1. Introduction

This paper is a critical engagement with Heiko Motschenbacher’s Linguistic Dimensions of Sexual Normativity (2022), which examines the discursive construction of sexuality through the lens of normativity. Relying on corpus and discourse analysis methods, the work makes no claim to being a diachronic study of the lexicon. However, it both is underpinned theoretically by and aims to empirically test, in the corpus data, ultimately lexicalist positions: that modern sexual normativity was triggered by the creation of late nineteenth-century labels (novel Foucauldian argument), and that those labels among others continue to determine the contemporary sexual normativities of individuals, in the form of identities (application of labeling theory). In this frame, lexicogenesis is held to be at the core of heteronormativization – an argument that thus invites a historical lexicological reading.

I am in agreement with the author that study of the lexicon is a central task for linguistics in problematizing how lexemes (or labels) relate to social meaning, including sexual norms. After all, the lexicon is the one linguistic domain where meaning precedes discourse, so it makes sense that collective histories of lexical use may inform and be informed by norms. Yet the inherently diachronic nature of studying the lexicon means that it must employ the appropriate methodologies, starting with rigorous sociolexical historiography of which corpus methods serve as only one potential tool. In Lo Vecchio (forthcoming), I lay out some general principles for a sociolexicology of queer naming, focusing on structured lexical variation across time and space and centering the agentive lexical practices of language users (and their metadiscourse). This companion piece now applies those principles more specifically in offering a sociolexicological response to the approach adopted in Linguistic Dimensions.

2. Overview and theoretical basis of the book

In line with the author’s previous appeals to bring more empirical study to queer linguistics (Motschenbacher, 2011), the bulk of Linguistic Dimensions of Sexual Normativity is devoted to deploying corpus linguistics as a means of empiricizing queer linguistic data, via seven case studies of several national varieties of English.1 Critical discourse analysis supplements the corpus methods, resulting in a hybrid of both quantitative and qualitative analysis of queer sexuality-related language use.

Three of the case studies use or adapt existing corpora (Chapter 6, on sexual labels, uses the Corpus of Contemporary American English, or COCA; Chapter 7, also on sexual labels, draws on a Sketch Engine corpus; a custom sample is extracted from LexisNexis for Chapter 8, on discourse related to Ricky Martin’s coming out). More innovatively, four chapters employ custom historical corpora, created specifically for the author’s LIDISNO project, which involved the digitization of a sample of prose publications from the nineteenth and twentieth centuries (Chapter 9, on language use before and after Stonewall; Chapter 10, on gay men’s metalinguistic comments pre-Stonewall; Chapter 11, on sexual speech acts; Chapter 12, testing the supposed 1890s “desire-identity shift”). These are bookended by chapters focusing on introductory, theoretical, and methodological issues (Chapters 1 to 5, 13). The concluding section (Chapter 14) departs argumentatively from the rest of the book in proposing guidelines for inclusive language use, with indications on particular LGBTQ or gender labels.

Two theoretical currents run through the work, simultaneously underpinning the argumentation and in the aim of testing it empirically. One is the well-known premise, attributed to Michel Foucault, that there exists a then and a now distinguishing past conceptualizations of sexuality from the current one, following a discursive rupture in the nineteenth century: this is the so-called “desire-identity shift”, which is assumed throughout (LDSN, pp. 5, 13, 29, 30, 145, 206) as well as tested specifically in Chapter 12. The other is the linguistic application of sociological “labeling theory”, which is developed at length in Chapters 4 and 13.

2.1. Foucauldian paradigm shift

Foucault’s philosophy has famously structured several generations of scholarship on sexual theory and LGBTQ studies, and could now even be said to form part of the doxa.2 The treatment in Linguistic Dimensions merits commentary. From a lexicological point of view, the main concern is that the presentation of Foucault’s theory is erroneous in lexical terms while also attributing world historical importance to its premises.

In the introduction, Motschenbacher notes that the “conceptualization of sexuality in terms of identities has, according to Foucault (1978 [1976]), evolved fairly late, predominantly as a consequence of the creation of the terms heterosexual and homosexual in Western medical discourse at the end of the 19th century” (LDSN, p. 5). The author further elaborates:

A classification of people into sexual types, however, was unknown before the 19th century. This does not mean that sexuality was not talked or written about, but that this was done in different ways than today. What was highlighted at that time was not what kind of sexual being somebody was, but rather language was used to express sexual desires. (LDSN, p. 5)

While this sort of qualification is fairly common in sexuality studies, it should be contextualized from a lexicological perspective, because something similar could be said about all abstract categories, discursively framed “in different ways” over time. Appending this statement to discussions of sexuality, yet not to all other historical phenomena, entails applying a distinct epistemological criterion. In contrast, lexical researchers have been careful to question assumptions about the transhistorical existence of more banal categories – for example, of “parenthood” or “anger” (Wright, 2023, pp. 17, 65-66, citing Nevalainen and Diller).3 Additionally, this presentation assumes homogeneous discursive practices now and in contrast to the past, though the recourse to agentless passive-voice wording (e.g., “language was used”) and a broad yet unspecified timeframe (“at that time”) makes it difficult to situate those practices in time and space.

