Roche, Michael. (2005). Chapter 9: Historical Research and Archival Sources. In (Eds.

), Qualitative Research
Methods in Human Geography (pp. 133-148). Oxford University Press.

This reading is being made available to you under fair dealing for the purpose of education in your Toronto
Metropolitan University class. This is considered a reasonable amount of a work and is being delivered to you for the
purpose of instruction by Toronto Metropolitan University. Please do not redistribute
1-----
I
I,
(
1;'
1
Historical Research and
\I
, Archival Sources
l \ Michael Roche

CHAPTER OVERVIEW
Despite heightened levels of interest in qualitative research methods, human geographers have
tended to overlook one of the oldest of qualitative techniques; historical approaches based on
the study of primary documents. The interrogation of archival sources is an essential technique
for most historical geographers and while other human geographers have been willing to incor-
porate a historical dimension into their work they do not necessarily have any archival research
skills (Driver 1988). Human geographers more generally can benefit from having some of these i I
I
skills, for they can be deployed for contemporary as well as historical research.

;/!
INTRODUCTION 'f
':

After twenty-five years of working as an historical geographer I still relish the opportunity to
undertake archival research. There is a continuing sense of delving into the unknown, of engag-
ing in academic detective work trying to make sense of inevitably fragmentary and partial sur-
viving rerords, of striving to make sense of the evidence you are scrutinising. And yet, my own 'i i
introduction to archival research was orchestrated only at the graduate leveL To a large degree I 1
was able to learn by a process of trial and error, following the tendency of historical geographers
to regard ~chival research as part of their craft, something to be acquired on the job. Good
archival scholarship was to be inferred from reading journal articles or books by leading geog-
raphers working on related topics and from points of interpretation arising from discussions
with supervisors. In many ways this was a laudable model, one that allowed me to develop my
skills and understanding at my own 1'ace, but geography srudents of today wishing to explore
archival sources can benefit from a more overt discussion of fundamental steps in exploring
arch.ives. This is also the case because archival research needs to engage with larger disciplinary
theory and the research ethics that are also a part of historical inquiry. Even so, like other

133
methods. archival skills can [0 some extent only be learned by doing archival research. Expertise
improves with experience and this is not easily reduced to a checklist of best practice.

WHAT IS ARCHIVAL RESEARCH?

Archival scholarship at irs best. ir seems to me, is an ongoing. evolving interaction between the
scholar and rhe voices of the past embedded in the documents (Harris 200 I, p . .3.32).

Archival sources are a subset of what historical geographers and historians refer to as primary
sources. These include non-current records of government departments held in public archives
but can be extended to include company records and private papers. As well as documents.
handwritten and typed, these can embrace personal letters, diaries, logbooks, minutes of meet-
ings, as well as reports, plans, maps, and photographs. This chapter focuses on oHicial papers,
including manuscript and typescript files, maps, and photographs. I Most of the comments are
also applicable to company archives and private papers. With the target readership of this book
in mind this chapter concentrates largely on government archives, the more recem past of the
last century or so, and on the 'New World'. This simplifies the discussion as much of this doc-
umentation is typewritten and the language of the more recent past is relatively easy to com-
prehend. As a repository of unique, single documents, created contemporaneously with the
events they discuss, the materials lodged in archival repositories provide a particular window

Archives New Zealand

<http://www.archives.govt.nzlindex.html>
library and Archives of Canada
<http://www.collectionscanada.ca/index-e.html>
National Archives of Australia
<http://www.naa.gov.au/>
National Archives of Ireland
<http://www.nationalarchives.ie/>
National Archives of Scotland
<http://www.nas.gov.ukl>
National Archives of South Africa
<http://www.national.archives.gov.za/>
Public Record Office (England and Wales)
<http://www. nationa larchives.gov. ukI>
US National Archives and Records Administration
<http://www.archives.gov/>

..
-
Historical Research and Archival Sources I 135 II
I
I
I
on the geography of earlier times. As such they are a major source of valuable information for
geographers. Historical approaches applied to archival sources will not enable all of the research
questions of human geography to be addressed; however, they do provide a means to answer
questions about the recent and more distant past that are not recoverable by the other tech-
niques and sources available to human geographers.
More often than not the researcher will make use of public archives, a government agency
charged with the preservation of non-current records. On other occasions small regional col-
lections, such as those associated with some museums, may be targeted. Sometimes access to
the records of private organisations may be sought. Michael Williams (1992) provides a
concise summary of a range of archives from the national to the local. The Web increasingly
provides the initial contact point between research and archival collections. Some major
repositories and their 'www' addresses are listed in Box 9.1.
For Mayhew (2003), historical geography has a two-fold significance to the discipline as
a whole that lies beyond increasing the understanding of the geography of the past: to
include re-evaluation of taken-for-granted concepts, and to develop a comparative perspec-
tive so that as geographers we might more fully appreciate what is distinctive about today's
world and how we understand it in disciplinary terms. Historico-geographical research
based on archival research underpins both of these objectives.

ADVICE ON CONDUCTING GOOD HISTORICAL AND

ARCHIVAL RESEARCH

The first step in reconstruction of past stages of a cultural area is mastery of its written docu-
ments (Sauer 1941, p.l3).

It is an oversimplification to believe that the study of change through time by means of either
historical documents or field evidence does not require special training and skills (Perry
1969, p. 96).

So where to begin? Good archival research is difficult to reduce to a checklist of points.

