You are on page 1of 18

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/255731487

GIS: Tool or science? Demystifying the persistent ambiguity of GIS as "tool"


versus "science"

Article  in  Annals of the Association of American Geographers · June 1997

CITATIONS READS

144 961

3 authors, including:

Dawn J Wright James David Proctor


Environmental Systems Research Institute (ESRI) Lewis & Clark College
230 PUBLICATIONS   2,888 CITATIONS    50 PUBLICATIONS   1,152 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

World Population Estimate View project

EcoTypes View project

All content following this page was uploaded by Dawn J Wright on 11 September 2015.

The user has requested enhancement of the downloaded file.


Forum
GIS: Tool or Science?

Demystifying the Persistent Ambiguity of GIS as “Tool”


versus “Science”
Dawn J. Wright,* Michael F. Goodchild,** and James D. Proctor***
*Department of Geosciences, Oregon State University
**Department of Geography and National Center for Geographic Information and Analysis,
University of California, Santa Barbara
***Department of Geography, University of California, Santa Barbara

Is GIS a tool or a science? The question is clearly important in the day-to-day operations of
geography departments. Departments need to know if GIS is a tool that should be taught at the
undergraduate level, or a science and thus a legitimate research specialty of faculty and graduate
students. We summarize the debate on this question that was conducted on GIS-L electronic
listserver in late 1993. In evaluating this discussion it became clear that GIS could be understood
not by the two distinct positions taken by the GIS-L discussants but as three positions along a
continuum ranging from tool to science. These positions attach several meanings to “doing GIS.”
These are (1) GIS as tool, i.e., the use of a particular class of software, associated hardware tools,
and digital geographic data in order to advance some specific purpose; (2) GIS as toolmaking, i.e.,
the advancement of the tool’s capabilities and facilities (ease of use); and (3) the science of GIS,
i.e., the analysis of the fundamental issues raised by the use of GIS. Recognizing the importance of
understanding what is meant by “doing science” as well as what is meant by “doing GIS,” we
conclude that only one of these positions—“the science of GIS”—is a sufficient condition for
science. The “toolmaker” position is rarely able to meet the test of science; and the “GIS is a tool”
position involves “doing science” only if it yields progress on some substantive problem. The debate
is certainly problematic in light of the variety of perspectives on science and on GIS. The persistence
of the issue suggests, however, that the GIS community should continue to work toward a resolution.
Key Words: GIS-L, systems, geographic information science, geographic thought, nature and philosophy of
science, nature of geographic information systems.

W
hen Roger Tomlinson coined the term 1950s. We cannot be certain whether the GIS
“geographic information system” for debate is more or less intense than its precursors,
the Government of Canada in the but we can be certain that it is important given
early 1960s (Coppock and Rhind 1991), he can the interdisciplinary nature of GIS, albeit a nature
scarcely have imagined the impact that “GIS” in which geography is widely accepted as having
would have on the discipline of geography, or the a unique role (Morrison 1991; Kennedy 1994).
intensity of the debates that were provoked by Geography’s debates over GIS are thus unusually
this seemingly innocent three-letter acronym. Of exposed to general view.
course geography is no stranger to methodologi- The purpose here is not to review the various
cal debates, and some of the arguments over GIS appraisals of GIS, which range from GIS as savior
echo the arguments over quantification in the putting the “geographic Humpty Dumpty” back
Annals of the Association of American Geographers, 87(2), 1997, pp. 346–362
©1997 by Association of American Geographers
Published by Blackwell Publishers, 350 Main Street, Malden, MA 02148, and 108 Cowley Road, Oxford, OX4 1JF, UK.
Tool or Science? 347

together again (Openshaw 1991) to dismissal as science—a tension that ultimately defines what
“non-intellectual expertise” (Jordan 1988); from it means to be “doing GIS” in geography—we
excitement over GIS as positivism’s second com- hope to shed some light on these issues, which are
ing (Heywood 1990) to GIS as a last-ditch rally clearly important in the day-to-day operations of
by positivism’s battered survivors (Taylor 1990). geography departments. Departments need to
More interesting are the social implications of know if GIS is either a tool that should be taught
GIS—the message it sends, whom it empowers, exclusively at the undergraduate level or a sci-
and the responsibility its developers should bear ence and thus a legitimate research specialty of
for its eventual use (Smith 1992; Pickles 1994; faculty and graduate students. Are students who
Harvey and Chrisman forthcoming). In the U.K., “do GIS” doing substantive science? Is an associa-
these debates within the discipline have caught tion with GIS sufficient to ensure that research is
the attention of even so authoritative a source as substantive, or if not, what other conditions are
the Times Higher Education Supplement (Davies necessary?
1995). Much of the motivation for this paper derives
At heart, these debates arise from the ambigu- from a debate on the GIS-L electronic listserver
ity of GIS as a tool or as a science. While Tomlin- in late 1993. These electronic lists, or “invisible
son was clear enough in his definition of a GIS as colleges” (Crane 1972) as it were, span the barri-
a computer application designed to perform cer- ers between disciplines. Since its inception, GIS-
tain specific functions (Coppock and Rhind L has provided a forum for a variety of discussions
1991), it is not at all clear what is meant by “doing of GIS issues (Mark and Zubrow 1993; Thoen
GIS,” “the GIS community,” or “GIS research,” 1996). During October–November 1993, the
since in all these cases the etymological path topic “GIS as a Science” generated 64 postings
between acronym and phrase has become hope- from 40 individuals in 8 states and 6 countries
lessly muddied. At face value, “doing GIS” seems (Figure 1). The usual length and intensity of the
to imply nothing more than interacting with a discussion made it clear that the “tool versus
particular class of software: “the GIS community” science” debate sparked great interest among
is no more than a group of individuals with an many scientists, technicians, and practitioners,
intense interest in that software; and “GIS re- whatever their discipline. One of the objectives
search” seems an oxymoron. By examining the of this paper is to explore the relationship be-
tension between GIS as a tool and GIS as a tween the positions taken in this electronic de-

Figure 1. Graphical summary of GIS-L postings on the topic “GIS as a Science” during October and November
1993.
348 Wright et al.

bate and current debates within the discipline of cussion lists, the World Wide Web, electronic
geography. journals, and digital libraries (on the Internet’s
The “tool versus science” debate has received impact in oceanography, see Hess et al. 1993; and
little mention in the published literature of geog- likewise on traditional journals, see Odlyzko
raphy, which is surprising in view of the attention 1995). Subscribers to an electronic list such as
given to GIS the past decade. The closest the GIS-L are able to reach hundreds of colleagues
literature comes to the debate is Goodchild’s around the world to discuss an issue or ask a
(1992) paper on “geographic information sci- question, thereby crossing all of the traditional
ence,” Sui’s (1994) discussion on reconciling the structures of the research community (save per-
differences between GIS enthusiasts and critics, haps structures based on language) within min-
and the articles on “Automated Geography” that utes. While it is impossible to determine exactly
appeared in The Professional Geographer in how many individuals read GIS-L and with what
1993—the latter, a series of reflections on devel- level of interest, Mark and Zubrow (1993) re-
opments in the ten years since Dobson (1983) ported that at the time of their analysis, the list
announced that advances in analytical methods contained approximately 1100 individual In-
and computer technology had made it possible to ternet addresses and was redistributed to more
automate several aspects of geographical research than 30 additional lists worldwide.
and problem solving. Once registered as a subscriber to an unmod-
In the discussion that follows, the electronic erated electronic discussion list, any individual
debate on GIS-L serves as the point of departure with an Internet address automatically receives
for an exploration of “doing GIS” from the per- all messages posted by any other subscriber. The
spectives of both geography and the society in essential informality of this system of communi-
which the discipline is embedded. The paper cation is both a blessing and a source of difficulty
frames the tension between the positions of “GIS for anyone attempting to synthesize these discus-
as a tool” and “GIS as a science,” summarizes the sions. Many of the discussants do not have the
GIS-L debate, considers the implications of the time, the inclination, or perhaps the energy to
debate, explores the position adopted, and pro- research the positions that they adopt on such
poses a solution and discusses its implications for topics as, in our case, philosophies of science,
the profession. One note on terminology is nec- geographic methodology, or the interplay be-
essary. The term “geographic information sci- tween science and technology. Electronic com-
ence” has appeared with an increasing frequency ments posted to a discussion list are not as
in the geographic literature (Goodchild 1992; carefully thought out as writing in the scientific
Rhind et al. 1991; Rhind 1992; Abler 1993; literature. A written synthesis is thus perhaps
Cromley 1993; Dobson 1993; Fedra 1993). Good- more akin to the proceedings of a workshop, in
child (1992) has argued that GIS has done much which useful ideas are expressed but not yet con-
to remove the traditional isolation between the solidated or put into perspective.
fields of photogrammetry, remote sensing, geod- Another challenge is how best to present the
esy, cartography, surveying, geography, computer discussion; in other words, how to properly cite
science, spatial statistics, and other disciplines communication from an electronic conference.
with interests in the generic issues of spatial data, As on-line newspapers, journals, libraries, and
and it is these disciplines that constitute geo- data archives become more prevalent on the “in-
graphic information science; hence it makes formation superhighway,” and it becomes neces-
sense for the research community to decode the sary to refer to information that may exist only in
GIS acronym in this way. Be that as it may, in this electronic form, formal methods of citation will
paper every reference to the acronym GIS is to have to emerge that are as robust and persistent
“system” not to “science.” as conventional methods. The use of Universal
Resource Locators (URLs, the electronic World
Wide Web addresses that commonly begin with
The GIS-L Debate “http://”) for citations to information available
through the World Wide Web is already causing
Documenting Electronic Discussions problems with “broken URLs,” which occur
whenever information is deleted or moved from
Scholarly interaction is being revolutionized by its existing site, or the name of a server or its file
the Internet applications of electronic mail, dis- structure is changed. Until better methods are
Tool or Science? 349

