The story behind Miriam Van Waters’s burning of her personal correspondence illus­trates the problem of interpreting sexual identities historically. Like the working-class lesbians in prison, the administrators and staff of women’s reformatories were vulner­able to charges of sexual deviance. Long before I became her biographer, I had heard rumors about Van Waters’s personal life, so the question “Was she a lesbian?” lurked in the background as I wrote. Each discovery of new sources contributed to my analy­sis. When I finished the book, I wrote this essay to retrace my interpretive process. I wanted to show not only what the evidence revealed about Van Waters but also how I came to reframe the question of lesbian identity by breaking it into historically useful components. This essay, more than my earlier work, acknowledges the inseparable layers of subjective, discursive, and social meanings that constitute sexuality.

on a clear June morning in 1948, the controversial prison reformer Mir­iam Van Waters made a painful and momentous decision. For months, she had been embroiled in a political struggle with conservative state officials who wished to dismiss her as the liberal superintendent of the Massachu­setts Reformatory for Women. Local newspapers headlined the claims that Van Waters coddled prisoners, hired ex-inmates, and condoned homosexual behavior in prison. Investigators from the Department of Corrections inter­rogated her staff and seized inmate files as evidence.

As she sat before a glowing fireplace in her home that June morning, Van Waters fueled the blaze with some of her most precious possessions. “The Burning of Letters continues,” she wrote in her journal that day. “One can have no personal ‘life’ in this battle, so I have destroyed many letters of over

Previously published as Estelle B. Freedman, “‘The Burning of Letters Continues’: Elu­sive Identities and the Historical Construction of Sexuality,” Journal of Women’s History 9, no. 4 (Winter 1998): 181-200. © Journal of Women’s History. Reprinted by permission of The Johns Hopkins University Press.

22 years.” Van Waters had met her patron and romantic partner, Geraldine Thompson, twenty-two years earlier. Since the late 1920s, they had corre­sponded almost daily, sharing their thoughts, their activities, and their love during the weeks between their regular visits with each other. All but a few of the daily letters Thompson had addressed to her “Old Sweet,” her “Dear­est Dearest Love,” went up in flames that day. As she burned the letters, Van Waters recorded her sense of loss: “They might have been inspiration, history, joy, style — to me in ‘old age.’” Instead, she resolved to keep their message within herself: “The letters are bone and sinew now in my carnage. Doubtless my character has been formed by them.”1

As Van Waters’s biographer, I felt deeply mixed emotions when I read this passage in her journal.2 On the one hand, my empathic pain for her loss quickly turned to anger at the politicians and the society whose condemna­tion of love between women forced her to destroy this personally meaning­ful correspondence — and, in the process, prevented me from reading this significant historical evidence. On the other hand, uncovering the revela­tion that she had destroyed the bulk of her correspondence from Thompson provided in itself an important clue in my quest to understand Van Waters’s elusive sexual identity. “The Burning of Letters” passage offered a smoking gun of sorts to explain the absence of evidence documenting the intimacy of the Van Waters-Thompson relationship. Furthermore, that threats of political scandal had prompted the burning helped confirm my interpreta­tion of Van Waters’s ordeal as a precursor of the sexual surveillance of the McCarthy era.

Van Waters’s decision to destroy the record of her intimate relationship with another woman was not unique among professional women of her gen­eration. Born in the late nineteenth century, just as the modern concept of homosexuality emerged within European and American medical literature, educated women like Van Waters bridged two eras in the social construc­tion of sexual identity. Although they came of age in a world that still valued female intimacy, over the course of their lifetimes, romantic friendships lost the sexual innocence they had once enjoyed.3 That so many women in public life destroyed their correspondence with close female friends or partners in­dicates their awareness of the process of stigmatization that took place in the early twentieth century. As a result, self-censorship — apart from familial or historical suppression of evidence — has created historical silences that speak worlds about the relationships between such women as Miriam Van Waters and Geraldine Thompson, Alice Paul and Elsie Hill, Molly Dewson and Polly Porter, M. Carey Thomas and Mamie Gwinn, and Frieda Miller

and Pauline Newman, to name a few who burned all or part of their per­sonal papers.4

Van Waters’s conscious effort to conceal her relationship with Thompson represents only one of the many challenges I faced as a feminist biographer committed to interpreting private as well as public life. Two related problems complicated this task: evasion and contradiction within existing historical sources and the weaknesses of my own preconceived historical categories.

These problems surfaced even before I became Van Waters’s biographer. Years ago, a historian who knew of my interest in women’s prisons suggested that I write about Miriam Van Waters. When I seemed uninterested, he added the tantalizing comment, “She was a lesbian, you know.” Although I did not undertake a biography at the time, I did become intrigued a few years later when I looked up the heading “lesbian” in the Schlesinger Library card catalog and found a lone reference to files in the Van Waters Papers. Catalog drawer in hand, I walked into the archivists’ office, mentioned that someone had once told me that Miriam Van Waters was a lesbian, and asked, rather naively, “Was she?” An archivist responded — I paraphrase slightly—“We don’t say that about anyone without proof.” The implication, in tone and words, was that I was making an unpleasant accusation.