However, it is useful to explore chronologically the sorts of things you might need to do and
that you might encounter in undertaking archival research. To begin, like any other research
project your work ought to be informed by some in-depth reading on current scholarship
around the topic but accompanied by an openness to discovery. The point is well made by
the highly accomplished Canadian historical geographer Cole Harris (Box 9.2).
Archival research begins before you arrive at the archive. You ought to be familiar with
the existing literature before beginning any search for archival material. For instance, for a
project involving state agencies, it is important to read any previously published institudonal
histories, the annual reports of that agency, and to look at parliamentary debates (or their
equivalent). You should always write to any archives you intend to visit well beforehand and,
if possible, get a letter of support from an academic supervisor. This will help you make best
use of what is inevitably too little time. Outline your research topic in reasonable detail,
indicate if it is for a small project or for a thesis, and signal what you are seeking to do and
 r:;;,',rl""':!.':'.~-----------"'-
! 136 'Doing' Qualitative Research In Human Geography

, Ii
!~dII APPROACHING ARC~IVAL RESEARCH
"-;';:';7'" <.. J:- -}. L., _,,_~" "";'--'-"'-'~"'- -_-C. ... .::>,,;.·!.t!Cn_
~! i

I,It
i r
The first point to make about archival research is that it cannot be contained within a single

Ii :, methodology, Any sizable archive holds a vast array of material, and even if one's research Ques-
!! II tions are fairly specific, the chances are good that there will be far more potentially relevant doc-
! ~ ! l;
'I! i! uments than there will be time to examine them, .•
~ I~ ... [A]rchival research tends to gravitate towards one of two polar reactions-neither, I think
! I j1 particularly helpful. It is easy enough to be taken over by the archives, to attempt to read and
ii :,~,
d :1
record all their relevant information. In this way months and perhaps years go by, and eventu-
.l t ally the investigator has a vast store of notes and usually, rather weak ideas about what to do
:i
"

with them, Afraction of the archives have been transferred from one location to another, while
the challenges of interpretation have been postponed ...
I,
'I I.,l!: In effect the archives have swallowed, the researcher. At the other pole are those who come
to the archives with the confidence that they know precisely what they want. They have concep-
Ii II' tualized their research thoroughly in advance. They pretty much know how they will argue their
~t
"II
/1
case and what their theoretical position is. But they do need a few more data, which is why they
return to the archives. As long as they cleave to their initial position. either they will find that
,!
data they need and leave fairly quickly or they will not find them and also leave. Fair enough for
i
h
1:1
certain purposes. But they are imposing their preconceptions on the archives. They have solved
j ~ the problem of archival research by, in effect. denying the complexity of the archives and the myr-
~;
iad voices from the past contained in their amorphous record (Harris 2001, p. 330-1).

how much time you have at your disposaL The archivist. with an intimate knowledge of the
both the ways in which the materials are organised and of their contents, will be able to help
you identify relevant materiaL
Archives are not like libraries though they may have some similarities. They may oper-
ate on quite restricted hours. Some will require that you sign up for a reader's ticket (you
may be able to do this electronically). You may also need to have someone vouch for you. If
you can, you should check out the specific characteristics and requirements of the archives
before you visit.
What should you take to the archives? Take related research noteS and pencils and paper.
Most archives operate on a 'pencil only' rute, to minimise damage to the original documents
should any be accidentally marked, albeit that many researchers now bring their own laptop
computers. Although most public archives will have supplies of scrap paper I would also
advise you not to depend on the archivists to supply stationery. Do not expect a small regional
archive to supply either pencils or to have a convenient power point for your computer.
Sometimes ingenuity is called for in that research questions may only be able to be
addressed obliquely. For instance, recently I have become interested in the discharged soldier
settlement scheme implemented in New Zealand after 1916 in order to recognise the contri-
bution made by those who served in the armed forces during World War I. This scheme was
portrayed in the 1930s as being a failure. In order to look for evidence of failure and success I
have had to understand what these terms meant at the time and search for sources of evidence.

....
 -~' f. ·1
Historical Research and Archival Sources I 137
I
I
Ultimately I apprcciatcd that the registers listing f~lrm land taken up by discharged soldiers
contained incidental material about the percentage of the dairy company milk pay-out appro-
priated directly by the government to repay deferred rents. This provided a due to levels of
indebtedness and the extent to which the ex-soldier dairy farmers were succeeding or strug-
gling. This information was particularly useful where the individual farm files have not sur-
vived. The archivists themselves can sometimes provide helpful, expert advice about record sets
that you may not have considered useful. Typically, as a new user you will be given the oppor-
tunity to explain what you are researching and why. The archivist will tell you how the find-
ing aids work and will be able to offer suggestions abom where to start looking. Their
experience and expertise can often prove invaluable but it is important to remember that they
may have limited time available to provide help to individual researchers.
A crucial difference between a library and an archive lies in the way in which each stores
material and by the nature of the finding aids. Libraries typically catalogue books and jour-
nals by either the 'Dewey Decimal' or 'Library of Congress' classification systems that group
together all books with similar subjects. In contrast, public archivists seek to maintain the :1

integrity of the record sets they obtain from government departments or other agencies in Ii
terms of preserving specific files place in the broader record set, maintaining the original 11

ordering of documents in the file, providing storage conditions that will ensure the long II
t
term survival of the records, and making them available to the public. Archivists place great F
emphasis on the provenance of the fLIes; the actual order of the material within the files in I
Ii
itself tells the researcher something about the situation that prevailed when the file was being rt,j!
created. Thus, whereas in a library you can refer to a catalogue to find a book on a particu- ,:
,I, i i
lar subject on an open shelf, in an archive basic finding aids take the form of sequential ?:
n
i1
,I
'series lists' of all the files held by particular agencies. '~
:lj
In many major archives electronic searching of the collections is now possible. This
il
means that specific items can be searched for so that the process, superficially at least, .~