devised, material posted on electronic discussion the same sense that Curran (1987) defines re-
lists and bulletin board falls into the realm of mote sensing as a technique. From this perspec-
“personal communication,” but unlike oral com- tive, GIS on its own is meaningless; its gains
munication, electronic mail provides a more or meaning only by its goals, which generally involve
less permanent and verbatim record of commu- the application of knowledge by scientists, but not
nication and is precisely quotable. In this paper science itself (McCauley 1993; Moll 1993; Skelly
we follow the electronic citation style proposed 1993b). In the GIS-L discussion, those who de-
by Li and Crane (1993). fined GIS as a tool did so in the sense of a physical
A final issue is confidentiality. For this study of entity and also as a technique (Crepeau 1993a;
the GIS-L discussion, all participants were noti- Feldman 1993b; Halls 1993; Moll 1993). Viewed
fied of our intent to present and synthesize their in this way, GIS may belong more to the field of
comments in a published manuscript, and given engineering than to science (Feldman 1993d;
the option of having quotations and references to Skelly 1993c, 1993d). Discussants identified en-
their comments removed. Before the manuscript gineering as a problem-solving activity, while sci-
was submitted for publication, participants were ence was linked to discovery and problem
sent a draft version for review and comment. It understanding (Al-Taha 1993). That said, the
should be noted that the views expressed by GIS- boundaries between the two are often muddied,
L discussants do not necessarily reflect the views particularly at the level of basic research where
of their institutions or organizations. engineers may use scientific methods to identify
and understand the problems they will eventually
attempt to solve (Al-Taha 1993).
Summary of the GIS-L Discussion Some of the “tool side” of the issue seemed to
feel that if GIS had any scientific aspect, it derived
The GIS-L exchange (summarized in Table 1) from GIS’s place within the discipline of geogra-
began as an offshoot of a discussion about one phy (Crepeau 1993a; Feldman 1993c; Halls
important aspect of the scientific enterprise: the 1993). GIS is thus a tool applied when going
value of peer-reviewed literature within the sci- about the business of geographic science (Halls
entific community, particularly with respect to its 1993). If “doing geography” is a science, then
suitability as reading material for students. With “doing GIS” amounts to a science (on the “geog-
the mention of “GIS as a science” in the context raphy as science” issue, see Couclelis and
of peer-reviewed literature, some correspondents Golledge 1983; Hart 1982; Johnston 1979, 1986;
responded that “GIS is a tool, perhaps even a Smith 1992; and Unwin 1992).
problem-solving environment, but it is most cer- Those on the “science side” of the GIS-L dis-
tainly not a science” (Skelly 1993a). The elec- cussion spoke mainly about the use of GIS as a
tronic debate launched a discussion that unfolded method or body of knowledge for developing and
between October 28 and November 28, 1993 testing spatial theories (Brenner 1993; Laffey
(Figure 1). Table 1 is organized chronologically 1993; Sandhu 1993b; Wright 1993a), not about
because the debate was dynamic and evolved via the physical entity GIS itself. While they agreed
a series of responses to particular statements, that the “toolbox” view of GIS was accurate, it
claims, etc. The evolutionary course of the debate was at the same time very limiting (Bartlett
was also very important in shaping its contents. 1993b; Sandhu 1993a; Wright 1993b). As impor-
Space does not permit a presentation of the entire tant as are the hardware/software components of
discussion, which when downloaded from GIS-L GIS, it is the conceptual elements of GIS (e.g.,
and printed out comprises fifty pages of single- the rules governing the creation of spatial models
spaced text. The table is organized according to for GIS, the measurement and modeling of error
the two major positions taken during the debate propagation through a GIS, or proofs of theorems
(“GIS as tool” vs. “GIS as science”) as well as by on data structures) that enable GIS to claim a
the participants’ general comments about the place as a science (Bartlett 1993a; Carlson 1993a;
scientific enterprise. Wright 1993b).
Many of those who argued on the “tool side” Some discussants raised more fundamental
of the issue could not see how a computer appli- questions: “what exactly is science?” and “what
cation could be described as a science (e.g., specifically allows us to call GIS a science?” (Feld-
McCauley 1993; Moll 1993; Skelly 1993a, man 1993c; Piou 1993). Although no simple
1993b). They saw GIS as a tool or technique in consensus on science emerged, discussants ac-
350 Wright et al.

Table 1. Excerpts from the GIS-L Discussion, October–November, 1993


General Comments
GIS as Tool GIS as Science about Science
GIS is a tool, a technology, a GIS as a science should move Investigating data models and
problem-solving environment, away from the technology to- methods of database design is an
but definitely not a science wards the fundamental aspects of applied area of interest, perhaps
(Skelly 1993a, 28 Oct. 01:55 modeling spatial phenomena applied science (Crepeau 1993b,
PST). (i.e., the conceptual underpin- 29 Oct. 14:04 PST).
nings of CHARACTERIZING
How NOT to generate erroneous spatial phenomena). How NOT Why is computer science called
spatial data is a problem but not to generate erroneous spatial data “science,” not engineering?
a science (Skelly 1993b, 28 Oct. is a science in and of itself What exactly is a science? (Piou
20:12 PST). (Carlson 1993a, 28 Oct. 17:34 1993, 31 Oct. 18:39 PST).
PST).
GIS will become a science when Beware of having too high a re-
it divorces itself from geography. See the research initiatives of the gard for science, especially in
If GIS is a science, statistical soft- NCGIA. Aspects of GIS as a sci- terms of believing that it provides
ware and its use are science. ence include the study of spatial the “truth” (Britton 1993, 1 Nov.
Theoretical knowledge that data uncertainty and error, data 12:29 PST).
forms the design of a model is lineage, and how GIS is adopted
science, not knowledge of how to by agencies (Wright 1993a, 28 Note carefully the implications of
run software. In short, geography Oct. 22:44 PST). your definition of science (Feld-
is the science, not GIS (Crepeau man 1993a, 1 Nov. 13:25 PST;
1993a, 29 Oct. 09:33 PST). The component of GIS that ad- Feldman 1993d, 8 Nov. 06:47
dress the fundamental nature of PST).
GIS in and of itself is a tool. The digital geographic information is
science label is merely to attract a science. Part of the process of The success and acceptance of a
research money and/or enhance design, developing, and improv- given activity in a historic time or
the status/power of GIS users. Im- ing tools is science (Cooper 1993, for some social groups may well
portant to realize WHY people 29 Oct. 9:16 PST). depend on whether that activity
are making the science argument is considered as either scientific,
(Petican 1993, 29 Oct. 12:54 GIS may be a discourse which philosophical, political, or statis-
PST; Groom 1993, 1 Nov. 14:03 would include both the science tical. Different groups in different
PST). and tool views. A discourse is a times emphasize the importance
methodology for studying some- of one or another perspective
Many are understanding spatial thing. The tool view doesn’t give (Scalise 1993, 1 Nov. 18:36
phenomena through the USE of enough credit to geography, the PST).
GIS. GIS is a tool USED by scien- science view takes GIS away from
tists (McCauley 1993, 29 Oct. geography to stand on its own Philosophy of science is an impor-
13:59 PST). (Brown 1993, 29 Oct. 16:22 tant thing to consider for this dis-
PST). cussion. For example, “basic
Perhaps a LITTLE bit of science laws” are part of science only in
goes into GIS as some science Does investigation of the role of the positivist approach to sci-
goes into most technology. EN- scale in GIS and its relationship ence, not in approaches such as
GINEERING DESIGN, not sci- to map area, the importance of realism (Feldman 1993b, 2 Nov.
ence, primarily goes into GIS spatial integrity when all layers of 10:05 PST; Feldman 1993d, 8
(Skelly 1993c, 31 Oct. 14:04 information come from the same Nov. 06:47 PST).
PST). base map, or the appropriateness
of a spatial control for a particular Consideration of the *motivation*
The science of what we are about scale warrant a science? Are an- for labeling something as science is
is in the spatial ordering and in- swers available from another sci- also important for this discussion.
terrelationships of objects and in- ence? (Carlson 1993b, 31 Oct. When we label something as “sci-
formation. The tool we are 10:48 PST). ence” without looking critically at
concerned with is GIS. GIS is the what we mean, we are in fact laying
application of spatial science to The answer depends on who is the groundwork for the arbitrary
the study of earthbound objects, involved. For example, develop- exclusion and delegitimization of
etc. (Halls 1993, 1 Nov. 03:01 ers of GIS see it as a sophisticated certain reasonable forms of scien-
PST). science, salesmen and students tific knowledge (Feldman 1993b, 2
recognize it as tool (Rao 1993, 1 Nov. 10:05 PST, 1993c, 5 Nov.
Nov. 02:22 PST). 12:56 PST).
Tool or Science? 351

