F< g>I g<,o~~ ik ~

gc) Ac5MARY GILMARTIN

ACI<NOWLEDGEPv'lENTS

Feminist Epistemology in geographyThanks to Vincent Del Casino, Carole Gallaher, John Paul ]ones, Dervila Layden,Pamela Moss, Nancy O'Donnell, and Priya Rangan for their comments on earlierdrafts of this chapter. Meghan Cope

- RES'EARCM TIP,:::j:-:',;-',;.:::-'::::::

-: :-".:::;,, theory, rnethoclalogy, data'collection, analysis.- : : : : . . ' ' ' :. : , : : :: :

~':::.pote-jour ideas as they evolve'"and.your thoughts"-orI diffeierit:,,':l.: aspects of the research topic.."--.: ".",',';:-'.='-":-::::',,::

,

Record "ah-has," ."oh; yeas," and ',".of courses"" as they,;corrIe to:-; jyou. . . : :

:

: " ,: : '; . , ,: ' . , :: : ; : -'- : : : , ' : ' , : :, : : - :.

:::,,'I:~P-m',,:",;:„.',",: .,„;,:;:".;;-",-";:-" '.'

Develop a system that works conveniently and thoroughly.-;for your.';-,''::

research purpos'es,'::::::-:--':..":- "

Organize you(:;journal, e,g, use color,,coding, index tabs --'recipe,.",:-"cards, or-different binders to organize..your material.Subdivide" your jouinal into categoriessuch as'research question, ,' :

,.:- + ..:.,Keep rriore thari one jourrIal';: e.'g "'one'for method~logy, one for' .=

: the'ory-,' c rIe fdf" detaiIs,, :::::::,-',-::I'-';:,::.:.""::;:::,:,+ . Make notes on lInks amorig" data„"th'eory', and method;-.

Enrich your jouinaj,b j..drawtrl'gs, figur'es„,'.or clIppIngs, Don t IesttIct -- ,:,:

yourself to,.written 'fexti~<':-'-:':::-:,'::::::::Maintain 'a refererIce Iibrary-(bibliographic,:-",software carI be useful). : ,' ,

Keep a detailed recorcl:of names, numbers;-'.and addresses.

:: and. to the project.Date even ent

- '- i,.::,: Make a riote of the. commitments you' ve made to your participants

Some type of epistemology undcrlies every research project, yet it is notalways (or even often) made clear by the authors of scholarly work. Anepistemology is a th.eory of knowledge with specific reference to the limitsand validity of knowledge. More simply put, an epistemology helps usanswer the question "How do I k now w ha t i s t r ue?" (McDowell andSharp, l999, p. 75). For example, if we decide only to validate knowledgebased on direct observation, we have identified an epistemology: we cansay "we know X i s t r u e because we can observe it." This part icu]arepistemology (called empiricism) is fairly simple, yet already we run intoproblems in many areas of research. For instance, it may bc di f i cul t toobserve phenomena such as racism, the global economy, or democracy, yetwc know they exist because of their effects on society, economics, andpo i iti cs.

In order to understand what epistemology is and how i t i n fl uencesresearch, we first need to recognize that knowledge is humanly constructedor "produced." That is, knowledge is not just "out there" wait ing to berevealed to us; rather, we are active participants in producing what countsas knowledge. Once we understand this point, we can move to recognizingthat if knowledge is indeed produced by human actors, there must bemultiple and even contradictory perspectives, interpretations, and uses ofknowledge. I<nowledge is not something that we can passively or activelyacquire because we are always invo]ved in its production and interpreta-tion. Similarly, knowledge production is never a "value-free" or unbiased

A third point ar ising out of the two premises above is that there arcmany epistemologies possible in the research process. Researchers' individ-

What is an "Epistemology" ?

process.

MEC HAN COPE FEMINIST EPISTEMOLOGY IN GEOGRAPHY44

ual perspectives will inevitably influence their privileging of difierent typesof 1<nowlcdge (for example, a researcher will value some sources, forms, orkinds of data more than others). Privileging certain types of knowledgeshapes the researcher's understanding of the results and in turn influencesthe way that others wil] understand the published interpretations of resulfs.

