Kate O’Riordan. The Journal of Sex Research. Volume 49, Issue 4. 2012.
What kind of life does the gay gene have, and what does it mean to be addressed by it? This question is inspired by the persistence of the gay gene, its repetition across news media, and its new lease on life in online databases. The gay gene is not an object that can be examined under a microscope or visualized through imaging technologies. It has no biochemical makeup. It cannot be tested or experimented on. The original hypothesis by Hamer, Magnuson, and Pattatucci (1993), that there is a correlation between the X chromosome genetic marker—Xq28—and gay male sexuality, has never been replicated. However, it has a kind of life and it exists as an idea, a repetition, a discursive pattern, an emotional effect, a label, and a hypothesis. It appears in scientific journals, newspaper articles, popular science writings, databases, and Web sites. It has been featured on t-shirts and as poetry and song lyrics. It has become a biomedical artefact through a lively media materiality that has strengthened over time. At the same time that its presence has strengthened, the identity politics of sexuality have become depoliticized (Hennessy, 2000). This article examines how the gay gene seems to live by looking at its early reception in the 1990s, its presence in a selection of media forms, and its place one decade later in the early 21st century.
The core argument here is about the recursive, material, and lively nature of representation as a feature of biomedicalization (Clarke, Mamo, Fishman, Shim, & Fosket, 2003). The gay gene has gained a materiality of address through the persistence and aggregation of references, residues, and traces. When I say it has an address, I mean that it speaks to people as a media object in the world with the power to inform identities. Although this address primarily remained in print forms, its life might have been limited, but new leases on life have been offered by digital culture as it reconstitutes the gay gene through a technological structure of information management. Once subject to such indexes of biomedicalization, the gay gene, it seems, cannot be retracted, and even news of its failure to replicate in scientific studies becomes part of an aggregation of noise that increases its presence as a biomedical signifier. The more noise there is about the gay gene, including noise that deconstructs it, the more its signal strength grows. Clarke et al. (2003) argued that the redefining of old health-related identities, although with new techno-scientific modes of interaction such as the database, is a central dimension of biomedicalization. The medicalization of sexuality was challenged through the release of the civil rights movements and identity politics of the mid-20th century, only to be re-instantiated by the depoliticized databases and digital culture of the 21st century, and this article represents this process. It also provides a platform from which to demonstrate that the construction and linking of biomedical databases requires cross-disciplinary collaboration.
Research Questions and Reflections on Method
My research has followed the life of the gay gene since its identification by Hamer et al. (1993) through news media and online databases, and has examined one main trajectory: its instantiation as Xq28. I borrowed from the idea of the biography of the object, mainly from Celia Lury’s (2004) influential work on branding, to look at what might be thought of as the life of the gay gene since its first instantiation in 1993. Lury’s work on branding is useful in this case because the gay gene’s life could be thought of as having similarities to the life of a brand. The materiality and reality of a brand is similarly fraught. Following a brand as an object throws up similar challenges to that of the gay gene—namely, what is the “it?” How can an object that lacks a referent be followed?
To pursue this, I have followed the terms gay gene and Xq28 through U.K. news media and popular science writings. I used the online news database Lexis to conduct searches and to map the appearance of the key terms. I have followed all instances where these terms appear in the U.K. news archives from the mid-1990s onward. These stories have, in turn, identified further media in the form of popular science writings and biographies of scientists. I have also pursued the gay gene in online media, and here I have been much more selective. The radical proliferation of the term gay gene once it gets online is too unruly for this scale of research to map, so I have narrowed this trajectory to Xq28. What emerges from this is a story about the life of Xq28 annotated as the gay gene. In these stories about Xq28, homosexuality and gay are used interchangeably. For example, Science used the headline “homosexuality gene” about Hamer et al.’s (1993) work, whereas Hamer used the phrase “gay gene” in his 1994 book title. The Online Mendelian Inheritance in Man (OMIM) database, which I refer to later, gives this research the notation HMS1 as an abbreviation for Homosexuality 1, but these references are not used elsewhere. The terms of the object, then, are Xq28, homosexuality, and gay gene as they come together.
