Wouldn’t it be nice to be able to see what was underground without having to go through the time-consuming, expensive and destructive process of digging it up? There are some nice geophysics techniques in archaeology for doing just that, though none are yet a substitute for excavation. I noticed the other day (and you were quizzed on it) that there is a very comprehensive recent report online (45 meg PDF) by Anna Marie Prentiss and colleagues, on work at the middle Fraser pithouse village of Bridge River (EeRl-4). This village lies in the territory of Bridge River Band (Xwisten) and the St’át’imc Nation. While there is a huge amount of archaeological interest across the 350 page report as a whole, it was the use of geophysics on an interior pithouse village which got my attention.
First, some readers may need a bit of background. For the last 4,000 years or so, the semi-subterranean pithouse was a common wintertime dwelling on the interior Plateau of the Northwest. Such pithouses were generally circular, were excavated several feet into the ground, and covered by a heavy-timbered, earthen roof. Upon abandonment, the wooden elements largely disappeared and the roof collapsed, leaving behind an array of circular, crater-like features behind, each of which might be from 8 to 15 or even more metres in diameter
With their prominent surface visibility, and working from the principle that each house was a coherent spatial unit of analysis which reflected a true social unit, then pithouse archaeology was an early, and leading element of a major archaeological movement of the past few decades known as “household archaeology”. While the first, pioneering pithouse excavation projects are closely associated with David Sanger and Arnoud Stryd, it was the energetic foray of SFU’s Brian Hayden into the field at Keatley Creek in the middle 1980s which really ushered in the true social archaeology of Plateau Pithouses.
Hayden blazed an original and compelling theoretical path and spawned a veritable flower arrangement of work on that large village site. More recently, there has been a very active debate between Hayden and Anna Prentiss over the timing and meaning of the major cultural historical events at Keatley Creek. Meanwhile, as noted the other day, the empirical basis for some of Hayden’s conclusions, such as the confident speciation of salmon remains, has also been undercut by new methods. So Hayden’s original conclusions have frayed somewhat as new researchers have followed his path. It must be said, though, that when following this debate from my little foxhole on the coast, it is notable how these newer researchers have noted the collegial co-operation of Brian Hayden as they seek to challenge and in some cases refute his conclusions. Letting other archaeologists have access to your own data, specimens and “your own sites”, knowing that they might reach different and perhaps more persuasive conclusions, is the true mark of a good colleague and that has been refreshing to see. (I mean, maybe this is just people making nice in their acknowledgments but it doesn’t read that way). (Also, Hayden wrote one of the best and accessibles books about BC Archaeology to date).
OK – so, pithouses are extremely challenging to excavate, moreso than you might think. The earthen roof is a site-formation nightmare, poising older, dug-out sediments above the more recent occupation levels. What goes up, must come down. Further, unless you like living under a creaking, 10-ton roof, you will need to renovate your structure, perhaps every 20 years or so. In so doing, you may take down the support posts, rip apart the roof, re-dig the floor, and re-assemble and re-cover the roof. Basically, you, the 2,000 year ago St’át’imc ancestor, thoughtlessly rips the heart out of future archaeologists. Awkward.
This process is nicely illustrated by a diagram by Hayden you can see above – click for larger but not great image. If you follow it through you get the idea about the shifting target of the house floors and the possibility for changing size of house between occupations – both of which were central issues in the debates over the last 5 years or so between Hayden and Prentiss in the pages of American Antiquity and elsewhere. Hayden’s long-term research was based on the complete excavation of entire Housepits, which surely allowed him to decode many site formation processes and therefore make sensible analytical units of materials which belonged together, in the sense of having been used by the same group of people at about the same time. However, this limited him to work in a select number of houses, and complete excavation is time consuming and also makes subsequent re-investigations difficult.
