EN
This article analyses the use of the active and middle diathesis in the Homeric speech conclusions of the type “so (s)he/they spoke” with the simplex verb φημι. It starts by listing the instances and discusses the metrical aspects of the different formulae. The article then proceeds to the previous suggestions that have been made to explain the co-occurrence of active and middle forms. The first explanation was that the middle forms were the result of a so-called Diathesenwechsel by which a middle past tense form could be used besides an active present form. The second one stated that the forms were used because of metrical necessities. The article shows that both explanations cannot be sustained, because many middle forms have a metrically equivalent counterpart with an active form and vice versa, and argues that the distinction between active and middle can be explained by the fact that the middle is used to stress the involvement of the subject, whereas the active is not. This explains why the middle is preferred when the subject is mentioned and is never used with undefined subjects.