In some places, (hetero)normativization is described as a kind of autonomous macro-level process, although the individual linguistic mechanisms driving it are unaddressed beyond the attribution to Foucault: “As discussed by Foucault (1978 [1976]), the normativization of sexuality gained momentum at the end of the 19th century in the Western world” (LDSN, p. 13); “Foucault (1978 [1976]) has shown that heteronormativity itself is a fairly recent discursive formation that gained greater momentum only at the end of the 19th century, when terms for sexual person types were created” (LDSN, p. 29). In one case, the Foucauldian “shift” is described as being merely “believed” or “assumed” by researchers (LDSN, p. 145). In the most explicit case, Motschenbacher writes:

The desire-identity shift in the conceptualization of sexuality has been famously postulated by Foucault (1978 [1976]) in his History of Sexuality. Foucault discusses how the creation of the terms heterosexual and homosexual in the medical field of the 1890s has caused a shift in the way sexuality is (predominantly) conceptualized. Whereas before this time, sexuality was commonly framed in terms of what people did and desired sexually, the medical treatment of the subject has initiated a conceptualization in terms of certain sexual person types (today widely known as “sexual identities”). (LDSN, p. 206)

This presentation distorts Foucault’s argument. Since no page numbers are cited, nor is the translator specified,4 it is hard to know which passages in Histoire de la sexualité Motschenbacher is referencing. The philosopher’s hypothesis does not articulate a “desire-identity shift”. Foucault used French homo-/hétéro- terminology minimally, without metadiscursive commentary and without mention of their “creation” in the 1890s. Instead, Foucault attributed, almost at random (1976, p. 59), the birth date of modern homosexuality – as a concept – to the year of publication of Westphal’s psychiatric paper (1870 [1869]), in German, which did not use any word from the homo/hetero5 series but instead the coining conträre Sexualempfindung (“contrary sexual sensations”, in Hurley’s translation), which would serve as the basis for the later innovation sexual inversion – first in Italian, then French, then English. (The lexical implications of this will be discussed in further depth in Section 3, infra.) The citations presented here announce a new and novel lexical argument over and above Foucault’s texts, more specifically by applying Foucauldian notions to the English language. This would have merited in-depth theoretical engagement with Foucault’s writings and the body of work they have generated since.

Then there is the issue of the empirical basis of the “desire-identity shift” attributed to Foucault. If such a conceptual shift occurred, it must be describable based on language practices as visible in the historical documentation – a question the author poses in this work. Motschenbacher rightly notes that the “linguistic consequences” of the desire-identity shift “have never been empirically studied”, clarifying that, “If indeed a conceptual shift has taken place, traces of it should surface in the way people use language to talk and write about sexuality” (LDSN, p. 207; see also LDSN, p. 49). In my view, it would be more accurate to state that, if the shift has taken place, it can only exist in the way people talk about sexuality: by definition, a discursive shift occurs in discourse. In other words, there are not “linguistic consequences” to the shift; the linguistic material is all there is, out of which the conceptual shift should have been, in principle, deduced. Yet indeed this claim has never been interrogated linguistically.

So Chapter 12 sets out to test the desire-identity hypothesis seeking evidence in a custom corpus created using Anglo-American prose narratives pre‑ and post‑“shift” (1830-1899 vs. 1900-1969). The results are mixed: “the three classical corpus linguistic methods of comparing two corpora [keywords, key parts of speech, key semantic categories] failed to verify the shift, while an in-depth analysis of a specific semantic category yielded convincing evidence for it” (LDSN, pp. 223–224). The overall conclusion is that “the basic premise of the shift can thus be confirmed in light of the linguistic evidence” (LDSN, p. 223). In this discussion, Motschenbacher justifies why some corpus methods failed and why others succeeded:

We can, therefore, interpret the fact that we had to dig deeper to find linguistic patterns suggestive of the shift in this earlier period as another piece of evidence that it was the events revolving around Stonewall and the LGBT rights movement that caused the shift to reach its full effect. (LDSN, p. 224)

Likewise, in Chapter 9, when the data “challenges the traditional view that the desire-identity shift […] took place at the end of the 19^th century”, the author reasons that the shift must have begun then in “medical discourse” but only took “effect […] in everyday language use” after Stonewall (LDSN, p. 161). Leaving aside methodological issues, we can see that some of the data in two case studies actually contests the hypothesis of the Foucauldian shift; yet the author’s conclusions appear to reinterpret the data to fit the teleology in a novel way. In this reading, an autonomous normativizing force, initiated in the 1890s in pathologizing out-group discourse, “reach[es] its full effect” at the outset of queer liberation in in-group speech. Discourse patterns are exterior to the individual subjectivities languaging them. A simpler way to interpret these findings, detached from any extradisciplinary theoretical framework, is that identity expression is discursively salient at and following one momentous event in modern LGBTQ liberation.

There remain methodological questions about how the corpus studies were conducted, or what they make possible to infer. The “shift” corpora studied contained 4 million tokens in total, distributed between 66 British and American prose sources for the nineteenth and twentieth centuries – all told, a minimal data set to test a major claim. Is it valid in sociopragmatic terms to compare such vastly differing textual/social realities while assuming common discursive practices among them, and then extrapolating out to generalize for the English language as diasystem? The corpus is made up of love and sex prose narratives; medical texts, representing the type of discourse addressed in Foucault (1976), were not included. In the spirit of Foucault’s hypothesis of the discursivization (mise en discours) of sexuality,6 the import of the second half of the nineteenth century was more the fact that particular people were talking about sexuality in particular ways – not all people, but a distinct diastratum in the Judeo-Christian West, discoursing in diaphasically circumscribed conditions, who had the institutional power to influence how dominant normativities were publicly framed: the white male medical establishment above all. Due to the sheer volume of medical texts published on the subject, the early 1890s do represent a key moment in the discursive construction of sexualities in some Western societies. However, this does not mean that all people in all (Western/European) societies shifted their discursive practices at once – an assumption Motschenbacher’s corpus methods seem required to make.