becomes more like using a library catalogue. In large archives this can increasingly be done "~t
.~
from distant places. thus reducing the time spent in the archives themselves searching for 'I
material and releasing extra time for reading the materials found. Do not forget about prove- !l
;~
nance. What survives in the file is likely to be only a fragment and it may be quite partial in
terms of insights that it provides about the past. In many cases, however, you will have only
the list of files to guide you. The files remain organised along the lines that the original cre- ,!
ating agency devised. Inevitably you will find that some of the records staff have been more
thorough than others. The name of a file may not always be a clear guide to its contents;

1
.;
••
material may be misfiled; and some files may have been lost or destroyed. i
In the case of some national collections. precious and fragile originals may have been
electronically scanned or photographed and made available online or as microfilm copies.
Again, it is important to check in advance whether you will be reading originals or copies of
the documents that you plan to consult. Some sources have been much studied and are
I
![
:t[
,i
i;
I·
I~
available in published form (e.g.• Powell 1973). However. remember to use these sources
critically for they may have been edited reflecting the conventions and morals of the age. For
'I
ill
example, well-known New Zealand historian Bill Oliver provides a light-hearted reflection I

on this serious matter when observing that the published correspondence of the Richmond
and Atkinson families, who were prominent in nineteenth-century New Zealand politics.

I I
PfiOFERTY OF
RYERSON UrJIVERS/TY LIBRARY

•
r;;:rr-:-'- - - - - - - - - - -.....
I I 138 'Doing' Qualitative Research in Human Geography

"r
!
I

i 'struck me as curious in places; upon inspecting the originals I found that the editor had not
only made mistakes in the transcription (a venal sin) but has defaced the manuscript with
, I'
Ij !, overwriting and instructions to his typist a (a mortal sin or if it isn't it should be)' (Oliver
2002, p. 107). Just because some primary source material has been published in printed
df!!i
II
form does not mean that you can relax your cridcal judgement.

d~
1,[ i
Other files held may have restricted access. Personnel files typically fall within this cate-
gory. In different national contexts the period during which restricted access applies may

iJ
~1
I,'
vary, but thirty years after the closing of the file is typical. In some instances a lesser degree
of restriction applies where permission to look at files may be granted by a senior archivist,
!~
,1 I
government official, or someone associated with the organisation that created the files. A
,! q formal written request outlining your research project may result in access being granted;
however, this may have some conditions attached, for example being able to read only a par-
ticular portion of the material and having the remainder of the file physically sealed. In other
cases the researcher can do nothing but ~ait patiently until the material is released.
Unlike a library, archived files are not kept on open shelves. They are not necessarily kept
on the same site as the reading room and may only be delivered from storage hourly upon
request, so be prepared to have several tasks to go on with (for example, searching the find-
ing aids for other files to call up while you wait). Photocopying of material is usually possi-
ble, but this can be comparatively costly and you may have to pay in advance of delivery.
Some material may be deemed to be too fragile or, if bound, too difficult to photocopy. It
is therefore advisable [0 find out what the policy is beforehand. Large plans, maps and charts
in excess of A3 size can be copied by means other than photocopying, but this is sometimes
quite expensive. However, it may be the only means of obtaining a copy of an essential
document.
When the file is produced for you to work on, you will find in most cases that new items
are on [Op of the older material, particularly if the material is secured by paperclip. You will
probably need to work from back to front. Will you answer any of your research questions?
It may be immediately obvious that the material is relevant to your inquiry or it may appear
only tangentially relevant or even irrelevant. Sometimes it is difficult to make a judgement
at first glance and you may have to recall material previously examined whose significance
was not appreciated at the time. Alan Baker, a British historical geographer, has offered some
guiding thoughts on evaluating primary sources including archival.materials (Box 9.3).
Baker's words seem to me to be crucial for those using archives as qualitative sources in
human geography. It is essential to understand as fully as possible the original purpose of the
document, who created it, and how and when it was made.
On the basis of their extended experience working with documents, historians have
identified some generic questions to pose when assessing them in order to better understand
their significance (Box 9.4).
These questions provide a useful starting point for geographers as well as historians
although I would make three qualifying points. First, it is possible to extend 'document' to
include maps and plans '(see Harley, 1992). Second, this approach tends to privilege the
ideas behind actions. That is to say, the past is being understood in idealist terms where the
thought behind the action is regarded as providing the understanding necessary to interpret
 Hlstoncal Research and Archival Sources I 139
rr
I '

No source should be taken at face value: all sources must be evaluated critically and contextually.
The history and geography of a source needs to be established before it can legitimately be utilised
and incorporated into a study of historical geography. The historical sources we use were not com-
piled and constructed for our explicitly geographical purposes; they were more likely to have been
prepared, for example, for the purposes of taxation and valuation, administration and control. We
also have to understand not only the superficial characteristics of a specific source but also its
underlying motivation, background and ideology of the person(s) who constructed it. In order to
make the most effective and convincing use of a source we must be aware of its original purpose
and context and thus from its limitations and potential for our own project (Baker 1997, p, 235).

