The Son is here [Ignat. Ad. Eph. vii.] declared to be gennetos as man and agennetos as God, for this is clearly shown to be the meaning from the parallel clauses. Such language is not in accordance with later theological definitions, which carefully distinguished between genetos and gennetos between agenetos and agennetos; so that genetos, agenetos respectively denied and affirmed the eternal existence, being equivalent to ktistos, aktistos, while gennetos, agennetos described certain ontological relations, whether in time or in eternity. In the later theological language, therefore, the Son was gennetos even in his Godhead. See esp. Joann. Damasc. de Fid. Orth. i.8 [where he draws the conclusion that only the Father is agennetos, and only the Son gennetos].
There can be little doubt however, that Ignatius wrote gennetos kai agennetos, though his editors frequently alter it into genetos kai agenetos. For (1) the Greek ms. still retains the double [Greek nun] n, though the claims of orthodoxy would be a temptation to scribes to substitute the single n. And to this reading also the Latin genitus et ingenitus points. On the other hand it cannot be concluded that translators who give factus et non factus had the words with one n, for this was after all what Ignatius meant by the double n, and they would naturally render his words so as to make his orthodoxy apparent. (2) When Theodoret writes gennetos ex agennetou, it is clear that he, or the person before him who first substituted this reading, must have read gennetos kai agennetos, for there would be no temptation to alter the perfectly orthodox genetos kai agenetos, nor (if altered) would it have taken this form. (3) When the interpolator substitutes ho monos alethinos Theos ho agennetos...tou de monogonous pater kai gennetor, the natural inference is that he too, had the forms in double n, which he retained, at the same time altering the whole run of the sentence so as not to do violence to his own doctrinal views; see Bull Def. Fid. Nic. ii.2 § 6. (4) The quotation in Athanasius is more difficult. The mss. vary, and his editors write genetos kai agenetos. Zahn too, who has paid more attention to this point than any previous editor of Ignatius, in his former work (Ign. v. Ant. p.564), supposed Athanasius to have read and written the words with a single n, though in his subsequent edition of Ignatius (p.338) he declares himself unable to determine between the single and double n. I believe, however, that the argument of Athanasius decides in favour of the nn. Elsewhere he insists repeatedly on the distinction between ktizein and gennan, justifying the use of the latter term as applied to the divinity of the Son, and defending the statement in the Nicene Creed genneton ek tes ousias tou patros ton huion homoousion (De Synod.54, 1, p.612). Although he is not responsible for the language of the Macrostich (De Synod.3, 1, p.590), and would have regarded it as inadequate without the homoousion, yet this use of terms entirely harmonizes with his own. In the passage before us, ib. §§ 46, 47 (p.607), he is defending the use of homousios at Nicæa, notwithstanding that it had been previously rejected by the council which condemned Paul of Samosata, and he contends that both councils were orthodox, since they used homousios in a different sense. As a parallel instance he takes the word agennetos which like homousios is not a scriptural word, and like it also is used in two ways, signifying either (1) To on men, mete de gennethen mete holos echon ton aition, or (2) To aktiston. In the former sense the Son cannot be called agennetos, in the latter he may be so called. Both uses, he says, are found in the fathers. Of the latter he quotes the passage in Ignatius as an example; of the former he says, that some writers subsequent to Ignatius declare hen to agenneton ho pater, kai eis ho ex autou huios gnesios, gennema alethinon k.t.l. [He may have been thinking of Clem. Alex. Strom. vi.7, which I shall quote below.] He maintains that both are orthodox, as having in view two different senses of the word agenneton , and the same, he argues, is the case with the councils which seem to take opposite sides with regard to homousios. It is clear from this passage, as Zahn truly says, that Athanasius is dealing with one and the same word throughout; and, if so, it follows that this word must be agenneton, since ageneton would be intolerable in some places. I may add by way of caution that in two other passages, de Decret. Syn. Nic.28 (1, p.184), Orat. c. Arian. i.30 (1, p.343), St. Athanasius gives the various senses of ageneton (for this is plain from the context), and that these passages ought not to be treated as parallels to the present passage which is concerned with the senses of agenneton . Much confusion is thus created, e.g. in Newman's notes on the several passages in the Oxford translation of Athanasius (pp.51 sq., 224 sq.), where the three passages are treated as parallel, and no attempt is made to discriminate the readings in the several places, but "ingenerate" is given as the rendering of both alike. If then Athanasius who read gennetos kai agennetos in Ignatius, there is absolutely no authority for the spelling with one n. The earlier editors (Voss, Ussher, Cotelier, etc.), printed it as they found it in the ms.; but Smith substituted the forms with the single n, and he has been followed more recently by Hefele, Dressel, and some other. In the Casanatensian copy of the ms., a marginal note is added, anagnosteon agenetos tout' esti me poietheis. Waterland (Works, III., p.240 sq., Oxf.1823) tries ineffectually to show that the form with the double n was invented by the fathers at a later date to express their theological conception. He even "doubts whether there was any such word as agennetos so early as the time of Ignatius." In this he is certainly wrong.