The gay gene also has an academic life in the arts, humanities, and social sciences, and I have followed other scholarship on the gay gene, which maps much of its life in the 1990s (Fausto-Sterling, 2000; Kitzinger, 2005; Roof, 2007; Rosario, 1997; Rose, 1995; Terry, 1999). Jenny Kitzinger’s work has been particularly helpful in thinking about the close relations between reporters and lesbian, gay, bisexual, and transgender (LGBT) activism in the 1990s, the sense of heightened interest in both sexuality and genetics, and the role of the science journals in promoting gay gene language. Particularly compelling is her conclusion that many of the problems attached to gay gene stories in the 1990s were not generated by journalists, but by press releases from scientific journals:
If we wish to problematise the gay gene as a concept then the media representation is not the main problem. The role of heterosexism and the place of science in society is a more crucial target for challenge; the press can, in some cases, act as a forum for exposing prejudice and pursuing critique and resistance. (Kitzinger, 2005, p. 117)
Anne Fausto-Sterling (2000) also noted, writing of the same period, that, although Hamer et al. (1993) were very cautious in their use of language, “Such caution did not however, extend to other pages in the same issue of Science” (Fausto-Sterling, 2000, p. 236).
Given the criticism and resistance in the 1990s to the gay gene as a concept in news media, and as a scientific entity in research and scientific commentary, the question remains as to how and why it is alive and well now. One factor is certainly this lack of caution to which Fausto-Sterling (2000) referred, which can also be seen as the use of sexuality to gather publicity for research. The readiness of multiple actors, including journalists, LGBT activists, and scientists to give the gay gene life has had a lasting impact. However, another important aspect is the changing context of identity politics and their de-politicization in the late 1990s through co-modification (e.g., Hennessy, 2000). In this context, the gay gene, unmoored from any original accountability, is undergoing a new lease on life through online databases and, as an object of digital culture, it threatens to be given indefinite license.
A final reflection on method relates to the nature of object. The gay gene is a controversial object. It is easily disavowed; of course, sexuality is controversial, and it might not be the best example to answer questions about broader issues. More compellingly for me, however, is the question of how a complex, but largely nonexistent, entity can be given a dimension of reality and life. What are the dimensions of such a life, and what does this tell us about how reality and science are made? The gay gene is important because it is a border case causing trouble at the edges of biomedicine and identity, and it is precisely this trouble that might help to illuminate some of the broader issues.
Gay Genes in News Media
Sexuality and the genome stories, which have been almost (but not quite) exclusively about gay men, have appeared in the U.K. press in a variety of forms since the early 1990s. The first study to make news was the now classic study by Dean Hamer et al. in 1993. Hamer announced the results of his genetic study of 40 pairs of gay brothers in the high-level journal Science, and published with his collaborator Peter Copeland in their book on the topic, The Science of Desire: The Search for the Gay Gene and the Biology of Behavior (see Hamer & Copeland, 1995), the following year.
However, although Hamer et al.’s (1993) study was the first to find a correlation to the Xq28 marker, thus making a clearly defined genetic association with homosexuality, there were also several other biomedical studies in the early 1990s that were connected to these results in news reporting. In 1991 and 1993, twin studies in the United States, carried out by Michael Bailey and Richard Pillard (1991; Bailey, Pillard, Neale, & Agyei, 1993), reported associations between genetics and sexual orientations (Terry, 1999). Bailey et al.’s (1993) study on homosexual women and twins also concluded that female homosexuality was “familial.” These studies were storied as news items in the U.K. press showing that there was “something” in the genes that was linked to homosexuality.
In the news articles at this time, the Xq28 genetic association study was also linked to Simon LeVay’s (1991) research, which had concluded that the size of a part of the brain (INAH3) was linked to sexuality. This was not a genetic study. He claimed that homosexuality had a “biological substrate” (p. xii). He published his book, The Sexual Brain (LeVay, 1993), the following year, and his review of science and homosexuality, which reprises his own research, Queer Science (LeVay, 1996). The way that stories about genetics repeatedly include these other biological associations demonstrates that, in the news in this area, there is significant slippage between biological research of any kind and genetic studies in their particularity. That genetics was over-determined in understandings of biology in the 1990s is reinforced by research in media studies and the anthropology and sociology of science (e.g., Herbert, 2005; Nelkin & Lindee, 2004). This slippage allowed brain research to be linked to genetic research in the news, but this association also extended to the other places in which genetic research was recorded, including catalogues. The readiness of connection between genetics, on the one hand, and biology in general, on the other, is one of the conditions that helps to give the gay gene a biomedical life.