There are many levels at which we can identify ways that the underlyingajstemology biases and shapes research processes and results. Researchers

make decisions about how to f orm their research questions, how to co lectdata, how to analyse data, and how to communicate results to a wideraudience. At every point in th is process, the beliefs and interests of theresearcher matter to the outcome. In the past few decades there has been apush to reveal and make explicit the underlying cpistemologies of all areasof research because of the growing recognition that all research is rife withbiases that researchers bring with them to the research process: culture,

ce gender class and other forms of difference. While we recognize thatthere is no such thing as bias-free research, there is a growing commitmentto minimize harmful biases of omission or discrimination (e.g. ignoring theinfluences of. race or gender on disease statistics) and to freely acknowledgeother biases (e.g. coming from a middle-class background in a study onpoor families).

So V/hat is a Feminist Epistemology?

11

duction of knowledge. What do we as a society consider to be legitimateknowledge? And what role does gender play in the production of knowl-edge? How is our understanding of knowledge production influenced if westar f fhinking of multiple genders along a spectrum rather than as a binaryof women/men. I f t h c p r o duct ion o f k n o w ledge is an active processinvolving differently situated human actors, we would expect that people' svarjous experiences, identities, and social locations will influence what theycount as knowledge and how t hey pa r t ic ipate in i t s p roduction andlegj f<mIzatron.

Gender influences thc ways that people experience the world, interactwjfh others, and what opportunities or pr ivi leges are open or closed tothem. One of the most important elements of gender relations is the waythat they solidify h ierarchies and relationships of power in a s o c iety,fhr()ugh various means of, on one hand, oppression (violence, discrimina-

IT)algjnaljzatjon ) and, on the o ther, pr iv i lege (preferences, favors,power over others). Therefore, in the processes of producing knowledge,we would expect those who are oppressed to have di f ferent roles inconstructing and legitimating knowledge than those who are privileged andjn power. Most likely, the views and thoughts of the former group will besubsumed under those of the latter group. This is an example of consideringhow gender relations influence the production of knowledge: women' sactive participation in what "counts" as knowledge has historically been.seen as less significant than men's through the mechanisms of power-basedgender relations. That is not to say that women are less active in or capableof producing knowledge; rather their ro les in t h is p rocess have beenunrecognized or discounted due to the exaltation of men's roles.

A gendered analysis of knowledge production is more complicated thanmerely looking at the roles of men and the roles of women in makingknowledge. Gender affects societies deeply and in mult iple ways that arenot always easily identificd, separated, or categorized. Gender as a set ofrelationships influences the production o f k n o w ledge th rough manyavenues: media, the socjaljzatjon of children, religious and cultural values,and political and economic processes. Part of the task of understanding theinfluence of gender on the production of knowledge is to try to identifyand tease apart these many influences while remaining conscious of the'vays that these are constantly changing and affecting each other. Further,by opening our inquiries to include multiple genders and not just those ofwomen and men, we introduce another level of complexity to a cr i t icalapproach to the ways that gender relations influence the production ofknowledge. Overall, a feminist epistemology takes gender as central tounderstanding the production of knowledge and thus influences the natureof research performed and interpreted f rom this perspective.

If we understand an epistemology to be a theory of knowledge that isparticularly concerned with how knowledge is produced and its inherenfbiases, we can identify epistemologies coming from various perspectivessuch as an anti-racist epistemology, a Marxist epistemology, or — as isexplored here — a feminist epistemology. There are two aspects to a feministepistemology according to El izabeth Anderson (1995). The first involveshow the consideration of gender influences what "counts" as knowledge,how knowledge is legitimized, and how knowledge is reproduced andrepresented to others. For example, the recent push in feminist scholarshipto "listen to women's voices" challenges the ways that previous researchersdiscounted women's words as qualifying as "knowledge." To take thisnotion a step further, a feminist epistemology involves not only hearing"women's voices" but also thinking about how gender as a set of socjarelations affects both men's and women's responses jn the research frame-work, how the genders of interviewers and respondents might affect thedata, and how research results are circulated to academic audiences andthe public.

The second aspect of feminist epistemology requires thinking about howsocially constructed gender roles, norms, and relations inf l uenc the pro-

46 MECiHAN COPE

Challenging Masculinist Science

47

tlie use of aspirin for preventing subsequent heart attacks (the results werewidely circulated via television and magazine advertisements), the initialstudy was done entirely on men. The researchers failed to consider theapplication of this treatment for women and yet the results were pub}ishcdas if they applied equally to everyone.