In retrospect, I realize that my penchant for naming and the library’s reluctance to do so were equally problematic. Each of us acted upon certain assumptions about lesbian identity. My modern “lesbian feminist” impulse sought to reject what Blanche Cook has called “the historical denial of lesbi­anism” in favor of restoring the lesbian element to women’s history, without questioning the historical meaning of the term “lesbian.” Or perhaps I was unconsciously adopting Adrienne Rich’s concept of a lesbian continuum, which generously labeled a range of woman-centered behaviors as lesbian.5 The archivist, it seemed to me, adopted a modern homophobic response, seeking to protect historical subjects from accusations of identities that might have offended them (or, I suspected, her). The library’s policy, I later learned, was that only self-identified lesbians would be categorized as such.

Eventually, my simplistic question “Was she a lesbian?” would evolve into a more complex inquiry into Van Waters’s sexual subjectivity. By the time I made the commitment to write a biography of Van Waters, I recognized the need to historicize my own inherited categories, particularly those — such as homosexuality—that derive from modern, Western notions of the self, identity, and politics. The very category of sexual identity often rests upon concepts of a unified self and describes the consciousness of a bourgeois (male) historical actor, for whom the taken-for-granted privileges of gender,

class, and frequently race permit the foregrounding of sexual subjectivity. For many women, class, race, or ethnicity may be more salient than sexu­ality in the formation of modern identities. Moreover, as scholars such as Earl Lewis and Dana Takagi argue, for many women and people of color, multiplicities of identities are perpetually uncertain and in flux. Whether in the case of visible ethnicity or invisible sexuality, identity categories mask the complexity of incoherent identities. Recognizing this fact, however, does not preclude the need to understand the historical construction of identities, which in practice wields significant power on the levels of both individual and social relationships. Studies of sexuality thus face a dilemma similar to those that address ethnicity: should scholars and activists discard or claim social categories that we admit to be unstable?6

I believe that exploring the concept of sexual identity remains a critical task for historians, despite the difficulty of pinning down our subject. Let me briefly survey the historiographical landscape to contextualize my ap­proach. While the historical denial of homosexuality has recently given way to extensive scholarly inquiry, competing theories continue to character­ize the literature. At one extreme lie essentialist notions of a transcendent “gay” identity that disregard historical and cultural specificity. These theo­ries range from the biological/hormonal/genetic to the historical argument that “gay” people have always existed, from the Greeks to the moderns, as a unique category. At the other extreme lies a postmodernist tendency to frag­ment identity into a category so fluid and unstable that it threatens to return, curiously enough, to the denial with which we began. In short, we cannot study lesbians in the past because the term is too modern and limiting, too falsely universalizing, and too much a figment of discursive imagination.7

A middle ground, where I am most comfortable, draws on the model of social construction first articulated by the British sociologist Mary McIn­tosh. This account of the historical transition from homosexual acts to ho­mosexual identities emphasizes the impact of economic and social relations within Western industrial societies. John D’Emilio and I applied elements of this model to our survey of American sexual history, arguing implicitly against theories of the discursive creation of homosexual identity by modern medical and psychiatric authorities. Rather, we suggested, within an emerg­ing capitalist society, the separation of reproduction and heterosexuality, along with the growth of wage labor, provided economic and social spaces for the emergence of self-conscious same-sex relationships, which were then

labeled homosexual. The possibilities for these relationships, outside tradi­tional family controls, in turn made possible, and likely, the formation of gay identity and eventually modern gay politics.8

As feminist scholarship reminds us, however, the timing and meaning of this transition differed by gender. Trying to fit women into the (traditional)- acts-to-(capitalist)-identity model raises problems, given women’s continu­ing reproductive responsibilities and their economic dependence within both middle – and working-class families in the nineteenth century, when this transition was occurring. Nonetheless, distinct female constructions of same-sex relationships did emerge in response to new economic roles. These included but were not limited to women who passed as men and married other women, usually associated with the working class; romantic friend­ships, usually associated with middle-class women; and mutual households, known as “Boston marriages,” largely associated with educated women but adopted by working women as well.9

Only in the twentieth century, however, as the U. S. consumer culture in­creasingly sexualized women of all classes, did an explicitly lesbian identity emerge, first within black and white working-class cultures and later among middle-class feminists. As I have suggested elsewhere, racial and class con­structions of female sexuality deeply influenced this process. Given the sexual objectification of African American and working-class women and the denial of sexual agency to middle-class white women, the lesbian label was often applied to black and working-class women who were already as­sociated with criminality and prostitution.10 Despite their practices of same – sex love, white middle-class women rarely claimed lesbianism as an identity before the 1940s, with the notable exception of some artists, writers, and bohemians.11

This social-constructionist account of the emergence of lesbian identity remains, I feel, too abstract. Its very insistence on the social, as opposed to individual, meaning of sexuality risks overlooking the fact that identity for­mation is in fact as much an individual as a social phenomenon. Indeed, the two are often in conflict, especially in periods of transition, when individu­als raised with the sexual categories of an earlier culture partake in the social changes that redefine their behaviors. Historians too rarely acknowledge either the complications of contradictory and competing identities in their subjects or the dilemmas we face in handling elusive identities/2

In the interest of becoming more explicit about how historians grapple with the elusive subject of sexual identity, I want to reflect on my own pro­cess of research and interpretation by recalling three kinds of evidence that

combined to shape my understanding of Miriam Van Waters’s sexuality: first, what her professional writings reveal about her intellectual catego­ries — that is, the authoritative discourses she encountered and may have internalized or modified; second, how personal sources such as letters and diaries provide clues to her subjective experience of sexuality; and third, how rumors, accusations, and assumptions enter the historical record and influence the historian. Together, these sources help explain why the burn­ing of letters continued in Van Waters’s lifetime and what it meant for her identity.