THE. HISTORIAWS STANDARD QUESTIONS OF DOCUMENTS J,l

lr i
" '(~fte"Kitso" Clark 1969) 'I,
oiAiiiialiiiil .a. _~_::. ~~ ~~ ·'"-$; ...... t7 .. 'M,,~- .,. ?7t7C"''"~ I!
I I

What was the writer's bias? ;1 i

2 What was the writer's situation and intention at the time of writing?
3 What were the writer's opportunities for knowledge?
4 What were the writer's general standards of truthfulness?
5 What powers of critical observation did the writer usually bring to bear (was the
writer credulous)?
6 What was the writer's framework of ideas and what did certain words mean when
used by the writer?

these events. Historical geography can be written legitimately from a viewpoint other than
that of contemporary observers (Baker 2003). Third, the documents themselves cannot be
read in isolation but must be understood in their wider context and even then any conclu-
sions will be provisional rather than absolute.
To some extent all archival researchers develop individualised strategies for note-taking
from archival materials. However, there are two basic strategies. This first involves collecting
material by topic, noting specific details and suitably referenced quotations. Classically, his-
torical researchers have made use of large index cards for this purpose, though many now
use computers to organise these notes. New topics can be added on new cards as more files
are read and new research questions formulated. The alternative approach is to record
chronologically any pertinent information from each file and then subsequently to identify
themes that emerge across the files. Both strategies have advantages and disadvantages. The
former depends on identifying key topics at the beginning of the project within which to
collect information. Such an approach offers the capacity to add new topics or to identify
dead ends and to see how themes merge or diverge. My personal view is that, while this
approach means that many diverse sources are brought together, it can blur a researcher's

d
I;

,/11
I: .:..
 140 'DOing' Qualitatrve Research In Human Geography

capacity to make good inductive judgments. The latter method is more sensitive [() the
provenance of files and can give a clearer sense of the role of particular officials or depart-
ments. It does however involve a degree of double-handling in that evidence that has already
been collected by the researcher needs to be reorganised after each visit to the archives and
I perhaps annotated further. It is important to follow up other research questions that may
Ii'! emerge from this re-sorting process. The latter approach is one that I have used over many
I; i , I
years. It suits me and, as a full-time academic, I can incorporate it into my way of working.
J I' I would acknowledge that it probably works best when researching in an area where both

;I, the secondary literature and the archival sources are familiar, even if the specific contents of
the files are unknown. Students with a relatively short period for archival research may pre-
,1
III i
'I ",I fer to work to the first model outlined above.
Having located and extracted archival evidence, this must be adequately cited in the writ-

Iii.)I
II
ten products of the research. The first step is carefully recording the specific document descrip-
tion and file reference. For example, the personnel file for Edward Phillips Turner, Director of
Forest in New Zealand from 1928 to 1931, is located in the New Zealand Forest Service files
II at Archives New Zealand in Wellington. The specific reference is 'F Acc W2338 821113 E. P.
Turner'. The 'F' refers to the Forestry files, 'W2338' is the accessions number for that collec-
,II'.
I ,
'J

• , r tion of documents, 82 is the series, and 113 the specific file number, while E. P. Turner is the
1
descriptive label (a little confusing in that the individual in question always used the double-
!r!:1
, j barrelled surname). There is considerable variation among filing systems. You need to quickly
become familiar with the system used by the agency whose records you are working on, and
,! I
jI ,
here the expert assistance of the archivist can be invaluable. The crucial thing is to record the
details carefully. This is important for two reasons. First, it enables you as a researcher to keep
track of where you found specific infurmation. Second, it enables a subsequent researcher to
I,,;
l I'"' , t relocate the m,uerial. The idea is simple enough but given the nature of archival material it is
somewhat more exacting than, for example, the standard bibliographic requirements of author,
date, tide, and publisher/place for a book in the reference list of a thesis. Citing archival mate-
rials correctly can also pose problems in that human geography has tended to adopt versions
of in-text citation systems, such as Harvard (see Hay 2002). Most archival sources sit uncom-
fortably within this framework and are generally better referenced in footnotes or endnotes,
typically used by historians. Students undertaking archival research may need to negotiate a
variation from their university's social science oriented formats for referencing.
The archive does not constitute the only source for historical research. For instance.
newspapers, private papers. and unpublished memoirs may provide valuable material for
cross-referencing with the archival record. Likewise, once archival work is completed, the
researcher may need to follow up on unfamiliar key actors by checking old editions of Who}
Who, or newspaper obituaries. This 'post-archive' work can of course help to shape and
inform the purpose of subsequent trips to the archive.
Moreover, files are not the end point of research, as US cultural geographer Carl Sauer
reminded historical geographers over 60 years ago:

Let no one consider that the historical geographer can be content with what is found in
archive and library. It calls, in addition, for exacting fieldwork. One of the first steps is the
ability to read the documents in the field for instance of an account of an area written long
 Hlstoneal Research and Archival Sources I 141

ago and compare the places and their activities with the present, seeing where the habitations
were and the lines of communication ran, where the rtHeS[S and the tlcld stood, gradually get-
ting a picture of the former culturalland,cape behind the present one (Sauer 1941, p. 13).

While Sauer's words llIay indicate nostalgia for a pioneering rural past and the focus could
just as easily be urban and social, his challenge remains pertinent and remains one only par-
tially taken up by subsequent generations of geographers (for example, Raitz 2001).