The mss. of early Christian writers exhibit much confusion between these words spelled with the double and the single n. See e.g. Justin Dial.2, with Otto's note; Athenag. Suppl.4 with Otto's note; Theophil, ad Autol. ii.3, 4; Iren. iv.38, 1, 3; Orig. c. Cels. vi.66; Method. de Lib. Arbitr., p.57; Jahn (see Jahn's note 11, p.122); Maximus in Euseb. Præp. Ev. vii.22; Hippol. Hær. v.16 (from Sibylline Oracles); Clem. Alex. Strom. v.14; and very frequently in later writers. Yet notwithstanding the confusion into which later transcribers have thus thrown the subject, it is still possible to ascertain the main facts respecting the usage of the two forms. The distinction between the two terms, as indicated by their origin, is that agenetos denies the creation, and agennetos the generation or parentage. Both are used at a very early date; e.g. agenetos by Parmenides in Clem. Alex. Strom. v.14, and by Agothon in Arist. Eth. Nic. vii.2 (comp. also Orac. Sibyll. prooem.7, 17); and agennetos in Soph. Trach.61 (where it is equivalent to dusgenon. Here the distinction of meaning is strictly preserved, and so probably it always is in Classical writers; for in Soph. Trach.743 we should after Porson and Hermann read ageneton with Suidas. In Christian writers also there is no reason to suppose that the distinction was ever lost, though in certain connexions the words might be used convertibly. Whenever, as here in Ignatius, we have the double n where we should expect the single, we must ascribe the fact to the indistinctness or incorrectness of the writer's theological conceptions, not to any obliteration of the meaning of the terms themselves. To this early father for instance the eternal gennesis of the Son was not a distinct theological idea, though substantially he held the same views as the Nicene fathers respecting the Person of Christ. The following passages from early Christian writers will serve at once to show how far the distinction was appreciated, and to what extent the Nicene conception prevailed in ante-Nicene Christianity; Justin Apol. ii.6, comp. ib. § 13; Athenag. Suppl.10 (comp. ib.4); Theoph. ad. Aut. ii.3; Tatian Orat.5; Rhodon in Euseb. H. E. v.13; Clem. Alex. Strom. vi.7; Orig. c. Cels. vi.17, ib. vi.52; Concil. Antioch (a.d.269) in Routh Rel. Sacr. III., p.290; Method. de Creat.5. In no early Christian writing, however, is the distinction more obvious than in the Clementine Homilies, x.10 (where the distinction is employed to support the writer's heretical theology): see also viii.16, and comp. xix.3, 4, 9, 12. The following are instructive passages as regards the use of these words where the opinions of other heretical writers are given; Saturninus, Iren. i.24, 1; Hippol. Hær. vii.28; Simon Magus, Hippol. Hær. vi.17, 18; the Valentinians, Hippol. Hær. vi.29, 30; the Ptolemæus in particular, Ptol. Ep. ad. Flor.4 (in Stieren's Irenæus, p.935); Basilides, Hippol. Hær. vii.22; Carpocrates, Hippol. Hær. vii.32.
From the above passages it will appear that Ante-Nicene writers were not indifferent to the distinction of meaning between the two words; and when once the orthodox Christology was formulated in the Nicene Creed in the words gennethenta ou poiethenta, it became henceforth impossible to overlook the difference. The Son was thus declared to be gennetos but not genetos. I am therefore unable to agree with Zahn (Marcellus, pp.40, 104, 223, Ign. von Ant. p.565), that at the time of the Arian controversy the disputants were not alive to the difference of meaning. See for example Epiphanius, Hær. lxiv.8. But it had no especial interest for them. While the orthodox party clung to the homousios as enshrining the doctrine for which they fought, they had no liking for the terms agennetos and gennetos as applied to the Father and the Son respectively, though unable to deny their propriety, because they were affected by the Arians and applied in their own way. To the orthodox mind the Arian formula ouk en prin gennethenai or some Semiarian formula hardly less dangerous, seemed always to be lurking under the expression Theos gennetos as applied to the Son. Hence the language of Epiphanius Hær. lxxiii.19: "As you refuse to accept our homousios because though used by the fathers, it does not occur in the Scriptures, so will we decline on the same grounds to accept your agennetos ." Similarly Basil c. Eunom. i., iv., and especially ib. further on, in which last passage he argues at great length against the position of the heretics, ei agennetos, phasin, ho pater, gennetos de ho huios, ou tes autes ousias. See also the arguments against the Anomoeans in [Athan.] Dial. de Trin. ii. passim. This fully explains the reluctance of the orthodox party to handle terms which their adversaries used to endanger the homousios. But, when the stress of the Arian controversy was removed, it became convenient to express the Catholic doctrine by saying that the Son in his divine nature was gennetos but not genetos. And this distinction is staunchly maintained in later orthodox writers, e.g. John of Damascus, already quoted in the beginning of this Excursus.