‘Let Me Tell the Story Straight On’: Middlemarch, Process-Tracing Methods and the Politics of Narrative
| Author | Timothy J. Ruback |
| DOI | 10.1111/j.1467-856X.2010.00425.x |
| Published date | 01 November 2010 |
| Date | 01 November 2010 |
| Subject Matter | Article |
doi: 10.1111/j.1467-856X.2010.00425.x
B J P I R : 2 0 1 0 V O L 1 2 , 4 7 7 – 4 9 7
‘Let Me Tell the Story Straight On’:1
Middlemarch, Process-Tracing Methods
and the Politics of Narrativebjpi_425477..496
Timothy J. Ruback
Much thought has been put into developing rationales for the process-tracing method, but propo-
nents of this narrative method have been agnostic about the criterial demands for ‘writing up’ a
case study. This article addresses that lack through a double reading. First, I show that ‘good’
process-tracing prose mirrors a narrative voice found in Victorian fiction, most notably in George
Eliot’s Middlemarch. Then, in the second reading, I critique this narrative approach through a
close reading of Middlemarch. In doing so, I explain how this style attempts (and ultimately fails)
to mask its own pre-theoretical political commitments. For process-tracing to seem effective, its
practitioners must turn a blind eye to the theoretical consequences of narrative style and must
remain silent on the instability inherent in their prose.
Keywords: IR theory; qualitative methods; narrative; process-tracing
Who shall tell what may be the effects of writing?
George Eliot (2003 [1871], 412)
Why aren’t the writers working in academic International Relations (IR) more
aware of how they write, and the effects that their narrative style2 may have on
what they write? After all, writing is a vocation for most IR theorists. Yet, although
there has been a ‘virtual explosion of interest in narrative and in theorizing about
narrative’ (Kreiswirth 1992, 629) in the human sciences,3 when it comes to our
own words, most IR scholars have huddled safely beyond the blast zone.
This lack of reflection is puzzling. I am reminded of a moment during a qualitative
methods training institute when a class on process-tracing4 methods led to a
fascinating exchange. During the session, much was made of the social scientific
rationales for using process-tracing in a single-case context. These reasons include
the ability to unearth micro-causal relationships, to consider major actors’ decisions
in context, and to trace events from a static pre-causal point to their eventual
outcome in cases with a small or unique n (where neither statistical inference nor
the comparative method would do).
But in the subsequent discussion, many attendees focused on one question: not why
one might use process-tracing, but instead how to report the method. Once having
chosen process-tracing as a method, how do we go about constructing our narrative
in a methodologically sound manner? What should the process-trace look like? The
answers seemed like platitudes: write it well. Make the prose convincing. Copy
© 2010 The Author. British Journal of Politics and International Relations © 2010
Political Studies Association
478
T I M O T H Y J . R U B A C K
others who have successfully used process-tracing.5 It appeared as if, for these
methodologists, writing was something considered after method as simply a way to
relate one’s methodological findings.
It seemed curious that little emphasis had been placed on how scholars narrate
narration because, as a method, process-tracing is narration—storytelling in the
service of relating a causal logic. Therefore, we should pay attention to the form of
causal stories. Convinced of the promise of process-tracing, many attendees fretted
over using the method without a model of how it should be properly employed.
Later, in a session on publishing, some advice on prose arrived. Don’t write like
Faulkner, attendees were told. Try to be young Hemingways instead. This is rea-
sonable advice. Effective non-fiction prose ought to emphasise clarity. Hemingway’s
economical sentence structure does this. Faulkner’s stream-of-consciousness
detours do not. Yet, it is unsurprising that Hemingway wrote such crisp prose. His
subject matter was often equally clear. Most scholars working on case studies will
discover that Hemingway’s style does not fit the subtlety and complexity of tracking
causal sequences. But given the goals of clarity and simplicity in service of our
arguments, we do not want to write like Faulkner6—or so the argument goes. Are
we left again without an answer for how we should write our process-tracing case
studies?
