17 Who Was Manifested in the Flesh? A Consideration of Internal Evidence consider the possibility of accidental corruption, though ὅς at first seems more likely to have given rise to the other readings, it must be recognized that an exemplar in poor condition, where specific letters are easy to confuse, as in the case of MSS F and G, can also explain the confusion of Θ̅C̅ and OC. While accidental changes seem to be the least subjective and the easiest to identify, determining which changes are easier as scribal creations can in no way be described as an objective process. One can construct scenarios for nearly any variant by merely presupposing the attitude and aptitude of a scribe. A most conscientious scribe, faithfully copying what is before him will be prone only to accidental errors. The theologian may fall to the temptation to “clarify” a text with Θεός in a way that a gram- marian would not, while the Greek-speaking scribe may not believe that sacred writ could contain a pronoun disagreeing with its antecedent’s gender. It is not always possible, therefore, to decide this variant on the grounds of “most difficult reading” without unwarranted and improvable speculations. Determining the most appropriate reading for the style of the author or the genre of the text seems to have more promise. From the data pre- sented, it is obvious that neither the Christ hymn as a form, nor Paul’s use of it in his texts is adequately described by saying of the relative pronoun “It is a typical way to introduce a hymn ... and it is not necessary to locate an antecedent in the text.”47 Hymns were introduced in a variety of ways, and it is indeed tenuous to dogmatically state that any Pauline use of a Christ hymn must be introduced with a relative pronoun. There is more than sufficient evidence that Θεός would have been appropriate here. Dr. J. Keith Elliott, in a different context, himself asks the broader question with which we must conclude this discussion: just how far can thoroughgoing eclecticism go in determining the original text of the NT?48 We might translate this question to ask, in light of the ambivalence with which the data may be interpreted, is it possible or responsible to ignore external evidence in the practice of textual criticism? From the data pre- sented for 1 Tim 3:16a, we conclude that it is futile to do so. As Gordon Fee points out, “... very often internal questions either are indecisive or sometimes collide.”49 For example, it may be said that choosing a variant that is most in line with authorial style by definition must inherently conflict with the reading that is the most difficult for the scribe, unless Mounce, Pastoral Epistles, 214, n. d. 47 Elliott, The Greek Text of the Epistles, 10. 48 G.D. Fee, “Rigorous or Reasoned Eclecticism – Which?”, SBL Seminar Papers 2 (ed. 49 George MacRae) (Missoula, MT 1975) 44.