CHALLENGES OF ARCHIVAL RESEARCH

There are two sorts of challenges facing researchers working with archives. The first of these
is intellectual and the second technical. When dealing, for instance, with the file materials
contained in an official government archive it is important to bear in mind the sorts of
power relations inherent in the surviving materials. This is rather more than just acknowl-
edging that the surviving files are fragmented and partial. The records are those created by
politicians and officials. They reflect the outlooks and understandings of the dominant
groups in that national context at the time they were created. Duncan (1999) writes of these
concerns in terms of complicity stemming from use of the 'colonial archive'. Thus for much
of the nineteenth century and well into the twentieth century they are records created largely
by men in the upper echelons of society, and in a colonial situation such as in New Zealand
are predominantly the records of colon ising British settlers. Summarising the contents of
files from the archives merely reproduces these uneven power relations rather than inter-
preting them. The records of nonofficial community or sporting groups may provide a way
into understanding the concerns and aspirations of those who had no position in the pub-
lic political sphere. Likewise oral histories for the recent past may provide insights into gen-
dered and minority concerns. Furthermore an awareness of the power relations within the
archival material may allow the researcher to reinterpret surviving materials. For instance,
what I once mapped as examples of illegal felling of forest in New Zealand in the 1870s I
would now be inclined to understand as a resistance on the part of Maori forest owners to
the imposition of authority by the Crown and of flaunting of government regulations by
timber cutters who had limited other means of supporting themselves.
The most fundamental technical difficulty relates to the ability to actually read the doc-
uments retrieved in the archive. For the first half of the 19th century many official docu-
ments were handwritten in copperplate script. This looks elegant but can take some time for
novices to learn to read proficiently, a situation that may be exacerbated when officials wrote
both across and along a sheet of papers in order to save paper. Perseverance will payoff.
Archives for Australia and New Zealand are from the latter nineteenth century and are gen-
erally written in a 'modern hand'. They are generally readable with a bit of effort, the main
problems being encountered with faint letterbook copies. However, original manuscripts
concerned, for example, with the early settlement of North America in the period before
1700 may be written in 'secretary hand'. This was the script of professional scribes of the r'
I,
,,
time and is difficult to read without specialised instruction. Unless the material has been I
transcribed and printed. its translation requires additional palaeographical skiJ1s. In any case, f
I

II
r
i J

~ !I
i!
,I!;
 142 'Doing' Qualitative Research in Hurnan Geography
•
all kinds of handwritten documents made in the past tt:nd fO bt: difficult to decipht:r t:spt:-
, cially when the investigator is trying to read only a binr letter book copy of rhe original. for
! instance, in the mid-nineteenth century the 'long s' written similar to an T W;L, fretluendy

I used in official corrcspondt:nce. This means that words with a doubk 'ss', t()r instance 'lesson'

,I are rendered as what looks to us like 'Ieston'. Not only do spellings change as you move hack
in time but so does rhe very construction of the English language." This makes it more of a

I!
II
challenge [0 understand the world-view of these earlier times. from arollnd the I BROs, type-
written material becomes more common in government files, but important marginal <lnI10-

I
1
t,Hions will be handwrirren and often cryptic in meaning.
A good example of some of the challenges posed by handwritten documents is provided
by the correspondence of Captain Campbell Walker, the tlrst Conservator of Forests in
I 1876-77 in New Zealand. His letters were an important source for my PhD research into state
forestry but were nearly impossible to decipher. I decided to photocopy and later transcribe
them bur only managed to do so after painstakingly working out on a Ietter-by-Ierrer basis
how he wrote the alphabet. Anyone contemplating the use of archive material as a source for
qualitative research in human geography must be prepared to be patient and resourceful; the
use of documentary evidence is rarely easy and generally requires a great deal of time.
The units of land area and currency may also be different from those in use today (for
example, acres rather than hectares), This raises the issue of whether to convert every meas-
urement to the current system or to give a general conversion factor and use the units of the
period (generally, I prefer the latter). Some facility with the original units is useful.
Appreciating that there are 640 acres to a square mile makes it possible to recognise, for
example, that the apparently precise data on the forest areas in Otago Province in Kew
Zealand in the 1867 are actually only estimates to the nearest quarter mile or 110 acres.
More specialised measures may also need to be understood depending on your field of
research. For example, throughout the British Empire from the nineteenth century quanti-
ties of sawn timber, for instance, are often given in superficial feet (colloquially referred to a
'superfoot'), that is: twelve inches by twelve inches by twelve inches (thirty cm by thirty em
by thirty cm). However, in North America the term 'board fom'-which is not a surfing
term-was more often used for the same unit.
If you are working through a file and time runs short you may find yourself copying
whole documents that could be of use-but do not have the time to read them carefully and
decide. In the end you find yourself with page after page of material that may subsequently
prove to be marginal to the research. This is particularly important in that when returning
to your research material it is too easy for the photocopied docum::nt to overshadow your
handwritten notes so that you end up with the situation Harris refers to as the archive 'swal-
lowing the researcher' (see Box 9.2).
Mistakes in interpretation can and do occur. I once mistook the numbered applications
for the position of Director of Forests in New Zealand for the rankings of candidates. The
result was an apparently nonsensical list of candidates. On closer subsequent investigation
of the date stamps showing the receipt of application it became clear that the numbers
related only to the order in which they had been received. Retrospectively, I can draw three
points from this episode. First, scrutinise documents carefully. In this case, the answer was
there in the documents but I did not see it first time around. Second, if you are uncertain
 Historical Research and Archival Sources

abollt what the docllmclHs indicate acknowledge this and do nO£ make too definite a claim
143
[!
I'
i'
11I
!

of the surviving evidence, Third, by 'learning the ropes' as an undergraduate or graduale stu- ii !
dent YOll can avoid some of the more obvious pitfalls of interpretation before you have any-
thing published,