General Comments
GIS as Tool GIS as Science about Science
What are the accepted scientific The science of GIS is not the devel- HYPOTHESIS TESTING is nei-
methods of a science of GIS? Soft- opment of tools (i.e., what the ESRI ther necessary nor sufficient for
ware development is engineering, programmers do). GIS theory is sci- science. Science is concerned
not science. Math and statistics ence assuming that the develop- with understanding, which is
are not science either. Both spa- ment of algorithms, proper often an aid to invention, but in-
tial analysis development and its methodology, is science (Brenner vention is not science (Feldman
application through GIS assist 1993, 1 Nov. 07:04 PST). 1993d, 8 Nov. 06:47 PST).
science, but like math they are
not science (Skelly 1993d, 1 Nov. The toolbox view limits GIS. De- Could it be the “authority” or
16:50 PST). vising a representation of spatial “stature” suggested by the term
data, developing an algorithm “science” that causes its adoption
GIS is a tool that describes proc- (method) to solve a spatial problem in naming an emergent area?
esses relating to the manipulation and applying it to test a theory are “Science” lends a sense of valid-
of spatial data; GIS is an entity part of science. GIS embeds the ity, one that may not always be
that describes organizations; GIS knowledge a scientist has about a warranted (Elliott 1993, 19
is a technique employed by a sci- region and provides ways to test Nov.).
entist, not a science unto itself theories and alternatives. Devising
(Moll 1993, 2 Nov. 09:17 PST). new methods to test a theory may Engineering is problem-solving
constitute a PARADIGM shift in while science is discovery and
GIS may be a revolution-driving Kuhnian terms (Sandhu 1993a, 1 problem understanding. At the
tool rather than a scientific revo- Nov. 18:14 PST). professional level, the distinction
lution in and of itself. For exam- is quite clear (e.g., geologist and
pl e , the t e l es cope is the Developing new methods of using civil engineers studying soil me-
revolution-driving tool whereas GIS is part of the larger science of chanics for different reasons). At
optics and metallurgy are the sci- geography, computer science, etc., the research level, the boundaries
ences that extend it (Murphy not a separate science. GIS does are less apparent (e.g., engineers
1993, 2 Nov. 14:27 PST). not represent a PARADIGM shift may use scientific methods to
in Kuhnian terms or a scientific identify and understand the prob-
Is GIS sufficiently distinct as to be revolution comparable to Quanti- lems they will eventually attempt
separate from other sciences? tative Revolution of the ‘60s or the to solve) (Al Taha 1993, 29
Does it have formal properties onset of Humanism and Marxism Nov.).
that are distinct from those inves- in the ‘70s and ‘80s (Feldman
tigated in statistics, computer sci- 1993b, 2 Nov. 10:05 PST). A GIS database is a structured
ence, cartography, etc.? For GIS subset and abstraction of reality.
to be considered a separate sci- GIS is part of a broader spatial in- The rules governing the creation
ence, the problems it addresses formation science. GIS in these of spatial models for GIS were
must be unique to GIS, and the terms is not merely the hard- formulated on the basis of scien-
extension of GIS must be towards ware/software component but the tific thought and experimenta-
a greater understanding of GIS ISSUES surrounding its use (e.g., tion, and now that they are in
problems rather than towards spatial data uncertainty, its meas- place, they form a PARADIGM
solving specific technical issues urement and modeling). The tool- whereby GIS users can seek to
(Feldman 1993c, 5 Nov. 12:56 box view is limiting. GIS analyze the real world. Every time
PST). encompasses the way in which geo- one applies these data modeling
graphical info. is collected, per- rules, one is TESTING THE HY-
PARADIGM shifts occur when ceived, managed, and used (Wright POTHESIS that it fits reality
the limits of a design are reached. 1993b, 2 Nov. 19:05 PST). (Bartlett 1993b, 6 Nov. 10:21
The problem of deriving a new PST).
design is treated as science—the Are people who look for newer
design PRINCIPLES of GIS are ways to analyze data not doing GIS is a science which, like geol-
“scientific,” even though the science? This would involve de- ogy for example, draws from other
process of building one is engi- veloping theory on entities such sciences to create new ways of
neering (Geissman 1993, 6 Nov. as time and spatial phenomena. looking at data and analyzing it.
12:00 PST). For the scientist using it, it is a The spatial aspect draws from
tool, for the producer of GIS, it is math, geomorphology, and other
the engineering of a tool, for the disciplines to get at the business
scientist extending GIS, it is sci- of data (Calef 1993, 28 Nov.).
ence (Sandhu 1993b, 3 Nov.
10:27 PST).
352 Wright et al.

General Comments
GIS as Tool GIS as Science about Science
Complex technology that evolves GIS is an environment as well as
is a field of engineering, not sci- a method used to discover, ex-
ence. Abstractly considering data plore, and test spatial theory. It is
that exists in an n-dimensional also a tool of the hand and mind
space and how such data can be (Laffey 1993, 4 Nov. 16:56 PST).
joined, intersected, etc. in this
n-space would be a scientific A technocentric view of GIS is
problem. Still, this does not make very limiting. As important as the
GIS a science (Feldman 1993d, 8 hardware and software are the
Nov. 06:47 PST). conceptual elements of GIS (data
modeling, visualization, commu-
Unless we can find some scien- nication, legal aspects, etc.)
tific object that is distinct to GIS, These theoretical and conceptual
then GIS is an area of application aspects give GIS at least a foot in
combining elements of computer the realm of the sciences. GIS is
science and geography, and not a a technology, so is the paper map.
science unto itself (Feldman Denying that aspects of GIS are
1993d, 8 Nov. 06:47 PST). science is like denying that as-
pects of cartography are science
(as well as art). The concepts that
the tools seek to facilitate, auto-
mate, and develop are strongly
rooted in science (Bartlett 1993a,
5 Nov. 11:13 PST).