The slippage between genetics as a particular area of study and biology in general in sexuality research is not limited to science news media. In the biomedical database OMIM, studies that are not strictly genetic (e.g., brain research) are also recorded in the same entry as HMS1. This demonstrates that the slippage between terms like genetics, heritability, familial, and biological is not only a feature of science news or popular science writings, but also of scientific databases, maps, and catalogues. This slippage, as others have noted, contributes to a depoliticization of sexuality as it appears as a biological and medical category. However, such biomedicalization generates a political question about how such changes in knowledge production occur and by what authority.
Of the scientific sexuality studies in the early 1990s, Hamer’s (1994; see also Hamer et al., 1993) and LeVay’s (1991, 1993, 1996) received the most U.K. press coverage. However, studies on brain size, chromosomes, and biochemistry continued to be linked to the gay gene in news stories about research by Blanchard and Klassen (1997), Camperio-Ciani, Corna, and Capiluppi (2004), Hu et al. (1995), Mustanski et al. (2005), Rice, Anderson, Risch, and Ebers (1999), and Savic and Lindström (2008). All of this research, regardless of its genetic dimension, appeared with links to the gay gene in news stories. Although Rice et al.’s work directly contradicted the Hamer et al. (1993) study by finding no correlation with Xq28, this was still reported as a gay gene story. In each new telling of subsequent sexuality research, references to the critique of the gay gene became diminished. As attention to the critique diminished, the idea of the gay gene conversely became stronger. In the reporting on later research in the early 21st century on the work of Camperio-Ciani et al., Mustanski et al., and Savic and Lindström, the focus on any critique of the gay gene disappeared and, in these new stories, it became a footnote to the research, part of the historical context, and, in some contexts, assumed as fact.
As the gay gene has gained more of a footnote status for later research, it has become assumed as fact in many instances of science media. One clear example of this is that of the news in the U.K. press of the “gay paradox” in 2004. In this case, Camperio-Ciani and Francesca Corna (see Camperio-Ciani et al., 2004), at the University of Padua, claimed that the female relatives of gay men were more “reproductively successful” (p. 2217) than the general population. This study was linked to the gay gene reporting and popular science books of the 1990s in news articles. There was a reprisal of all previous gay gene association studies in the reporting of the new study. However, in news of this study, an additional narrative link to the idea of the gay gene was further developed because the study was offered as an explanation for what Sam Jones (2004) of the U.K. broadsheet, The Guardian, referred to as “The ‘Darwinian paradox’ of gayness.” Jones’s article was entitled “Genes for Gay Men Make Women Fertile,” and claimed that a paradox “has puzzled scientists for decades. If male homosexuality is genetic, and homosexuals reproduced less than heterosexuals, the trait should eventually disappear” (p. 5). The addition of this paradox to the news media as a new item—framed as always already there—produces a narrative structure that depends on a gay gene for intelligibility. Thus, this new structure reinforces an assumption of the gay gene’s existence. Ironically, the use of this story to reinforce the existence of the gay gene echoes Risch, Squires-Wheeler, and Keats’s (1993) critique of Hamer et al.’s (1993) work in a letter to Science in 1993 in which they tried to refute that study.
Risch et al. (1993) suggested in their letter that finding a major gene linked to homosexuality was unlikely because of the selection pressures against its survival. This Risch et al. commentary about selection pressures (lack of genetic reproduction) is echoed in the 2004 reporting (Jones, 2004) with a new provenance, but this time as a paradox that has puzzled scientists for decades, rather than a speculative objection to Hamer et al.’s (1993) work. The flexibility of this reference to be reformulated such that a 1993 critique against the gay gene is used in a 2004 story arc as evidence for the gay gene is a feature of this area. It is through such decontextualized borrowings and intertextual shifts that the specificity of much of the critique of gay gene stories in the early 1990s has been lost and much of the assumption that it does exist has been gained. An imagined research arc where successive genetic links to sexuality have been found and, therefore, reinforced the original gay gene research has been given a life in science news media through these references.
