What is validity? A simple question but apparently one with many answers, as Paul Newton highlights in his review of the history of validity. The current definition of validity, as entertained in the 1999 "Standards for Educational and Psychological Testing" is indeed a consensus, one between the classical notion of attributes, and measures thereof, and Cronbach & Meehl's (1955) proposition that validity should be concerned with interpretations of test scores and not with the tests themselves. Newton is certainly right when striving for a clearer and more precise definition of validity than the "Standards'" definition; and in many ways Newton succeeds in achieving this goal. In both Newton's proposed clarification as well as the current "Standards'" definition, validity is closely associated with "measurements" of certain "attributes". For many researchers in psychology, these terms are intimately associated with the measurement model that underlies the generic latent variable model. In earlier work, the author has postulated a new theoretical framework in which psychological phenomena like "major depression" (MD) are "networks" in which the symptoms are not caused by a latent attribute but, instead, are directly related. In this article, the author discusses validity from a network perspective. (Contains 2 figures.)