Third point: I’m not arguing that there’s a missing verse that says “faith alone,” and I’m not under the influence of Martin Luther’s German translation. Rather, verses that say things like “everyone who believes” will be saved should be read as referring to “everyone who believes.” And Luther didn’t slip those everyones into the text.
(PS — As a matter of history, Luther’s insert of “alone” (in Luther’s German, it’s “alone” allein not “only” nur) in Rom 3:28 is simply not and never has been the basis of the Zwinglian position that faith is sufficient to save prior to baptism. Luther’s own defense of his translation makes for an interesting read, as well as an excellent introduction to the challenge of dynamic and formal equivalence in translation.)
Fourth point: Those who argue that James overrules Paul understand neither James nor Paul. James writes,
(Jam 2:14 ESV) What good is it, my brothers, if someone says he has faith but does not have works? Can that faith save him?
The “that” is missing in the KJV, but it’s very present in the Greek, whether you follow the Textus Receptus or the critical text (the exact opposite of the Luther argument). “That faith” refers to faith that doesn’t produce works — and so on throughout James’ argument. “That faith” is a false faith, an alleged faith, because faith includes faithfulness and trust — enough trust to be faithful — as I’ve previously shown. So Paul and James do not contradict in the least. You just have to take the trouble to define your terms.
The use of the article both points back to a certain kind of faith as defined by the author and is used to particularize an abstract noun. …
In particular, the author examines two kinds of faith in 2:14-26, defining a non-working faith as a non-saving faith and a productive faith as one that saves. Both James and Paul would agree, I believe, with the statement: “Faith alone saves, but the faith that saves is not alone.”
Daniel B. Wallace, Greek Grammar Beyond the Basics (1996).