[239] The interior plateaus of southern India may be fairly characterized as the heart of the distribution of megaliths in South Asia; notwithstanding other isolated examples, it is in the southern states of India that these features are found in the highest density and diversity of form. Although megaliths were among the earliest archaeological features to attract popular and scholarly attention, and although hundreds, perhaps thousands of megaliths have been excavated, we still know distressingly little about their actual temporal range, functions, and life-histories. This paper cannot, of course, address all of these concerns, though we point to several key themes for the development of a revised understanding of south Indian megaliths. Here we draw on results of ieldwork from three major projects: irst, the Vijayanagara Metropolitan Survey (Morrison 2009; Sinopoli and Morrison 2007); re-analysis of Mortimer Wheeler’s Brahmagiri excavations in central Karnataka (Morrison 2005); and third, recent detailed excavation of both domestic contexts and megalithic features in a region adjacent to the Tungabhadra River, particularly at the early town site of Kadebakele (Morrison, Sinopoli and Gopal 2005, 2009; Sinopoli 2003; Sinopoli, Morrison and Gopal 2008). These new data, along with the results of other scholars working in the region (Bauer 2010; Sugandhi 2008), allow us to begin to formulate a more detailed and speciic sense of what south Indian megaliths looked like, when they were built, and – critically – how they were used, re-used, maintained, and abandoned. Megaliths in Northern Karnataka: Time and Variability Although known primarily for its spectacular Middle period remains, data recovered from northern Karnataka afirm a long and complex occupational history for this area (Fig. 01). Both surface remains and excavation data indicate that this region, part of the Tungabhadra/ Krishna river corridor through the dry interior of the peninsula, was home to numerous settlements, ash mounds, and other features of the Southern Neolithic and saw particularly intensive use during the Iron Age and Early Historic periods (Allchin 1963; Bauer 2010; Morrison 2009; Sugandhi 2008). We focus on these latter periods here, although we note that critical continuities between these and both earlier and later periods make this separation problematic. That is, although the developmental sequence for South India runs from Neolithic, to Iron Age, to Early Historic, to Early Medieval, what are ostensibly consecutive time periods are not only poorly deined and dated, but are also in many cases overlapping, representing typological schemes other than simple chronological ones. Here we merely point out that the Iron Age and Early Historic periods are very poorly differentiated on the basis of ceramics and in the absence of radiocarbon dates. 1 Thus, although the Kadebakele excavations are well-dated, one should exercise some caution in drawing comparisons to other sites and especially to surface remains. megalithS anD memory – eXcaVationS at KaDebaKele anD the megalithS of northern KarnataKa K athleen D. m orriSon , m arK t. l ycett , anD m uDit t riVeDi 1. Black-and-Red Ware (BRW) is common to both periods (although painted BRW is generally seen to be earlier and arguments have been made that BRW differs in fabric through time) and Russet- Coated Painted Ware (RCPW, Wheeler’s Andhra Ware) to the Early Historic. Following this scheme, we may provisionally assign a date of ca. 1200 to 300 BCE for the Iron Age and 300 BCE to 500 CE for the Early Historic, making this a very long stretch of time.