What specifically allows us to call


GIS a science? Maybe we should
think in terms of formal science
(purely abstract thought as in
math and computer science) vs.
substantive science (phenomena
that exist outside of thought).
GIS may have a scientific compo-
nent as a formal science (Feld-
man 1993c, 5 Nov. 12:56 PST).

cented the conceptual elements mentioned These attempts at defining science naturally
above, along with activities such as “obtaining led to the question of whether GIS is significantly
theoretical knowledge to form the design of a distinct from sciences such as computer science
model,” “development theory on entities such as or geography. In other words, if GIS is a science
time and spatial phenomena,” and “development in some respects, is it a science unto itself, with
algorithms to test a theory.” These were thought its own unique, logically coherent object of
to be parts of the scientific enterprise (Crepeau knowledge (Carlson 1993b; Feldman 1993d;
1993a; Sandhu 1993a; Wright 1993b), and hence Skelly 1993d)? Hence Dobson’s (1993) query: “Is
a possible basis for testing the scientific status of GIS prompting a scientific revolution? The most
a given activity. There was also a distinction made severe test would be whether there are hypothe-
between “formal science” (purely abstract ses and theories that can only be conceived and
thought, as in mathematics) and “substantive tested through GIS.” GIS is special in that it is
science” (phenomena that exist outside of uniquely visual and able to make explicit the
thought) (see Feldman 1993c). Accordingly, one implicit features of data. Those on the “science”
must consider the implications of his or her defi- side or in the middle of the GIS-L discussion,
nition of science so as not to arbitrarily exclude however, did not seem to require a separate body
or delegitimize certain reasonable forms of knowl- of knowledge for GIS. Instead, they viewed the
edge (Feldman 1993a, 1993c, 1993d). science of GIS as a subdiscipline of geography or
Tool or Science? 353

computer science (the way biogeography or geo- These so-called -isms are defined variously by
morphology are sciences within the larger field of geographers. Johnston (1986) uses the terms
geography, or paleontology is a science within the “positivism,” “humanism,” and “structuralism” to
broader field of geology) (Bartlett 1993a; Calef describe human geography’s three main scientific
1993; Wright 1993b). Discussants voiced strong approaches; Haines-Young and Petch (1986) ac-
agreement that the connections between GIS and cent “empiricism,” “positivism,” “relativism,” and
the science of geography are the strongest, and that “critical rationalism,” and Cloke et al. (1991)
GIS is not merely a subset of computer science focus on “Marxism,” “humanism,” “structuration
(Bartlett 1993a; Wright 1993b). It was pointed out theory,” “realism,” and “postmodernism.” Thus
that many of GIS’s early pioneers were geographers, to ask the simple question, “Is GIS a science?,” is
e.g., Coppock, Rhind, Bickmore, and Unwin in usually to presume the superiority of one or an-
Britain; Tomlinson, Garrison, Berry, Tobler, and other approach to generating knowledge. For ex-
Marble in North America (Bartlett 1993b); and ample, the GIS-L discussion pointed out that the
that geographers more than anyone else actually concept of “basic laws” is part of science only in
identified, conceptualized, and formalized the the positivist approach, not necessarily in ap-
initial connections between spatial concepts and proaches such as realism or humanism (Feldman
computer technology (Bartlett 1993b). 1993b, 1993d). Many argue that the positivist
approach is privileged with regard to GIS (Hey-
wood 1990; Taylor 1990; Smith 1992; Lake 1993;
Putting the Debate into Shepherd 1993), but Goodchild (1994) sees in
Perspective: Definitions of Science the growing literature on the epistemology of GIS
(e.g., Pickles 1991; Wellar et al. 1994) the entire
A lengthy foray into the philosophy and soci- spectrum of approaches, from the positivist to the
ology of science is beyond the scope of this paper, postmodernist. If in the past certain approaches
but some considerations of these matters is un- to science have been preferred in GIS, GIS need
avoidable in order to know what scientists do, the not preclude other approaches in the future.
significance of what they do, and the relationship If aspects of GIS are to be considered as “sci-
of science to other knowledge-generating mecha- ence,” according to what philosophical approach
nisms. There is one caveat at the outset: there are are they scientific? This issue was raised briefly in
probably as many definitions and viewpoints of the GIS-L discussion (see Table 1; Feldman
science as there are scientists (Feibleman 1972), 1993b, 1993c) but was not examined in detail. In
and not all of these are necessarily correct! A the longstanding debates that have occurred in
concise definition of science cannot hope to cap- geography over the appropriateness of different
ture the full meaning of the term. Science encom- approaches, “positivism” and the “critical ration-
passes a wide range of fields that differ widely from alism” of Karl Popper (1959) are conventionally
each other in philosophy, knowledge content, and associated with “science” (Haines-Young and
methodology. The term “science” may be viewed as Petch 1986; Johnston 1986), but the rigorous
shorthand for a logical and systematic approach to collection and evaluation of data in the produc-
problems that seeks generalizable answers. This is tion of knowledge are not exclusive to positivism
the position by Robinson et al. (1984) in describing or critical rationalism (e.g., Keat and Urry 1975;
how cartography employs “the scientific method” in Johnston 1986; Sayer 1992). It is not our wish to
constructing its products. Given Robinson et al.’s downplay the explanatory power of these alterna-
emphasis on logic, most computer applications tive, nonpositivistic approaches; indeed various
would pass the test of being “scientific,” though it philosophies of science have succeeded in under-
leaves unanswered the question of whether “doing mining positivism’s claims to being a superior
cartography” is “doing science.” Nonetheless, many method for understanding the world (e.g., Willer
participants in the GIS-L debate were probably and Willer 1973; Hindess 1977; Couclelis and
unaware of the finer shades of meaning conveyed Golledge 1983; Sheppard 1993).
by the term “science,” or that many users of GIS
might think of themselves as “scientists” in the
unqualified sense of that term. Why Does Science Matter?
Depending on one’s inclination, there are sev-
eral different approaches to science, each with its Why should one care whether GIS is a sci-
own ontology, epistemology, and methodology. ence or not? The technological (toolbox) face
354 Wright et al.

of GIS is widely successful in government, busi- answers on other matters have no bearing on how,
ness, and education, and it appears to have af- and how well, we are able to think about human
fected and improved the lives of far more people purpose, free will, and other such things.
than have many theoretical advances (e.g., the Clearly it does matter whether or not “doing
theories of spatial data and of data structures, GIS” is “doing science,” if for no other reason
datamodels,andalgorithms).Technologyingen- than that “doing science” is often regarded as a
eral has the potential to contribute greatly to code-phrase for academic legitimacy. We will now
society and culture. argue that “doing GIS” may express at least three
Science, though, is often held in high regard, meanings that are represented by three positions.
and labeling a field as a science may sometimes Our strategy, therefore, is to examine the role and
help to ensure it a place in the academy or to legitimacy of each of these three positions within
secure it greater funding and prestige. “Science” the academy in general and within the discipline
is often used as a generic synonym for “research,” of geography in particular.
particularly research of a basic, systematic, and
generalizable kind. Thus “science” often func-
tions as a rather crude but convenient shorthand Three Positions on GIS
for academic legitimacy; if “doing GIS” is “doing
science” then its claim to a place in the academy, In synthesizing the general themes of the GIS-
as a topic of research and graduate-level instruc- L discussion, it became clear to us that GIS could
tion, is clearly strengthened. be understood not by the two distinct positions
In the GIS-L discussion, some participants im- taken by the GIS-L discussants but as three posi-
plied (Groom 1993; Petican 1993) that those tions along a continuum from tool to science.
arguing for “GIS as science” might be driven by These positions focus on the several meanings
ulterior motives. Some participants warned of attached to “doing GIS” rather than to GIS alone.
having too high a regard for science, especially in These are (1) GIS as tool; (2) GIS as toolmaking;
believing that it provides the “truth” (Table 1). In and (3) the science of GIS. It seems clear from
the opinion of these correspondents, science does the GIS-L discussion that the label “GIS” is sim-
indeed greatly influence our everyday lives and plistic, since it fails to indicate by itself whether
our ideas about the world, but does it deserve the research involves fundamental scientific
special reverence? Is there something special questions and hypotheses, or whether GIS merely
about science and the contributions it has made? adds gloss to the research through the use of a
One must strike a balance between debunking complex and sophisticated tool—whether it de-
science and dragging it off its pedestal, on the one codes as “science” or “system.” We have derived
hand, and falling into scientism (the claim that three positions on GIS from the GIS-L discussion;
the scientific method is the only true method of these represent a “fuzzy” continuum of opinion
obtaining knowledge), on the other. The point which recognize that positional labels are not
has already been made eloquently by Bauer perfect. Although these three positions do not
(1992:144): capture all of the nuances of argument during the
GIS-L debate, they do represent the principal
That science does not have all the answers does not
mean that it has no answers. That science now has
points of view along a “tool-science” continuum.
inadequate answers in some areas does not mean The “GIS is a tool” position sees GIS as the use
that the answers will not become adequate in the of a particular class of software, the associated
future; in fact, history teaches that science’s answers hardware tools such as digitizers and plotters, and
become better and better as time goes by. That digital geographic data in order to advance some
science is fallible does not mean that science is specific purpose. The tool itself is inherently neu-
entirely fallible or that it is as fallible as such other tral, its development and availability being largely
modes of human knowledge and belief as folklore, independent of its use, which is driven by appli-
religion, political ideology, or social science. That cation.
science has no answers in some matters—such as The “toolmaking” position sees GIS as con-
the value of human life or the purpose of liv-
ing—does not mean that it has no answers in other
cerned with advancing the tool’s capabilities and
areas—those areas that are within its purviews, mat- ease of use. Besides using the tool, toolmakers
ters of forces and substances and natural phenom- regularly promote the adoption of GIS, play a role
ena. And that science has no direct answers on in educating users, and work to ensure responsi-
matters of human purpose does not mean that its ble use.
Tool or Science? 355

