Since the beginning of prehistoric archaeology, various methods and approaches have been developed to describe and explain stone artifact variability. However, noticeably less attention has been paid to the ontological nature of stone artifacts and the adequateness of the inferential reasoning for drawing archaeological interpretations from these artifacts. This dissertation takes a scientific perspective to rethink critically the ways that current lithic approaches generate knowledge about past hominin behavior from stone artifacts through experimentation (Chapter 2), and further, to explore the use of controlled experiments and uniformitarian principles for deriving inferences. The latter is presented as two case studies about Late Pleistocene Neanderthal behavior in southwestern France (Chapter 3 & 4). Archaeological reasoning is inescapably analogical, and archaeological knowledge is bound to be established on the basis on modern observations. However, simplistic treatments of archaeological analogs often result in inferences of questionable validity. In this dissertation, it is argued that greater attention is required to consider the implication of experimental design, variable control, and analogic reasoning in the construction of archaeological inference from stone artifacts. It is argued that the ability to move beyond the constraint of modern analogs in archaeological knowledge production lies in the use of uniformitarian principles that operate independently from the research questions archaeologists wish to evaluate. By examining the uniformitarian connection between platform attributes and flake morphology, the first case study explores how the production of unretouched flakes can be altered in ways that increase their relative utility, as reflected in the ratio of edge length to mass. Application of this relationship to Middle Paleolithic assemblages shows two modes of flake production pattern, possibly related to different ways Neanderthal groups managed the utility of transported tool-kits. The second case study applies a geometric model to assess the lithic cortex proportion in the Middle Paleolithic study assemblages. An excess or deficit of cortex relative to artifact volume provides an indication of possible artifact transport to or from the assemblage locality. Results show correlation between assemblage cortex proportions and paleoenvironmental conditions, suggesting possible shifts in Neanderthal artifact transport pattern and land use during the late Pleistocene.