Academic journal article
By Petruso, Karl M.; Ellwood, Brooks B.; Harrold, Francis B.; Korkuti, Muzafer
Antiquity , Vol. 68, No. 259
Albania, isolated from Europe for nearly half a century, was closed to absolute archaeological dating during that time. New dates from an unusual large cave-site in southern Albania go beyond the single first radiocarbon date published for the country in ANTIQUITY in 1991, and permit the establishment of a radiocarbon and archaeomagnetic sequence.
Following the recent thaw in relations between Albania and the West, the Sarande Project, involving the Institute of Archaeology of the Academy of Sciences of Albania and the University of Texas at Arlington, began in 1992. Investigations at a cave near Konispol, Sarande District, southern Albania, seek to document the prehistoric occupation of this part of ancient Epirus (Petruso et al. 1992; Ellwood et al. 1993). Konispol Cave is a solution cavity c. 50 m long, 6 m deep, and 6 m in maximum interior height, near the summit of a limestone ridge some 400 m above sea level (approx. 20 [degree] 10' E, 39 [degree] 40' N). Its entrance, which faces south, commands a view of the plain below and the Strait of Corfu.
Following its first study in the late 1980s (Korkuti & Shabani 1989; 1990), we explored the cave's stratigraphic sequence and began collecting data on its palaeoenvironment and palaeoeconomy. We excavated two 2 x 2 m trenches near the rear wall, just east of the entrance, to a depth of more than 3 m; in neither was bedrock reached by season's end. Below Eneolithic, Bronze Age, and Iron Age remains, we recovered a substantial stratified sequence of pottery of the Neolithic period (c. 6000-2600 BC) representing all three major recognized phases. Although a few sherds suggest contact with Macedonia, Thessaly and Serbia, the bulk of the ceramic material is indigenous, and is paralleled at other Neolithic sites from this part of Epirus.
The radiocarbon determinations
The half-century of official xenophobia which shut Albanian scholars off from contact with radiocarbon laboratories led them to avoid citing absolute dates, relying instead on cross-dating with contiguous regions, based on pottery and other artefacts. Before our project began, a total of two radiocarbon dates had been obtained from all Albania, only one of which has been published (Guilaine & Prendi 1991). We submitted samples of carbonized material from Konispol Cave to Beta Analytic Laboratory for analysis. The results, the first series of radiocarbon samples from a stratified site in Albania, are given in TABLE 1.
The dates are internally consistent and of the right order of magnitude. The results are consistent with the radiocarbon determination published for Copper Age Maliq (Phase IIa), 5530 [+ or -] 110 b.p. (Ly-4975); Guilaine & Prendi (1991) place the Late Neolithic of Maliq (Phases Ia-Ib) within the period c. 5950-5650 b.p., which accords well with our Late Neolithic determination of 5810 [+ or -] 120 b.p. (Beta 56417). The earliest sample (Beta 56414) was taken from Trench VIII, and one of the lowest levels excavated in 1992; in this stratum, chipped stone tools were associated with mammalian fauna, but without pottery. Of the 380 lithic items recovered from the c. 1 m of aceramic deposits, most were flakes or shattered debris in chert of mediocre quality. As only 8 pieces (2.1% of the total) exhibit deliberate retouch, no informative comparison with other Balkan industries is yet possible. Three slender bladelets -- one bearing slight marginal retouch -- from the same archaeological unit as sample Beta 56414 are generally similar to the bladelets common in the late Upper Paleolithic and Mesolithic of nearby Greek Epirus, and more generally of Greece and neighbouring regions (e.g. Perles 1987: 202-9: 1989: 119-38; Adam 1989). While the lithic evidence alone is inconclusive, the TABULAR DATA OMITTED radiocarbon date would suggest that these are pro-Neolithic levels.
Our current understanding of the nature of the Mesolithic (not to mention the late Palaeolithic) in Albania is very incomplete. …