Authoring a PhD

Treatise on
Civil Government
, Book 4, Chapter 3, section 6, might appear as
(Locke,
TCG
, IV.3.vi). This kind of abbreviated reference is
perfectly neat to use many times over in your text, but is also
accessible enough once you have explained the convention
being used. Where your thesis revolves centrally around the
use of a set of primary sources, then it is often useful to 
discuss them in a Research Methods Appendix, and this is 
a good place also to explain the referencing conventions you
have followed.
◆
Unpublished and un-indexed sources, such as documents
located in a depository that is not a well-organized historical
or other archive with retrieval numbers, can be handled in a
similar way in the Harvard in-text reference. Establish and
explain your own referencing or naming convention as for
primary sources above. Include a set of convenient
abbreviations, ideally acronyms that will be intuitively
understandable (as with the Locke reference above).
◆
In-text references for interview material are also sometimes
cited as a problem for Harvard referencing, but are in fact
straightforward to handle. ‘On-the-record’ interviews should be
cited in a similar way to primary sources, by establishing a
convention including the interviewee’s surname, the fact that it
was an interview and the interview date, as: (Smithers, interview,
26 October 2000). Your Research Methods Appendix should then
W R I T I N G C L E A R LY
◆
1 2 9

---

include full details on who interviewees were; where and 
when you talked to them; how interviews were conducted 
(for instance, face-to-face, by phone, using a questionnaire or 
a dialogue mode, etc.); and how you recorded the material 
(for instance, taped or noted). If you want to cite evidence
from ‘non-attributable’ interviews then referencing issues do
not arise, because you cannot link particular points to any
definite respondent. Instead you need to find a way of
introducing phrases into your main text which give as much
useful contextual information about your informant or source
as possible, while yet fully preserving their anonymity.
Material from ‘off-the-record’ interviews cannot be cited or
referred to at all without breaching normal academic research
ethics. (Make sure that you carefully discuss with your advisor
any possible issues in referencing different kinds of interview
material at the examination stage.) In all these last three
respects there is no significant difference between the
difficulty or ease of citing sources under Harvard referencing
and using alternative systems like endnotes or footnotes.
A final issue worth noting about bibliographies concerns seg-
mentation. A single unified bibliography arranged in a strict and
predictable alphabetic ordering is best for all textual materials.
In some older works, and in the PhD regulations for a few more
old-fashioned universities, it is still possible to find bibliogra-
phies broken up into primary sources (such as unpublished doc-
uments) and secondary or published sources, or even separate
listings for books and articles. All such devices breach the one-
stop look-up principle, because from the in-text reference alone
readers normally cannot tell what kind of source they are being
directed to. With any kind of segmented bibliography they may
have to look in several places to find the reference they need.
Endnotes
The main viable alternative system for referencing consistent
with full citation and one-stop look-up are endnotes.
◆
The
in-text reference
is reduced to a minimal superscript 
number, as.
8
The numbers should restart at 1 with each new
chapter. The number is automatically entered in your text
1 3 0
◆
A U T H O R I N G A P H D

---

by the word-processing package when you create an end-
note. Note numbers should always be located at the end of
sentences, not in the middle. You should also avoid having
multiple note numbers at different points inside or at the
end of the same sentence.
◆
The