We propose a method for the encoding and the computer analysis of the macro
seismic effects deduced from historical sources allowing the complete forma
lization of the process of seismic intensity assessment. In the framework o
f historical sismology we make use of a multi-criteria decisions-support al
gorithm, based on the theory of the Fuzzy Sers. By analyzing the texts of t
he available sources for the 1919 Mugello and 1920 Garfagnana earthquakes,
we followed a classification criterion which is independent of any macrosei
smic scale: we "disarrange" each sentence reported on the sources into 5 sy
ntactic elementary components and represent it by a set of alphanumeric cod
es. This allows us to retain the maximum adherence to the original sources
and to avoid forced interpretations and losses of information due to the ne
ed of fitting a given description to each observed effect. Moreover this sc
heme also allows to gather equivalent effects by reassigning them the same
code, and to use this new classification in further processing. This proced
ure could even be seen as an attempt to define a new macroseismic scale on
the basis of a statistical counting of different effects occurrences. (C) 1
999 Elsevier Science Ltd. All rights reserved.