Finally, the “science of GIS” position insists on Scientists use many types of tools in their re-
a more intimate and reciprocal connection be- search. Some, such as typewriter or telephone, are
tween tool and science—one that involves re- generic in nature, with no particular association
search on a set of basic problems, each of which with any discipline. Others are developed strictly
probably existed prior to the development of GIS, for one discipline, or even for one project or for
but whose solution is more pressing now because one group of scientists. GIS falls somewhere in
of the technology. This practice of collecting sets the middle, being of interest, in principle, to any
of basic problems under new names has a long discipline dealing with the distribution of phe-
history in science. It occurred, for example, with nomena on the surface of the Earth. It seems
the emergence of computer science, when the neither a generic tool whose use is so ubiquitous
development of computing technology provided that one can reasonably assume universal famili-
the impetus for solving certain fundamental re- arity (word processor, calculator), nor the exclu-
search problems that had previously been associ- sive tool of a single discipline. Perhaps a useful
ated with mathematics. analogy is the tool of statistics, which in some
disciplines (e.g., agronomy) is close to universal,
while in others (e.g., anthropology) usage is mixed
Discussion of the Three Positions and its value is the subject of continuous debate.
on GIS For these less-than-universal tools, the acad-
emy traditionally provides the necessary infra-
structure in the form of technical courses and
GIS Is a Tool technical support. But in addition, the academy
satisfies the need for education in the associated
For those who take this position, “doing GIS” concepts. In the case of statistics, for example, it
amounts to making use of a tool to advance the would not be adequate to provide a laboratory of
investigation of a problem. If the investigation statistical tools without at the same time provid-
merits the label “research,” then “doing GIS” is ing courses to ensure that students have the
probably “doing science” as well, but the exist- necessary understanding of concepts. The same
ence and use of the tool are separable from the distinction between technical training in the use
substantive problem. The documentation and of tools and education in the underlying concepts
write-up of the research tend to focus on the applies to GIS. While the concepts of GIS may be
substantive problem, and indeed the tool may not familiar to professional geographers, they must be
be mentioned at all. In some cases, GIS may be taught anew to each generation of students.
only one of a number of tools used, each of which Without conceptual courses, the use of GIS is
has been selected strictly for its efficacy in the likely to degenerate to data management and map
research project. In these cases, the tools do not making, however complex the tool’s capabilities
drive the research. for scientific analysis and modeling.
If the research objectives are to some degree If GIS is a tool of particular value to geography,
“methodological,” then the rules of engagement and if geography has traditionally taught many of
with tools such as GIS may be somewhat different, the concepts that the tool implements, then it
as may the content of the paper documenting the would seem that formal courses in GIS are most
research. In these cases, the tool may assume a appropriately taught in geography. In the absence
greater role in directing the research, and hence of departments of geography, universities have
be given greater prominence in documentation, found a variety of solutions to the need for GIS
and case studies may be used to illustrate the instruction. In some cases, courses are taught by
technique rather than to provide generalizable faculty or staff in computing facilities; in others,
empirical results. In papers of this sort, phrases they are taught in departments such as surveying,
such as “The Use of GIS in . . .” may appear in civil engineering, or forestry (Morgan and Fleury
paper’s title, although the processes responsible 1993). But wherever they are taught, these
for the tool’s development are independent of the courses serve two purposes—they prepare stu-
substantive research problem. Because their pri- dents to do their own research, and they provide
mary motivation is to advocate the use of the tool, students with useful job skills.
methodological demonstrations of GIS are more While this technical line of argument provides
appropriately included under the second position a solution that is satisfactory for students, it cre-
discussed in the next section. ates problems for the faculty assigned to teach
356 Wright et al.

them, particularly untenured faculty. Technical At the research level, the view of the tool-
courses require very heavy commitments of fac- maker presumes critical analysis and reflection.
ulty time, and teaching them is unlikely to boost The result is an extensive research literature that
the instructor’s scholarly research career. The focuses on GIS as a generic tool (Goodchild et al.
time required to maintain the GIS teaching labo- 1991); at the same time there is remarkably little
ratory and to deal with students’ technical prob- published research on specific systems, perhaps
lems cuts down on the time available for because of the propriety nature of most GISs and
research—and for securing tenure. Some depart- because of scholarly fears that principles of aca-
ments deal with this issue by relying on sessional demic freedom and the First Amendment would
teaching staff or on technical staff, much as they not protect them against a suit brought by a
did in the past when offering courses in technical private-sector GIS vendor against publication of
fields such as cartography or remote sensing. a critical academic evaluation of a product.
Critical analysis and reflection extends beyond
the techniques of toolmaking to encompass ques-
GIS Is Toolmaking tions about the social responsibilities of toolmak-
ers and the social implications of the widespread
Advancing along the continuum between adoption of the tool (Smith 1992; Pickles 1991,
“GIS as tool” and “GIS as science,” we reach the 1994; Harvey and Chrisman forthcoming). This
middle position on GIS, that of toolmaking. For involves evaluating a tool that has applications
toolmakers, the tool is inseparable from the sub- spanning the full range of human activi-
stantive problem, i.e., “doing GIS” implies in- ties—from economy to politics to society. In this
volvement in the development of the tool itself. case, the issues become complex indeed. The
Geographers who are makers of the GIS tool scope of research is determined not by the tool’s
participate directly in its specification, develop- value to geographers, but rather by the multifari-
ment, and evaluation, as well as in its use. ous applications of GIS, to include all of the
In reality, the developers of GIS tools have societal effects of the computerization of geo-
backgrounds in many disciplines, including com- graphic information. Whether “GIS” can with-
puter science, engineering, design, and mathe- stand the stress of such varied usages remains to
matics, as well as geography. Few geographers be seen.
have the necessary technical skills to build major
software systems or to write “industrial-strength
code”; but for that matter, academics in general The Science of GIS
are not regarded as suited to the development of
reliable software. Indeed, most current GISs origi- It was evident from the GIS-L debate that GIS
nated in the private sector in companies employ- is widely viewed outside the discipline of geogra-
ing a mix of disciplines (GRASS and Idrisi are phy as a subset of geographical science. Although
notable exceptions). geography is a small, unevenly represented disci-
Geographers possess two unique and powerful pline, and doubts about its legitimacy in the acad-
abilities as GIS toolmakers. The first of these is an emy are widely held (Smith 1987), the recent
excellent understanding of the geographic con- growth of GIS and its affiliation with the disci-
cepts that form the primitive elements of GIS pline has meant increased visibility in the acad-
databases and processing and the ways that these emy. Moreover GIS is associated with clear
concepts are embedded in theories, methods of physical imagery, hence it is much easier to imag-
analysis, and models. Second is that geographers ine “doing GIS” than “doing geography” if one has
are trained in a discipline that integrates under- no familiarity with the latter. Geography’s affili-
standing of a wide range of processes influencing ation with GIS thus pairs it with the computer
phenomena on the earth’s surface. Both of these (however inappropriately). Computerization
abilities are essential to “doing GIS” if one adopts automatically confers precision, rigor, and repli-
the position that “doing GIS” is toolmaking. A cability in the popular imagination, all of which
GIS toolmaker thus requires a basic education in contribute to the flawed notion of GIS as a subset
geography together with technical courses that of geographical science.
emphasize critical analysis of the technology’s The very rapid growth of technology and the
capabilities. emergence of a technology-based society in re-
Tool or Science? 357

