established—the one being typified as religious and the other as more speculative.4 In our century, new life was brought to Luther studies by Karl Holl,5 who was aided immeasurably by the discovery of Luther’s lectures on Romans given in 1515 and 1516. One of Holl’s first conclusions drawn from these and other documents was that Luther taught an “analytic” justification. This shattered the comfortably established opinion of Ernst Troeltsch that Luther’s teaching was a mere “re-modeling” of medieval