Finding And Classifying Forgotten Sites

I haven't blogged much about my research lately. One reason is that I am only working with it at ~50% this academic year since I'm teaching in addition to my usual 25% editor's job. Another is that I'm in an intensive desk-based data collection phase, which gives rise to a lot of hypotheses and hunches but not much in the way of analytical conclusions. Here's what I'm doing.

I've got a great big database of about 400 Bronze Age finds from the Lakes Mälaren and Hjälmaren provinces. This sample is delimited thusly:

a) datable finds
b) that are not demonstrably from graves or settlements
c) which have enough location info that I know at least which hamlet's land they came from. In many cases I know what land parcel within a hamlet, or even the find-spot's coordinates to within a few tens of meters.

Most of these finds appear to be “deposits”, which may be called “sacrifices” if you decide to ignore a two-century-long debate over their interpretation.

What I'm doing now is to establish as exact location info as possible for each find, measure its distance to the nearest burnt mounds and rock art, and plot it on the Geological Survey's interactive map of ancient shorelines and lakes. Finding sites often takes historical detective work, since the only information I may have is the name of a 19th century meadow or a measurement from a named but long-deserted crofter's cottage that is no longer on the map. Most of these riddles I can solve on-line. Indeed, one reason that nobody's done this before is probably that it would have taken way too much time and travel between libraries and archives.

Having pinpointed a site, I classify it into one of a few groups that I've come up with inductively. I've now done all the sites with known coordinates, most of the parcel-level ones and some of the hamlet-level ones. At this moment I have located and classified 101 sites which have seen only one documented deposition event each. They break down as follows.

39% are in Bronze Age lakes or near their shorelines.
23% are in the Bronze Age Baltic Sea or near its shoreline.
20% are in Bronze Age rivers and streams or on their banks.
8% are in Bronze Age bogs.
7% are well pinpointed sites without any interesting characteristics that I can identify.
3% are on inland hill tops.
1% are in sources.

In addition to these sites I have four that I regard as the key to the whole area of research. They are cumulative deposition sites, that is, places where people have returned repeatedly to deposit objects despite time distances that would make them unlikely to retain any accurate information about earlier deposition events. Rather than reflecting memories of individual events, these sites in my opinion demonstrate long-lived traditional landscape rules for where sacrifices should properly be made. And they share some important traits.

During the Bronze Age, each of the four cumulative sites was in or next to a river at the point where it entered and/or exited major bodies of water. At least three sites were white-water gorges with rapids or waterfalls. And all four were in settled areas, 1–4 kilometres from registered burnt mounds and rock art.

I look at the single-use sites in the light of the cumulative ones. Most are as you can see above associated with water, and an important rule seems to have been that deposition was appropriate at sites where a river or stream changed states. Put differently, if the water interrupts its steady flow to do something interesting, then that's where you need to sacrifice an axe. This supports and extends the observation that my collaborator Christina Fredengren has arrived at by less data-crunchy methods and published last year: "... metalwork depositions were placed at exits of waters such as river mouths and the confluence (meetings) of different waters, sweet and salt".

Image from Lenas fotoblogg.