It is surprising that “Great Paradigm Shift” thinking has endured so long, considering how Eve Kosofsky Sedgwick already so eloquently critiqued the utility of postulating any “unidirectional narrative of supersession” because it “obscures the present conditions of sexual identity” (Kosofsky Sedgwick, 1990, pp. 44–48) – which can hardly be conceived of as discursively homogeneous any more so now than in the past. Linguists more than anyone ought to be attentive to the variation in sexual discursivity across speakers and situations. The heterogeneity of the corpus data studied in this book may be explained more simply, to cite Motschenbacher, by acknowledging “that sexual labels are not just used to categorize sexual identities but also other sexuality-related aspects such as desires, sexual practices, romantic relationships, political actions and academic debates” (LDSN, p. 227).

Citing Foucault has become de rigueur for many sexual or cultural theorists, to such an extent it may now be something like an “article of faith” (Eribon, 2012 [1999], p. 466, see also p. 19). The treatment in Linguistic Dimensions does not repeat Foucault exactly, or really even cite him, let alone engage with his ideas. It basically name-checks him, while building its argument around received notions attributed to him. A more convincing approach would have critically heeded Eribon’s observation: “Il ne s’agit pas de répéter Foucault, mais de l’utiliser”7 (Eribon 2003, p. 62).

2.2. Labeling theory

Adapted from Anglophone sociology, labeling theory locates deviance, and thus normativity, in the discursive act of labeling (or naming, denominating, categorizing) rather than in people or their behavior. The main idea is that “labeling constitutes a practice that does not describe a reality but creates it” (LDSN, p. 34), with concomitant effects in how people view themselves or behave due to those very denominations. Motschenbacher acknowledges that labeling theory is a somewhat dated concept borrowed from an earlier era of sociology, noting: “In sociology, it had its heyday in the 1960s and 1970s, at a time when same-sex sexualities were still considered pathological and deviant” (LDSN, p. 31). Despite its name, the theory “has so far not drawn on linguistic insights, even though it deals with an issue (‘labeling’) that is ultimately a language-based activity” (LDSN, p. 30). A main aim of the book is thus to dust the concept off and give it new life in linguistics, in large part by reframing it: viewing “labeling” not “as a deviance-producing activity but as a normativity-inflicting process” (LDSN, p. 38).

A discussion of labeling theory is not, in itself, irrelevant to linguistic theorizing of sexual categorization. Lexical categories certainly both generate and are generated by the social reality in which they are uttered – an idea compatible with, even indispensable to, modern queer or sexual theory, such as the Foucauldian insight that what makes sexuality “modern” is the very fact of its discursiveness (mise en discours as historicized processes) as well as the Butlerian insight that gender and sexual orientation are not just performative but performed discursively. However, several aspects weaken Motschenbacher’s application of labeling theory to queer sexual normativities.

First is the minimal connection of the theory to the corpus studies (see LDSN, pp. 140–141, 180) and, in reverse, the use of the corpus studies to back up the theoretical chapters (see LDSN, pp. 227–228). Presented as such, it feels more like an attempt to fit the data to the (preformulated) theory, rather than the other way around. Due to this treatment, one may question whether textual corpus analysis is the heuristically appropriate way to demonstrate what is fundamentally a macro psychosociological theory on the relationship between how people view themselves and how they name themselves, both as individuals and as collectivities. Corpus methods may be able to “shed more light on labeling from a linguistic point of view” (LDSN, p. 40) as in the case studies, but they have little to say – as presented here – about the psychosocial mechanisms that would substantiate the theory.

Sociolexical historiography has challenged assumptions that medical labels drove the creation of modern queer identities. In the next section (Section 3, infra), a number of examples will show how queer identity labels preceded medicalized ones, even as they were coopted by the medical establishment. Another author, not referenced in the book, long ago acknowledged this in arguing against labeling theory as applied to sexual normativities, in historical studies drawing on a rigorous queer philology (Chauncey, 1982/1983, 1985). In one study, Chauncey held that “[s]uch assumptions attribute inordinate power to ideology as an autonomous social force” (1982/1983, p. 115). In another, that author noted the “exaggerated” role of medicalization as emerging from a textual study of a military incident in Newport, Rhode Island (1919‑1921), showing how inoperable the medicalized identity framework was both among a working-class queer underground and among the military hierarchy (Chauncey, 1985, pp. 203–206).8 “[I]t also suggests how we might begin to refine our analysis of the relationship between homosexual behavior and identity” (Chauncey, 1985, p. 203) – a statement still applicable four decades later.

