Symbolism Without Symbols? The Unsoundness of the Artifact to Symbol Inference in Paleolithic Archaeology

Behavior mediated by symbols—commonly termed ‘symbolic behavior’—is widely recognized as a defining feature of human cognition. In deep history, its emergence is frequently inferred from artefacts or features that appear to lack utilitarian function and are thus interpreted as ornamental or representational artefacts potentially carrying symbolic meaning. However, such interpretations have been contested and have become a topic of increasing scrutiny due to the inherent ambiguity of both the archaeological data and the conceptual frameworks used to analyze them. Despite these critiques, few alternative models to the prevailing interpretations have been proposed. A central issue lies in the difficulty of defining and identifying ‘symbols’ in Paleolithic contexts. This paper explicitly outlines and interrogates this problem. We argue that dominant theoretical frameworks fail to establish a coherent inference chain from ‘artefact to symbol’. These approaches are often either ad hoc or grounded in Peircean semiotics, which relies heavily on the role of the interpretant, a process that poses significant challenges in prehistoric contexts, given its dependence on knowledge of social conventions. As such, we argue that archaeologists often fail to reconstruct symbols, but rather construct new ones. Moreover, if the criteria for identifying symbolic behavior were consistently applied across species, many non-human animals would qualify. This undermines the analytical utility of symbolism as a marker of humanness. We contend that the prevailing symbolic framework is symptomatic for a deeper issue, namely a longstanding research tradition marked by a problematic nature/culture dualism. To move beyond these limitations, we advocate for a longue durée perspective that integrates recent developments in evolutionary theory, particularly the Extended Evolutionary Synthesis, with insights from the philosophy of technology, especially the work of Simondon and Boëda. In doing so, we aim to move away from the fixation on discrete symbolic objects and toward a more processual and theoretically robust understanding of early human behavior.