Prehistoric Projectile Points


Statistical Cluster Analyses

Two studies have classified Baja California projectile points on the basis of numerical taxonomy or statistical cluster analyses. These studies were done by Eric W. Ritter (1979), using points recovered in the Bahía Concepción area, and by Kelli Carmean (1994), using points in the Castaldí collection from central and southern Baja California that had been previously reported and analyzed by William C. Massey (1966).

Eric W. Ritter (1979). Ritter's study involved analysis of a sample of 90 points (excluding small triangular side-notched or serrated points, and points without modified stems). About two dozen metrical and non-metrical attributes were recorded, and a dendrogram was generated, grouping the points statistically into seven clusters, as well as sub-clusters. However, several of the clusters were assessed intuitively as not being meaningful. Ritter (1979:173) concluded, "It seems clear this computerized technique as used here is not very helpful." Consequently, he adopted instead a more traditional, intuitive classification of the points into named types.

Kelli Carmean (1994). Carmean reanalized 137 points from the Castaldí collection. She settled on six attributes to use in the analysis:

  • point length
  • distal shoulder angle, based on the orientation of the segment of the point extending distally from the distal end of the stem (shoulder orientation)
  • proximal shoulder angle, based on the orientation of the segment of the point extending proximally from the distal end of the stem (stem orientation)
  • notch opening index, based on the angle between the segments extending distally and proximally from the distal end of the stem
  • neck width
  • stem length
Using these variables, Carmean distinguished eight statistical clusters. To these she added three more clusters (numbers 9 through 11, below), based on judgmentally selected nominal variables. The total of 11 clusters included:
  • Cluster 1 - "small, high-notched" (n = 26). (Note that this cluster included "small" points with measured lengths up to 5.9 cm and point fragments that were evidently originally even larger than that. It included a classic "Desert side-notched" point, as well as large, stemmed points.)
  • Cluster 2 - "small, leaf-shaped" (n = 49). (Note that this cluster included "small" points up to 7.5 cm in length, as well as triangular points.)
  • Cluster 3 - "medium, low-notched" (n = 11).
  • Cluster 4 - "stretched-diamond" (n = 2).
  • Cluster 5 - "large La Paz" (n = 3). (The designation of "La Paz" refers to the points' contracting, generally pointed stems.)
  • Cluster 6 - "large square-based" (n = 2).
  • Cluster 7 - "medium, high-notched" (n = 15). (Note that the "small, high-notched" and "medium, high-notched" clusters overlap substantially in their size ranges.)
  • Cluster 8 - "small La Paz" (n = 8).
  • Cluster 9 - "serrated" (n = 10). (This includes notched as well as serrated points.)
  • Cluster 10 - "prismatic blade" (n = 5). (Carmean noted the possibility that these points, which were all made from chert, were brought to Baja California Sur from mainland western Mexico during the historic period.)
  • Cluster 11 - "basal notched" (n = 4).
Carmean (1994:53) concluded, "I do not necessarily wish to suggest that all eleven clusters represent tools with discrete functions....As with Ritter...I do not consider all eleven clusters to be 'good,' reliable form groups that probably related to prehistoric behavior and perception."


References Cited
Carmean, Kelli. 1994. A Metrical Study of Baja California Sur Projectile Points. Pacific Coast Archaeological Society Quarterly 30(1):52-74.

Massey, William C. 1966. The Castaldí Collection from Central and Southern Baja California. Contributions of the University of California Archaeological Research Facility No. 2. Berkeley.

Ritter, Eric W. 1979. An Archaeological Study of South-Central Baja California, Mexico. Ph.D. dissertation, University of California, Davis. University Microfilms, Ann Arbor, Michigan.