Kaplan (1964) describes the now classic distinction between "logic-in-use," the more or less logical cognitive style and procedures used by the educational scientist, and "reconstructed logic," the scientist's after-the-fact, explicit formulation and idealization of the logic and procedures. However, most technical documents read as smooth, polished, error-free examples of Kaplan's reconstructed idealizations, rather than reflecting the logic-in-use which would reveal many of the most important incidents in the drama of science, which reconstructed logic leaves safely behind the scenes. The author further contends that educational inquiry suffers from a collective inability to be candid in public reports. Three examples are offered to illustrate those factors perceived as influential in suppressing honest talk about trial and error efforts in research: (1) Scientific truth is often not the preferred way of "knowing" in education; (2) Sponsors of research, development, and education are often naive about science; and (3) Interorganizational and interpersonal politics often impede or prevent full and honest reporting. (PN)