Stephen W. Frary 16 passage seems to be complete in itself, it is far too brief and its lines too interrelated to lend itself to a division into stanzas, ... Moreover, passives have not previously been encountered in these hymns in reference to the redeemer, and, except for Phil. ii, 9-11, the redeemer has always been the explicit subject.45 If these observations are decisive, and we are not dealing with a hymn in this passage, then we should not in fact expect a pronoun, relative or other, to introduce the passage, and the case for Θεός as the original read- ing is strengthened. On the other hand, if it is a hymn, in spite of these objections, then we should expect an explicit reference to the redeemer as the subject, with “God manifested in the flesh” being the most likely. The second objection concerns the provenance of the pronoun in those hymns which have relative pronouns with which they introduce the divine subject. When they are identified as typical of this genre, it is generally assumed that the pronouns are a part of the hymn itself. Gloer correctly notes though that their presence “...implies at least one preceding phrase which may have taken the form of a ‘eulogy’ (‘Blessed be God’ or ‘Christ’).”46 In other words, the pronoun may not be a part of the hymn at all, but rather an insertion by the author quoting it to relate the subject of the hymn to a person already under discussion in his own text. Thus each author using a hymn fragment may in fact have his own distinct way to join the hymn to his text. Reviewing the previous chart and disregarding 1 Tim, however, we see that Paul makes this integration in a variety of ways, using a pronoun three times, a noun five, and the person-number suffix of a finite verb once, showing no clear pattern. Conclusion Having considered the possibilities of accidental or intentional changes and the congruence of the possible variants with the hymn genre and Paul’s way of quoting them, what are we left to conclude about 1 Tim. 3:16a particularly, and the value of internal evidence in general? Clearly, the easy way in which most text critics dismiss the internal evidence (Fee, Metzger) or assume its attestation for ὅς (Elliott) is unfounded. There are far more data to consider, and their verdict is not unanimous. If we Sanders, Christological Hymns, 16. See table above for which passages Sanders con- 45 siders hymns. Gloer, “Homologies and Hymns”, 128. 46