Gregory Crane
Philology and the Future of Work
This paper describes how the methods of digital philology reconnects the study of
Greco-Roman culture with the developments not only in the Humanities but Computer Science
and even business. We are in a position to design programs in philology that train not only the
next generation of professionals who teach and study historical languages such as Greek and
Latin but that also produce students who have a wide range of professional opportunities
outside of academia. Such a redesign of our curricula involves, in part, more clearly articulating
the analytical, data-driven habits of analysis that our students already acquire in the deep study
of textual sources but it will demand as well an openness by philologists to observe the skills
that the world outside of academia demands. This in turn will challenge us to change the
curricula that our students pursue -- encouraging them, in many cases, to develop skills that we
ourselves may not be in a position to acquire.
At least two factors are at work changing -- and possibly transforming -- how humans produce
goods and services. First, Artificial Intelligence (AI) has begun to emerge as a credible (or, at
least, widely feared) force that can automate large segments of work that humans now conduct.
Such work goes beyond mechanical physical and clerical tasks, and promises (or threatens) to
include the intellectual processes of highly trained experts such as doctors and lawyers. This is
not the AI that seeks to fabricate consciousness, but a much more pragmatic practice that
exploits increasingly powerful hardware to detect patterns in large bodies of data, learning from
inputs and outputs how to replicate expert conclusions. Such machine learning, in turn, builds
upon, and provides a motivation to support, decentralized contributions: systems analyze the
initial conditions and the conclusions that humans make, whether these decisions involve
medical diagnoses or the choice of which movie to watch. The same hybrid network of machine
learning, citizen science, and expert knowledge is already at the core of Digital Classics and
Classicists have an opportunity to prepare their students for the future (or, increasingly, the
present) of work.
Furthermore, within this world of evolving work, the ability to analyze natural language at scale
has emerged as a critical barrier. “If Financial Institutions want AI,” states the title of an article in
Bank Innovative , “they need to crack natural language processing first.” Philologists 1 are much
more closely tied to language than are bankers. The subtasks of natural language processing
and text mining are central to the study of language and thus of philology. These subtasks
include morphosyntactic analysis, recognizing the names of people, places, organizations and
the relationships between them, automatically detecting topics in vast multilingual corpora,
recognizing changes in the sentiment of comments on a particular topic, tracing patterns of
quotation and intellectual influence across texts. The paper will describe the first iteration of a
course taught in Computer Science that frames these techniques within the tradition of
philology.
1 https://bankinnovation.net/2017/06/if-fis-want-ai-they-need-to-crack-nl…