What is the subject matter of the philosophy of computer science? I believe that the philosophical questions lurk everywhere, even beyond those areas conspicuously covered in the current discourse–artificial intelligence, the nature of formal systems, and technology ethics. I want to interrogate more widely, of computational objects, questions of how things work in the world vis-a-vis how things work in computation, and to ground those wider inquiries in terms of the conventional areas of modern philosophy—metaphysics and ontology, epistemology, and ethics.
To illustrate the broader scope, let’s take voting, an object of
computation that is the focus of considerable current interest.
Consider: What is a vote? In ontology, we can ask what is the nature
of this simple datum, firmly associated with its source (“MY vote”)
until that vote is cast, when it becomes… what? It becomes firmly
detached from its source (“my SECRET vote”). Tallying is merely
counting, the simplest of computations. In that process, aggregation
subsumes “MY vote”, its individual identity vanishing. Yet at the end
of tallying, for audit purposes, “MY vote” may again become salient, not
necessarily in its content, but in its metadata — time, place, and
crucially, verification of proper execution of intention (but not in its
connection to an identified voter). What are the features of this data
structure? Are they the same from start to finish? Is there any
conflict that would block a coherent treatment?
Epistemologically, we note that “MY vote” enjoys special status in the
mind of the voter, as–what? The epistemic object may be manifest not
only as a proposition, but as an act of volition, a choice, coupled with
another volition, the will to make sure that it counts. In that case,
we can ask epistemological questions that inform policy questions
regarding what voters want and need to know about the tallying
computation that will secure confidence in the system, and conversely,
under what epistemic states they should ask for more information and
verification. (“What happened to MY vote?”)
As for ethics, many inquiries present themselves. We can ask whether
particular digital representations of votes, or voters, uphold or
violate ethical principles. We can ask how to partition and allocate
the substantial burden of responsibility among the many parties involved
in the voting process; in particular, we can ask how a computer
scientist’s professional ethics should govern her hardware and software
products, in terms of not only specification, verification, testing, but
documentation and deployment.
These are questions that may reveal the fortuities afforded by the
myriad human interactions that we have taken for granted in election
protocols. These are questions that may help the discipline of
computational social choice to work out formal results. These are
investigations that may help political science mediate claims of recemt
systematic election fraud (bogus) and concerns about Internet voting
security (valid). These are questions about voting, just one slice of
the life-world, that emerge from our relations with data and
computation. What other questions await in human experience?
Robin Hill
Robin K. Hill is a charter member and the Secretary of the new
Association for Philosophy and Computing, which replaces the former APA
Committee on Philosophy and Computing. An earlier version of these
thoughts appears in the APA Newsletter on Philosophy and Computers issue
of Spring 2016 ("A Call for More Philosophy in the Philosophy of
Computer Science", online here and also in the
online blog of the Communications of the ACM.