Into the Labyrinth: Databases and Biodigital Culture
The gay gene is not confined to the news and popular science writings; a set of references in digital culture also underpin it. It is the layering of different references through different media forms that helps embed the gay gene as a reputable biomedical object at the same time that it fails to have biological materiality. References to the gay gene also reside in online databases. Putting the marker Xq28 into the Google search engine elicits results from Wikipedia (an encyclopedia) and the biomedical database OMIM as the first two hits, respectively. These are different kinds of databases, but it is worth drawing out a little about their respective politics because they both have entries on Xq28 linking this genetic marker to homosexuality. In fact, Wikipedia does not link it to anything else, although Xq28 is linked to 20 other entries in OMIM.
Wikipedia and OMIM are both online databases, constitutive of different contemporary knowledge cultures that also have some interesting similarities. These databases are both associated with an imperative to democratize or free up knowledge. They invoke communities of editors and a stewardship of knowledge production, and they both generate a much wider reading public. OMIM is more oriented toward biomedical knowledge, and Wikipedia toward general knowledge.
The Wikipedia entry on Xq28 was entered August 9, 2006, and in the first version, only the Hamer et al. (1993) study was recorded. In January 2007, the entry was expanded to add details of the Rice et al. (1999) study, which contradicted the Hamer et al. (1993) study by finding no correlation with Xq28. Simultaneously, a reference to Hamer’s (1994) rebuttal of Rice et al. (1999) in Science was also added. The entry remains stable with reference to the 1993 and 1999 studies. A reading of the article in this current form implies that a link between Xq28 and homosexuality has been proposed, refuted, and defended. However, the defense is not symmetrical, and the refutation has more overall emphasis in the article, as can be seen from the Wikipedia text:
Xq28 is a genetic marker on the X chromosome found by Dean Hamer and others in 1993. Hamer’s study led to his belief that they found a link between the Xq28 marker and male homosexuality,[1] but the original study’s results have been disputed [italics added]. (see http://en.wikipedia.org/wiki/Xq28)
The OMIM entry listed as 306995, which links Xq28 to homosexuality, is fuller and more ambivalent. Wikipedia and OMIM share something called an “edit history,” which shows the dates and authors of all edits. In Wikipedia, it is possible to recall earlier versions through the edit history and to see how contested entries have been settled. However, in OMIM, the edit history for 306995—HOMOSEXUALITY 1; HMS1—is only a list of dates and names, and does not allow the user to recall earlier versions. The edit history in OMIM runs from 1993 to 2011. The entry records the genetics of homosexuality research by Bailey and Pillard (1991) in the first paragraph, and then homosexual brain research by LeVay (1991) before entering details of Hamer et al.’s (1993) study. A second study linked to Xq28 is listed, and then the Rice et al. (1999) study, which did not find an Xq28 correlation, follows. Later studies up to 2006 that investigated the X chromosome, and one that did not find any links to Xq28, are also listed.
The OMIM entry gives the gay gene a similar lifeline to that of the news coverage. It provides a list of studies that make biological claims about sexuality in relation to genes, chromosomes, brain sizes, and hormones and curates them under a genetic banner. It lists all studies in a chronology, regardless of provenance or replication, as long as they have been published in a science journal. It also constitutes a lodging and material substrate for the gay gene in OMIM and by offering openings to other databases and audiences.
OMIM opens up to other databases across the U.S. National Library of Medicine and to the University of California, Santa Cruz (UCSC) Genome Browser. In addition to the comprehensive entry, OMIM provides external links to information about clinical trials and to the Web site of the Genetic Alliance. Genetic Alliance describes itself as “the world’s leading non-profit health advocacy organization.” Clicking out of OMIM from the 306995 entry creates an automated keyword entry in the Genetic Alliance database based on the keyword homosexuality. Genetic Alliance does not find anything on homosexuality, which is not surprising for a health advocacy database because homosexuality is not currently listed as a medical disorder in most contexts. However, it does link to another range of resources via a link to “The National Library of Medicine Resources for HOMOSEXUALITY.”
In the following example, the use of uppercase letters signals an automated entry generated by the keyword. Following the link for “The National Library of Medicine Resources for HOMOSEXUALITY” takes a user to the site of the National Center for Biotechnology Information (NCBI). OMIM is also part of the NCBI, so this is a circular move. A user starting off in OMIM, thus, returns to the source to find a page purporting to offer resources for people who may have just been diagnosed with HOMOSEXUALITY. This system is designed to aid with diagnostic information and to provide quick aggregations of information about inherited conditions and links to resources. Although OMIM primarily is a biomedical resource and, as such, is aimed at biomedical professionals, these links between different databases also anticipate patients or newly diagnosed individuals as audiences.
