Empiricism (especially in its behavioralist variety) is premised on the ability to collect data and draw inductive conclusions from the data. Epistemology is t

Empiricism (especially in its behavioralist variety) is premised on the ability to collect data and draw inductive conclusions from the data. Epistemology is the study of how things can be known, what can be known, or what the limits to knowability are. How can we justify a belief based on a collection of observations? davak began with the assertion that testosterone is linked to a number of health problems. mironathetin notes that studies of the negative health implications of testosterone were often performed on professional athletes, who are going to have high baseline levels of testosterone naturally, and who have in all likelihood supplemented with large doses of testosterone at various points in their career/training. But, mironathetin notes that positive lifestyle changes can increase your body's natural production of testosterone without the need for supplementation. golemotron is skeptical that naturally increasing testosterone will lead to different health implications than the studies indicated. They posit that there is a direct relationship between testosterone levels and health risks. vannevar responds that because the body is a system, non-linear and feedback effects are much more likely to control the health effects of testosterone than a simple linear relationship between testosterone and health risk. kingmanaz agrees and calls golemotron's position reductionist, which means that golemotron implicitly believes that one can understand a phenomenon by understanding and measuring each of its component parts (empiricism). An epistemological critique of golemotron's position would begin with an assertion that the whole is different from the sum of the parts, because the parts interact with each other, producing non-linear effects. One must understand the gestalt, not just the parts. Induction straight from data will often miss things that holism might not. For an accessible critique of data and inductivism, you might pick up a copy of Nassim Nicholas Taleb's Fooled by Randomness. If you are interested in systemic non-linearity in political science, check out Robert Jervis's System Effects.