Authoring a PhD

CORE
Focused
literature review
Sequence of chapters
Breadth of coverage
Figure 3.4
The compromise model

---

beginning to appreciate your research contribution. Your core
stuff thus comes across much earlier, leaving more space at 
the end of the thesis for you to do a couple of chapters or one
decent, long chapter of analysis. The first part of these con-
cluding materials might focus on bringing together and inte-
grating the conclusions from your core chapters, each of which
should cover a different component of your research. The sec-
ond part of these concluding materials can then do a more 
limited opening out from the results of your analysis back into
the wider literature. By saving much of the theory discussion
and literature discussion to handle at the end of the argument, 
you should be able to form a strong theoretical or broad-view
chapter. This way you can conclude your thesis on an upbeat,
confident and professionally salient note.
Four patterns of explanation
I have yet to see any problem, however 
complicated, which when you looked at it the
right way did not become still more complicated.
Poul Anderson
7
When you try to communicate a set of connected information
to someone else there are only a limited number of ways that
you can do it. If your chosen way cuts across the other person’s
expectations then crossed wires may occur in the communica-
tion. This problem is made worse when your audience does not
listen intently to every twist and turn of your account. For
instance, people of different genders famously tend to choose
incompatible modes of communication. Most women like to
give and receive process-organized explanations, often running
through the history of an event or an interaction from begin-
ning to end in narrative succession. But most men prefer to
receive ‘bottom-line’ information first. They want to know at
the start what the key point of a story is, and only then will
they be ready to listen much more selectively to the detail of
how the story’s outcome ended up as it did. Hence men easily
get annoyed by what they code as women ‘rabbiting on’.
Equally, women often get turned off by men’s overly terse and
6 2
◆
A U T H O R I N G A P H D

---

inaccessible explanations of complex phenomena. A variant of
this particular contrast in modes of communication (serialistic
rather than holistic) also runs through two of the alternative
patterns discussed here. My thesis is that in the humanities 
and social sciences there are only four fundamental ways of
handling long, text-based explanations, which I shall discuss in
turn. These organizing patterns are: descriptive; analytic; argu-
mentative; and a matrix pattern, combining elements of any
two of the other three approaches.
8
Descriptive explanations
Suppose that I am asked to give an account of the room where
I am sitting and writing these words, which is my home study.
Figure 3.5 shows the main features of the room, which are rea-
sonably complex. A descriptive mode of explaining something
is to take the way that things are organized 
externally
or exoge-
nously to me and to then use that pattern to structure the
sequence of what I say. For instance, in explaining about my
study I might start at some particular point, like the door, and
then decide to sweep my arm around the room in a particular
direction (clockwise in this case) listing everything that comes
into my line of sight as I do so, as shown in Figure 3.5(a). Here
I might say: ‘First there is a white door, and next to it in a
clockwise direction is a green painted wall, and a grey beaten-
up sofa, and above it a noticeboard with papers pinned on it, a
CD rack, then a series of long bookshelves with four-drawer fil-
ing cabinets underneath, and then a printer, an old desk-top
PC, and a new laptop on a desk surrounded by papers, then a
window with three frames …’ and so on.
This listing account already illustrates some obvious defi-
ciencies of a descriptive way of explaining things. The sequence
of objects being named is united in only one way, namely prox-
imity in the room. The things I list are next to each other. But
in every other way the different objects described together are
jumbled up randomly and unpredictably. The list may work OK
if readers get to see Figure 3.5 (I certainly hope so). But without
this visual support, the list could be very hard to take in and 
to visualize. The account I give of my study could also easily
P L A N N I N G A N I N T E G R A T E D T H E S I S
◆
6 3

---

(b)