Noting that librarians have long debated whether specific library functions may, or should, be scientifically examined, or whether they constitute an art not susceptible to scientific study, this paper reviews and critiques attempts to develop models of library material selection using the theories and methods of scientific investigation. First, the nature of scientific theory and its application to research methods are reviewed, and three levels of adequacy, descriptive, explanatory, and predictive, are discussed with illustrative examples. Several published models of the selection decision for acquisition are examined in terms of their theoretical adequacy, and the nature of theories describing potentially optimal practices instead of existing methods is examined. It is concluded that much of the research which has been done in book selection does not achieve any of the levels of adequacy described in this paper, and the hope is expressed that researchers will examine the nature of scientific theories and the criteria by which they are judged so that future work may be more scientific, and thus more productive. (29 footnotes). (Author/EW)