So what do we mean by "science" here? Science is perceived as system-atic, unbiased, neutral, rigorous, and ultimately, the best way to get to thegraph. Yet from the examples above, we know that these perceptions ofscience are open to question. Many of the roots of modern Western sciencelie in the European Renaissance era when philosophers and researchersatte]ilpted to use empiricism (direct observation) to throw off the veil ofmysticism prevalent prior to the Renaissance. The notion of t ruth devel-oped such that i t was assumed to be pre-existing (out there f<ir us todiscover) if only we could gather enough evidence and nieasure phenomenausing enough variables. However, feminists and p ostmodernscholars havecritiqued the empiricist approach of "science," claimingit produces nierelyone of many competing "truths." The very term "science" has had a longhistory <>f masculinism because it has represented a powerful force insociety that has consistently ignored or actively suppressed diverse forms ofknowledge production, the importance of gender and other sets of relation-ships on constructing multiple truths, and, finally, "science" has carriedwith it the assumpti<)n that its complete and exhaustive authority overknowledge cannot be challenged.

FEMINIST EPISTEMOLOCY IN GEOCRAPHY

The first task of understanding and adopting a feminist episten tology is toreveal the m<isculinist underpinnings of science, particularly the ways t}iatclaims to "value-free" or "neutral" science actually mask gender signifi-cance. Gillian Rose (1993, p. 4) identifies masculinist work as that whichclaims "to he exhaustive and it therefore thinks that no one else can add toits knowledge."

CorIsidcr a labor market analysis that looks carefully at all the obvi<>usvariables of where different jobs are located, what the qualifications ofpeople hotding those jobs are, journey-to-work times, and the increase ordecrease of job ava i labil ity <>ver several years. This analysis may bethorough in some people's eyes, but a feminist researcher would immedi-ately want to know how gender affects the workings of the job marketasking such questions as: Are women concentrated in certain jobs andexcluded from others? Do women and men earn the same aniount forcomparable work? Do women's greater household responsibilities affecttheir labor market participation? Are women's journeys to work shorter orlonger than men' s? Is there a gender division of labor? And, more broadly,how do gender relations saturate the practices of the labor market incountless subtle ways, ide<ilogica}}y, culturally, and poiiticaiiy? By ignoringthese questions (and others, such as what effects race and racism have onthe labor. market), the initial study seems inherently incomplete and f}awedbecause it was based on the assumptions that not only do i n d iv idualwomen and men behave the same way in labor markets, but also broadgendered social relations have no impact on employment. This is anexample of the limited vision of masculinist science. Although the research-ers und<)ubtedly felt they were being "neutral," objective, and thorough intheir research, their failure to consider a wide range of questions and issuesrelating to gender renders their analysis partial and narrow. A feministcritique of this research challenges its epistemology as masculinist becauseby faiJing to examine differences between nien and women and the impactof gender relations on w<irk and labor markets, the authors assumed thatall practical and ideological impacts on labor markets had been ful lyaccounted for in the analysis.

Another aspect of masculinist science is an assumptionthat men corn-prise the norm for. humanity. The pharmaceutical trials of aspirin in the USdemonstrate this p<>int clear}y. Pharmaceutical trials for new drugs and olddrugs for new use are considered to be some of the most r igorous and"bias free" in all of science. The US Food and Drug Administration is verycareful about allowing drug companies to make claims about the effective-ness of their pr<iducts based on their own research. However, in the case of

The Science Question

Sandra I-Iarding (1991) developed three critiques of masculinist science. Inone sense, there is the rnatter of simply doing "bad science" by creatingand maintaining highly discriminatory hierarchies within science disciplinesthat prevent women (and other people who are perceived as "unscientific" )from reaching the upper echelons of performing research. Second, Hardingconsiders "standpoint theory" in w h ich there is an understanding thatmultiple perspectives are valid in as ntuch as they are genuinely held bypeople coming from di f ferent "standpoints." From this perspective, "sci-ence" can be critiqued on the basis of the limitations <>f ma)e researchers'understandings due to their positions and experiences of power and privi-lege. That is, maje researchers cannot p<issibly understand. multiple viewsof a problem because the world is at their feet and there is no need to stepdown from a pedestal of power (or so the theory goes ). In a third critiqueof rnasculinist science, I-larding reviews the contributi<ins of feminist post-moderiiists who challenge the very premises <if science as merelyexpressi<>ns of power and oppression. That is, the whole project of science

