Abstract
This paper attempts to define the concept of placebo as it is used in the clinical context The author claims that X is a placebo if and only if X has such a property dp, that whenever in a therapeutic situation T a stimulus S appears, then in attending conditions A, it will cause a beneficial reaction R in the patient. Formally, the same structure may be used to define any pharmacologically active drug. The main difference between the drug and a placebo is in the range of possible substitutions for X and the property d. For the active drug there is only one possible substitution for X and property d and it can be scientifically explained why, and how the drug works. In the case of a placebo a set of possible substitutions for X and d is open, and so far it is impossible to offer any scientifically valid explanation of the action mechanism of placebo