'j

ETHICS AND ARCHIVES

It is all too easy for archival researchers [0 dismiss ethical issues as something relevant only Ii
to geographers working on present-day topics with other qualitative or quantitative method- ~ i

i :!:
ologies. Actually, archival researchers also have some ethical obligations, accentuated by the I ;;1

fact that the individuals who created--or are the subjects of-the records in question are, in
all likelihood now deceased and unable to defend themselves. Moreover, the deceased may !
1,
l!
!
have living descendants, so sensitivity for their sakes is required, \'¥'here more recent records
are being used the individuals involved may still be living. The issue is one of making an
!
i !
informed and defensible judgement about past events and the actors involved therein, !i I
Scrutiny of social science research projects by universiry human ethics committees is now
increasingly the norm, The ethical aspects involved in archival research have less to do with 11
~
safety, harm, and risk to the researcher or research panicipams and more to do with retain-
I"
ing the imegriry of the material contained in the files and the preservation of the archives I:
themselves. Those using archives should, after all, regard access to the material as a privilege. r
~
Historical records are precious and often irreplaceable. All researchers are under an obliga-
~
tion to look after archival material and to ensure that it is preserved in good order for any II
~~i
subsequent scholars.
The US National Council of Public History identifies three guidelines for using archive
material. 3 We can reasonably substitute 'geographer' in each of these. , '

"r~

r
Historians work for the preservation, care and accessibility of the historical record, The ij

unity and integrity of historical record collections are the basis of interpreting the past.
2 Historians owe to their sources accurate reportage of all information relevant to the sub-
ject at hand, ~ "

3 Historians favour free and open access to all archival collections, (National Council on Ii'

I
Public History 2003.)

Those who are living can take a libel action against you if they feel you have defamed them.
More often in historical research you are dealing with those no longer living. While you can- II I
not libel the dead, what sensitiviry do you need show to them and their descendants? Some
:
of these issues can be illustrated by way of example, As part of the research on World War I I
\
soldier settlement I referred [0 earlier, I have viewed the soldiers' personal files for informa- ,r
tion on age, previous employment details, residential information, and the like. Byexamin-
ing the hospitalisation details it is possible to identify soldiers who were treated for venereal
disease. Do I need to mention this? Ifit is not relevant to my research questions then arguably
it can be omitted. Even over eighry years on it may be embarrassing to descendants, but is
this sufficient reason to omit the information? If it becomes relevant in that soldier was too
ill [0 successfully farm then you may decide it is necessary to mention this point. You might
144 'Doing' Qualitative Research In Human Geography
---lB.
do this with varying degrees of subrlety. However if your research topic was focused explic-
itly on the medical geography of World War I soldiers then you might have to confront these
sorts of sensitivity issues direcrly. Ultimately these decisions rest very much with the indi-
vidual researcher and are nor to be resolved by presenting it checklist bur can be worked
through in terms of thinking about the ethics of research (Hay 2003a, pp. 46-7).

PRESENTING THE RESULTS OF ARCHIVAL RESEARCH

There is no single correct way of presenting the results of archival research. The theoretical
foundation of the research project, the sorts of empirical information retrieved, and the writ-
ing style of the researcher all shape how the research project or thesis is expressed. 4 Typically
however, archival researchers will make use of direct quotations from key documents in
order to demonstrate their case (see also chapter 17 for a discussion). They also make use of
case study material to illustrate points. On occasions good use may be able to be made of
cartographic or pictorial material.
Adept researchers are often able to move easily from specific points of detail to sketch a
much larger picture and to relate this to what is known about related topics. I would rec-
ommend critical examination of writings by leading historical geographers (for example,
Dennis 2001; Harris 2002; Meinig 1986; Powell 1988). However, it is not just a matter of
identifYing key quotations, but rather of building an argument. This obliges you to select
ideas in a logical way from the pre-existing literature and then to use this to provide an
informed discussion based on what has been found in the archives. The desirable end point,
however, is to be in command of the SOUTce material. Rather than merely reproducing a
chronicle of part of what is contained in the archive strive, to make your writing a synthe-
sis of specific detail and informed interpretation.

CONCLUSION
Although it has much to offer human geography in general, archival research tends to have
been neglected by other than historical geographers. As a research method, historical
research using archival sources;

1 calls for creative thinking in identifYing source materials relevant to your research problem
2 needs patience, precision, and critical reflection in collecting and evaluating material
3 requires a sense of historico-geographical imagination in interpreting source material
whereby theorisation does not outstrip the evidence
4 is partial and requires that you relate archival material to other contemporary sources of
a textual and pictorial sort that may be held in other collections
5 asks researchers to continually negotiate between the theoretical and the empirical.

Archival work can be extremely time consuming, and superficially, at least, frustrating in
that the information retrieved may offer only partial answers, particularly where you find
yourself under time pressure to complete a research project. Archival work done properly
 HistOrical Research and Archival Sources I 145

takes time and patience. Rarely will the surviving archival material provide 'full' answers ro
the questions you pose. In the case or
public archives the surviving material rypically says
more about politics, economics, the concerns of elites, and men than it does about social
and private spaces, women, and minority groups. Likewise, surrendering to the temptation
to merely summarise the content of files, a trap into which inexperienced archival
researchers can fall, is another way of being-as Harris terms it-'swallowed by the archive'.
It is all too easy in discussing archival research ro create the impression that there is no
room for novices when, in fact, more human geographers need to be encouraged to incorpo-
rate archival work into their research programs. I would simply describe archival research as
somewhat akin to confidently accepting the challenge of working on a jigsaw puzzle where
you can be reasonably certain that pieces are missing and that the box cover with the picture
of the completed puzzle will never be found. Good archival research can be extremely satis-
fying, both in learning the skills to conduct it and in the presentation of resuits.

archives 'pencil only' rule

archivist provenance
files records
finding aids

1 What sorts of research questions can be addressed using archival sources?

2 What are the problems of a researcher being, as Harris terms it, 'swallowed by the archive'?
3 What is meant by provenance and why it is important to archival researchers?
4 What steps are required in assessing historical evidence?