Most significant is the negative frame through which Motschenbacher’s labeling theory proposes to define normativity. This is hard to reconcile with self-identification in an era when queer and trans liberation is so tied to emancipationist and self-determinative language practices. The author is aware of this limitation, stating: “Future work should try to extend labeling theory to study not just negative, stigmatizing labels but also positive ones” (LDSN, p. 36) – yet it is only briefly addressed in this book. In a section on the profusion of “self-labeling practices” (Section 13.3) where Motschenbacher acknowledges the difficulties affirmative in-group practices present to the theory, the author essentially reduces all social categorization to normativization, as if new norms were inherently harmful:

At the same time, the new labels create their own new standards for inclusion and exclusion and, connected to this, normative expectations about what a member of a certain category typically is (or should be) like. What needs to be viewed critically is that they still adhere to the notion of sexual identities and that their use is, therefore, likely to create and uphold sexual normativities. (LDSN, p. 233)

Motschenbacher also notes that the “proliferation of sexual identity categories, therefore, contributes little to the deconstruction of sexual identities” (LDSN, p. 233); elsewhere, “sexual identities” are said to enforce “normative regimes” (LDSN, p. 81).

An alternative approach would foreground the liberatory, self-determinist, collectivist aspect of in-group naming and neology, granting queer people – or any social group – equal linguistic agency in asserting their place in the social structure. This is not to naively deny the normatively oppressive forces that may oppose minoritized groups, but to see their constant neologizing and metadiscourse practices as intrinsically linked to their status as social actors within the “linguistic marketplace”, in the sociologist Pierre Bourdieu’s well-known terms (Bourdieu, 1982, pp. 99–100, 142; see also Boutet, 2016 [2010], pp. 85–89; Elchacar & Salita, 2018, pp. 140–141; McConnell-Ginet, 2020, pp. 8, 24; Lo Vecchio, forthcoming). For “labeling” queer people does not refer to a discrete reality so much as it reflects a multidirectional power relation based in heterocisnormative social hierarchy.

3. Treatment of lexical history

A central claim of Linguistic Dimensions of Sexual Normativity (and the LIDISNO project more broadly) is that corpus methods make it possible to empiricize queer language data. But corpus methods are not the only way of empirically studying the discursive construction of sexuality. Indeed, the lexicalist arguments advanced in this book require qualitative philological substantiation. If the creation of two lexical items altered the course of history, themselves determining what the book sets out to prove – discursive normativity – it would have been necessary to operationalize a mechanism to demonstrate how lexical creation translates into a predominant conceptualization shared at the societal level.

Lexical historiography plays an indispensable role, by tracing word histories from their attested origins and through subsequent spread, informed notably by explicit intertextuality and metadiscourse commenting upon lexical creation and use in real time. Historical dictionaries are a requisite starting point – in particular the Oxford English Dictionary for the language under study. However, no dictionaries or lexicological works are cited in this book, which gives the impression that word histories were accessed through introspection. Of course, dictionaries themselves participate in the creation of linguistic (and societal) norms, as their human authors are subject to the same ideological pressures as the rest of society. Any use of the reference dictionaries must involve a critical metalexicography, acknowledging their importance in empirical research while also keeping in mind potential errors, distortions, or omissions – particularly important for queer lexis (Nossem, 2018; Turton, 2020, 2024; Lo Vecchio, 2021). Independent consultation of all historical corpora (especially untagged archives such as Gallica, Google Books, or Hathi Trust) thus acts as an independent check on the lexicography, as does targeted research in primary and secondary sources.

As study of the historical documentation will demonstrate, the late nineteenth-century lexis of queer sexualities and genders proves extraordinarily complex, constituting a space of neological experimentation bridging law, medicine and the (pseudo)sciences, literature, and incipient queer activism. Meanwhile, reading Catholic Church doctrine or government policy in some places (e.g., “sodomy laws”), then and later, can be a shock lesson in translinguistic discursive continuity – not change – whose textual basis can be traced back centuries to censorious medieval ecclesiastic discourse in Latin and its translation to the vernaculars. To focus on only one domain, medicine, raises the question of what legitimates privileging this one over others. There is, to be sure, an argument to be made about the major role that medicalized discourse, and more specifically its lexical use, played in consolidating sexual norms across fin-de-siècle Europe and the West. This cannot be assumed, without substantiation from the textual documentation. Nor can it be detached from the voices of queer people who not only dialogued with the medical profession but were – in the most significant of ways – actually responsible for some of that lexical creation themselves.

Here I will address a number of problems with Motschenbacher’s historical assumptions from a lexicological perspective, focusing on the late nineteenth century along with a few more recent examples. All details and examples described below may be found in the corresponding sections in Lo Vecchio (2020, 2021).

3.1. Chronology

The homo/hetero series in all European languages can ultimately be traced back to the German, where the earliest known uses date to 1868-1869, with subsequent attestations in the decades following. Other European languages then began to adapt the homo/hetero series starting at the tail end of the 1880s, and the 1890s was indeed decisive in terms of its diffusion in some languages. The book’s chronology, limited to the 1890s, raises questions of Anglocentrism head-to-head with linguistic determinism (whereby the linguistic creation of a word determines the extralinguistic reality related to the concept referred to). If a desire-identity shift occurred in the 1890s with the “creation” of these terms in English, how do we problematize – to take just one example – the discursive normativity for German-speakers who had been using Homosexualität, Uranismus, conträre Sexualempfindung, and related terms in the three decades prior?

