Quite a while ago, I wrote a little about my first efforts to learn more about triangular tablets used for tablet weaving. A monograph reporting archaeological finds in the English town of Winchester has given me new food for thought on this issue. The citation is:
Section 3: The Small Finds, by H.E.M. Cool, in B.M. Ford and S. Teague, et al., Winchester, A City In the Making: Archaeological excavations between 2002 – 2007 on the sites of Northgate House, Staple Gardens and the former Winchester Library, Jewry St (Oxford Archaeology 2011).Section 3 can be downloaded from this page (scroll about halfway down and look for the PDF labeled "Small Finds").
Although no textile finds are described in section 3, there was some interesting information about textile making equipment, particularly particularly triangular weaving tablets. The report briefly reviews the history of Roman-era tablet finds in Great Britain, and observes, that, since 1970, most such finds have been triangular, not square. The author concludes, though the evidence is "scanty", that "square plates may well be a late Roman introduction in Britain." (page 16).
Even more interesting is the Winchester find itself, which consists of four square tablets and a triangular one. The author notes that the usual tablet find is of one tablet; sets are rare. Moreover, there are other British finds consisting of four-tablet sets, suggesting that a four-tablet grouping may have had a particular function in textile manufacture:
Interestingly where groups have been found together in Britain, they generally consist of four pieces. A set of four triangular tablets dateable to the late 1st century was recovered from the vicus rampart at Malton (Greep 1997, 145 nos. 15-8) and in discussing them Greep notes a similar unpublished set from Lincoln; a set of four large square tablets was found unstratified at Cirencester (Wild 1986, 114 nos. 218-221). This set from Winchester continues this pattern. Tablet weaving can make use of use [sic] large deck of tablets, a Viking ship burial at Oseburg had a part-woven linen band threaded onto a set of 52 wooden tablets (Walton Rogers 2007, 35), and four seems a rather small number for a set. The regular recovery of groups of four suggests that this number was indeed sufficient for some purposes, and it may be significant that Wild has hypothesised that some tubular selvedges that were tablet woven may have needed four tablets to create. (page 16)The author goes on to speculate about the purpose of the lone triangular tablet in the Winchester set of otherwise square tablets. This particular triangular tablet has four holes at the "base" and one at the "apex". The author consequently suggests that the four warp threads for weaving were fastened to the four holes along the base of the triangular tablet, one per hole, and the hole at the apex of the triangular tablet was used to fasten the entire assemblage to a post or the weaver's belt (probably using as separate string though the text doesn't so specify) to tension the threads for weaving.
This sounds like a plausible theory, though the comment I quoted above that mentions a different find consisting of a set of four triangular tablets suggests that sometimes triangular tablets might have been used for the actual weaving process and not just as a thread anchor. Perhaps it was routine, before the Romans introduced square tablets to Britain, to use triangular tablets to make selvages for larger pieces of cloth. When I get a chance, I should obtain four triangular tablets and do some experimenting.