MEGHAN COPE FEMINIST EPISTEMOLOGY IN GEOGRAPHY 49

is just a mechanism to keep large numbers of people under the thumb of afew elites. Of course, these three critiques are overlapping rather thanmutually exclusive — standpoint theorists may well agree that science is amechanism of the e l i te, for example. The important point is that theconcept of science is challenged on the basis of its masculinism throughmany different routes, both practical and ideological.

But where do wc go from here> Is science salvageable in any form? Ormust we all just wal low in a sea of subjectivity and relativism with nouniversal truths upheld by thc public? Harding and others argue that thekey to saving science is to recognize that objectivity cannot be increased bysome pretended "value neutrality." Rather, we need wide and r igorousinclusion of multiple perspectives, including those of the oppressed andmarginalized. Shc and others have advocated a "strong objectivity" thatmaintains the goals of rigorous scientific inquiry but requires a wider arrayof questions, interpretations, d i f ferent perspectives, and inc lusion ofresearchers and subjects from marginalized groups to strengthen the claimsof t l uth.

As another response in challenge to claims of value neutrality, DonnaHaraway uses the concept of " s i tuated knowledges" (1988). By this,Haraway means that we must reject the all-encompassing "truth" not ionin favor o f c o n t ex t-specific and s i tuat ion-sensitive knowledges. Forinstance, rather than searching for universal statements that apply every-where to everyone (and therefore really apply nowhere and to no one), itwould be better fo r u s t o a c k nowledge the b iases, perspectives, andcontextual factors such as political systems and cultural values inherent inthe research project and move forward from that point. Harding, Haraway,and others are attempting to demonstrate that the cloak of "neutrality" inscience is actually hiding the true complexities of research in our real worldof messy and complicated phenomena and masking the perpetually andnecessarily partial views that we as researchers hold.

Collecting data

Before a feminist researcher can even go about collecting data, shc or hemust formulate thc research question (s) that guide the project. For example,a general intcrcst in how people use and experience a public park woulddepend on various factors including the ages, race, class, gender, sexuality,ability, ctc. of the users, and on the purpose of being in the park (to play,to sleep, to meet people, as a short-cut to work, to buy drugs). These issuesmay then spur the feminist researcher to recognize that relations of power,such as gender, make a difference in how/when the park is used. Followingthis, the researcher may consider. the broader question of how constructionof gender norms, expectations, and relations influence perceptions of publicspace. This is how epistemology helps shape a research question — byraising issues of the significance of gender right from the start.

Forming research questions

Feminist Epistemology in Practice

Gathering data for a research project involves many steps as well, which,again, are influenced by epistemology. First, the feminist researcher needsto consider what kind of data would be appropriate to connect to her/histheoretical framework. For example, someone doing a study of housingissues for Hmong refugees in California would need to consider people' sown preferences, forms of d iscrimination in the housing market, loanavailability, job l ocat ions, t ransportation,cost of housing, ci t izenshipissues, language barriers, etc. The researcher would then need to determineexactly how to measure or create indicators for each of these categoriesand figure out where the data would come from — the census, a mail survey,in-person interviews, participant observation, bank data, and so on. Epis-temological questions in data gathering come through the consideration ofwhat "counts" as data. In feminist research, data issues are influenced byan explicit emphasison legitimizing women's knowledge, exploring genderas a set of p o wer r e lat ions, valuing gender as a central var iable inquantitative studies, and considering ways in which social constructions ofgender influence the production of 1cnowlcdge (Professionat Geographer,1994). Indeed, in many ways feminist scholars have pushed the boundariesof what "counts" as data, using diverse sources such as diaries, letters,photographs, songs, and a r tworl< to b r o aden our understanding ofwomen's lives and gender relations, particularly when few other sourcesare available for capturing their voices. For example, these have beenespecially helpful in historical work because women's history is less oftenr ecorded, arId rarely in t h ei r ow n words because masculinist history/

How does a feminist epistemology make a difference in a research project?First, we need to acknowledge that thcrc arc many possible feministepistemologies; that is, there is no one right way tn do feminist research.Just as there are multiple forms of feminism, there are multiple feministcpistemologies that may, even in a given project, complement, contradict,or build on each other. In a general sense, any research project involvesforming research questions, collecting data, choosing methods, analysis andinterpretation of data, and reprcscnting the results — in each of these stagesa feminist epistemology matters.

