Asking about Data: Experimental Philosophy of
Information Technology
Brian Ballsun-Stanton
School of History and Philosophy
The University of New South Wales
Sydney, Australia
brian@ballsun.com
Abstract- This paper explores recent research done into the
philosophy of data. The research utilized experimental
philosophy ideas combined with Information Technology
methodologies to assess participants’ philosophies of data.
Reusing the concept of the data flow diagram, I suggest a
methodology of experimental philosophy that allows participants
to categorize flows into data, information, and knowledge. This
allows me to explore their practical understanding instead of
their theoretical understanding. My research has found three
philosophies: “data as bits”, “data as hard numbers,” and “data
as recorded observations.”
I. INTRODUCTION
Many people believe that data is a technological construct,
that we encode information and knowledge inside data when
interacting in electronic systems. Other people believe that
data are the basis of science: hard numbers as the product of
experiments. That data must be objective, reproducible, with
the limits of precision known. Still other people believe that
data are an observation of some kind. That data can be
qualitative or quantitative, so long as it is a recorded
observation. People may use data in the singular or plural, not
as a grammatical error, but as a reflection of how one
understands this ultimately socially constructed concept.
These beliefs are incommensurate and largely incompatible.
They influence thought, analysis, and self-reflection, and are
strongly influenced by someone’s background and workplace.
In studying these different philosophies of data, held by
people who work with data every day, I found it difficult to
ignore my own philosophy of data. To state a philosophy by
fiat destroys any possible evidence for multiple philosophies.
For that reason, it seems better to set aside one’s own
philosophy of data and use the techniques and research results
described below to question that philosophy.
This paper explores my experimental research into the
philosophy of data. My research has two goals: to create a
methodology to probe the philosophy of data for practical use,
and to see if people really do have different philosophies of
data.
The research had two primary goals, phrased as statements
of interest to guide the abductive process of rapid hypothesis
forming. I seek to explore the statement: “People have
different philosophies of data” and the statement “My
methodology can probe people’s philosophies of data.” These
statements serve to focus attention and define a universe of
discourse for the investigations.
II. JUSTIFICATION OF RESEARCH
Creating a philosophical basis for data requires significant
justification. Information Technology (IT) researchers tend to
spend a great deal of time and effort chasing after quite
worthwhile new technologies and without giving much
consideration to the philosophical implications of those
technologies [1]. IT practitioners must serve as an interface
between computing and people. They must understand what
people actually want and must understand the reality from
which they desire that thing. If IT people cannot understand
the needs of the users and the reality that they live in, they
cannot do their jobs. However, if they cannot then understand
the philosophies encoded into programs via the many socially
constructed protocols that a computer requires to be useful,
they cannot understand what problems a computer system
thinks it solves.
An understanding of the philosophy of data is not merely an
academic question. Siloing in organizations, the practice of
small groups talking mostly amongst themselves [2], may be
partly due to different understandings of the nature of data.
Imagine someone with one of the other philosophies described
above talking to “a busy expert” about what he or she thinks
the needs of a system should be. Without an awareness of the
different definitions of data, the amount of effort needed to
create a linguistic trading zone and actually communicate with
this person about their infological needs is far more effort than
a simple dismissal.
Exploring the philosophy of data is a gateway question. I
seek to help IT practitioners to accurately model clients’ views
of reality, and then to entice them into other philosophical
thoughts. The difficulty of modeling is that the client seldom
explicitly states their understanding of reality. By building
tools to probe those models, this experimental philosophy is
both a vehicle for discovery and something that allows us to
start feeling our way into the philosophy of IT.
III. PHILOSOPHICAL LITERATURE
A dominant philosophical theme in my research is the
concept of a trading zone: two groups, not sharing a common
language, come to a place where they can evolve a locally
functional language [3]. It is a way of communicating
concepts between two groups without forcing either group to
change what they know to be true. Both groups understand