In the chapter that tests the “desire-identity shift”, the author explains that the period 1890‑1900 was chosen as the “historical dividing line […] allowing a ten-year period at the end of the 19th century, in which the newly created terms heterosexual and homosexual had a chance to make their way into everyday language use” (LDSN, p. 208). Here, the “everyday” use of these words is assumed. No historical attestations are cited. This assumption also runs counter to the assumption, more central to the book, that these words began as medical terminology in the stated decade. If homosexual and heterosexual subsequently underwent a process of determinologization, as Motschenbacher assumes here, that is a second step that must be described upon the basis of attested uses.

3.2. Lack of discussion of other lexical items

The book argues for a lexico-discursively constructed conceptual shift based on the creation of homo/hetero terminology. However, this argument fails to take into account all the other vast lexicogenesis, in various languages, attested in the late 1800s. There were many other lexical items concurrent with or preceding the homo/hetero series that informed debates on sexuality, both in medicine and wider society, including among queer people themselves. For example, consideration could have been made of the numerous creations, starting in 1862, of prolific neologist Karl Heinrich Ulrichs, a self-proclaimed Urning (uranist) who saw themself as a third sex (drittes Geschlecht), a woman trapped in a man’s body. Ulrichs’s lexis of uranism competed with the homo and inversion series, all three of them widespread in various languages in Europe in the late 1800s; like the homo series, uranism began in activist in-group discourse and was coopted by science – telling evidence of their sociolinguistic complexity and contradictory to labeling theory. The translinguistic origins of sexual inversion, based on Westphal’s neologism cited by Foucault, also precede the international adoption of the homo/hetero series. The selection of only one lexical series, homo/hetero (in English), as the defining hinge point in the supposed discursive shift retrospectively reads current normativities and metalexical intuitions back into the past.

It may be legitimate to broadly consider the 1890s as a kind of turning point in some European/Western societies in terms of medical discourse on queer sexuality, yet attributing the creation of homo/hetero terminology to the import of this moment gets the lesson backwards: the emergence of this one particular set of terms is symbolic of, reflective of, but not determinative of the discursive dialogue then active, crucially in continuity with previous and later decades. Neither qualitatively nor quantitatively the most consequential set of terms at the time, but as the one to later largely supplant others (inversion, uranism alongside pederasty, sodomy, tribadism, sapphism, etc.), homo/hetero terminology in the 1890s only becomes salient in hindsight.

3.3. Lexical origins and use between queer and medical discourse

In German, homo/hetero terminology did not originate in medicine. It is first attested in the field of law, by an author, himself identifying as a Normalsexualer (heterosexual), who penned open letters to a Prussian minister pleading in defense of homosexual rights. In English, early attested uses are found in medicine in the 1890s. Yet they are also found, concurrently, in the writings of early gay activist John Addington Symonds, who had great lexical influence as he corresponded and collaborated with the physician Havelock Ellis on Sexual Inversion, the most important sexological work in English by that point (see Crozier, 2008; Brady, 2012; Regis, 2016). Symonds’s memoir does figure in one corpus in the book, but apparently no use was made of it besides to retrieve a data point (LDSN, p. 201).

In-group self-referential use of all the terms mentioned so far, including homo words, is observed in the textual record throughout the twentieth century (see also Turton, 2020, pp. 225‑228; McConnell-Ginet, 2020, pp. 8‑9). Motschenbacher does acknowledge this briefly, explaining such early in-group use of homosexual as arising “probably for want of a better term” (LDSN, p. 160); compare also, much closer to the present day, how the author struggles to understand Ricky Martin’s self-referential use of homosexual when he spoke publicly about his coming out (including this contestable overgeneralization unrooted in the corpus analysis itself: “The term homosexual is nowadays generally used as an outgroup label by heterosexual language users who describe gay people in a distancing fashion” [LDSN, p. 137]). Past or present, the lexical item is analyzed through a negative prescriptivist lens (consonant with the book’s concluding chapter; see infra, Section 4.1). From a sociolexicological point of view (see infra, Section 5), it is possible to see much more complexity in self-referential use of homo terms, including today, notably in languages other than English, even in an emancipatory activist context.

To again stress its importance, the uranism series originating with Ulrichs was a militant queer creation. Now that new historical documents have come to light or been interpreted more incisively in recent decades, we know the significant if uneasy role that Ulrichs played in inspiring Krafft-Ebing’s theories amid their long-running dialogue (Kennedy, 2002, Chapters 5, 11, 12; Eribon, 2012 [1999], pp. 425‑428), just as Symonds’s (and Edward Carpenter’s; see Rowbotham, 2008, Chapter 10) views shaped Ellis’s. A more nuanced, and lexically justified, argument could be made that it was the asserted queer subjectivities themselves on and against whom the medical establishment based its new theories and terminology, not – as labeling theory would have it – that queer subjects found their identities only after the labels, in part created by themselves, were refiltered through the medical gaze.

