Thursday, 15 January 2015

Beyond the "Fringe Archaeology" — part two: the archaeology of thought

Michel Foucault
free-use image
I was drawn to reading Michel Foucault's The Order of Things: An archaeology of the human sciences after discovering the act of classification to be a destructive activity. I had wanted to create a more detailed typological study of the coins of the ancient Celtic Coriosolite tribe of Brittany by looking at the evolution of the coin-die designs. It had struck me that the current classification was based on the observance of several very noticeable design motifs, yet the designs on the coins had a very numerous collection of motifs. I reasoned that what was most noticeable today might not have been most noticeable to an observer in the time that the coins were made: even a visually small device might have had deep significance to someone who understood its visual language. The modern classifier, having no such knowledge must stick with what is seen but not necessarily understood  — devices that are visually dominant; devices that are commonly repeated; devices that seem to be grouped together with some relationship to the chronology. All other considerations are ignored.

Thus, classification becomes like the serpent Ouroboros eating its own tail: replacing the thoughts of the creator with the thoughts of the classifier. I knew that I would have to make contact with the thoughts of the creator of the work to avoid this subjectivity otherwise I would just be seeing horses in the clouds.

It was no easy task and I almost gave up on it, but I found an important clue in a set of minor variations — a clue as to how the die engraver was thinking about the designs. If you change one element within a composition, the whole composition will change and other compositional problems will emerge, and then also have to be changed. By discovering a small section of the complete work that exhibits a chronology that can be checked against other features, and found to be true, the reasons for changes can be inferred to have been due to certain tenets: that of the art (as the series of coins expresses that art) and of religious belief (as the devices have mythological connections). I looked at these subtle changes as my Rosetta Stone.

The biggest surprise to me was that as I began to arrange the dies in their order of manufacture based on sets of these subtle changes which proved their validity by overlapping each other, I not only began to understand the die engraver's thought processes more clearly, but also realized that what I was being left with was a total negation of the very idea of classes within a classification system. There was nothing left but the physical remains of a time period in the die engraver's thought processes. Each example of these remains was embodied within a single set of coin dies. There was no such thing as an objective class unless that class could be understood as the product of a modern classifier. The only valid units in the system were individual objects (coin dies which have left the evidence for their existence in their products). A classification system defines the classifier and the not the object being classified.

Foucault's work confirmed my suspicions, but where do we go from there? Stay tuned!


  1. I believe too much effort has been expended on classification of coins during the last 25 years, and even the more wasteful activity of mapping findspot distributions.

    I recall I became interested in Celtic coins because of the fantastic imagery.

    Years ago I asked a friend, a well-known archaeologist, where he thought the study of Celtic coins should be directed in the future. His answer "I would like to know something about the images".

    This reminded me of my original interest.

    But I don't think better classification schemes are the goal. We need to get at the MEANING of the images. I think the answer lies within the discipline of Semiotics.

    1. I couldn't agree more, Bob. When I first started working on Coriosolite coins, I just wanted to refine the chronology to a point where it could give us answers about distribution patterns etc. When I discovered that discovering meaning in their images was necessary to that goal, it actually replaced it. It's not an easy goal, however...