robert z selden jr, archaeology, archeology, geometric morphometric, ceramic, analysis, mathematics, statistics, 3d, 3d scan

Geometric Morphometrics – Ceramics

The study of form may be descriptive merely, or it may become analytical. We begin by describing the shape of an object in the simple words of common speech: we end by defining it in the precise language of mathematics; and the one method tends to follow the other in strict scientific order and historical continuity.  — D’Arcy Wentworth Thompson (1917)

My most recent work, and latest obsession, has focused upon the geometric morphometric analysis of Caddo bottles. The vagaries associated with the analysis of ceramic vessel shapes have long consumed analysts, and rigorous statistical analyses of Caddo vessel shapes have been a topic of considerable interest for some time. In addition to vessel shape, it has also been observed that a greater number of small vessels were included in adolescent sepultures, hinting at the possible utility for analyses of allometry.

Interactive citation network for applications of geometric morphometrics in archaeology. Access and explore the interactive version here.

Defined as “a vessel with a spheroid or oval body, surmounted by a slender, cylindrical neck,” Caddo bottles were initially seen as a somewhat homogeneous ceramic form; some with shapes and decorative motifs so similar that they were deemed to be the work of a single maker. Utility ware bottles consisted of simple bottle forms that were plain and unpolished, likely designed to meet performance needs with a coarser paste, rougher surface treatment, and thicker walls on the body than the fine wares. Harrington notes considerable variation in the size of Caddo bottles, an assertion confirmed by Jackson, who divided Caddo bottles into three groups based upon their size. Further, Harrington proposed that some bottles may have been produced using a gourd or existing bowl as a mold.

While connections have been advanced that would link the Caddo with Florida tribes, the artifact classes used to support that argument do not include ceramics. However, Krieger posited a similarity in bowl shape between the Texarkana and (possibly) McCurtain foci, and those at Pecos during Glaze periods III to V. Additional support for this argument can be seen in his prior study of Caddo carinated bowls. While not a direct correlation to bottles, if Caddo potters were indeed using existing bowls as a mold for the base and body of hand-built bottles as suggested by Harrington, it is possible that Caddo bottles—as well as the many other ceramic shapes—may bear additional evidence that could be used to test that hypothesis.

Caddo vessel forms vary through time and among groups, and reflect stylistic, functional, and social change. Elevated to a high art by female potters, Caddo ceramics “had no superiors short of the Pueblo country.” A recent evaluation of Caddo vessel shapes proposes a division of the Caddo bottle category into 27 shapes for the northeast Texas region; each with distinct temporal and spatial distributions. Novel deployments of geographic information systems are also aiding in the refinement of the probable geographic extents for known types based upon decorative motifs. One of the bottles used in this study, a Belcher Engraved bottle from Belcher Mound curated at the Louisiana State University (LSU) Museum of Natural Science, can be viewed and manipulated in the interactive 3D figure below.

Landmarks and Semilandmarks

A spline was used as the framework upon which the points would be added, and was inserted using a mesh sketch of the widest vessel profile. That spline was subsequently split at the horizontal tangent on each rim, and the remaining sections that continue into the bottle interior were deleted. The second split was added at the intersection of the spline and reference vector (center of base). Four additional splits were added at the juncture of the base/body and body/neck on each side of the vessel at the points of highest curvature. For those bottles with everted rims, the spline was also cut at the point of highest curvature; for those with direct rims, each spline was cut at the intersection of a horizontal line 10mm below the rim. The point of highest curvature used to split the spline was identified using the curvature function in Geomagic Design X, and does not represent an arbitrary location.

Spline splits for discrete components (rim, neck, body, and base) used in the GM analysis (left) segregated by landmarks (blue), with equidistant semilandmarks (white) populated between (right).

A total of nine landmarks and 46 semilandmarks divide each bottle into four discrete components corresponding to the rim, neck, body, and base. Landmarks and semilandmarks are populated along the spline, always beginning on the side of the profile determined to include the widest point. Divisions between each component articulate with those of the spline splits, where landmarks were placed at each of the points, with a series of equidistant semilandmarks between.

