The association of Tomb II at Vergina, Greece, with Philip II initiated a debate concerning the use of barrel-vaults in Macedonian tombs. The accepted theory at the time held that, since no Macedonian tomb was dated prior to the last quarter of the 4th century BC, Macedonians copied the barrel-vault from the Persians after the military campaign of Alexander the Great in Asia, and therefore Tomb II should be dated to a later period. After an intense dispute, fresh archaeological evidence proved that this theory was false. This article examines the ways we structure knowledge in archaeology from hypothesis to theory that can develop to consensus, and how later consensus exercises a conservative influence on the production of new knowledge. New evidence that contradicts consensual theories is approached with stronger hostility and is confronted with higher demands of confirmation. I suggest that the same amount of scrutiny should be applied to the established theories, which are not unchangeable representations of reality, but conventionally shared property of archaeologists.
