The use of ethnographic analogy to interpret archaeological remains has produced many misunderstandings, which must now be corrected. 1. Ethnology is traditionally oriented towards the analysis of the thought systems of the populations under study, and believes all too often that this type of discourse in natural language is an acceptable explanation for the observed empirical phenomena. The scientific discourse built by the ethnoarchaeologist must not imitate the distinctions made by the people under study, given that their constructs seek to satisfy different objectives. 2. The construction of inference rules must be subjected to the requirements of all scientific research. In consequence, one must not merely collect CAS D'ESPECE, but also assemble numeric data which are sufficiently representative for statistical treatment to be carried out. 3. It is necessary to define, in each case, the actualisation context of the rules, that is, the spatial and temporal universe wherein the proposed rule is applicable. 4. The only way to validate a rule resides in the precautions taken during collection, mobilisation and treatment of empirical data. 5. Successful application of an actualist rule to archaeological data does not mean that the latter has been validated. The only way to confirm an interpretation is by applying the principle of result convergence by independent methods.