Evidence 87
ing as either late medieval or early Renaissance. Before the
middle of the nineteenth century, buildings of the fi fteenth
century were not considered Renaissance because the cate-
gory had yet to be introduced into the tools of (then) cultural
history. Cohen’s study offers the valuable reminder that the
‘old’ task of understanding the past through the empirical
study of artefacts has hardly been exhausted by work done
to date. Historians can interpret new data and recalculate
proportional relationships in light of other kinds of knowl-
edge of the building, its procurement, design and fabrication.
Such studies as these force us continually to reconsider the
broad classifi cations of architectural history. As an example
speaking to the forms of evidence considered in the previous
section, this case treats new measurements of the building as
procedural evidence, which is considered in light of contex-
tual factors and is presented in full awareness of its concep-
tual implications.
Cohen’s article raises the problem of how a fi ne-grained
re-appraisal of existing evidence (through re-measurement,
in his case) throws light upon a category that has implications
for the historiography of an entire epoch. Our next example
concerns a specialist problem in the œuvre of Piranesi: how
and in what order did he compose his Vedute di Roma (pub-
13 Measured drawing and geometrical study of San Lorenzo,
Florence, by Matthew A. Cohen. By permission of Matthew A.
Cohen.