1 files changed, 0 insertions, 33 deletions

diff --git a/sandbox/ralessi/driscoll.tex b/sandbox/ralessi/driscoll.tex
deleted file mode 100644
index 105a9fd..0000000
--- a/sandbox/ralessi/driscoll.tex
+++ /dev/null
@@ -1,33 +0,0 @@
-Attributing inscriptions to masons on the basis of their style is
-useful for dating and reconstructing texts. A team of computer
-scientists, in collaboration with the epigrapher Stephen Tracy, has
-developed digital methods to quantify letter shapes, which can then be
-compared statistically across texts to determine likely
-authorship. Promisingly, the results have so far agreed entirely with
-Tracy’s own attributions. Digital stylometry, however, must not be
-fetishized as a fountain of “essentially more objective,” “hard
-evidence” or of “correct and unambiguous” facts (Papaodysseus et
-al. 2007: 749; Tracy and Papaodysseus 2009: 101; Panagopoulos et
-al. 2009: 1404; also see Tracy 2003: xviii). Computerization does
-eliminate the inconsistencies of human perception and enable great
-precision, but any promises of scientific objectivity should be
-scrutinized. This paper attempts to provide such scrutiny by advancing
-three sets of claims. First, epigraphy possesses a troubled history
-with supposedly objective criteria: the now-discredited
-three-bar-sigma dating rule, for example, held the field in thrall for
-much of the twentieth centurybefore being disproved (Chambers et
-al. 1990); the episode’s clear moral is that we should be skeptical
-about other, similar claims to objectivity. Second, comparison with
-the connoisseurship of Athenian pottery underscores the relative
-poverty of epigraphical connoisseurship: Tracy is the only recent
-Hellenist to devote sustained effort to identifying epigraphic hands,
-resulting, inter alia, in no robust scholarly consensus on the
-criteria or standards of judgment. Finally—and most important—major
-epistemological issues simply remain unconsidered. Essentially,
-Tracy’s method has been refined into a set of algorithms. But the fact
-that analysis takes place in a computer does not guarantee that the
-method is correct, nor does it make its results “objective” in any
-deep sense. The method and its calibration remain exercises in human
-judgment, grounded ultimately in our visual experiences. Digital
-imaging calls for reinvigorated, updated engagement with conceptions
-of style and connoisseurship, not their repression.