cent years have prompted new groupings and reenergized by the development of GIS. Although
priorities within science. Few would have pre- the capabilities of GIS are improving, geographers
dicted, for example, that the development of the who use it still look forward to the stage at which
digital computer would eventually lead to the all geographic concepts and procedures are im-
discipline of computer science, or that informa- plemented digitally (Dobson 1983, 1993; Cou-
tion would itself become the basis of a scientific clelis 1991). In the interim, GIS research will
discipline. Four conditions seem necessary for the most likely implement those concepts and proce-
emergence of a science out of a technology: first, dures that are the simplest, most logical, and most
the driving technology must be of sufficient sig- rigorously defined, i.e., the most primitive and/or
nificance; second, the issues raised by its devel- the most scientific. These include issues of recog-
opment and use must be sufficiently challenging; nition and measurement in the field, the choice
third, interest in and support for research on between alternative representations, the roles of
those issues must be inadequate in the existing generalization and multiple representations, the
disciplines; and fourth, there must be sufficient representation of uncertain information, meth-
commonality among the issues to create a sub- ods of analysis and modeling, problems of describ-
stantial synergy. ing the content of geographic data and evaluating
Two terms have evolved to describe the emer- its fitness for use, and methods of visualization.
gence of a science based on GIS. The first is These sorts of issues underscore the multidiscipli-
geomatics, a term favored in many countries be- nary nature of geographic information science.
cause of its simplicity and its ease of translation Besides geography, it includes such traditional
into French; the second, geographic information geographic information disciplines as geodesy,
science, a term that is well-known in the English- surveying, cartography, photogrammetry, and re-
speaking world. The latter is used here. mote sensing along with spatially oriented ele-
Geographic information science, the science of ments of such other disciplines as information
GIS, is concerned with geographic concepts, the statistics, cognitive science, information science,
primitive elements used to describe, analyze, library science, and computer science.
model, reason about, and make decisions on phe-
nomena distributed on the surface of the earth.
These range from the geometric primitives of Evaluation
points, lines, and areas to the topological relation-
ships of adjacency and connectivity through the In light of these three perspectives on GIS,
dynamic relations of flow and interaction to do- what can be said about the significance of “doing
main-specific concepts such as neighborhood, GIS?” If “GIS is a tool,” then its use has little to
geosyncline, or place. In their current state of do with the legitimacy of the research; in this case,
developments, GISs are comparatively crude significance derives strictly from the progresses
digital systems for representing and manipulating made on the substantive research problem. In this
geographic concepts, capable of handling only sense “doing GIS” is not necessarily the same as
the most primitive of these concepts. But while “doing science”; the latter depends on the meth-
current technology may constrain the science of ods deployed on the substantive problem, i.e., are
GIS, it does not limit its development, just as they scientific? Courses in GIS are more likely to
computer science is not limited by the current be offered at the undergraduate level and reflect
state of computer technology. Indeed the re- their essentially technical, service orientation. A
search problems raised by GIS and their solutions geography department using GIS on this basis
will help to define the future form of GIS tech- probably would not claim GIS as a research
nology. Perhaps the most crucial of these prob- specialty, nor would it encourage its students
lems for geographic information science is the to regard GIS as a substantive subfield of the
limitation of digital representation, i.e., are there discipline.
geographic concepts which can never be repre- The toolmaking position confers a more sig-
sented in or manipulated by GIS? nificant status on GIS. In this case, GIS includes
The digital representation and manipulation of case studies that demonstrate the methodology,
geographic concepts raise a number of fundamen- advocacy of GIS usage, and, perhaps also, the
tal research issues, many of which, though long- development of software. In the absence of indis-
standing in traditional disciplines, have been putable instances of scientific insights uniquely
358 Wright et al.

attributable to the use of GIS, toolmaking will Whether GIS is a geographical science in and of
remain more akin to engineering than to science. itself depends on both the rigor with which the
Consequently, tests of toolmaking’s progress tool is employed and the scope of the tool’s func-
would be based on indicators of improvements in tionality given the nature of the substantive prob-
the tool’s utility. Critical reflection on and the lem. These issues clearly must be resolved on a
evaluation of GIS are also included in the tool- case-by-case basis. Therefore the use of GIS is not
making position. While these are clearly legiti- a sufficient condition for science.
mate activities of the academy, they are not as
easily characterized as “doing science” (or “doing
engineering”). Conclusion
A department adopting the toolmaking posi-
tion would probably offer a range of undergradu- Goodchild (1993:445) notes that “an encour-
ate and graduate courses in GIS, including aging recent trend has been the willingness of a
courses in the toolmaker’s tools—programming broad spectrum of geographers to see GIS not as
languages—and the faculty would regard GIS as a tool that they can use in their own research, but
a research specialty and encourage students to as a phenomenon on which they can reflect and
make significant contributions as toolmakers. But comment.” We have chosen to reflect and com-
such a department might also expect continuing ment on “the phenomenon” of GIS because, as
tension between research and teaching in GIS noted earlier, some of the interest in the GIS-L
and in more substantive fields (i.e., where re- debate may have stemmed from the uncertainty
search is measured by the accumulation of knowl- faced by young scholars in particular over
edge rather than by the improvement of tools). whether departments will accept “GIS” as a topic
The third position, “science of GIS,” is con- for scientific research. In these situations, because
cerned with the analysis of the fundamental issues the label “GIS” is not altogether perfect, what is
raised by the use of GIS in geography or in other probably needed to fully describe the entity “GIS”
disciplines. As noted earlier, these issues may not is a shift from “black-and-white” boxes of descrip-
be unique to GIS, but rather are remotivated by tion to “fuzzier” continua. We find that GIS rep-
it; many of these issues continue to be regarded resents just such a continuum between tool and
as problems in cartography or surveying or spatial science. The technologies of GIS (i.e., the tools
cognition. A department taking this position with and the toolmaking) clearly have the potential to
regard to specialization in GIS would recognize it motivate a host of interesting and fundamental
as a substantive research field on a par with other scientific research questions. The science based
such fields and would measure progress based on on GIS (i.e., geographic information science) may
the accumulation of research results and contri- advance the tools and toolmaking of GIS, as well
butions to human understanding, rather than as scientific research done with aid of using a GIS.
from improvements in the tools themselves. This Surely the desired end from all perspectives is the
position is therefore the only one that provides building of an intellectual foundation for GIS by
sufficient grounds on which “doing GIS” is “doing geographers and members of allied disciplines
science,” and for the legitimacy of GIS as a re- alike, which will ensure its survival long after the
search field in the academy. novelty of the technology has worn off.
Proponents of this position, however, must be Debates arising out of the ambiguity of GIS as
careful not to confuse the use of GIS itself (e.g., tool or science must be understood within the
entering a sequence of spatial analysis com- context of broader trends in science and in society
mands) with an analysis of the issues surrounding generally. Older notions of science as the equiva-
the use. Some may try to derive legitimacy from lent of “hard science” are being replaced by a
the proposition that GIS is so uniquely funda- more open view. Warning against conflating sci-
mental to geography that to do GIS is necessarily ence and its positivistic expression, Johnston
to do science—or, more extremely, that to do GIS (1986:6) proposes a more generous view of sci-
is to do geography scientifically. This argument is ence as “the pursuit of systematic and formulated
somewhat flawed because it implies that GIS is knowledge, and as such [it] is not confined to any
vastly more effective than it currently is, and particular epistemology.” In this context, GIS
because it ignores the limitations of current GIS may represent a new kind of science, one that
in dealing with time, scale, interactions, and a emphasizes visual expression, collaboration, ex-
host of other sophisticated geographic concepts. ploration, and intuition, and the uniqueness of
Tool or Science? 359