MEGHAN COPE50

geography has been privileged in the production of knowledge, in part byprivileging certain data sources over others (see Norwood and Monk ,1987; I<atz and Monk, 19.93; Domosh, 1996; Cope, 1998).

The actual acts of collecting data. are also implicated in the episternol<ig-ical basis of feminist research. For example, who is in tcrvicwed, whatquestions are asked and h<iw they are phrased, thc gender of the interviewervis-a-vis the gender of the respondent, the nature of the interview (formallist of questions versus an open-ended conversation ), and even where theinterview takes place(a formal of f i c , a co f fee shop, a public space, aprivate home ) are important elements to the construction of the researchpr<iject. These issues all have implications for epistemologybecause, again,choices about performing research in certain ways and nnt others indicatewhat we consider to be the limits and validity of knowledge. These factorsare also important to acknowledge in the representation of results becausethey have a great deal of in f luence <in the findings. For instance, M<iss(1995b) showed that her gender facilitated. research with the h<iusc cleanersshe interviewed and observed (though her class status complicated itsomewhat), whrle Fngland (scc chapter 12, this volume) found that hergender hindered her in terviews with male banking CEOs in a h i gh lypatriarchal setting.

In terms of da ta, the feminist imperative t<i challenge "masculinistscience" requires en'tertaining a w ide a r ray o f ex p lanations for one' shypothesis or theory, and then th inking about how the data collectionstrategies reflect that commitment to feminist inquiry. Attention must bepaid to all stages of data "mining," collection, and validation.

Analysis and interpretation

Similarly, thc methods used for data analysis should be appropriate tnrthe purposes <if the research. If the purpose of the research is to understandthe social and economic implicati<ins of women as subsistence farmers theneel fain methods may be m<ire appropriate than <ithers — for exampleif"offr'cial" data sources show very few women engaging in farming butcasual observation suggests that the practice is in fact very common, thendependence on large, government inf<irmation and statistical methodswould not be adequate. And this data mismatch might suggest a goodplace to start a research project!

FEMINIST EPISTEMOLOGY IN G EOGRAPHY

Choosing methods

Analysis refers t<i more than merely the methods used to analyse the data,whether statistical regression or qualitative text analysis. There is also acomponent of in terpretation, reflection, and re-evaluation that involvesthInking about the meanings and impl ications of the data rather thanmerely the results. First, a feminist analysis is sensitive to gender differencesin social, political, and economic relati<ins for both women and men. Forcxamplc, a feminist epistemology requires that we acknowledge that menand women navigate labor markets, transit systems, health care, farmingpractices, and cultural traditions differently and leads us to seek to under-stand the root ot these differences and their impacts on daily l i fe andbeyond.

Second, feminist researchers are sensitive to using gender as a "problem-atic"; that is, seeing gender as the central hinge of the research on whicheverything else pivots. This level of analysis involves c<insidering differencesas part of a larger system of gender relations that are deeply embedded insocial, cultural, political, and economic pr<icesses and maintained throughcvclyday practices, beliefs, and expectations, as well as structural forcessuch as laws and institutions.

FinaIly, analysis in terninist research is sensitive to h<iw a g enderperspective influences thc productIon of knowledge. Inhere, analysis becomesa highly ref lectiv process in which a researcher acknowledges her or hisown gendered perspective and how that shapes the interpretation of results.This is akin to I-Iaraway's concept nf a "s i tuated knowledge" where thecontext of the researcher, the subjects, and the place (b<ith social andphysical) are taken into account in the analysis to understand how genderinfluences the producti<in of knowledge: who produces "lcgitirnate" knowl-edge, how it is produced, whether and h<iw that pr<iduction is contested,and in what broader context knowledge is created and re-created.

