David BIBBY
(Uni Konstanz, Germany)
Most sites are made of any number of partially independent stratigraphic series, for wich many permutations are possible. These so called multilinear stratigraphic sequences are best visualised by the Harris Matrix, which can be advantageously employed as a tool for coping with this difficult phenomenon. Without the introduction of non-stratigraphic dating or equating information (finds, scientific dates, other a priori knowledge about the site and its features) an almost infinite number of relative chronological constellations are possible, even for a relarively small site. Dating/equating information fixes the position of individual contexts on separate lines of the multilinear sequence (i.e. those contexts with no physical contact of immediate stratigraphic relevance to each other) in an absolute chronological relationship, without overtuning the sequences on the individual lines. In this way groups of contexts across the multilinear sequence can be created, and attributed to phases and periods.
This subject was discussed in a paper given in Vienna in 2002, the emphasis then being on the mathematics of stratigraphic multilinearity. Whilst this mathematics is fascinating in itself, the emphasis here is on how these concepts can help the archaeologist and stratigrapher to interpret his/her site and divide it into meaningful phases and periods. How far can the stratigraphy of a site be phased and is this phasing objectively quantificable? Can the phaseability and in turn the interpretability be meaningfully compared between sites?
During the workshop the author would like to discuss the possibilities computer technology might offer both for calculating the maximum number of permutations for any multilinear stratigraphic sequence and for its semi-automatic “reduction” into phases and periods.