place over more traditional concerns for mathe- clelis and Alan Brenner significantly improved the
matical rigor, hypothesis testing, and generality manuscript. The National Center for Geographical
(Goodchild 1992; Kemp et al. 1992; Rhind 1993; Information and Analysis is supported by the National
Fedra 1993; Muller 1993; Burrough and Frank Science Foundation under Cooperative Agreement
SBR 88-10917.
1995).
As a discipline, geography has long struggled
with the tension between the general and the References
particular (Bunge 1962). Maps and geographic
data capture the essence of the geographically Printed References
particular, the boundary conditions that in-
fluence the outcome of physical and social Abler, R. F. 1993. Everything in Its Place: GPS, GIS,
processes; in that sense, GIS illuminates the par- and Geography in the 1990s. The Professional Ge-
ticular. But unlike maps, the purpose of GIS is to ographer 45:131–39.
maintain geographic data in a state(s) that may Bauer, H. H. 1992. Scientific Literacy and the Myth of the
be transformed, processed, and analyzed in ways Scientific Method. Urbana: University of Illinois
that are geographically uniform. Thus GIS is a Press.
technology of both the general and the particular, Bunge, W. 1962. Theoretical Geography. Lund, Sweden:
implementing the former in its formalized algo- C. W. K. Gleerup.
Burrough, P. A., and Frank, A. U. 1995. Concepts and
rithms, concepts, and models, and the latter in Paradigms in Spatial Information: Are Current
the contents of its data sets. GIS as a technology Geographical Information Systems Truly Generic?
seems uniquely appropriate for geographic re- International Journal of Geographical Information
search and, more specifically, for transforming Systems 9:101–16.
geographic knowledge of processes into predic- Cloke, P.; Philo, C.; and Sadler, D. 1991. Approaching
tions, policies, and decisions. In this sense GIS Human Geography: An Introduction to Contempo-
captures geography’s tensions between basic re- rary Theoretical Debates. New York: Guilford Press.
search and application, and between the geo- Coppock, J. T., and Rhind, D. W. 1991. The History of
graphically general and the geographically GIS. In Geographical Information Systems: Princi-
particular. ples and Applications, vol. 1, ed. D. J. Maguire, M.
F. Goodchild, and D. W. Rhind, pp. 21–43. New
The demands for basic and applied knowledge York: John Wiley and Sons.
are several in the new worlds created and encoun- Couclelis, H. 1991. Requirements for Planning-Rele-
tered by GIS. Whether GIS serves as a techno- vant GIS: A Spatial Perspective. Papers in Regional
logical means to acquire and develop knowledge Science 70:9–19.
or as an end for scientific inquiry in its own right, ———, and Golledge, R. 1983. Analytic Research,
these systems will undoubtedly play a central role Positivism, and Behavioral Geography. Annals of
in knowledge making in the future. But it is the Association of American Geographers
important to understand what is meant by “doing 73:331–39.
science,” as well as what is meant by “doing GIS.” Crane, D. 1972. Invisible Colleges. Chicago: University
This paper has identified three well-defined posi- of Chicago Press.
Cromley, R. G. 1993. Automated Geography: Ten Years
tions on this matter, only one of which confers the Later. The Professional Geographer 45:442–43.
kind of academic legitimacy associated with “do- Curran, P. J. 1987. Remote Sensing Methodologies and
ing science.” In other cases, “doing GIS” is more Geography. International Journal of Remote Sensing
akin to using a tool (to be evaluated by the 8:1255–75.
appropriateness of the tool to the substantive Davies, J. 1995. Perspective: Chart Hits? Times Higher
problem) or to engineering better tools (to be Education Supplement, October 27, p. 19.
evaluated on the degree of improvement in the Dobson, J. E. 1983. Automated Geography. The Profes-
tool). In such cases, GIS appears not to constrain sional Geographer 35:135–43.
its users to any particular epistemological stance. ———. 1993. The Geographic Revolution: A Retro-
spective on the Age of Automated Geography.
The Professional Geographer 45:431–39.
Fedra, K. 1993. GIS and Environmental Modelling. In
Acknowledgments Environmental Modeling with GIS, ed. M. F. Good-
child, B. O. Parks, and L. T. Steyaert, pp. 35–50.
The authors wish to thank Pete Peterson, Kristin New York: Oxford University Press.
Lovelace, Steve Behnke, and Ray Smith at UCSB for Feibleman, J. K. 1972. Scientific Method. The Hague:
fruitful discussions. The critical reviews of Helen Cou- Martinus Nijhoff.
360 Wright et al.

Goodchild, M. F. 1992. Geographical Information Sci- Morgan, J. M., III, and Fleury, B. B. 1993. Academic
ence. International Journal of Geographical Infor- GIS Education: Assessing the State of the Art.
mation Systems 6:31–45. Geo Info Systems 3:33–40.
———. 1993. Ten Years Ahead: Dobson’s Automated Morrison, J. L. 1991. The Organizational Home for GIS
Geography in 1993. The Professional Geographer in the Scientific Professional Community. In Geo-
45:444–45. graphical Information Systems: Principles and Appli-
———. 1994. GIS and Geographic Research. In cations, vol. 1, ed. D. J. Maguire, M. F. Goodchild,
Ground Truth: The Social Implications of Geographic and D. W. Rhind, pp. 91–100. New York: John
Information Systems, ed. J. Pickles, pp. 31–50. New Wiley and Sons.
York: Guilford. Muller, J.-C. 1993. Latest Developments in GIS/LIS.
———; Rhind, D. W.; and Maguire, D. J. 1991. International Journal of Geographical Information
Introduction (Section II: Principles). In Geo- Systems 7:293–303.
graphical Information Systems: Principles and Ap- Odlyzko, A. 1995. Tragic Loss or Good Riddance? The
plications, vol. 1, ed. D. J. Maguire, M. F. Impending Demise of Traditional Scholarly Jour-
Goodchild, and D. W. Rhind, pp. 111–17. New nals. International Journal of Human-Computer
York: John Wiley and Sons. Studies 42:71–122.
Haines-Young, R. H., and Petch, J. R. 1986. Physical Openshaw, S. 1991. A View on the GIS Crisis in
Geography: Its Nature and Methods. London: Paul Geography, or Using GIS to put Humpty-Dumpty
Chapman Publishing Ltd. Back Together Again. Environment and Planning A
Hart, J. F. 1982. The Highest Form of the Geographer’s 23:621–28.
Art. Annals of the Association of American Geogra- Pickles, J. 1991. Geography, GIS, and the Surveillant
phers 72:1–29. Society. Papers and Proceedings of the Applied Ge-
Harvey, F., and Chrisman, N. J. Forthcoming. Spatial ography Conference 14:80–91.
Technology in the Making: GIS as Social Practice. ———, ed. 1994. Ground Truth. New York: Guilford
Progress in Human Geography. Press.
Hess, B.; Sproull, L.; Kiesler, S.; and Walsh, J. 1993. Popper, K. 1959. The Logic of Scientific Discovery. New
Returns to Science: Computer Networks in York: Basic Books.
Oceanography. Communications of the Association Rhind, D. 1992. The Next Generation of Geographical
for Computing Machinery 36:90–101. Information Systems and the Context in Which
Heywood, I. 1990. Geographic Information Systems in They Will Operate. Computers, Environment and
the Social Sciences. Environment and Planning A Urban Systems 16:261–68.
22:849–54. ———. 1993. Maps, Information and Geography: A
Hindess, B. 1977. Philosophy and Methodology in the New Relationship. Geography 78:150–59.
Social Sciences. Hassocks, NJ: Harvester Press. Rhind, D. W.; Goodchild, M. F.; and Maguire, D.
Johnston, R. J. 1979. Geography and Geographers: An- J. 1991. Epilogue. In Geographical Information
glo-American Human Geography since 1945. New Systems: Principles and Applications, vol. 2, ed.
York: Wiley. D. J. Maguire, M. F. Goodchild, and D. W.
———. 1986. Philosophy and Human Geography: An Rhind, pp. 313–27. New York: John Wiley
Introduction to Contemporary Approaches. London: and Sons.
Edward Arnold. Robinson, A. H.; Sale, R. D.; Morrison, J. L.; and
Jordan, T. 1988. The Intellectual Core: President’s Muehrcke, P. C. 1984. Elements of Cartography,
Column. AAG Newsletter 23:1. 5th ed. New York: John Wiley & Sons.
Keat, R., and Urry, J. 1975. Social Theory as Science. Sayer, A. 1992. Method in Social Science. London: Rout-
London: Routledge & Paul. ledge.
Kemp, K. K.; Goodchild, M. F.; and Dodson, R. F. 1992. Sheppard, E. 1993. Automated Geography: What Kind
Teaching GIS in Geography. The Professional Ge- of Geography for What Kind of Society? The
ographer 44:181–90. Professional Geographer 45:457–60.
Kennedy, M. 1994. Review of Geographical Information Smith, N. 1987. Academic War over the Field of Ge-
Systems: Principles and Applications, ed. David J. ography. The Elimination of Geography at Har-
Maguire, Michael F. Goodchild, and David W. vard, 1947–51. Annals of the Association of
Rhind. Annals of the Association of American Ge- American Geographers 77:155–72.
ographers 84:172–73. ———. 1992. History and Philosophy of Geography:
Lake, R. W. 1993. Planning and Applied Geography: Real Wars, Theory Wars. Progress in Human Geog-
Positivism, Ethics, and Geographic Information raphy 16:257–71.
Systems. Progress in Human Geography 17:404–13. Sui, D. Z. 1994. GIS and Urban Studies: Positivism,
Li, X., and Crane, N. B. 1993. Electronic Style: A Guide to Post-positivism, and Beyond. Urban Geography
Citing Electronic Information. Westport, CT: Meckler. 15:258–78.
Mark, D. M., and Zubrow, E. 1993. Join the GIS-L Taylor, P. J. 1990. GIS. Political Geography Quarterly
Electronic Community! GIS World 6:56–57. 9:211–12.
Tool or Science? 361

