31

" The search for major demographic characteristics of Ancient populations based on studies of skeletons extracted from archeological sites has encountered a number of obstacles. Paleodemographers have attempted (though rarely together) to avoid these problems through methodological adaptations or innovations. The recent publication of books, round table meetings, and conferences underline the renewed vitality of this discipline, thus rendering an evaluation of sources, tools, as well as past and previous hypotheses adopted by paleodemographers in France and abroad all the more interesting. First and foremost, the evolution of this discipline has been shaped by a longstanding transatlantic debate relative to the issue of estimating the age of death of adults. The validity of results presented by paleodemographers depends on the precision of data used. However, no matter the indicator, the estimated age of death is always accompanied by a troublesome margin of error. To avoid this difficulty, French paleodemographers proposed, twenty years ago, to work from a "collective age" and no longer from an individual age. Their statistical method, adopted by most European paleodemographers, is refused by American colleagues. This article offer a critical analysis of diverse methods used, examines proposed adaptations, and denounces abusive practices. In so demonstrating the impossibility of attributing an exact age based on any type of bone characteristic, the authors contribute complementary evidence to existing methods and pursue their research from a statistical approach. In recognizing the imperfections of the anthropological document it is grounded in, and clearly defining its limits, paleodemography may provide original information there where traditional demographic methods can not venture. New methodological approaches should benefit from the important corpus of data provided by anthropologists. "