DISCUSSION
The Smithport Landing Site is one of a number of village and mound sites along the Red River valley and its tributaries in northwestern Louisiana ([Fig. 1]) at which varying amounts of Alto Focus pottery, whole vessels or sherds, have been found. The mound sites shown are within the river flood plain, with exception of Thigpen Mound and Village Site, which are on a terrace immediately overlooking the valley; Gahagan, Curtis, Mounds Plantation, and Belcher mounds are on old river channels near the present stream. The burial vessels at Gahagan were Alto types—five Holly Fine Engraved, three Hickory Engraved, one Kiam Incised—and 12% of the 76 sherds from the surface are the distinctive Alto types (Hickory, Carmel and Holly Engraved, Davis Incised, Crockett Curvilinear Incised, Weches Fingernail Impressed and Pennington Punctated-Incised). The Thigpen Site is preponderantly Bossier, but included in the scant collection of 102 sherds are one Weches, five Dunkin, and five Wilkinson. We have only a few sherds from the Curtis Mound (Sunny Point in Moore’s 1912 report) but Hickory Fine Engraved is included. At the Belcher Mound Site (Webb, 1959) the premound level had sherds and burial vessels of both Alto and Haley types.
The Mounds Plantation (Pickett Landing in Moore’s 1912 report) Site has recently been explored with some intensity (McKinney, Plants and Webb, to be reported). Twenty-six percent of the decorated sherds in the previous surface collection were of the distinctive Alto types, 4.15% Coles Creek. A trench through one of the mounds showed intrusive Belcher Focus burials but the fill, habitation, and premound level sherds were Coles Creek and Alto, with admixture at all levels but increasing amounts of Alto in the top levels. Alto types are Davis and Harrison Bayou Incised; Hickory, Holly and Carmel Engraved; Pennington, Crockett, Wilkinson, and Weches in the punctated and punctated-incised categories. Coles Creek types were Coles Creek, Chase and Beldeau Incised; Rhinehart Punctated; and the shared types Hardy and Sanson Incised. Deep burials in a second mound had scant pottery but the two vessels were Holly Engraved and a bowl with Crockett and Pennington designs, both black and polished.
The non-mound village sites shown in [Figure 1] are on hills fronting the valley or on tributaries and lakes. All of those shown have Alto and Bossier pottery types, most have Coles Creek-Troyville, all have a good representation of the shared types Hardy-Kiam, Dunkin-Manchac, Harrison Bayou and Sanson Incised, Wilkinson Punctated, and Rhinehart-atypical, Pennington Punctated-Incised. Omitting these shared types and using only distinctive types, the Allen Site has 7% Alto, no Coles Creek; the Wilkinson Site has 10.5% Alto, 0.3% Coles Creek; the Chamarre Site has 14% Alto, 1.5% Troyville; Williams Point has 4% Alto, no Coles Creek; East Smithport has 8% Alto, no Coles Creek; the Colbert Place has 1.6% Alto, 5.45% Coles Creek; Greer has 6.7% Alto, 1.8% Coles Creek; Pease and Sinner are strong Bossier sites but have 0.3% and 0.5% Alto, respectively; Swanson’s Landing has 4% Alto and 4% Coles Creek; and Harrison Bayou has 1% Alto.
Not all of the mound or village sites in this same area show this kind of representation of Alto or Coles Creek; there are as many or more which are well developed Bossier sites and have little or no Alto. For example, we have 230 surface sherds from the Vanceville Mound in Bossier Parish with no Alto or Coles Creek types; the 3942 sherds from the lower and premound levels of the Oden Mound include two questionable Hickory Engraved, no other Alto or Coles Creek types; 1275 surface sherds from the Marston Village Site show no Coles Creek, one Holly Fine Engraved, and three Pennington Punctated-Incised. In these same sites, as the distinctive types drop out, the shared incised and punctated types like Dunkin-Manchac, Rhinehart-Pennington variants, Wilkinson, Harrison Bayou Incised and even Hardy-Kiam Incised are almost completely replaced by Pease Brushed-Incised, Belcher Ridged, and the brushed types (Webb, 1959). Large projectile points and heavy scraper types also disappear, replaced by small arrow point types and thumbnail-size, triangular and rectangular flake scrapers (Webb, 1959: Fig. 126).
The Smithport Landing Site shares with the other hilltop or hill slope village sites of this earlier Caddoan period the carry-over of late Archaic dart points, especially types like Gary, Ellis, Kent, Carrollton, Palmillas, San Patrice, Evans, Maçon, and Pontchartrain. Large as well as small scrapers, pitted stones, manos of hand size, oval metates, small drills, large and small celts, brown and white sandstone hones, hammerstones, and crude choppers are usual at these sites. Triangular and ovate knives, recurved-edge (Copena-like) knives, stone beads and polished stone problematicals (boatstones, bannerstones, gorgets) or plummets are all missing or very rare, although stone beads and problematicals occur in the late Archaic. The slate bead from Smithport, a recurved (Copena) blade fragment from the Thigpen Site, and a two-hole gorget from a small site north of Wallace Lake (Webb, 1948: Pl. 16, 9) are exceptions. Small projectile points, generally of Alba and Colbert types, about equal the number of large ones at these sites. Ear ornaments, shell and bone tools are infrequent.
In conclusion, the Smithport Landing Site is one of the larger village sites of the earlier Caddoan (Gibson Aspect, Alto Focus) period along the Red River valley in northwestern Louisiana. It shares with a number of other village sites of this period evidences of a carry over of late Archaic projectile points and stone artifact traits. It also shares with numerous village and mound sites evidences of admixture of Coles Creek ceramic types and influences with the Alto pottery types as the earliest pottery at these sites. It seems increasingly clear that the advent of Coles Creek and Alto Caddoan peoples and/or ceramics, arrow points, and riverine mound building into this area were virtually simultaneous occurrences. Out of this blending developed the subsequent Bossier Focus ceramics and other cultural manifestations over a wide portion of northwestern Louisiana, extending into eastern Texas and southern Arkansas.
It is possible that the large ceremonial mound groups, like Gahagan and Mounds Plantation, served as ceremonial centers for a number of villages, including those in the adjoining hill areas, accounting for the frequency of specialized burials, with ceremonial copper and polished stone objects, pipes and ornamentation, and highly developed burial ceramics, in the mound sites, in comparison with the paucity of these objects in the hill villages.
Considerable research is needed (1) to establish the nature of the relationships between mound sites in the valleys and the villages in the hills; (2) to trace the extent of Coles Creek and Alto contacts and the process of amalgamation of these two strong cultures over the wide area from central Louisiana into Arkansas, Oklahoma and eastern Texas (this must have been friendly, as it is inconceivable that Caddoan peoples would have supplanted Coles Creek almost overnight in hundreds of villages); (3) and the development out of this amalgamation of Bossier, Plaquemine, and other later cultures.