It's also the case that Church only knew that the notion of function defined by the lambda calculus was coherent once the Church-Rosser property was proven, and that was not published before Turing submitted his article.