent reader is in order at this moment: what follows will be a bit technical (and will even contain some untranslated Latin); my simple aim however is to illustrate what can go wrong in the transmission of conjectural emendations, and what it takes–sometimes–to untie the knot.
My standard procedure in such cases is to check the sources, and then the sources’ sources, and so on. In due course, almost invariably the case will become clear, albeit almost as invariably with some surprising elements.
In this case, van Manen’s source—which he usually deals with rather uncritically—is Bowyer’s collection of New Testament conjectures in its fourth and last edition of 1812, edited by John Nichols.2 And there indeed (p. 191) it is:
Beza takes the words ἐν ἀνθρώποις εὐδοκίας to be an interpolation; which has been refuted by Jac. Hase, Bibl. Brem. Fasc. V, p. 713.Thus at least in this case, van Manen is not to blame for the error; at most he should have thought twice about the likelihood of such an emendation.
But what is the value of Nichols’s note? As it occurs only in the 1812 edition, and contains an—incomplete—reference to a German journal (the Bibliotheca Bremensis—see below), it is most likely taken from the additions Schulz made to his German translation of Bowyer’s second edition of 1772.3 And there indeed (p. 123) it is:
Beza hält die Worte ἐν ἀνθρώποις εὐδοκίας für eine Glosse, welches Jac. Hase, Biblioth. Fasc. V, S. 713 f. widerlegt hat.As you see: the same conjecture and the same source. Nichols be thanked for adding “Brem.”, though he omits “f.” after the page number.
So the next step is to look up this “refutation” by Jakob Hase (1691-1723), professor of Philosophy at the Bremen gymnasium and brother of the more famous Theodor Hase (sons of Cornelius Hase), and who contributed some articles to the Bibliotheca Bremensis (also known as the Bibliotheca historico-philologico-theologica).
Hase’s article itself is not immediately found, for Schulz’s reference only mentions the issue (“fasciculum”) but strangely enough omits a far more essential piece of information: the volume (“classis”).
In the end—journals do not have eternal life after all—it turns out to be the following: “De glossematibus quorundam locorum Novi Testamenti quae textui, causa explicationis adscripta, hinc in ipsum verborum ordinem intrusa esse existimantur, dissertatio”, BibBrem 1.5 (1718), pp. 687-738. And indeed, on p. 713 and following Jacob Hase discusses Luke 2:2. He mentions some exegetical difficulties, and a way out, but that is only the introduction; then on p. 714 he writes:
Sed quod huc maxime facit, disco ex Clar. Hombergii ad h. l. annotationibus eundem Bezam ad Marc. IX. 43. et 45. haec notasse: ultima illa verba ἐν ἀνθρώποις εὐδοκία a paraphraste quodam in margine posita ad explicationem primum, postea in textum irrepsisse. Unde et hunc locum inter illos merito retulimus, qui glossematum postulati sunt.So that would be the source: Beza’s annotation on Mark 9. Well, I know those notes, and there is nothing on Luke 2 there, let alone Luke 2:14. There is just a note on Mark 9:43, where Beza suggests that the words εἰς τὸ πῦρ τὸ ἄσβεστον may be such an interpolation from the margin.4
Hase discovered this as well, and notes that he cannot locate the citation in the 1594 edition of Beza’s annotations. He therefore speculates that it may be found in an earlier edition:
Sed, fateor, me in mea editione annotationum Bezae, quae Genevae A. M D XCIV. prodierunt, haec istic locorum non reperisse, imo ne vestigium quidem, neque vel in notis ad hunc ipsum textum. Oportet itaque quo haec legantur in prioribus Bezae editionibus, quas saepius per δευτέρας φροντίδας edidit castigatiores incomparabilis Theologus.The point is well made, for Beza indeed did a lot of revisions over the years. But the conjecture does not occur in earlier editions.
To get closer to the solution, Hombergk zu Vach, referred to by Hase, and in particular his New Testament annotations5 will have to be consulted (the sources’s sources). Hase gives no further reference, but the work happens to be well known and even conveniently organised in the order of the New Testament texts. And indeed, on p. 122, Hombergk writes on Δόξα ἐν ὑψίστοις θεῷ etc.:
Varia excogitata sunt a Viris Doctis, quo vel excusarent hiulcum illud, vel ἀσύνδετον huius versus partem aptarent reliquis verbis. Quorum pleraque videri possunt apud Bezam in Annotationibus, et alios. Nobis quidem haec locutio et defectus copulae in hoc loco non admodum insolens videtur; Unde etiam eo devenire nolumus, ut dicamus, ultima illa verba ἐν ἀνθρώποις εὐδοκία, a paraphraste quodam in margine posita ad explicationem primum, postea in textum irrepsisse. Licet praeeuntem habemus Bezam, magnae auctoritatis virum ad Marc. 9. vers. 43. 45. Et alibi saepius. ...So here it all starts, but what does this note mean? Well, it does not mean that Beza made the conjecture; nor even that Hombergk made it in Bezan vein; no, it means that Hombergk did not want to make this conjecture, though he could point to Beza’s example for other texts such as Mark 9:43. Thus: there is no Bezan conjecture on Luke 2:14, and even no conjecture at all. However one may speculate that Hombergk’s presentation betrays a fair amount of dissimulatio; in that case the conjecture is brought in the open while Hombergk himself cannot be accused of temerity. The only reason for it, by the way, would be the problematic asyndeton in ἐπὶ γῆς εἰρήνη ἐν ἀνθρώποις εὐδοκίας.
In any case, the connection with Beza is established. The first to misunderstand Hombergk is Hase, who read Hombergk as if he cites Beza’s opinion on Luke 2:14 from his notes on Mark 9, and hence a Bezan conjecture, instead of giving an example of a critic sometimes assuming interpolations such as one might perhaps assume for Luke 2:14. Hase’s article is then picked up by Schulz, and the way he does so illustrates a typical disadvantage of such collections: important pieces of information are omitted, such as the link with Hombergk, the reference to Beza’s opinion on Mark 9:43, and also Hase’s doubt. And finally, via Bowyer’s 1812 edition, it all boils down to a short note in van Manen’s collection:
ἐν ἀνθρώποις εὐδοκία is later ingevoegd. Beza. B.
ἐν ἀνθρώποις εὐδοκία is a later addition. Beza. B[owyer].Quod non. One down, many to go.