This article is an empirical study of two science and health policy controversies — "to screen or not to screen" with ultrasound in pregnancy and with mammography for breast cancer. In each case, conflicting experimental results have been published. Which of the results have been accepted within the medical science community? The article is also a theoretical and methodological study of three views of science — an institutional view, an interests view, and a semiotic view. How might each approach scientific publications as evidence? Could they be eclectically combined in a more complex view of science discourse?