General
1. Method. Book Two was translated from 300 dpi page images of the 1921 printing of Beilage IV, page by page rather than in continuous batches, and verified against those images after translation. The reason for working page by page is given in note 2.
2. Why the Gospel is harder than the letters. In the Apostolikon, Harnack usually has one principal witness per verse and can quote him at length. In the Evangelion he is frequently arbitrating between four or five witnesses at once, and his prose compresses accordingly: a single clause can carry a reading, the witness for it, a competing witness, and his own verdict. At that density the first casualties in translation are negations and clause boundaries, which is why no page here was rendered without the image in front of the translator.
3. Lukan chapter numbers are locators, not a claim of priority. Harnack orders and cites the Evangelion by the chapter and verse of canonical Luke, because he held Marcion’s Gospel to be an abbreviation of Luke. That numbering is preserved throughout, since removing it would make the reconstruction impossible to cite against the scholarship of the last century. It should be read as an index, not as an argument. The chapter titles in this edition name the Evangelion first and give the Lukan range in parentheses for exactly that reason.
4. Harnack’s verdicts are kept as verdicts. Where Harnack writes that a reading is probable, or that Marcion “offered only” a given phrase, or that a witness is groundless, the strength of his claim is carried across without softening or strengthening. Where he registers his own uncertainty, the uncertainty is translated too.
5. Greek and Latin. Greek is reproduced as printed, with an English rendering in brackets at first occurrence in a note. Latin citations from Tertullian, Adamantius, Epiphanius and the rest are reproduced in italics as printed, with a translation supplied immediately after, so that a reader can see what the witness actually said and what Harnack drew from it.
Notes on the reconstruction
6. Where the Gospel begins. The reconstruction opens not at a nativity but at a descent: the fifteenth year of Tiberius, the days of Pilate, and Jesus coming down to Capernaum. Harnack notes that Marcion deleted the datings by the Jewish authorities that stand at Luke 3:1, and that both “in the days of Pilate” and “from heaven” are probable readings because Tertullian paraphrases them. This first page is the single most consequential in the volume, and it is also the one where the witnesses disagree most sharply about the destination city; Harnack’s reasoning is translated in full rather than reduced to his conclusion.
7. The witnesses do not weigh equally. Tertullian argues against Marcion’s text and therefore reports it selectively but often verbatim. The Dialogues of Adamantius preserve citations at second hand. Epiphanius works from a list of excerpts rather than a copy, which is why Harnack repeatedly finds his claims about whole letters and passages groundless. The introduction to Book Two sets this out; readers meeting a surprising reading should check which witness stands behind it before treating it as settled.
8. The Investigations are Harnack’s argument, not his edition. The two Investigations chapters and the Diatessaron appendix are continuous scholarly prose, and Harnack there defends conclusions the Assembly does not share, beginning with his opening insistence that Marcion’s Gospel is a mutilated Luke. They are translated in full and without comment. The place to answer them is a separate volume, not a footnote in someone else’s edition.
Questions left open
9. (a) Harnack’s Prolegomena, to which his manuscript sigla are keyed, are not part of the appendices translated here; sigla therefore remain unexpanded, as in Book One. (b) Several especially compressed Tertullianic clauses in the Gospel apparatus admit alternative construals; these are flagged inline where the ambiguity materially affects the reconstruction. (c) Where the printed source itself carries an apparent slip in a citation number, it is reproduced and marked “sic” rather than silently corrected, on the principle that a reader deserves the page as Harnack left it.