To God. — Our English translation has evaded the difficulty of this verse by leaving out two words. The Greek stands literally thus, “To the only wise God, through Jesus Christ, to whom be glory for ever.” “To whom,” if it refers to God, as it is decidedly more probable that it was intended to refer, is ungrammatical. If it is inserted, the words “To him that is able... to God, the only wise,” are left without government. This might, indeed, under ordinary circumstances be got over, as such broken constructions are frequent with St. Paul, but it is somewhat different in the last solemn words of an Epistle, and would be especially so if this doxology were composed by itself separately from the rest of the Epistle. There would not then be the usual excuse of haste; and for so short a passage it may be doubted whether the Apostle would even employ an amanuensis. The difficulty is heightened when we ask what is meant by the phrase, “through Jesus Christ.” Separated, as it would then be, from the ascription of glory, and joined to “the only wise God,” it would seem to be impossible to get any really satisfactory sense out of it. “To God, who through Christ has shown Himself as the alone wise,” is maintained, but is surely very forced. Our conclusion then, prior to the evidence, would be that there was a mistake in the reading, and that the words “to whom” had slipped in without warrant. And now we find that a single uncial MS., but that precisely the oldest and best of all the uncials, the Codex Vaticanus, with two cursives, omits these words. The suspicion would indeed naturally arise that they had been left out specially on account of their difficulty. But this is a suspicion from which on the whole, the Vatican MS. is peculiarly free. And, on the other hand, it is just as natural to assume that another common cause of corruption has been at work. Doxologies so frequently begin with the relative, “To whom be glory,” &c., that the copyist would be liable to fall into the phrase, even in places where it was not originally written. The probabilities of corruption may therefore be taken to balance each other, and it will seem, perhaps, on the whole, the most probable solution that the relative has really slipped in at a very early date, and that the English version as it stands is substantially right. There are some exceptions to the rule that “the more difficult reading is to be preferred,” and this is perhaps one.
The subscription in its present form hardly dates back beyond the ninth century. The earliest form of subscription up to the sixth century was simply “To the Romans.”