I congratulate Paul E. Newton on a thoughtful and evenhanded contribution to test validity theory. I especially appreciate the evident care that went into interpreting the various authors whose work Newton discusses. I found many useful insights along with the few minor points with which I might quibble. I comment on three aspects of Newton's article: (a) the construct/attribute distinction, (b) the idea that a decision entails an attribute, and (c) the rejection of degrees of validity and ongoing validation.