One. of the most striking points about the critical discourse on film noir is that the starting dates of the films are no more challenged than are the other criteria of definition, even by those (Paul Kerr for example) who contest the latter. Certainly they are stretched back a bit (to 1941 for the father, The Maltese Falcon, or even 1940 if Stranger on the Third Floor is recalled) and forward a bit (to 1958 if Touch of Evil is to be included). A few ancestors such as Scarface or Marked Woman (on account of the final shot and of Bogart) can be found, but the charm of the genre must be respected and no one goes any further back than these few forerunners. My purpose here is to show that this way of breaking up time has no real validity whatsoever. 1945 marks the end of World War Two, which is not a cinematographic event, and 1955 marks either the appearance of the book by Borde and Chaumeton or the year that the RKO studio sold its stock of films to television, events that are totally incommensurable with the historical weight of the first date. To put it another way, what is completely strange in discourse on film noir is that the more elements of definition are advanced, the more objections and counter-examples are raised, the more precision is desired, the fuzzier the results become; the closer the object is approached, the more diluted it becomes. The result is that the energy deployed passes entirely into refuting or circumventing objections and not into searching for a more solid foundation. The cause of this situation seems to me to be a triple lack, historical, aesthetic, and theoretical: a historical lack concerning the American production of the detective film and the appearance in France of the notion of film noir; an aesthetic lack concerning the image and mise en scene (film stock and lighting, but also composition of the image and stock scenes); finally, a theoretical lack concerning the plot structure and characters of these detective fictions, including those found in the detective novel.