So, to finally get to the point – with such large, complex sites, any extra edge you can get prior to digging is a huge advantage, especially if it means you can dig less. It is therefore welcome to see Prentiss using magnetometry, electrical conductivity, and ground penetrating radar at the Bridge River site. Just taking the first of these, magnetometry depends on the principle that materials in the ground may have varying magnetic properties, which produce extremely small, but measurable, anomalies in the local magnetic field of the earth. These can be measured with hand-held instruments carried over the surface. A positive anomaly (red in the map) indicates a stronger magnetic presence, perhaps the result of fire-alteration of fine sediments, hearths, or, in some other contexts, the presence of ferrous metals or fired clay bricks. A negative anomaly may be the absence of those positive influences vs. baseline levels, or an excess of such very-low magnetic substances as carbonates.
Magnetometry has been in use in archaeology for over 50 years, and yet, in common with other geophysical methods, it has barely been seen in British Columbia. I think Arcas used it at Tswwassen in about 1990, and RG Matson used it in the middle 1990s at Shingle Spit, but geophysics is a long way from being the routine part of archaeological impact assessment that it could be, probably should be in British Columbia, and indeed is elsewhere in the world.
As you can see in the above, different geophysical methods give different results. To the left, you can see a strongly-defined housepit rim as well as internal structure reflecting, I think, post-holes. From the report:
In both instances, HP25 displays the typical pithouse magnetic signature. Rim deposits are clearly delineated by anomalous positive-valued gradients surrounding interior floor areas that are largely characterized by negative-valued gradients. Typically, one or more positive magnetic features are indicated within the central area of the house floor, surrounded by mainly negative gradients that are particularly strong over southern-southwestern areas of the house floor and extending into the interior flank of adjacent rim deposits. In the case of HP25 (Figure 5), magnetic gradient readings acquired over the central floor area reveal a particularly interesting pattern of anomalously positive gradients and there is evident correlation between these features and apparently related conductivity features.
These specific, cautious results are probably the result of not many comparable studies being done and I haven’t read the entire report to get a full appreciation of how these data and method intersect through the years of Prentiss’ research. I note that Prentiss et al. attribute to targeted geophysics their very high success rate in getting appropriate samples for their research goals, despite not excavating entire houses. It is this last point which matters most. As the authors note:
While Hayden’s data are superb the lengthy time required to excavate entire houses was costly and it limited the number of houses that could be explored. Further, it effectively prevented future field investigations associated with these housepit floors. These problems can be avoided by using geophysical methods to identify activity zones associated with major hearth and cache pit features and subsequent excavation sampling.
In other words, geophysics allows one to do more, with less: less money, less labour, and above all, less destructive impact on the archaeological record itself – and yet recover the same amount or more of the information needed to answer questions.
And, put into a Culture Resource Management context, by creating a road map of the subsurface structure of a site, these methods may allow pre-emptive navigation around crucial parts of important sites. While at Bridge River the housepits are obvious on the surface, at many sites subsurface features to quite some depth which have no surface visibility could nonetheless be mapped. You can’t easily find sites with these methods, but you sure can map some of their contents and coarse structure. Furthermore, as you use it more and more, interpretation gets better and better through experience and by comparison of the geophysical map to the material recovered in the ground. In particular, geophysics on the NW Coast and interior may be especially well-suited to finding human burial features and thereby avoiding them, as well as finding hearth features, ditches, large posts, and house perimeters. All of these are pointers to high cultural and archaeological significance.
It is for this reason that I think that we need to have a very serious debate about geophysics becoming routine and expected parts of CRM work in this province, certainly any time there is proposed substantial impact onto known sites. The onus should be to explain why these methods were not used, and not the expectation of praise on those few occasions that they are. Were geophysical methods applied yet at Glenrose Cannery, for example? Were they considered, and rejected for some reason? In many countries and jurisdictions, these methods are, indeed, routine and I wonder why they are not in British Columbia.
Reference: Brian Hayden, 2005: The Pithouses of Keatley Creek (2nd edition) SFU Archaeology Press.