~I
:;
!

Baker, A. R. H. 2003, Geography and History, Bridging the Divide, Cambridge University Press,
Cambridge.
_ _ 1997, 'The dead don't answer questionnaires: researching and writing historical geography',
, I
Journal of Geography in Higher Education, vol. 21, no. 2, pp. 231--43.
Cameron, L. 2001. 'Oral histoty and the Freud archives: incidents, ethics and relations', Historical
Geography, vol. 29, pp. 38--44.
Hall, C. 1982, 'Private archives as sources for historical geography', in A. R. H. Baker and M. Billinge
(eds), Period and Place: Research Methods in Historical Geography, Cambridge University Pres,
Cambridge, pp. 274-81.
Hanlon, ]. 2001, 'Spaces of interpretation; archival research and the cultural landscape'. Historical
Geography, vol. 29, pp. 14-25.
!
I
 r. ,. ." .,.~ .\... > .."...,

145 'Doing' Qualitative Research in Human Geography

Harris, C. 200 I, 'Archival fieldwork', GeogmphiuZ/ Rellu'u', vol. 91, nos 1&2, Pl" ,'\28-34,
Harris, C, 1978, 'The historical mind and the practice of geography', in D, Ley and ~1. Samuel (elh),
Humanistic Gl'ography, Croom Helm, Chicago, pp. 12.3-37.
Kurtz, M. 200\, 'Situating practices: the archives and the fIle cabinet', Historical Geogl'llpby, vol. 2'),
pp.26-37.
,\1ayhew, R. 2003, 'Researching historical geography', in A. Rogers and H. Viles, (cds), JlJe Studellts
companion to Geography, 2nd cdn, Blackwell, Oxford, pp. 260-5,
Ogborn, M. 2003, 'Knowledge is power: Using archival research to interpret state formation', in A,
Blunt, I~ Grufford, J. P. May, M. Ogborn and D, Pinder (eds), Cultural G{'ogmplry in PMctice,
Arnold, London, pp. 9-20.
Powell, J. M. 1988, An Historical Geography of Modern Australia: The Restive Fringl" Cambridge
University Press, Cambridge,
Sauer, C. 1941, 'Foreword to historical geography', Annals of the Association ofAmerican Geographers,
vol. 31, no. 1, pp.1-24.

Notes
For the purposes of this discussion 1 will leave to one side the whole question of archival material cre-
ated in electronic form,
2 Indeed you may be dealing with records in anmher language, For insrance there is good deal of cor-
rdpondence in Maori in Archives New Zealand: much, but not all, of which is accompanied by an
English version prepared by official translators,
3 These are expanded on in the American Historical Association's Statement of Standard ofProt'essional
Conduct.
4 Given that this volume focuses on qualitative methods I have omitted discussion of how a researcher
might extract and present in tabulated form quantitative information derived from archival sources.

..
.Using Questionnaires
in Qualitative
Human Geography
Pauline M. MCGuirk and Phillip O'Neill

CHAPTER OVERVIEW
This chapter deals with questionnaires, an information-gathering technique used frequently
in mixed method research that draws on quantitative and qualitative data sources and analy-
sis. We begin with a discussion of key issues in the design and conduct of questionnaires. We
then explore the strengths and weaknesses for quatitative research of various question for-
mats and questionnaire distribution and collection techniques. Finally, we consider some of
the challenges of analysing qualitative responses in questionnaires and we close with a dis-
cussion of the limitations of using questionnaires in qualitative research.

INTRODUCTION
Qualitative research seeks to understand the ways people experience the same events, places,
and processes differently as part of a fluid reality; a reality constructed through multiple
interpretations and filtered through multiple frames of reference and systems of meaning-
making. Rather than trying to measure and quantifY aspects of a singular social reality, qual-
itative research draws on methods that reveal and interpret the complexities, context and
significance of people's understanding of their lives (Eyles and Smith 1988). Within this
epistemological framework, how can questionnaires contribute to the methodological reper-
toire of qualitative human geography? This chapter explores the possibilities.
Commonly in human geography, questionnaires pose standardised, formally structured
questions to a group of individuals, often presumed to be a sample of a broader population
(see chapter 5). Questionnaires are useful for gathering original data about people, their
behaviour and social interactions, attitudes, and opinions, and awareness of events
(Mclafferty 2003; Parfitt 1997). They usually involve the collection of quantitative and
qualitative data. Since such mixed-method questionnaires first appeared with the explosion