A feminist epistemologydoes not require the use of specific methods, but itdoes require critical reflection on the usc of al l methods ot analysis andinterpretation. The combination of a set <if methods with a par t icularepistemology is commonly rcfcrred to as the "methodology" of a project.As with any research, the methods should be appropriate for the data andfnr the research questions. Interviews and personal logs may be m< ircappropriate for learning about women's experiences of gender discrimina-tion in employmcnt while large-scale data bases and statistical analyses aremore suitable for demonstrating the broad effects of labor market biases(Lawson, 1995). However, b<ith qualitative and quantItative methods canbc imbued with a feminist epistemology that shapes the research questi<ins,sees gender as an importantset of relations that has deep repercussions onthe lives ot both mcn and women, and considers the ways that genderinfluences the prnducti<in of knowledge.

FEMINIST EPISTEMOLOGY IN G EOGRAPHY 53MEC HAN COPE

Representation of resultsexploitative relations and the experiences nf the processes of marginaliza-tion ( 1995b, p. 83; emphasis in original). That is, just because feministresearchers try to be inclusive in the writing or production process and tryt„engage at a pol i t ical level with cmpowering their "subjects" does notniean that power and oppression are eliminated. While various researchersapproach this problemdifferently, common strategies include maintaininga self-critical reflexivity (questioning one's own actions and motivations intlie research) and using existing privileges of the researcher as ways tnfoster change (e.g. using the "authority" o f a u n i versity setting to cal lattention to oppression, abuses, and exploitation). The work of a cr i t icalgeographer is never done!

The communication of the results of data gathering, methods, and analysisto a broader audience is not a simpletask free of biases. Whether it is in awritten paper, an oral presentation, or through some other form nf media(film, website, policy report, etc.), the representation of results is complexand epistemologically significant. Feminists have long challenged the notingof "experts" who spread "the truth" in al l realms of science; such expertsare critiqued as validating only some forms of knowledge, maskingorignoring their b iases, and using the power of t hei r posit ions to avoidopposition in t h e c o nstruction o f t h e i r t r u ths. I n r e sponse, feministresearchers have developed various strategies that have been used tn disruptthe traditional power dynamic between researcher and subjects, such assharing and confirming the results of analysis with the respondents andcvcn co-authorship of the final presentation. These strategies are not freefrom problemseither, but they are worth considering for certain types ofresearch. A key po in t i s t ha t th e r epresentation of results should beapproached with as much care and self-critique as any other step in theresearch project. Feminist epistemologies make a difference to the represen-tation stage; in effect, the representation of research is producing knowl-edge and is therefore highly influenced by the underlying epistemologicalconcerns of the authnr (s).

However, even herc we need to d i s rupt nur t i dy n o t ions of w h a tconstitutes "representing results." Feminist researchers have been on theforefront of exploring alternative research strategies, which influence alls tages of the research process including representation of results. fo rexample, co-authoring with subjects, participatory action research, andother forms of collaboration destabilize the traditional model of researcher-researched and create openings fnr new forms nf communicating findings.

Finally, the media through which the data are shared and the audienceto whom they are targeted are important considerations. While much ofacademic research is published solely in disciplinary journals (oft'en for thcbencfiit only of the researcher who gets a degree or a promotion based onpublications!) or prescntcd at academic meetings, there are other forumsthrough which research results can bc cffcctively spread — and perhapsmake a greater difference in people's l ives — such as press releases tonewspapers, television news spots highlighting special events, reports topolitical decision-makers, and internet sites. The process of communicatingresearch findings is always politically and epistemnlogically complicated.As Pamela Moss has written, '"wc Ias rcscarchersJ must be able to come toterms with differences in the zuay pozuer is constituted. . . be t ween theresearcher and the ' researched'.. . . D ifference embodies oppressive and

Feminist epistemologies in geography

Feminist geographers have, in the past twenty years, dealt with all nf theissues discussed above both by actually doing research and writing aboutit, and by reflecting on the research process and writ ing about that. Interms of actually doing geography from the position nf a feminist episte-rnOoOy, I would l ike to p rov ide two examples by reviewing two qui tedifferent projects that use gender as a central problematic.

