Mapping Deliberative Systems with Big Data: The Case of the Scottish Independence Referendum
| Author | John Parkinson,Sebastian De Laile,Núria Franco-Guillén |
| DOI | http://doi.org/10.1177/0032321720976266 |
| Published date | 01 August 2022 |
| Date | 01 August 2022 |
| Subject Matter | Articles |
https://doi.org/10.1177/0032321720976266
Political Studies
2022, Vol. 70(3) 543 –565
© The Author(s) 2020
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/0032321720976266
journals.sagepub.com/home/psx
Mapping Deliberative Systems
with Big Data: The Case of
the Scottish Independence
Referendum
John Parkinson1,2 , Sebastian De Laile3
and Núria Franco-Guillén4
Abstract
Deliberative systems theorists have for some time emphasised the distributed nature of deliberative
values; they therefore do not focus exclusively on ‘deliberation’ but on all sorts of communication
that advance deliberative democratic values, including everyday political talk in informal settings.
However, such talk has been impossible to capture inductively at scale. This article discusses
an electronic approach, Structural Topic Modelling, and applies it to a recent case: the Scottish
independence debate of 2012–2014. The case provides the first empirical test of the claim that a
deliberative system can capture the full ‘pool of perspectives’ on an issue, and shows that citizens
can hold each other to deliberative standards even in mass, online discussion. It also shows that,
in deliberative terms, the major cleavage in the ‘indyref’ debate was not so much between Yes and
No, but between formal and informal venues.
Keywords
deliberative systems, big data, topic modelling, Scotland, independence
Accepted: 30 October 2020
Introduction
For the last decade, some branches of deliberative democratic theory have been shifting
away from institutions and norms that are thought to provoke idealised deliberation
towards thinking about how complex constellations of institutions, norms and practices
realise deliberative values in democratic societies as a whole (Dryzek, 2010; Parkinson
and Mansbridge, 2012). However, the challenges of empirical research into ‘deliberative
1 Department of Philosophy, Faculty of Arts & Social Sciences, Maastricht University, Maastricht, The
Netherlands
2
Centre for Deliberative Democracy and Global Governance, University of Canberra, Bruce, ACT, Australia
3Independent Data Scientist, Brisbane, QLD, Australia
4Department of International Politics, Aberystwyth University, Aberystwyth, UK
Corresponding author:
John Parkinson, Department of Philosophy, Faculty of Arts & Social Sciences, Maastricht University, P O Box
616, 6200 MD Maastricht, The Netherlands.
Email: j.parkinson@maastrichtuniversity.nl
976266PSX0010.1177/0032321720976266Political StudiesParkinson et al.
research-article2020
Article
544 Political Studies 70(3)
systems’ thus conceived are considerable (Fleuß et al., 2018; Owen and Smith, 2015).
Systemic perspectives emphasise the distributed nature of deliberative values; they there-
fore do not focus exclusively on ‘deliberation’ but on all sorts of communication and
practices that advance deliberative democratic values, including the importance of every-
day political talk in informal settings (Mansbridge, 1999; Neblo, 2015); organised, coor-
dinated confrontation, contestation and performance (e.g. Curato et al., 2019); as well as
the formal deliberation of legislatures and the institutions of ‘middle democracy’
(Gutmann and Thompson, 1996). But researching such complexity at the scale of mass
democracy always entails a classic empirical trade-off: we can either focus on the micro-
dynamics of everyday communication in particular sites but then lose sight of overall
patterns, or we can stand back and examine patterns of communication at the large,
national and transnational scale, but lose sight of detail we think important.
There are several ways of handling this trade-off in the deliberative literature. The
dominant strategy is to narrow the range of sites and institutions under study, usually by
focusing on the connections between one or more small-scale, citizen-centred forums
and a wider or more formal democratic process (e.g. Boswell et al., 2016; Hendriks,
2016), as in the growing literature on constitutional deliberative democracy (Levy et al.,
2018; Suiter and Reuchamps, 2016). Such work embraces more communicative detail
but often stays focused on deliberation, the noun, rather than the distributed, adjectival,
deliberative quality that systems theory suggests is so important (Parkinson, 2018), and
is only systemic in a limited way, focused on a small number of connected sites, often
just two or three.1 Another, surprisingly rare, approach is to look at institutional net-
works and relationships and draw inferences about communicative flows based on those
linkages (e.g. Cinalli and O’Flynn, 2014), and while such studies approach a more sys-
temic view of democratic complexity, examining dozens of sites, they do not study com-
munication directly.
An important alternative approach is taken by Stevenson and Dryzek (2012, 2014),
who take a macro view not so much of institutions but of communication, focusing on
discourses both as coordinators of and resources wielded by people in collective action:
in their case, counter-hegemonic battles over transnational climate governance. Stevenson
and Dryzek (2014) start with discourses conceived in a particularly macro way – few in
number and persistent over decades – while they ‘reconstruct’ such discourses using
methods ‘broadly consistent with’ Fairclough’s (2003) critical discourse analysis (CDA),
techniques which still require a great deal of human, deductive judgement to create, sam-
ple and make sense of the body of texts (Baker et al., 2008). While the idea of ‘discourse’
is macro, the cases they work on are more middle range, focusing on activist and non-
governmental organisation (NGO)-led engagement in governance extracted from obser-
vation and samples of text – the sampling is not discussed – of four ‘organized spaces’ for
civil society discussion of climate change response leading up to and including the 2009
Copenhagen climate summit. In short, this approach is significantly more systemic com-
pared with the dominant strategy, but it examines a small number of venues, does not
capture everyday informal talk and uses methods that are less inductive than might appear.
Of course, no method can be fully inductive, no method can capture every element we
want to study and every result needs interpretation. Rather than rejecting what has gone
before, this article suggests a complementary approach that provides more radically
inductive grounding for large-scale, complex, systemic analysis. It examines large-scale
communicative patterns, but extracts those patterns more directly from everyday textual
data, using recent developments in electronic social science. In particular, it applies a tool
To continue reading
Request your trial