3.4. View of “modern” homo/hetero conceptualization

Even within the international medical sphere, none of the new nineteenth-century terms had one fixed meaning tied to a single dominant paradigm. Instead, they were used variously, interchangeably or synonymously, in the cacophonic (pseudo‑)scientific discourse as new conceptions – plural – of same-gender behavior and love were then being defined. This profusion manifests in many ways, but is best exemplified by the conception behind Westphal’s original psychiatric coining conträre Sexualempfindung, since it is the fateful term invoked by Foucault to operationalize a starting date for homosexuality within the framework of his hypothesis.9

Westphal’s lexical creation did serve as the model for later inversion and was the only major neologism that actually began in medicine (in contrast to uranism and homosexuality). But Westphal’s paper was criticized for its incoherence – such as by Tamassia (see Beccalossi, 2012, pp. 52‑54), responsible for Italian inversione sessuale and, ultimately, the international spread of this series – and its legacy is much more lexical than conceptual. In the paper, Westphal explored two quite different case studies: a woman who sexually desires other women, feelings that began at a young age; a man who frequently dresses up as a woman, has sexual relations with women, and likes to cross-dress while with women. Westphal set out to describe these two distinct cases as a particular condition (Zustand) but concluded that they represented a possible symptom of some undetermined pathology (a view partially legible in the article’s title alone). Does this recognizably represent an identity-based conception of homosexuality, same-sex sexuality, “non-heterosexuality”, or whatever modern label you wish to apply? Yet Westphal’s idiosyncratic and barely recognizable conception was Foucault’s birth date of modern homosexuality.

Writing in 1976, Foucault may not then have known about Ulrichs’s pamphlets celebrating proto-queer liberation, or even if he did, he could not have known that it was the out Urning Ulrichs to whom Karl-Maria Benkert was writing on 6 May when he penned the earliest known use of German Homosexual (that letter was discovered later; Herzer, 1987). The in‑group/activist origins of “homosexuality” in 1860s Germany may not have altered Foucault’s thinking anyway, because he was making a cultural critique – not a linguistic argument – about homosexuality as seen as a medicalized construction, not in and of itself, but as, among other sexual realities, one particular social and, yes, dominant instantiation, within the Judeo-Christian European tradition. Yet Foucault also made clear that normative domination could not be seen as going unidirectionally from dominators (dominateurs) to the dominated (dominés), but arises out of a complex and unstable multiplicity of power relations (rapports de force) (1976, pp. 123-129, 132-135).

3.5. Speaker agency

While the author does “note that language users are not helpless victims” and that “[t]here is usually some space for agency” in their “discursive structures” (LDSN, p. 12), awkward wording in several passages minimizes the lexical creativity of agentive individual speakers.

Most seriously, Motschenbacher misrepresents gay lexical history as follows:

The term “gay men” is problematic in this context [pre-Stonewall], because, as we will see in the analysis, “being gay” was generally not an available discourse in the pre-Stonewall era. The phrase “gay men” is here therefore used as a (simplifying) shorthand for men, before and after Stonewall, who experienced same-sex attraction or engaged in sexual practices with other men. (LDSN, p. 146)

In fact, starting in the 1920s or 1930s in urban centers in the U.S., English-speaking gay people began adopting gay precisely because it asserted linguistic self-determination: it was an in-group innovation, created by gay people themselves, initially as a code word that simultaneously indexed queer solidarity and permitted dissimulation among the uninitiated. By the 1950s, gay was already widespread in the U.S.,10 attested even in out-group homophobic discourse, although after Stonewall it would be brandished more openly and proudly, to be sure. The self-deterministic in-group adoption of gay (among other terms) presents a challenge to labeling theory (see remarks to this effect pp. 36–37, 39, 233-234), but this ahistorical presentation of gay’s origins neglects the role of speakers in this major lexical development with global resonance.

One chapter is devoted to “folk linguistic” practices, cautioning there on the need to distinguish speakers’ popular perceptions from actual language use (LDSN, p. 163). Yet the author appears to cede to some folk notions of gay lexical history as described in the corpus (see LDSN, pp. 170–171, attributing the innovation to the 1960s; and compare to pp. 180–181, where Motschenbacher calls attention to possible 1930s use of gay but does not verify it, such as could be done in the OED). In one explication of a corpus data point (LDSN, p. 171), Motschenbacher endorses a narrator’s metalinguistic comment:

It is also noteworthy that the narrator does not ascribe the linguistic changes to the agency of gay men as social actors (The language changed). This points to a broader discursive development that is significantly co-produced by wider society. (LDSN, p. 171)

This statement is consonant with the later remark that the “individual language user is normally not able to create new means of expression” which “can only evolve as discursive formations, intertextually and across language users” (LDSN, p. 237). Ultimately, this is the crux of the matter: a book that argues for the lexico-discursive construction of normativity provides no coherent way to problematize the relation between individual speakers, their lexical creations, and societal norms.

4. Queer authority

Because this work both inscribes itself in a queer project and aims to counter normativity, it is legitimate to pose some frank questions about the political ramifications of its execution.

4.1. Norm enactment via prescriptive guidelines

Linguistic Dimensions of Sexual Normativity concludes with a chapter of “non-heteronormative language guidelines”, a set of fairly general principles proposed “for English” (LDSN, pp. 242–243). These do not follow from the detailed corpus analyses that form the central portion of the book. The motivation for concluding on this note appears, then, to be an activist impulse linked not argumentatively to the rest of the work but instead ideologically to the spirit of the project, holding “that the world should become better in certain ways” (LDSN, p. 238).

Motschenbacher explains that these guidelines “do not aim to intervene at the (structuralist) level of the language system” but at the “discursive level” to create a kind of “competition of discourses”, thereby arguing that they do not constitute “normative rules that tell people how they have to use language” (LDSN, p. 241). The stated goal is “to offer advice to language users who do not wish to support the status quo and want to contribute to social change” (LDSN, p. 241). All the same, a clear value system is established, framing their use as a choice of being either on the right side of history (“the world should become better”; “contribute to social change”) or the wrong (“support the status quo”). Thus, the argument that these are not, or are somehow less, prescriptive is unconvincing.