The first is Susan Hanson and Geraldine Pratt's mult i-year projectdesigned to examine the connections between work and home for womenand mcn in Worcester, IVlassachiisetts, in the late 1980s and carly 1.990s(Hansnn and Pratt, 1995 ). In this project, gender is scen as a centraldynamic through which the social and the spatial arc mutually constructed.They argue that the gendered practices of the household division of laborand the labor market's occupational segmentation are inherently dependenton each other — women's socially constructed duties in the home serve toput both spatial and time limits on their job-search activities, and, simul-taneously, cultural gender expectations mean that employers are morelikely to hire women for part icular jobs and under particular conditions(e.g. clerical work at part-time for lnw pay and no benefits). In this case, afeminist epistemology made a d ifference to the research in so m anyinterlocking ways that it is nearly impossible to conceptualize what theirstudy would have been without it . Certainly we would nnt have such anelegant exploration of the deeply embedded nature of gender in labormarkets, nor such a thorough demonstration that the links between horneand work are extremely important for. constructing both women's andmen's lives. From the first moment of conceptualizing a research question("How are gender differences constructecl spatially?" ) to thc final writ ing("Our argument isthat social and economic geographies are the mediathrough which the segregation of large numbers of women into poorly paid

54 MEGHAN COPE FEMINIST EPISTEMOLOGY IN GEOGRAPHY 55

jobs is produced and reproduced" (Hanson and Pratt, 1995, p. l.)), theauthors highlighted women's and men's different experiences, used genderas a central set of relations and constructions, and considered the ways thatgender matters for the production of knowledge.

The second project involves a grassroots polit ical organization,Kensington Welfare Rights Union (KWRU), that is concerned with economic human rights violations, particularjy their impacts on poor wontenand their families in the kensingtonarea of Philadelphia. This project iscurrently being conducted by Me l issa Gilbert and M ichele Masucci atTemple University (Gilbert and Masucci, 1999). The researchers classifytheir project as "action research" to acknowledge its significant componentof political activism. The project's feminist epistemology is demonstratednot only through the centering of how gender roles, relations, and identitiesinfluence thc situations of poor people, but also through the positions of-the researchers and their students as facilitators of political change. There-fore, all stages of the research process are intertwined with act ivism,teaching, the functions of the KWRU, and empowerment of poor people.For instance, Gilbert br ings materials about the I<WRU to professionaltajks, Gilbert and Masucci co-teach a "service learning" course in whichstudents get experience with and simultaneouslylearn about communityorganizing and uneven development in c i t ies, the researchers' studentsrecord poor people's tcstimoniajs to document abuses and conditions,Gilbert and Masucci have used their expertise and access to technology toset up an information management system to catalogue these testimonials,an internal website was developed to display data for other poor people touse, and several workshops have been run by KWRU with assistance f romthe researchers and their students (Gilbert, personal communication,November, 2000).

In this project, it is virtually impossible to separate out the many waysthat a feminist epistemology has saturated the research or even to identifyclearly what thc "representation of resuJts" means. By sharing the processof knowledge production between researchers, students, KWRU members,and poor. people themselves in a fluid and recursive manner, Gi)bert andMasucci demonstrate that there is a great potential for the rich nuances ofdiverse experiences to be explored and represented, not by a s i n g le,privileged authorial voice, but through a more collective effort resulting inreflective narratives, po l icy d i r ect ives, theory-building, and p o l i t i calempowerment.

Feminists concerned primarily with gender quickly discovered that it wasvery difficujt to talk about the oppression of women without also makingrejeren<-e to racism, heterosexism, and oppression based on disability,religion, age, culture, class and other forms of difference. Fighting oppres-s;on on one axis makes little sense when there are multiple forces at workand the effects of each. are impossible to separate. Is a particular womanliving in a poor neighborhood and work ing at a low-wage job because

woman? Because she is Latina? Because she and her family havealways been poor? Because she has a disability that l i tnits her job andhousing options? These things are impossible to separate out in a singlewoman's life because they are so deeply intertwined in hcr identity andher experiences of these things arc inseparable due to mult iple forms ofoppression. Consider your own identity — are you able to separate yourgender f rom your race, from your c lass, from your rel igion, f rom yoursexuality, from your age? Or are they all mixed up together in your self?This is the realization that feminists have come to (slowly) and has forcedus to think not only of how to challenge gender oppression, both polit i-caljy and academically, but also how sexism is just one aspect of a kal-eidoscope of oppressions that require a united strategy of analysis andcritique.

