Authoring a PhD

Anatole France
11
Fairly or not, doctorates are notorious for being over-referenced.
This aspect of authoring often absorbs a disproportionate
1 2 0
◆
A U T H O R I N G A P H D

---

amount of time and attention amongst PhD students, especially
in the humanities and social sciences. There is no reason that it
should. I examine the principles on which your referencing
needs to be built; how to choose an appropriate system; and two
standard systems that do the job simply, Harvard referencing
and endnotes.
Principles for referencing
The ‘need to know’ criterion provides the basic rationale for
what should be sourced, and in how much detail. Two, three or
four readers, the examiners, have particular responsibilities to
guard the portals of the PhD against incorrect or stolen work.
Meeting their needs does impose a much higher standard of 
referencing than is common in academic books or even most
journal articles. For instance, in these sources authors exten-
sively use ‘whole book’ citations, where they designate a book as
a source without specifying where to look within it, as (Foucault,
1995). Doctoral authors should strictly avoid this approach,
because in theory the examiners should be able to check every
source referenced. Obviously it would take them a long time if
they had to read the whole of Foucault’s book to find the one
point which you say is in there. So thesis references must always
be fully precise, ideally sourcing citations to particular pages, as
(Foucault, 1995, pp. 56–9), or at worst indicating a specific chap-
ter, as (Foucault, 1995, Ch. 4). In practice the examiners will
very rarely follow up references, unless they have reason to
think either that you have misquoted another researcher or per-
haps that there is ‘unacknowledged quotation’ (plagiarism) in
your text, which is a quick way to instantly fail your doctorate.
Yet they will rightly become a bit suspicious about your schol-
arly qualities if they see that you are providing less than full and
precise details for every citation and quotation.
Your referencing system also needs to reflect a more general
principle of good authoring, namely that it should prove a 
one-stop look-up
facility. Readers should have to go only to one
place to follow through the sourcing of all quotations and 
citations. They must never be asked to look in two or more
places in order to find out which source is being referred to. 
W R I T I N G C L E A R LY
◆
1 2 1

---

For instance, it is still very common to find books where an
author uses footnotes or endnotes in the main text, but when
you turn to the note there is a Harvard referencing system 
in use, showing only Smith (2001). This means that you have
to look further on again in the bibliography at the end to find
out which source this is. Another example of two-stage look-up
occurs often in humanities disciplines (like history) where
many authors still use pointless and anachronistic Latin abbre-
viations. Here you may find in an endnote or footnote a refer-
ence such as ‘White, 
op. cit.
’ or ‘White, 
loc. cit.
’, where the Latin
bits mean ‘the same work as when White was last cited’. You
then have to embark on a complete magical mystery tour of
looking back through dozens of previous notes, trying to find
the last time White was cited. The most careless and discourte-
ous authors will pursue this 
op. cit.
logic across several chapters,
asking you to ransack possibly hundreds of notes to find the
last time White was referenced. This approach would be a very
rash one to adopt with PhD examiners or assessors. Both these
examples illustrate the dangers to you as an author of not using
a one-stop look-up system. If readers have to dig around in sev-
eral different places to track down where you got a point from,
they will form a worse view of your text and of your compe-
tence as an author compared with if you make their task
straightforward.
Within the two principles of meeting the examiners’ need to
know precise sources for