An analysis of the G. E. Arnold survey of east Texas




Im, Hyo-jae

Journal Title

Journal ISSN

Volume Title





The purpose of this paper is to examine and explain cultural similarities and differences in east Texas through time by analyzing surface collections made by Gus E. Arnold, who was Research Archaeologist at University of Texas (UT) at Austin. Arnold collected large samples of artifacts in east Texas counties under the auspices of the UT-Works Progress Administration during 1939-40. Unfortunately, no detailed report of this work has been published. Survey notes and collections are on file at the Texas Archaeological Research Laboratory, UT at Austin. According to Arnold's survey records, a total of 255 sites was found in the sixteen counties. The majority of the sites are located along major rivers such as the Angelina, Neches and Sabine and their tributaries, or near natural lakes. As indicated by the site distributional maps (Figs. 1 and 2), recorded archaeological localities are congregated in the certain areas. For convenience, the area under consideration has been divided into three arbitrarily defined blocks of counties which conform in part with geographical zones: (1) northern section (Cherokee, Gregg, Rusk, Nacogdoches, Houston, Angelina, San Augustine, and Sabine counties), (2) middle section (Polk, Newton, Jasper, and Tyler counties), and (3) southern section (Hardin, Orange, and Jefferson counties). According to Arnold's notes, the sizes of the sites vary, but, in general, village sites and burial sites are common in the first two sections, while shell mounds are numerous in the southern section. Artifacts found from 255 sites include potsherds, dart points, arrow points and a few miscellaneous clay and stone objects. There are limitations in the interpretation of Arnold's data. First, as his collection was made on the surface, the quantity of the data are no doubt much affected by the degree of erosion and amount of vegetation on the site. Second, as he collected mainly ceramics and worked stones, other kinds of specimens such as flint flakes and remains useful for ecological studies were disregarded. Third, sampling sites from the surface are by their very nature biased as they generally do not uncover most types of archaeological features. For these reasons, Arnold's finding cannot be expected to reflect completely the prehistory of east Texas. Accurate reconstruction of this prehistory ultimately depends upon further investigation in the future