Andcrs<>n (199S) suggests that we must extend our v iew f rom justgender to include the many and var ied forms of o ppression that aresocially and spatially constructed. This extension forces us to ask researchquestions differently ("H<)w are gender azzd race twined together in peo-ple's experiences?" ), suggests a rc-evaluation of our data collection pro-cedures and the methods we use for analysis and interpretation, demandsthat we represent our research in ways that are sensitive to all the multi-farious forms of oppression that in fluence the processes, pcopje, andevents we study, and, u l t imately, makes us reconsider the ways thatknowledge and truth are socially and spatially constructed, produced, andcritiq ued.

Beyond Gender to /MultipleForms of Oppression

ACKNOWLEDGEMENTS

The author thanks Pamela Aloss for insightful comments, patience, an<i supportc(uring the writing of this chapter. Thanks too to Melissa C'ilhert for giving me awindow into her and Michele Masucci's Kensington project, and to anonymousreviewers for helpful comments. In memory of Sara and Emily, who would h lvebeen very good feminists,

MEGHAN COPE

RESEARCH. TIP..

The Difference Feminism Mal<es:Researching Unemployed Women

::Gaining'':-Access to..ReseaI'ch Participants

»:.",.Make a "co!:!d-call ': • . --: ..Use. your contacts (rIetworl(ing)..-:».'..: Obtain documentation::ou! tlInIn!g institutio!

n! al.:,support. :for your

"':.;.,=-:"-'.' p roj ect.".::::::-:.»,.."-::.Prepare a brief description':of,, your research':project..for quick,'

:.,".:-'.-,-: distribution.»,",:- Be:.prepared to substantiate your credentials.»'.",'='Follow: up on initial forays into" making research contacts, i.e. don' t'

::::;-,;. leave..someone hanging and don't make undeliverable promises.::.'.:»"',, :Produce information requested by interv'ie!wee promptly,» -' Customize contact letters.

Demonstrate the value of potential intervI'ev'vee'5' participation in-".:.your project.Gather appropriate infoI.mation.'.to demonstrate.awareness of....!the-:".fit of the potential particip! a !nt In.'t!he research p!'!r'oject. "'-

in an Australian RegionLouise C. Johnson

:. !». . Be persistent.

l am a geographer and a f eminist wh(i has been d(iing feminist geographyfor more than ten years. This has involved documenting the patriarchaleconomy of Australia's textile industry (Johnson, 1990), reading the gen-dered. spaces of suburban houses and shopping centers (J ( )hnson l993,~000) and uncovering the pol i t ics of d i f ference on Melbourne's urbanfringe (Johnson, 1994a, 1994b). Fach of these feminist and geographicalprojects has been structured by a further theoretical perspective: socialist,postmodernist (ir postcolonial f'eminism. The research in each case — thequestions asked, the methods adopted, the conclusions reached and thcstyle of presentation — all varied as a result, but their ferninisrn remainedconstant. As such, each research project was underpinned by some passion-ately held assumptions: that gender matters, that women are both differentfrom men and oppressed hy them, and there is an obligation not only todescribe this situation but a lso to change it . Doing feminist geographyresearch for me therefore has inv(ilved a set of c lear and prescriptiveguiclelines that have no doubt colored the research process. The fore-gr(iunding of such assumptions along with the active place of the researcherwithin the process are two (if the great strengths of a feminist approach.However, as a result of assessing feminist research grants and by reflectingon my (iwn recent work, I am increasingly convinced that such assumptions— or subject positions — present pnifound l imitations. Assessing feministresearch applications has led to a spirited questioning by others — usuallyn(in-feminists — of the predetermined nature of this research such that thereappears to be no real testing of patriarchy if it is already assumed. And ithas proven difficult to dismiss such arguments, raising real questions aboutthe rigor and veracity of such feminist research. Reflections on my currentresearch into women's unemployment have raised similar questions on thepolitics of doing feminist research — the power relati(ins involved and the

' !! f !