By W. M. Ramsay
The Mediator (Gal 3:20)
“The Law was ordained through angels by the hand of a mediator. Now, a mediator is not of one, but God is one.”
The precise meaning of the argument that lies in the words of Gal 3:20 is very difficult to catch; and I shall not attempt to add one to the 250 or 300 interpretations that have (according to Lightfoot) been proposed for this passage. We have in § XXXVI found a case where Paul sees the right result, and yet attains it by an argument founded on the generally accepted, though mistaken, view of that period, that grammatical forms had a deep philosophical meaning (usually assigned on arbitrary and capricious grounds to suit some individual instance). Is it not the case here also, that he aims at a right result, but reaches it by a bad process of reasoning?
Paul is evidently emphasising a certain contrast that exists between the free grace of the Promises and the indirect character of the Law — the Law being merely a means to an end beyond itself, and not being the sufficient and ultimate gift of the grace of God. The distinction is undeniable and of immense importance. In this paragraph, therefore, he does not use the word Diatheke to indicate the “covenant” made with Abraham. In accordance with the distinction drawn in § XXXIV, it is necessary for him to use the word “Promise,” ἐπαγγελία, in order to emphasise the character of freedom and grace in the covenant made by God with Abraham and his seed. Accordingly, the words “Promised” or “Promises” occur three times in the short paragraph Gal 3:19, Gal 3:21-22: the Greek text has the verb instead of the noun in Gal 3:19, where the English translation, if literal, would be “the Seed to whom it hath been promised”.
The Law did not come immediately and directly from God to men. It was conveyed from God by angels; and a mediator, viz., Moses, carried it down from the Mount to the Hebrew people. This method is far less gracious and kind than the direct communication from God to Abraham; and brings out the consciousness of an impassable gulf separating God from even the chosen people. The allusion to the angels seemed founded more on Rabbinical interpretation and later tradition than on the text of the Books of Moses; but the words of Stephen (Act 7:53) and of Herod in Josephus,1 as quoted by Lightfoot and commentators generally, seem to imply that the common belief of the time supposed the ministry of angels.
A mediator implies one who goes between two parties to an agreement, and therefore to a certain degree might seem to diminish the absolute authority and completeness of the one party in this case. Can this, then, be the sense of the last words of Gal 3:20, “but God is one”. So Lightfoot thinks, and so it may be. But it seems an unsatisfactory form of expression; and I cannot avoid the suspicion that Paul here is betrayed into a mistake, and is thinking of the other and infinitely more important sense of the words, “God is one” — as in Rom 3:30. — “He is one and the same God in all His acts, one God makes both the Promises and the Law.” The argument would then be a fallacy, “a mediator implies (two parties), but God is one”. I may probably be wrong; but, if one speaks, one must say what one thinks. Here, while Paul aims at a great truth, he reaches it, I think, by a mistaken argument.
We have here, as recognised in the translation (repeated by Zöckler and others, and not disputed by Lightfoot, but, seemingly, recognised by him as the obvious sense), a clear and apparently undisputed example of a participle used in the sense of καί with a finite verb: “The Law was added because of transgressions, till the Seed should come to whom the Promise had been made, and it was ordained through angels etc.,” where the Greek has merely the participle “being ordained”. But, distinctly, the giving of the Law by God is the first step, and the carrying into effect by means of angels is the following step. This is one of the many examples justifying the construction διῆλθον . . . κωλυθέντες in Act 16:6, in the sense which I have pleaded for, “they traversed . . . and were prevented,” That loose usage of the participle is common in the later Greek and Latin.
 Ant. Jud., XV 5, 3.