Samuel Messick (1931–1998) was an American psychometrician and long-time Distinguished Research Scientist at Educational Testing Service (ETS) in Princeton. His 1989 chapter "Validity" in Educational Measurement (third edition) reorganized the modern understanding of what test validity is, and its influence reaches well beyond language testing into education, psychology, and the social sciences.
Messick's argument — now standard — is that all validity is ultimately construct validity. The older typology of "content, criterion, and construct" validities was not wrong so much as misleading: content and criterion evidence are types of support for the central construct claim. He added a further, more controversial move: validity has a consequential face, concerned with the social consequences of test use, not just the psychometric qualities of the instrument.