The first question that comes out of this concluding chapter is: Who are these guidelines for? Prefatory remarks qualify the recommendations as applying to “public language use” as well as to “private communication involv[ing] non-heterosexual, non-binary or trans-identified people” (LDSN, p. 243). Yet formulating broad guidelines clashes with the fundamental principle, acknowledged by the author (LDSN, pp. 241–243), that the “communicative realm” of application must be taken into account. Aside from contravening the principle that linguistics does not purport to tell people how to speak (even if all linguistic description itself participates in the creation of norms), the main disciplinary problem with these guidelines is that they decomplexify something extremely complex: language use in its social context.

Many of the recommendations may seem uncontroversial but, in broader context, are problematic. For instance, homosexual is described as “unacceptable” in part because it “originated in medical discourse and has retained pathologizing and, therefore, stigmatizing connotations” (LDSN, p. 249; see Motschenbacher, 2021: “‘homosexual’ is a bad word”). I have just explained why this origin is inexact. While clinical connotations of homosexual are certainly relevant to speakers, as noted in frequent metadiscourse, word origins themselves need not be the decisive factor when speakers determine what is and is not acceptable; instead, it is the axiological or sociopragmatic value as perceived by the linguistic community at the moment of use (McConnell-Ginet, 2020, Chapter 8).

More significant is the question of sociopragmatic context. Is homosexual “bad” or “unacceptable” in all “public use” and by all people? There are very many attested uses of homosexual – in various languages – including ones by self-identifying homosexuals past and present, that can be described more accurately and interestingly than in these disapproving terms. While some case-by-case exceptions to the guidelines are allowed (e.g., “within the framework of a more radical queer language policy” [LDSN, p. 242]; for “certain sub-communities of the LGBT community” [LDSN, p. 249]), the very fact of ignoring sociopragmatic context inherently assigns fixed axiological values to lexical items at the abstract diasystemic level and overlooks the extreme complexity of the “linguistic marketplace”, again in Bourdieusian terms, whereby it is rare that any given speech community or communicative situation will be perfectly delineated between public and private, in group and out, conservative and radical.

Other questions arise based on lexical structure. Diachronically, English homosexual was not created alone, but in tandem with heterosexual. If homosexual is unacceptable, why then is heterosexual acceptable? Motschenbacher does not proscribe heterosexual and in fact recommends and candidly uses it. Even though the claim about the homo/hetero medical origin is misleading, if heterosexual also began in medical discourse, shouldn’t heterosexual be banned too? No explanation is given for this incongruity, other than to state that heterosexual “remains the default term” (LDSN, p. 250). (Motschenbacher counsels “care” when using straight, due to its “slang” and “informal” connotations [LDSN, p. 250] – implicitly subordinating the latter to an undefined standard). Likewise, the homo- combining form is employed unproblematically: Motschenbacher uses it in homonormativity, homonationalism, homophobia. It would be interesting to hear why bound-morphemic use of homo- is deemed unhomophobic, despite its “origins”, but the question is not addressed. On the other hand, clipped homo as a free morpheme is “derogatory” and “should be avoided in general” (LDSN, p. 249). Such structurally inconsistent recommendations will offer little help to speakers who want some sort of pattern to rely on and they are incompatible with a rational sociohistorical description of the lexis.

Queer is described as “challenging”, noting that in-group users “have reappropriated the originally abusive term” (LDSN, p. 249). Again, the reasoning cited is the origin, which is incomplete: so far as the textual documentation allows, queer was created by queer people themselves as an in-group term of solidarity in the U.S. in the early twentieth century, with out-group uses showing up later in the documentation. The pertinent fact, once again, is not the origin, but the axiological value as perceived by the community; so, at the point that queer was rehabilitated in the early 1990s, the injurious use was salient, allowing for the political force of reclamation. Motschenbacher argues that queer is a “questionable choice for heterosexually identified language users” (LDSN, p. 249). Another linguist (cited elsewhere in the book) came to a different conclusion. McConnell-Ginet (2020) accurately notes the early history of the term: “although queer was a preferred self-designator for many gay men in the early twentieth century, it also was consistently used outside the target group to derogate” (McConnell-Ginet, 2020, p. 168). She more convincingly argues that as a straight woman she can now unproblematically use queer, due to the linguistic situation now, following considerable social change in the U.S. in the last thirty years: “Yet I do not need special standing to use the relatively neutral lesbian or the once slurring queer” (McConnell-Ginet, 2020, p. 166). In short, a description of language as a complex, evolving thing, with empirically based historical context and a metalinguistic comment not meant as verbal hygiene for some unidentified other, but merely speaking for herself.

Any set of prescriptive guidelines would be open to critique by people with differing views – most especially so in this dynamic, constantly renewing lexical field where speakers of many languages are in sustained metadiscursive dialogue and discord about the words they use to define themselves. This is a crucial point. If every linguist drafted such recommendations, no two would look alike. Upon what basis would any one rank above another?

4.2. Self-reflexivity and the politics of language norms

In choosing to step into the role of prescriber, Motschenbacher raises questions about the role of linguists as agents of social change. What is the basis of the authority claimed? Amid all the extensive contextualization within the guidelines (LDSN, Sections 14.1‑14.3), no mention is made of authority in language and the author’s relation to it, beyond the claim that the guidelines are not meant to be prescriptive (LDSN, p. 241).