While sliding semilandmarks were initially thought to be a good option, there does appear to be a potential issue with the sliders introducing features–in this case, an everted section of an otherwise direct neck–to the Hickory Fine Engraved and Smithport Plain vessels that do not exist (below). Due to this issue, the analysis moved forward with equidistant semilandmarks, minus the sliding (above).

Archaeology Semilandmarks geometric morphometric Caddo bottles
Comparison of mean consensus configurations for bottle types included in the GM analysis.


The mean consensus configuration and Procrustes residuals were calculated using a Generalized Procrustes Analysis (GPA) for the aggregated sample. This initial view of the data demonstrates the degree of variability in Caddo bottles that occurs across the ceramic types. As an exploratory measure, GM methods—to include GPA—aid in clarifying shape differences, and in the production of novel a posteriori hypotheses.

Results of generalized Procrustes analysis (GPA).

Principal components analysis (PCA) was conducted on scaled, translated, and rotated landmarks, and demonstrate that the first two PCs account for 76 (PC1) and nine (PC2) percent of the variation in Caddo bottle shape. Together, PC1 and PC2 account for over 86 percent of shape variation for this sample of Caddo bottles, with all remaining PCs representing less than five percent of the variation. The first two PCs are plotted below, where warp grids represent the shape changes along PC1. This plot indicates that shape changes associated with PC1 articulate most readily with neck and body shape.

Results of Principal Components Analysis (PCA).

Additional analyses were included in this undertaking, and are omitted from this post. To see the results of those analyses, please download the posters and articles linked below.


This comparison of Caddo bottle shapes supports the hypothesis that some bottle shapes vary by type (ANOVA and pairwise comparisons). It also provides additional support for the notion that the Caddo were building at least some of their vessels with the knowledge of those decorations that would be applied (morphological integration). Further, it is proposed that calculations for morphological disparity might be gainfully employed as a measure of standardization to complement discussions of craft specialization and communities of practice. Morphological disparity may also have some utility in highlighting the development of skill through time should the earlier types be found to occupy a greater range of morphospace than those of the later types. Analytical results and hypotheses can easily evolve based upon the addition and incorporation of more data, and are not meant as a replacement for the current type definitions associated with decorative motifs.

This allows us to look at additional manufacturing and design trends concerned with specific elements in aggregate (i.e., how did bottle rim, neck, body, and base shape evolve through time?). The shape and form of ceramic vessels are assumed to be products of the same social parameters and craft traditions that influenced decorative elements. To what extent the variation in ceramic shape and form may or may not correlate with elements of decoration and design is a question that previously laid out of reach, but might now be considered using this systematic and rigorous approach.


Should it prove possible to identify a morphological transition, that could point to several interesting social possibilities including, but not limited to, (1) identifying the locus of a specific innovation, (2) the spatial and temporal dynamics of morphological variation for specific elements (neck, body, base, etc.) of ceramic design, (3) identifying or refining social networks used by specific Caddo polities/groups during temporal periods previously defined—primarily—through design-based seriations, (4) intra/inter-polity/group variation of shape and form for ceramic design, (5) potential trade relationships based upon the presence of a specific shape/form of vessel outside of known (assumed) social boundaries, and (6) the power or influence that shifted among and between polities through time. These considerations could theoretically be couched in discussions of communities of practice, craft specialization, ceramic technological organization, politics, religion, and—possibly—inter/intra-polity disputes/warfare. Furthermore, this research design has the capacity to inform greatly upon the evolution of ceramic design as it relates to the shape and form of ceramic vessels, and by adding related qualitative measures to our results we might just have the potential to bolster evidence for human behaviors associated with ceramic production and use within the ancestral Caddo territory.

Initially a development in the biological sciences, the study of geometric morphometrics in archaeology will no doubt include some interesting discussions regarding the various analytical and theoretical components that are most appropriate for a cultural system versus a biological system. There remains plenty of thinking left to do on this subject, but based on the preliminary results, the capacity for geometric morphometrics to inform upon issues related to material culture and cultural systems could be enormous.

Posters associated with this research program can be accessed here.

Publications associated with this research program can be accessed herehere, and here.

Additional publications and posters can be downloaded from ScholarWorks.

To view the beginnings of a similar program of analysis for bifaces and projectile points, click here.