By Eleanor Breen / Project Manager, Archaeological Collections Online
If you excavate a domestic feature on a historic site and use waterscreening and floatation to recover small finds, you are likely to find plenty of straight pins. This is especially true for the South Grove Midden, where archaeologists uncovered 1201 complete pins and pin fragments made of copper alloy with wound wire heads.
Straight pins were made by hand, and imported from England to the American colonies in large quantities in the eighteenth century. Pins were coated with tin, resulting in a shinier appearance before the copper core was exposed through post-depositional processes. The process of pin making was extremely laborious and included a dozen steps from pointing the shank to spinning the heads. Pointers, the workers who had the job of making the pin points, experienced what an awfully dangerous job pin making was. The pointing of shanks resulted in fine copper alloy particles that, when breathed into the lungs, caused serious damage to gums, hair, teeth, lungs, and stomach, and ultimately led to premature death.
The pin assemblage from the South Grove Midden contains 458 complete pins, 1 head, 309 heads with partial shanks, 257 shanks, and 176 tips with partial shanks. The minimum pin count from the midden contexts (calculated by adding up complete pins, heads, and heads with partial shanks) is 768! A quick comparative search of pin counts from other archaeological sites suggests that this is an extraordinarily large number of pins.
Of the 458 complete pins, there are 381 where the lengths and widths (or gauges) could be measured. As you’ll recall a from previous post, George Washington ordered different types of pins for his plantation, as did Alexander Henderson for his store. Mary Beaudry’s research on pins suggests that pin size is an indication of function. Therefore, by analyzing pin size in the South Grove assemblage, we may be better able to determine what activities led to their preponderance in this midden.
By making a simple scatterplot of width versus length (graph 1), the results are intriguing. The individual dots on the graph seem to cluster together into three groups, suggesting that most pins from the midden site fall into three different size categories. A histogram (graph 2) showing counts of pins by length yields the same results.
What’s really cool (and I can’t take credit for this, we have Wendy Miervaldis, Professor of Statistics and longtime Mount Vernon Archaeology volunteer to thank!) is if a statistical analysis called k-means clustering is used, we see three groups of pins: cluster 1 with a mean length of 28.2mm; cluster 2 with a mean length of 23mm; and cluster 3 with a mean length of 18mm (table 1).
If we revisit the orders for straight pins placed by George Washington, he consistently ordered three types: corking or larger whites; short whites; and miniken. Could those types relate to our three different pin clusters with corking measuring about 28mm, short white measuring 23mm, and miniken measuring 18mm? If so, then we have successfully linked the archaeological record to the documentary data!