Sandal relationships

Collection Inventory and Technical Descriptions

Current Straw Sandals Phylogenetic Tree

Current Straw Sandals Minimum Spanning Tree

Introduction

The purpose of the analysis of sandal relationship provided here is to enable scholars to describe footwear in a consistent, quantitative fashion that permits the application of mathematical tools to rigorously estimate the global relationships among specimens.  In the future, it will be possible to map the sandal construction relationships established by these methods onto genetically determined routes of ancient human migrations. The site was developed collaboratively by Lawrence E. Hightower, Ph.D. in Biochemistry from Harvard University, and Barry G. Hall, Ph.D. in Genetics from the University of Washington.

The Straw Sandals Project began when Larry Hightower, an emeritus professor of molecular and cell biology, connected his growing interest in the genetics of human migrations with his experiences visiting the ancient Silk Road in China. Early human migrations occurred over the rugged terrain of the high passes of the Pamir Mountains and across the Tibetan Plateau into the Gobi Desert. Dr. Hightower wondered “What did they wear on their feet that allowed humans to cross such forbidding terrain?” That curiosity led him to collect and study woven shoes and straw sandals. He is currently Curator of the Straw Sandals Project located at the University of Connecticut.

Straw sandals tend to be described in detailed prose and related to local cultures and materials, but there is little that permits the comparison of footwear across time, locales, and cultures in a way that establishes a rigorous relationship between footwear and human migrations. Enter Barry Hall, an emeritus professor of biology and Director of the Bellingham Research Institute, with broad interests in biology including phylogenetic analysis. He noticed that efforts have been made to cluster sandals on the basis of those descriptions and to relate those clusters to cultural groupings (1) and population movement (2) in the U.S. Great Basin, but those efforts have not extended to the consideration of global relationships among sandals to show a quantitative relationship between footwear structure and human migration. If those descriptions can be both objective and consistent then similar features and details can be used to cluster specimens. In contrast, the detailed differences can be used to distinguish individual specimens.

A necessary prelude to assembling a database useful for phylogenetic analysis is the establishment of a set of characters and character states representing hand crafted, natural fiber footwear technology based on a standard terminology. Fortunately, there are terminology resources available including The Primary Structures of Fabrics (3) by Irene Emery, Basketry Technology (4) by J. M. Adovasio and “Anasazi Fibrous Sandal Terminology” (5) by Ann Cordy Deegan.

The character descriptions of individual specimens in the standardized database can then be converted to a metric that is a quantitative, numeric, measure of similarity that can be used to define the distance of one specimen from another. One can then use various mathematical methods to estimate the relationships among the specimens. Biologists deal with exactly this problem to estimate the relationships among genes and proteins in different species. Sandals, like proteins, inherit many properties and features, not because of genetics, but because of “cultural inheritance”, the tendency of handcrafters to make objects in much the same way that their predecessors did. Properties change over time, for example in response to the availability of weaving materials; in effect, adaptation.

(1) Barker, P. 2008 Woven Sandals as Boundary Markers Between the Great Basin and Southwest Culture Areas. in Past, Present and Future Issues in “Great Basin Archaeology: Papers in Honor of Don D. Fowler”, Bryan Hockett, Ed. Cultural Resource Series No. 20 2009 U.S. DEPARTMENT OF THE INTERIOR Bureau of Land Management NEVADA)

(2) Adovasio, J.M., with R.L. Andrews. 1983. Artifacts and Ethnicity: Basketry as an Indicator of Trerritoriality and Population Movements in the Prehistoric Great Basin. In Anthropology of the Desert West: Essays in Honor of Jesse D. Jennings. Carol J. Condie and Don D. Fowler, eds. Salt Lake City: University of Utah Press.

(3) Emery, Irene. 1994. The Primary Structure of Fabrics. The Textile Museum, Washington, DC, 2009 paperback edition published by Thames and Hudson, New York. This book elaborates a comprehensive system of terminology for fabrics. It has found broad acceptance and applicability.

(4) Adovasio, J.M. 2010. Basketry Technology. Updated edition. Left Coast Press, Walnut Creek, CA. This paperback book is a guide to the identification and analysis of a range of handicrafts classified as basketry. The terminology is essentially an adaptation of Irene Emery’s general fabric terminology.

(5) Deegan, Ann Cordy. 1993. Anasazi Fibrous Sandal Terminology. Kiva 59(1): 49-64.

(6) Andrews RL, Adovasio JM, Carlisle RC. 1986. Perishable industries from Dirty Shame Rockshelter, Malheur County, Oregon. University of Oregon Anthropological Papers 34.

Estimating Relationships Among Sandals

Relationships among sandals are based on the pairwise distances between the specimens. The pairwise distance is simply the fraction of characters at which the pair of species differ.
There are a variety of methods that have been developed, primarily by biologists, to estimate the relationships among individuals. Those that are most applicable to estimating relationships among sandals are Phylogenetic Trees and Minimum Spanning Trees

Phylogenetic Trees

Phylogenetic trees are familiar to anyone who has seen diagrams that illustrate the relationships among mammals, birds, reptiles, fish, insects, sponges, etc. Phylogenetic trees usually represent relationships among biological species and their hypothetical common ancestors. The tips of the trees represent the species, the vertical lines represent hypothetical ancestral species, and the horizontal lines represent branches connecting ancestors to their descendants. The lengths of the horizontal lines are proportional to the distances between ancestors and descendant.

Biological species inherit characters from their ancestors, some of which change by mutation as those species diverge from common ancestors. Sandals likewise inherit characteristics from preceding cultures as craftsmen pass on their skills to new generations. As a people with a particular sandal making tradition divide and migrate to different destinations, sandal making traditions change to meet new demands (terrain, climate) and to accommodate to availability of different materials. The same methods used to diagram relationships among biological species can be used to diagram relationships among sandals. We use the program MEGA 7 to estimate Neighbor Joining trees of sandals. See the current Straw Sandals Phylogenetic Tree.

Minimum Spanning Trees (MST)

A phylogenetic tree is only one way to represent relationships among sandals, a way that is based on the assumption of identity by descent. Phylogenetic analysis assumes isolation sandal making traditions after they have diverged from each other. That isolation seems unlikely. It is more likely that cultures sometimes exchange sandal making techniques after they have substantially diverged. That exchange leads to loss of phylogenetic signal and inaccuracies in the tree. Minimum Spanning Trees are an alternative way to represent those relationships based only on identity by state. Minimum spanning trees are not affected by exchanges of techniques among cultures. Neither way of representing those relationships should be thought of as “truth”. They are simply different ways of estimating relationships.  

MSTs connect specimens in such a way as to minimize the total length of the tree while ensuring that there is a single path from any specimen to any other specimen. The MST below is of a small set of sandals. The connections between individuals are called edges and are labeled with the distance between the individuals (decimal numbers). Sometimes there is more than one equally “minimal” tree. In that case we show a consensus of those trees in which edges that are present in < 50%of the trees are not shown, edges present in 50% – 89% are shown as dashed lines, and edges that are present in > 90% of the MSTs are shown as solid lines 

 

History of the Catalog Entry Form