speaking for myself, I (momentarily) create myself--just as much as when I speak for others, I create their selves--in the sense that I create a public, discursive self, a self which is much more unified than any subjective experience can support. . . . The point is that a kind of representation occurs in all cases of speaking for, whether I am speaking for myself or others, that this representation is never a simple act of discovery.
Telling my own story (representing myself) when describing how I came to this research was not a "simple act of discovery," but a very challenging task, as was representing these women. I began to feel more comfortable about the issue of representation, as I recognized that meanings drawn from these representations were "plural and shifting" ( Alcoff, 1994, p. 291). I was still concerned about generating theory because I was afraid that the plural and shifting nature of the meanings I had drawn from these representations would be forgotten, or simply interpreted as "true" or "the way it is."
Trinh Minh-ha wrote of theoretical or academic writing or jargon that "to many men's ears. . . [it] is synonymous with 'profound,' serious,' 'substantial,' 'scientific,' 'consequential,' 'thoughtful,' or 'thought-engaging'" ( 1989, p. 41). I felt torn between the academic push to create grounded theory and the desire to allow women to speak for themselves without analysis or representation. I selected a safe middle ground.
I decided that I would engage in some analysis, such as looking at similar themes, points of contention, and points of contradiction. I feared the power of my words as I created my arguments because language is powerful. Trinh Minh-ha ( 1989, p. 49) captured the power of language by quoting a "wise Dogon man" from a "remote village of Africa" who: "used to say 'to be naked is to be speechless.' Power, as unveiled by numerous contemporary writings, has always inscribed itself in language. . . . And language is one of the most complex forms of subjugation." Recognizing the power of language to "subjugate" women, I challenged every assertion I made in my analysis and questioned what right I had to make it. This should be part of all research where one is representing or speaking for another.
The predominant reporting method in social science drug research has been a quantitative approach that has often statistically controlled for and essentialized aspects of one's complex social location such as gender, race, and social class by examining the differences among groups and making generalizations from these findings that frequently reinforce existing stereotypes. One way to break down the existing race, gender, and social class stereotypes that continue to be reinforced in the quantitative research literature is to listen to many diverse voices about how gender, race, and class are experienced in drug use and how aspects of social location are taken up by people when they construct meaning about drugs. Aspects of social location such as power, social class, and race were problematic in this project, making representation challenging but neces-