Thoen, B. 1996. On-line Demographics: GIS Commu- Science [Discussion]. Geographic Information Sys-
nities Flourish. GIS World 9:48–51. tems Discussion List [Online].
Unwin, T. 1992. The Place of Geography. Essex, Eng- ———. October 29, 1993b. Re: GIS as a Science
land: Longman Scientific & Technical. [Discussion]. Geographic Information Systems Dis-
Wellar, B.; Cameron, N.; and Sawada, M. 1994. Pro- cussion List [Online].
gress in Building Linkages Between GIS and Elliott, M. A. November 19, 1993. Department of
Methods and Techniques of Scientific Inquiry. Geography, Northern Illinois University. Re:
Computers, Environment, and Urban Systems Why “Science”? [Discussion]. Geographic Infor-
18:67–80. mation Systems Discussion List [Online].
Willer, D., and Willer, J. 1973. Systematic Empiricism: Feldman, M. November 1, 1993a. Community Plan-
Critique of a Pseudoscience. Englewood Cliffs, NJ: ning, University of Rhode Island. Re: GIS as a
Prentice-Hall. Science and Value of Peer Review [Discussion].
Geographic Information Systems Discussion List
[Online].
———. November 2, 1993b. Re: GIS as a Science
Electronic References (All references [Discussion]. Geographic Information Systems Dis-
available e-mail: GIS-L @ cussion List [Online].
UBVM.CC.BUFFALO.EDU.) ———. November 5, 1993c. Re: GIS as a Science
[Discussion]. Geographic Information Systems Dis-
Al-Taha, K. November 29, 1993. Department of Geog- cussion List [Online].
raphy and Anthropology, Louisiana State Univer- ———. November 8, 1993d. Re: GIS as a Science
sity. Re: GIS as a Science [Discussion]. Geographic [Discussion]. Geographic Information Systems Dis-
Information Systems Discussion List [Online]. cussion List [Online].
Bartlett, D. November 5, 1993. Geography Depart- Geissman, J. November 6, 1993. Re: GIS as a Science
ment, Cork University, Ireland. Re: GIS as a [Discussion]. Geographic Information Systems Dis-
Science [Discussion]. Geographic Information Sys- cussion List [Online].
tem Discussion List [Online]. Groom, A. November 1, 1993. Christchurch City
———. November 6, 1993b. Re: GIS as a Science Council, New Zealand. Re: Tool or Science? [Dis-
[Discussion]. Geographic Information Systems Dis- cussion]. Geographic Information Systems Discus-
cussion List [Online]. sion List [Online].
Brenner, A. November 1, 1993. Environmental Protec- Halls, P. J. November 1, 1993. Simon Fraser University,
tion Agency. Re: GIS as a Science [Discussion]. Canada. Re: GIS as a Science and Value of Peer
Geographic Information Systems Discussion List Review [Discussion]. Geographic Information Sys-
[Online]. tems Discussion List [Online].
Britton, J. November 1, 1993. GIS and Cartography Laffey, S. C. November 4, 1993. Department of Geog-
Department, Sir Sandford Fleming College, Can- raphy, Northern Illinois University. Re: GIS as a
ada. Re: GIS as a Science [Discussion]. Geo- Science [Discussion]. Geographic Information Sys-
graphic Information Systems Discussion List tems Discussion List [Online].
[Online]. McCauley, D. October 29, 1993. Agricultural Engi-
Brown, M. October 29, 1993. Department of Geogra- neering Department, Purdue University. Re: GIS
phy, San Diego State. Tool or Science? [Discus- as a Science [Discussion]. Geographic Information
sion]. Geographic Information Systems Discussion Systems Discussion List [Online].
List [Online].
Moll, B. W. November 2, 1993. Martin Marietta Energy
Calef, F. J. November 28, 1993. Re: GIS as a Science
Systems. Re: GIS as a Science [Discussion]. Geo-
[Discussion]. Geographic Information Systems Dis-
graphic Information Systems Discussion List [On-
cussion List [Online].
line].
Carlson, C. L. October 28, 1993a. Northern Illinois
University. Re: Value of Peer Review [Discus- Murphy, L. November 2, 1993. Indiana University. Re:
sion]. Geographic Information Systems Discussion GIS as a Science [Discussion]. Geographic Infor-
List [Online]. mation Systems Discussion List [Online].
———. October 31, 1993b. Re: GIS as a Science Petican, D. J. October 29, 1993. University of Waterloo,
[Discussion]. Geographic Information Systems Dis- Canada. Re: GIS as a Science [Discussion]. Geo-
cussion List [Online]. graphic Information Systems Discussion List [Online].
Cooper, A. October 29, 1993. CSIR Information Serv- Piou, J. October 31, 1993. University of Maryland. Re:
ices, Pretoria, South Africa. Re: Value of Peer Value of Peer Review [Discussion]. Geographic
Review [Discussion]. Geographic Information Sys- Information Systems Discussion List [Online].
tems Discussion List [Online]. Rao, L. November 1, 1993. Re: GIS as a Science
Crepeau, R. October 28, 1993a. School of Social Ecol- [Discussion]. Geographic Information Systems Dis-
ogy, University of California-Irvine. Re: GIS as a cussion List [Online].
362 Wright et al.

Sandhu, J. November 1, 1993a. Environmental Sys- ———. October 31, 1993c. Re: Value of Peer Review
tems Research Institute, Redlands, CA. Re: GIS [Discussion]. Geographic Information Systems Dis-
as a Science [Discussion]. Geographic Information cussion List [Online].
Systems Discussion List [Online]. ———. November 1, 1993d. Re: GIS as a Science and
———. November 3, 1993b. Re: GIS as a Science Value of Peer Review [Discussion]. Geographic
[Discussion]. Geographic Information Systems Dis- Information Systems Discussion List [Online].
cussion List [Online]. Wright, D. J. October 28, 1993a. Department of Geog-
Scalise, A. H. November 1, 1993. United Nations De- raphy, University of California-Santa Barbara. Re:
velopment Program, Argentina. GIS as a Science Value of Peer Review [Discussion]. Geographic
[Discussion]. Geographic Information Systems Dis- Information Systems Discussion List [Online].
cussion List [Online]. ———. November 2, 1993b. Re: GIS as a Science
Skelly, C. W. October 28, 1993a. James Cook Univer- [Discussion]. Geographic Information Systems Dis-
sity, Australia. GIS and Remote Sensing Research cussion List [Online].
[Discussion]. Geographic Information Systems Dis-
cussion List [Online].
———. October 28, 1993b. Re: Value of Peer Review
[Discussion]. Geographic Information Systems Dis-
cussion List [Online].

Correspondence: Department of Geosciences, Oregon State University, Corvallis, OR 97331-5506, email


dawn@dusk.geo.orst.edu (Wright); Department of Geography, University of California, Santa Barbara, CA
93106-4060 (Goodchild and Proctor).

View publication stats

You might also like