The following text is taken from the page generated through an automated query to databases linked to NCBI. This is entirely automated, and generates a combination of pre-authored text fields with the keyword inserted in uppercase letters so the user following the trail from Xq28 is addressed as follows:
This site pulls together the latest research and explanatory materials about HOMOSEXUALITY from NCBI’s databases and other sources.
- Links to introductory material about HOMOSEXUALITY and genetics.
- NCBI Book sections and chapters about HOMOSEXUALITY and genetics.
- Recent scientific articles about HOMOSEXUALITY.
- Links to resources for screening, genetictesting, and directories of specialists.
- Who is this page for?
- Those who live with HOMOSEXUALITY.
- Those who care for someone with HOMOSEXUALITY.
- Those learning about HOMOSEXUALITY.
- Those keeping up with the latest research on HOMOSEXUALITY.
This is a template, and other pages like this would be generated in a similar way, irrespective of the keyword entered. Clearly, the keyword is supposed to be a medical condition of some kind (e.g., ataxia), and assumes a diagnostic or health-related reading position. The only keywords that are automated by this system are those already in the database. The uppercase term here cannot be entered manually. It is generated through the automated links between the database entry number in OMIM and the keyword to which such an entry is attached across the National Library of Medicine databases and beyond.
Thus, HOMOSEXUALITY is generated as a biomedical condition by virtue of being lodged in OMIM as such. In other words, now that OMIM is part of the architecture of biomedicalization (as described by Clarke et al., 2003), the chain of association among Xq28, homosexuality, and medical condition is materialized and generates new pathways through biomedical resources. Homosexuality in the genome, thus, has a life in the biotechnological and biomedical databases that have been generated around genomics. Databases in this area, from collections of journals to raw genome data, are designed to open into each other and to open out to new groups of users. OMIM is one example of the current expansion in porosity between databases, and the “Online” in the title of OMIM signals this. Originating as Mendelian Inheritance in Man (MIM), this was a print publication aiming to catalogue genes and genetic disorders, edited by Victor McKusick at Johns Hopkins University in the United States. McKusick’s collection went online in the 1980s and became an NCBI Web service in 1995. OMIM is maintained at Johns Hopkins University (where MIM originated), but is hosted at UCSC. Santa Cruz is also home to the Genome Browser for the human reference genome and an expanding group of genomic datasets.
OMIM’s network of links to other databases is an opening horizon. The expansion and aggregation of genomic information and attempts to make datasets “talk” to each other are growth areas in genomic bioinformatics. Projects in these areas are attached to a language of democratization because they are associated with ideas about opening up information and resources to broader audiences (Reardon, 2011). They are also attached to a language of workflow efficiency and better genomics. For example, Lincoln Stein (2002), in a speech to an Open Bioinformatics meeting also printed in Nature, proposed a model for “seamless interoperability between online databases” (p. 120). In this, he encouraged standardization between Web services as a solution to existing fragmentation between data standards in genomic bioinformatics. However, these new forms of interoperability raise the question of what the nature of operability is, or what is operating with what. At the moment, we can represent the “seams” or disconnections through a case like Xq28 where the biomedicalization of homosexuality appears almost unintended.
It may well be that database integration has something to offer what we might call democracy and that it will offer new kinds of efficiency. As Stein (2002) himself acknowledged, seamless integration is something of a fantasy. However, interoperability facilitates the flow of errors, as well as other kinds of information, and could even be thought of as a kind of persistence of errors. The proposal that Xq28 is linked to homosexuality in the genome remains hypothetical at best. The framing of homosexuality as a biomedical disorder is highly problematic. One experimental finding does not make a scientific fact or a biological entity. However, its entry in OMIM gives it materiality and a life in the flows of genome bioinformatics, which persist as OMIM opens up into databases across the National Library of Medicine’s NCBI and beyond. This life comes from OMIM, which is primarily oriented toward collecting biomedical conditions. Thus, this life story of the gay gene attaches homosexuality to a biomedical database, which, in turn, tries to offer resources for clinicians and newly diagnosed patients with the condition HOMOSEXUALITY generated by OMIM’s cataloguing system. One issue with this process of automation is that questions about the validity of the research, the categorization of sexuality, and accountability for knowledge production become erased.
