ECONOMISTS USUALLY TREAT THE KNOWLEDGE AND INFORMATION possessed by agents in a dichotomous and one dimensional way: either agents know something or they don’t. That they may surmise rather than know, feel uncertain and be reluctant to resolve this uncertainty by using certainty equivalents and a coefficient of risk aversion is sometimes discussed, but more often swept under the rug. But perhaps that is where it belongs. In cases where such a procedure allows us to resolve the questions we wish to answers (and the questions that others wish us to answer) it should be treated as a useful simplifying rather than as a simplistic procedure. And even in some cases where a one-dimensional treatment of knowledge or information does not generate satisfactory answers, it may still be the appropriate treatment because concern with the multifaceted nature of information and knowledge may just complicate the analysis without substantially improving the quality of the answer. But in some other cases, such as situations of asymmetric information, greater attention to “depth” of information may have a high pay-off. The only way to find out is to try and see. Methodological discussions can provide a useful complement but not an adequate substitute for such an empirical approach.