Clark, Anna. 2005. “Twilight Moments” in Journal of the History of Sexuality 14, nos. 1-2: 139-60.
Clark came up with the poetic label “twilight moments” to identify practices relating to sexuality that are not openly acknowledged--and certainly not accepted--but that are treated as temporary and “recoverable” lapses, even if they are lapses that happen repeatedly and throughout a society. She opens with the example of the Codrington divorce trial in 1863 (the subject of Emma Donoghue’s novel The Sealed Letter) in which the accusation that Helen Codrington had engaged in some sort of sexual relationship with her close friend, feminist Emily Faithfull, was simultaneously raised and hidden under layers of misdirection. While the nature of the accusation was never made explicit either in the trial record or in the gossip surrounding it, both women were forced to withdraw from society for a time, though Emily Faithfull’s reputation was eventually rehabilitated.
Clark points out that this incident and its consequences don’t fit with Foucault’s timeline of the understanding of sexuality. Per Foucault, in the 1860s, it should not have been possible for people to conceive of a “type” of woman who would be expected to have lesbian tendencies. And yet Emily Faithfull--perceived as “mannish”, suspect for her feminist activism, and with a close emotional relationship to Helen Codrington--was treated as if such a “type” clearly existed in people’s minds. Rather, Foucault’s theory would have it that, in the 1860s, people would only have the concept of specific stigmatized acts--and yet no such acts were in evidence, nor were they named in their absence.
Based on this and other similar contexts, Clark asserts that we need “additional conceptual tools to discus cases...which involved sexual desires, relationships, and practices that did not produce identities, that were half-understood, expressed only by oblique gestures, veiled in silence.” She notes that just because a desire was erased from public discourse didn’t mean that it wasn’t subject to social discipline. But the consequences of these “twilight desires” might be as nebulous and elusive as the understanding of the desires themselves. She uses the metaphor of twilight--of something that is neither completely hidden in darkness, nor seen in the light of day--to describe these desires and activities that people experienced and pursued, perhaps as an open secret, and that were not considered to result in a fixed identity, but that were still disapproved and might result in (temporary) negative consequences depending on context. Those consequences would involve shame or social opprobrium, but did not change people’s relationships or social standing on a permanent basis.
Such “twilight desires” might be considered normal or natural--such as the expectation that unmarried people might still desire sexual fulfillment--without being “approved” or “moral.” And society might go to some length to accommodate them, despite that disapproval, as with attitudes towards prostitution, based on a belief that some sort of outlet for men’s sex drives was required to maintain social stability. That doesn’t mean that these “twilight desires” all involved victimless activities. In many cases, it covered not-entirely-consensual acts involving power differentials, such as interracial sex with enslaved women, or pedophilia involving class differences. The key was that the consequences of the act were not considered to disrupt the social structure and so could be ignored.
Clark notes that there have recently (as of 2005) been some cracks in the Foucaultian edifice from authors such as Halperin who argue that some forms of “sexual identities” can be identified in the pre-modern period, such as medieval concepts of “the sodomite” as an identity. Or Ruth Karras’s argument that the category of “prostitute” was treated as a distinct sexual identity in some medieval contexts.
But Clark’s concept of “twilight moments” specifically excludes desires, practices, and relationships that were considered to constitute an identity, whether public or private, licit or stigmatized. And she notes that such desires/practices/relationships could shift between “twilight moments” and fixed identities as society required. Sometimes prostitution and sodomy might be treated as ignorable lapses, and then society would require a scapegoat for some disaster and might shift prostitutes and sodomites into the required slot.
Clark spends several paragraphs examining the concept of sexual desire and its relationship to “sexual scripts” -- scenarios that a particular culture authorizes for the experience of desire. These scripts are prescriptive and are learned from one’s culture. Even “forbidden desires” may be learned via established scripts. But the established scripts don’t always account for the possible range of personal experience (or even for common experiences, as when no allowance is made for sexual desire outside of marriage). Existing scripts may be adapted for alternate contexts, as when Romantic Friendship borrowed the terminology of marriage, with female partners calling each other “husband” and “wife”. [Note: I think this is a key insight when considering the intersection of women's same-sex desires and the practice of gender-passing. If a particular cultural context doesn't have a "script" for desire between women, and women who experience such desire borrow a hetrosexual "script", there will be no fixed guidelines as to how much of that borrowed script should/must be performed.]
Anxieties about sex that crossed barriers of class, gender, or race were most likely to come to the fore during times when other social boundaries were perceived as being eroded. It was during these times that the allowances made for “twilight desires” were more likely to be interpreted as creating fixed identities, and those identities persecuted. But an allowance for “twilight desires” could also help to prop up dominant power structures by diverting illicit acts into a context where they had no power to destabilize.
The “twilight” concept is compared to other ways of framing non-normative behaviors and social categories, such as stigmatized economic activities (prostitution, collecting human waste) that created fixed identities but were recognized as socially necessary. Another related concept is that of identified geographic spaces in which illicit activities were tacitly permitted (such as “red light zones” for prostitution, or neighborhoods where interracial socializing was permitted within segregated societies). As long as the behavior was restricted to the designated location, privileged people could visit those locations without alteration of their public status and identity.
Concepts that are less similar to “twilight desires” include that of subcultures (which do imply fixed identities) and allowances for otherwise illicit activities during liminal states of age or status (which implies a linear transition, not an ability to move in and out of the twilight space). But again, specific behaviors might be recategorized among these concepts over time or with changes in the larger social context. So, for example, an age-based allowance for same-sex activity among young men in early Renaissance Florence might move in and out of “twilight” status depending on the larger political context and the need for moral scapegoats.
Although many allowances for “twilight desires” were solidly gendered and accessible only to men, Clark notes early modern contexts where even strict notions of sexual honor imposed on upper class women might be treated as twilight events that need not result in a fixed status as “dishonored” or "ruined" if the proper forms were followed for maintaining the appearance of honor (including prompt marriages or the quiet disappearance of obviously illegitimate pregnancies).
The concept of “twilight desires” can also apply to language around those desires, where disapproved activities could be ignored as long as they were not publicly named in so many words. This could include censoring of explicit sexual material in popular medical literature, or backlash for victims of sexual crimes if they made accusations in undeniable language. Disapproved sexual practices could be tacitly ignored if discussed only in terms of “exotic tastes.”
Such tactical silences could even allow for people to deny the reality of acts they had participated in, if that reality were never expressed concretely. Participants might claim that they didn’t understand what their sexual partner was doing, or that they didn’t know that it was forbidden. Similarly, communities might be well aware of the illicit activities taking place among them, but only “discovered” them when some event made it impossible to ignore, or when there was an ulterior motive for taking notice of it.
Taken together, all these phenomena can help understand how identity categories can coexist with deniable “twilight moments” across a long span of time, making it possible to both deny and prove the existence of sexual identities during transition periods. This is particularly the case with women’s same-sex relationships, where the wide variety of framings for romantic/erotic desire made it possible to practice and represent those desires according to the needs of the moment, for good or ill. Thus, in the mid 19th century, it was possible both to deny the existence of lesbianism and to use the suspicion and stereotype of it as a social weapon.
(Clark includes many examples of “twilight desires” that are only relevant to heterosexual contexts which I’m glossing over here.)