147
148 'Doing' Qualitative Research In Human Geography

of behavioural geography in the 1970s (Gold 1980), rhey have been llsed increasingly w
gather more complex dara of a qualitative nacure in relation to maners as varied as the envi-
ronment, transport and travel, quality of life and community, work, and social nerworks.
\Xlhile there are limitations to rhe depth and C,.xtent of qualitarive data thar quesrlonnaires
are capable of gathering, they have numerous strengths. Firsr, rhey provide insights into rele-
vant social £rends, processes, and interprerations. Second, they are cosr-effecrive, enabling
extensive research over a large or geographically dispersed population. Certainly, they are one
of the more practical qualitative research tools. Third, they are exuemely flexible. They can be
combined very effectively with complementary, more intensive forms of qualitative research [0
provide more in-depth perspecrives on social process and context. For instance, Ruming, Mee
and M!.;Guirk's (2004) investigation of the impacts of the social mix policies of New South
Wales' Department of Housing on understandings of communiry combined key informant
interviews with housing officials, questionnaires with local residents, and follow-up in-depth
interviews with volunteers who had participated in the questionnaire. Qualitative data from
the questionnaire provided a framework for the in-depth interviews, allowing key themes, con-
cepts, and meanings to be teased out and developed (see Askew and M!.;Guirk 2004; England
1993; and Winchester 1999 for similar examples). In this mixed-method format parriculariy,
questionnaires can be both a powerful and practical research method.

QUESTIONNAIRE DESIGN AND FORMAT

\Xibile each questionnaire is unique, there are common principles of good design and
implementation. Producing a well-designed que~tionnaire for qualitative research involves a
great deal of thought and preparation, effective organisational strategies, and critical review
and reflection, as an array ofliterature suggests (for example, de Vaus 1995; Dillman 1978;
Fink and Kosecoff 1998; Foddy 1993; Fowler 2002; Gillham 2000; and see the relevant
chapters in Babbie 2001; Clifford and Valentine 2003; Flowerdew and Martin 1997; and
Hoggart et al. 2002). The design stage is where a great deal of researcher skill is vested and
it is a critical stage in ensuring the usefulness of the resulting data. Notwithstanding the
quality of the questionnaire devised, we are beholden as researchers to ensure that we have
sufficient reason to call on the time and energy of the targeted research subjects. The desire
to generate our 'own' data on our research topic is insufficient justification (Hoggart et al.
2002). As with any study, the decision to go ahead with a questionnaire needs to be based
on careful reflection on detailed research objectives, consideration of exisfing and alternative
information sources, and appropriate ethical contemplation (see chapters 2, 4, and 5).
The content of a questionnaire must relate to the broader research question as well as to
your critical examination and understanding of relevant processes, concepts, and relationships.
As a researcher, you need to f.uniliarise yourself with existing local and international work on
your research topic. This ensures clariry of research objectives, and will help you to select an
appropriate target population and relevant key questions. You need to be clear on the purpose
of each question, who will answer it, and how you intend to analyse responses. You also need
to be mindful of the limits to what people are willing [0 disclose, being aware that this varies
across different sodal and cultural groups in different contexts. Public housing tenants, for
 Using Questionnalf8s in Qualitative Human Geography I 149

insrance, might be wary about otTering candid opinions about their housing authority. Every
question, then, needs to be carefully considered, and have a dear role and purpose.
Begin by drawing up a li~t of topics that you seck to investigate. When you have clari-
fIed these, develop specific questions. De Vaus (1995) suggests that it is helpful to think
about four distinct types of question content:

1 attributes Attribute questions aim to establish respondents' characteristics (for example,

age or income bracket, owner-occupier or private renter)
2 behaviour With behaviour questions, we are interested in discovering what people do
(for example, recreation habits, extent of public transport use)
3 attitudes Questions abollt attitudes are designed to discover what people think is desir-
able or undesirable (for example, judgment on integrating social housing with owner-
occupied housing)
4 beliefs Questions about beliefs aim to establish what people believe to be true or false (for
example, beliefs on the importance of environmental protection),

A basic guiding principle for all these question types, however, is that you need to be sure
that your target population will both understand the questions and have the knowledge to
I f
answer them. Asking respondents to comment, for instance, on whether they believe gov-
ernment planning policies have contributed to local coastal degradation might be beyond i
their ability to answer with any certainty.
Apart from the typology of question content, there is a range of questions types on
which to draw. We commonly make a distinction between closed and open questions, each
of which offers its own strengths and weaknesses. Closed questions may seek quantitative
I
J

I'

information about respondent attributes (for example, level of educational attainment) or

behaviour (for example, how often and where respondents buy groceries). Some examples
are set out in Box 10.1. Closed questions can ask respondents to select categories, rank items
as an indicative measure of attitudes or opinions, or select a point on a scale as indicative of
the intensity with which an attitude or opinion is held. A major benefit of closed questions
is that their responses are easily coded and analysed, a bonus when interpreting a large num-
j

ber of questionnaires. Closed questions are demanding to design, however, as they require
researchers to have a clear and full understanding of what the range of answers to a question
will be. Respondents' answers are limited to the range of categories designed by the researcher
as an exhaustive and exclusive list of possible answers, and this can be a limitation. More-
over, closed questions rest on the assumption that words, categories, and concepts carry the
same meaning for all respondents and this may not always be the case. For example, how a
respondent answers the question 'How often have you been a victim of crime in the past two
years?' will depend on what the respondent includes in their definition of a crime (de Vaus
1995). A criticism of closed questions, then, is that 'one may learn more about the behav-
iour of the sample in responding to a set a categories ... than about the behaviour under
investigation' (Cox 1981, p. 264). This limitation can be lessened by offering an answer
option such as: 'other (please specify(or by using combination questions that request some
elaboration or explanation for the selection made in a closed question (see Box 10.1).
In general, open questions have greater potential to yield in-depth responses in keeping
with the thrust of qualitative research: to understand how meaning is attached to process.

. .. ::..:...11