A strong argument is to be made that linguists have no authority, by themselves, to prescribe language use. Any such authority, to take the case at hand, would come not from disciplinary expertise but from belonging as a queer person to an LGBTQ community: “Communities are the ultimate semantic authorities” (McConnell-Ginet, 2020, p. 242). Amid widespread language contestations pertaining to gender and sexuality across many societies, it is likely that some linguists will choose to engage with prescriptivist debates. It seems to me that two criteria ought to apply to any prescriptivist pursuit undertaken within linguistic disciplines: explicit self-reflection on the positionalized linguistic authority of the prescriber; and situating any prescriptions within a defined community of practice where that authority, rather than being assumed or imposed, is negotiated and collaborative.

A final point involves eliciting self-awareness about the role of the English language, and English-language researchers, in global information flows. A monograph in and on English, overwhelmingly drawing on Anglophone sources (99 percent of the vast bibliography), written by a prominent researcher in queer linguistics, reporting on a publicly funded research program, published by a major multinational scientific publisher, and susceptible to being cited widely for all of these reasons, cannot set out to impose unsituated new linguistic norms (“guidelines”) “for English”, no matter how well intentioned, without some authorial consciousness of how such an undertaking relates to the contested predominance of standardized Englishes in the most varied of situations across the globe.

A critical task for fellow Anglophone researchers of queer linguistics will be to engage in self-reflexivity about the political ramifications of the monolingual epistemologies and methodologies that still inform much academic production and impact.

5. Sociolexicological perspectives

Outside of the hypothetical, static, top-down, Western-centric discourse shift paradigm (re)affirmed in Linguistic Dimensions, outside also of any prescriptivist frame, lexicology can look at queer sexual categories as linguistic signs, which signify in the same way as any other linguistic signs: that is, subject to structured variation in time and space and across language users and situations.

Thus, a sociolexicology, which acknowledges no essentialist vs. constructivist binary – all lexical categories are socially constructed! – but accepts that the same epistemological criteria must be applied to queerness (however defined) as to any other phenomenon; and that all conceptualizations of all abstract phenomena may vary across speakers and groupings of speakers both synchronically and diachronically, even if some of those are more prototypical or, if the analytical framework demands, normative. Rejecting truth-conditional semantics that would tell us what the true-false conditions are governing when it is valid to say, for instance, homosexual or when it is valid to say queer, a sociolexicology is rooted instead in prototype semantics that allow for describing variable meaning in terms of a gradient of family resemblance and that also, crucially, take into account ideological or sociopragmatic factors governing use (e.g., homophobia vs. in-group activism). Rather than dismiss the lexicon as trivial to language structure, as has been the case in some sociolinguistic traditions (Lo Vecchio, 2024), a sociolexicology asserts that lexical meaning does not reside solely (if at all) in objective entities and their supposed properties, but in communities of speakers creating that meaning themselves (Bourdieu, 1982; Boutet, 2016 [2010]; Wright, 2023).

By their nature, corpus methods abstract away from the individual, away from the speech communities, away from the margins in search of larger patterns. It is not possible here to address the innumerable theoretical questions surrounding the compatibility of corpus linguistics with the study of queerness or other pragmatically marked fields, these being historically characterized as much by absence in the documentation as by the extralinguistic biases expressed explicitly within it. In Chapter 5, Motschenbacher makes many astute observations about such challenges, offering warnings that constitute, in my view, the most important part of Linguistic Dimensions, very much worth reading for students of corpus linguistics (in line also with critical perspectives in McEnery & Baker, 2017; Baker, 2018; Taylor & Marchi, 2018; Larsson, Egbert & Biber, 2022; Wright, 2023, pp. 5, 9‑11, 167). The custom English queer corpora created for the project (Chapters 9‑12) are a promising endeavor worth expanding on (and hopefully the data sets will be made public). Yet, as detailed here, neither these corpora nor any other are alone suitable to empirically investigating queer language history – as the author indeed acknowledges in this chapter, leading to some dissonance between the theoretical warnings and the “empiricizing” corpus methodology adopted for the case studies.

The qualitative analysis in Linguistic Dimensions of Sexual Normativity is presented as a fundamental component, though it is circumscribed within the quantitative approach argued as providing an “important counterweight” (LDSN, p. 41); that is, it is the controllable, quantifiable corpus basis that apparently gives justification to the qualitative component. Reassuringly if superficially, computational methods feed the logic of scholarship-as-deliverable, especially alluring among funding bodies and institutional mechanisms of support. As we speed further into an era where some humans are eagerly ceding intellectual and moral authority to the machines for their supposed objectivity or neutrality, we must relentlessly resist the scientistic conflation of empiricism with quantification and automation – and queerness is as compelling a concept as any to demonstrate it.

By now it should be clear that all scholarly inquiry can be viewed through the prism of ideology and that no empiricist discipline, not even in the hard sciences, can be dissociated from its larger historicized social context, with all the attendant biases and privileges of the individual(s) doing the interpreting. Qualitative, most especially philological, research methods may remain stigmatized in some corners of linguistics, yet the queer view reminds us that the ever more sophisticated quantitative manipulations of language data are not necessarily any less ideologically suspect.