This question has serious consequences for future process-tracing research. Practi-
cally, clarity about the structure of process-tracing prose would help to develop a
method that presents its findings primarily through narrative. It would help us
understand when we have sufficiently narrated a causal pathway. Epistemologi-
cally, it questions the extent to which we can separate methodology from our
written presentation. If method and prose are necessarily conjoined, then we must
be clear about the narrative requirements of the method.
In the following sections, I use insights from narratology to develop propositions
about how to write a process-tracing case study. Specifically, I consider three
questions: (i) to what extent can narrative prose be considered separable from and
secondary to the process-tracing method? (ii) If method and prose are bound
together in process-tracing, then what stylistic demands does process-tracing make
on our prose? (iii) What limits would process-tracing’s narrative style impose upon
how we think and write about politics?
I contend that we have a model for the prose style that ‘good’ process-tracing
should emulate: George Eliot’s Middlemarch. Her prose provides a model for process-
tracing narratives because of how it positions the narrator so as to call forth a
Victorian subjectivity,7 and how it is wilfully amnesiac about this performative
production. In other words, because process-tracing (like any method) requires
operational reproducibility—any methodologist looking at the same material with
the same method should come to the same conclusions—this means that practitio-
ners must narrate historical events unambiguously, without imposing additional
meaning or structure on to the events through their writing. They must turn a blind
eye to the possibility that the words they use are at least partially responsible for the
clarity, structure and meaning of their historical narrative. And they must share
these presumptions about language with their audience. Only then could narrative
© 2010 The Author. British Journal of Politics and International Relations © 2010 Political Studies Association
BJPIR, 2010, 12(4)
‘ L E T M E T E L L T H E S T O RY S T R A I G H T O N ’
479
be considered a representational tool used in the service of method. Prose like
Eliot’s offers a standpoint from which all this seems possible.
In this article, I will show why process-tracers are indebted to Eliot. I will also show
how an alternative reading of Eliot provides IR scholars8 with an opportunity to face
the productive elements of narration and thereby to question the separation
between method and prose. But first, I must provide a framework for thinking
about the structure of academic writing9 and clarify the place narrative holds within
case study research.
Addressing Narrative in Qualitative Methodologies
For the most part, our methodological turn towards narrative tends to focus on how
narratives can illuminate causal processes (Büthe 2002) rather than how they are
presented. Yet, while little focus has been placed on how we write what we write,10
qualitative methodologists do not ignore the effects of academic prose. John Gerring
offers a typical position on the relationship between language and method:
Perhaps the most important criterion is something that I shall call (albeit
vaguely) good writing. A work should be coherently organized, so as not
to endlessly repeat its points; writing should employ standard English and
a minimum of jargon; and so forth. In contrast to writing in other venues,
good social science writing should privilege clarity and simplicity ...
Indeed social science’s greatest failing may have little to do with ‘meth-
odology’, as conventionally understood, but rather with the more
mundane problems of effective communication (Gerring 2001, 112).
Gerring suggests that writing can seem to have significant effects on methodological
rigour, but it is actually separable from method and secondary to it. How we write
might cloud our methodological principles, but problems with writing are not
methodological problems per se. They are the ‘mundane’ difficulties of presenting
one’s research. For Gerring, writing’s task is to be transparent. It is used to make
clear the methods that were presumably fixed and coherent prior to writing.11
Methodology guides tend to focus on the requirements and goals of process-tracing,
rather than the prose. Gary King et al. (1994, 85–86) define process-tracing as an
account of causality provided through the identification of causal mechanisms.
Their methodological advice focuses on how to identify causal mechanisms rather
than how to describe them. These mechanisms are always assumed to be explain-
able in a language that exists prior to and independent of writing, detailed ‘to the
point where the events to be explained are microscopic, and the covering laws
correspondingly more certain ... a platitude’.12 Beyond this, most advice focuses on
how to justify the use of a single case in one’s research.13
Process-Tracing, Narration and Method
To make sense of how process-tracing uses narration, I must first explain what
process-tracing is. Process-tracing, a within-case...
To continue reading
Request your trial