Introduction - Burdens of Proof - page 9

of the modern state—from accounting to policy, economics, and plan-
ning—is suffused with this authority, “among the gentlest and yet most
pervasive forms of power in modern democracies.” 18
Cryptography's reliance on the authority of proof to justify the supe-
riority of digital signatures as evidence (and the appeal of this authority
within the legal community) can thus be viewed as another chapter in
the mutually beneficial relationship of state power and mathematics—a
relationship somewhat at odds with the self-identification of cryptogra-
phers as cyberlibertarians. I will argue that in fact, the yearning for the
moral authority provided by “provable security” has marginalized research
on phenomena less amenable to mathematical formalization, but with
the potential for greater social impact. 19 Thus, for cryptography, the domi-
nance of analytical proof as the most authoritative method of mathemati-
cal persuasion also comes with a price, that of a certain irrelevance, as the
objects it seeks to realize in the electronic realm—signatures, votes, cash—
are irreducibly tied to the fuzzy and inexact worlds of bodily performances,
human psychology, and long-standing social institutions.
And last, I will argue that the specific ways in which cryptographers
modeled signatures had significant implications for the practical success
of the technology. Here, I seek to foreground the role of modeling as a
repressed dimension of cryptographic practice. 20 Both algorithmics and
computational complexity rely on idealized models of computation that
abstract away its physicality—the particular arrangements of processing
power, storage, and connectivity that define concrete computational
devices. Yet cryptographers have little language to discuss the trade-offs
inherent in any representation of real world phenomena and few concep-
tual tools with which to approach the inescapable gaps that arise between
physical computational artifacts and their mathematical models. 21
These gaps manifested themselves on multiple fronts. As befits the mili-
tary roots of cryptography, the digital signature threat model posited adver-
saries with considerable resources and motivation to break strong
cryptographic schemes—in short, nation-states with considerable intelli-
gence capabilities. This assumption translated into a one-size-fits-all risk
analysis of documentary evidence, conflating together the costs and risks
of very different types of transactions—for example, signing a mortgage
and paying for groceries with a credit card. Cryptography's much touted
superiority—its ability to withstand military-grade computational fire-