Conclusion
The gay gene lives across a number of media forms, and I have emphasized the print and online news media and online databases and the similarities and differences between them. These forms include biomedical media, general news media, and popular science writing, and link to the proliferation of Xq28 and the gay gene across multiple sites. Since the first press release by Science in 1993 about Hamer et al.’s (1993) research and, at the time of writing, 18 years later, the gay gene has gained materiality. Its actual bioscientific provenance has not improved in the sense that Hamer et al.’s (1993) Xq28 findings have not been replicated; and, on the contrary, they have been contested. There is no more a biological materiality to the gay gene now than there was in 1993. However, despite this, the address of the gay gene has persisted. It has become embedded in science media cultures and lodged in databases that open up into information flows with greater porosity than ever. It is fed by aggregations of noise that contribute to the erroneous signal strength of the message that there is a gay gene.
The gay gene takes on forms of materiality by appearing as a reference in scientific journals, in news media, and in online databases. It also has further instantiations. It has become part of queer culture, and has been adopted as a logo and a domain name signifying, for example, Web sites catering to gay interests. It has appeared in poetry and in song lyrics, and it has been the focus of art and a community engagement project in the United Kingdom called, “How Gay Are Your Genes?” It has been celebrated and attached to gay pride, as well as to anger about biomedical research into homosexuality. Jennifer Terry (1999) suggested that there are class dimensions to these responses, with (in the United States) middle-class, White, gay men finding it easier to affirm genetic sexuality research and working-class teens finding it much more problematic.
The spectrum of responses to the gay gene can be signalled by two extremes. At one end of this spectrum, Xq28 is attached to anger and injustice, and appears as a harbinger of eugenics. One example of this is in the use of Xq28 as a 2003 song title by U.K. radical punk band, The Restarts with the following lyric: “Xq28, Genetic profiling from within the womb, Orwelian nightmare leading to doom.” The other end of this spectrum is at the humorous and celebratory end, indicated by the printing up of t-shirts reading “Xq28—thanks for the genes, Mom!” worn by gay men in San Francisco in the 1990s (Kitzinger, 2005). This life of the gay gene presents an address to those attached to queer lives. It is not surprising that in this context, people who identify with the term gay, or its semantic field, feel that an address is made to them particularly. Although responses to research about the gay gene have always been mixed, these are gendered, classed, location specific, and have changed over time. In very general terms, middle-class gay men seem to have more positive attachments to the idea of the gay gene as a “biohistory” (Rosario, 1997, p. 1), whereas women and working-class teenagers have been rather more suspicious (Terry, 1999). A celebratory mode was certainly attached to Hamer et al.’s original research in 1993 in the United States, and the excitement about genetics at this early phase of the human genome project was undoubtedly a contributory factor. At the time of writing, one decade after the human genome project, the excitement has waned, and the biomedicalization of the gay gene has increased. At the same time, genomics has become an everyday explanatory narrative that depoliticizes questions about identity by categorizing them as genetic.
In the last 18 years, the mediation of the gay gene has generated a biomedical media materiality that helps ground the concept as fact. The grounding of this materiality in genetic research and in biomedical databases is a significant factor in establishing this quality of facticity. This media materiality signals a convergence of media forms such that the borders between science and the media become very porous. I have claimed elsewhere that the media is a site in the making of science (Haran, Kitzinger, McNeil, & O’Riordan, 2008; O’Riordan, 2010), and the case of the gay gene provides a specific example of this general claim. Sexuality in the genome is made up of the diverse assemblage of sites where references to Xq28 and the gay gene are made. The combination of scientific journals promoting scientific research in this area, together with science news media, reviews of popular science writing, and online databases, provides a substrate in which the gay gene has materialized and has a life story of its own. This materialization operates as an effective address that speaks to people but lacks real accountability. At the moment, the reach of the gay gene is more a potentiality that creates affirmation or anger than a biomedical tool for individual- or population-level control. However, there is an inaccuracy about it from the start, which is problematic. The way in which the preservation of this in databases contributes to the biomedicalization of homosexuality indicates that a sense of responsibility for creating realities lies at the heart of database design, journalism, and science communication, as well as with scientific research and its audiences.