文摘
The thorough reconstruction of subsistence practices throughout human history remains one of the most challenging questions in archaeological research. Analyses of microbotanical remains recovered from archaeological artefacts have greatly contributed to our knowledge of past livelihood strategies. However, certain methodological issues are seldom addressed throughout these analyses, including the integration of multiple proxies, the comparison between samples and the interpretation of control samples. This paper addresses these methodological concerns through the analysis of phytoliths and starch grains from a total of 80 samples from grinding tools from four archaeological occupations (ca. 7150–1900 cal bc) in northern Gujarat (NW India). The results were compared with 26 control samples from the same sedimentary matrix from which the tools were recovered and 12 control samples from laboratory consumables. Multivariate statistics were applied to (a) compare control samples with grinding stones to assess sample contamination and representativeness, (b) compare samples from different sites, and (c) identify tool clusters within a site. This study stresses the importance of the integrated analysis of phytoliths and starch grains and the application of multivariate statistics, which allow for stronger interpretations on the use and post-depositional trajectories of grinding stones, thus offering a solid framework for the reconstruction of past subsistence strategies. Moreover, the results show that the inhabitants of northern Gujarat continuously exploited small millets throughout the Holocene and that pulses, secondary at first, became a fundamental part of their subsistence strategy with the advent of settled life.
