Canadian Historic Sites: Occasional Papers in Archaeology and History No. 2

An Archaeological Study of Clay Pipes from the King's Bastion, Fortress of Louisbourg

by Iain C. Walker

Statistical Dating

The Binford formula applied to the material from Casemate 1 Right gave considerable help when studying that material in relation to the archaeological sequence. When applied to the right face casemates, however, the formula resulted in dates ten or more years later than had been expected from the historical evidence, and, to some extent, from the study of the pipe material itself, as the figures above show.

As an experiment, the material was also measured with the grooved end in the two ways described in dealing with the Casemate 1 Right material. The same sequence of dates was observed, but in this case it was the earliest dates that provided the only ones acceptable on other evidence. The grooved end, loose fit date was a year or two earlier and the grooved end, tight fit date was ten or more years earlier still.

As an additional experiment, the material from Casemate 14 Right, which had been excavated during two seasons and had been given different catalogue numbers for each season, was divided into these two groups and the dates calculated. The 398 measurable fragments from the earlier material, catalogued 4F, gave a date (using the butt end of the drill) of 1740.55; the later material, catalogued 4X, and amounting to 329 fragments, 1736.68.

The Binford date is the theoretical median point in the occupation of the site which, for this area, would be 1726. Binford, however, pointed out several practical difficulties in the determination of the true median date, the most obvious of which in the area under discussion is that the material need not have accumulated at an even rate over the 12 or so years that the area was in use. Nevertheless, dates somewhere between 1720 and 1732 should have resulted from the analyses shown, while in fact the dates derived from the orthodox measurements were between six and ten years too late at a minimum.

From the historical evidence it seems reasonably certain that these casemates cannot have remained open beyond 1732. The pipe marks do not give any absolute evidence one way or the other, although the absence of the Gouda coat of arms on any of the Dutch bowls strongly suggests that they date to earlier than 1739-40; so we are thus left to conclude that the Binford formula, which is entirely a statistical method of deduction, is probably in error here.

48 Six examples of Oswald's type 9 English pipe shapes. Context: 1720-32.

Harrington's principle was based on the fact that bore diameters on pipes between about 1620 and 1780 diminished fairly steadily in size coinciding with the fashion of in creased stem length (cf. Oswald and James 1955a: 188). As Harrington noted (1944: first page), however, the short-stemmed "dudeen" in Ireland and the "cutty" in Scotland were in use by the end of the 17th century (Deane 1914; 5; cf. Jewitt 1863: 76-7; 1878, 1: 293; and Thursfield 1907: 163) and the use of long and short pipes appears to have been a matter of social custom — it certainly was so by the end of the 18th century in England (Fresco-Corbu 1964; 1,286) and appears to have been so in The Netherlands from the introduction of smoking (North Carolina 1960: 81-2; cf. Corti 1931: 188, Fig. 42, 189). There is no reason to think that these short-stemmed pipes differed in their bore diameters from the longer-stemmed variety; unless, because the pipes were cheaper, some of the out-of-date larger wires were used in their manufacture; but differing lengths of stem would reflect on the number of fragments found and therefore (in theory) on dating evidence. Actually, dudeens and cutties do not appear to have reached the New World in colonial times (Omwake 1965: 27). It has to be emphasized, however, that the trend towards smaller bores was a gradual process. Harrington found, for example, that the 6/64 in. bore, which achieved its maximum frequency (nearly three-quarters of all the material) during the period 1680-1710, first occurred in the period 1640-80 and was still found, though rarely, in the period 1750-80.

As remarked earlier, the amount of Dutch material in the pipes from these casemates is uncertain, but it may have been approximately half of the total. This may be the cause of the difference between the calculated dates and those suggested by other evidence, for if Harrington's observations for the 17th century hold for the 18th century, the application of his graph for English pipes would result in a later date when applied to Dutch material. Further deposits containing large amounts of Dutch material would have to be analyzed before a definite statement could be made.

The use of the Binford formula on the material in these casemates, therefore, seems to confirm the opinion expressed in the study of the material from Casemate 1 Right: it should not be regarded as a substitute for historical research, but for deposits containing little or no other evidence it is useful.