Thursday, April 21, 2011

Sampling Sex


Psychologists A. Dana Ménard and Christine Cabrera recently set out to
gain an understanding of how sex and sexuality are portrayed in contemporary romance novels and to determine whether these portrayals have changed over the last 20 years. It was hypothesized that most depictions of sexuality in romance novels would adhere to Western sexual scripts (Gagnon 1977; Gagnon and Simon 1973; Simon and Gagnon 1986, 1987) and that this would not change over time. The sample consisted of books that had won the Romance Writers of America award for best contemporary single-title romance from 1989 to 2009. A quantitative content analysis revealed that hypotheses were supported with respect to characterization of the male and female protagonists, characterization and context of the romantic relationships, and order and nature of sexual behaviours.
I'm a literary critic, not a statistician, but it seems to me that sample size does matter if one is going to try to draw conclusions about an extremely large genre. The findings which Ménard and Cabrera report in "‘Whatever the Approach, Tab B Still Fits into Slot A’: Twenty Years of Sex Scripts in Romance Novels" are based on an analysis of
20 books and a total of 46 sex scenes. It is possible that some findings were non-significant because the relevant statistical tests were simply under-powered.
My own essay about bodies and sexuality in romance novels (which I co-wrote with Kyra Kramer), draws on only 26 primary texts; obviously I do think it's possible to use a sample of this size to demonstrate that certain character types and patterns of behaviour have been present in the genre for a very long time.

If, however, one wants to draw more precise conclusions about how common they are, I think one might want to use a much larger sample. Another important factor to consider is whether the sample is representative. Ménard and Cabrera state that
Rita award winners were chosen because it was thought that these novels might be considered especially representative or prototypical examples of the genre. They may also represent ideal romance novels that other authors might strive to emulate. In addition, the books included in this research sample all sold numerous copies [...]. Selecting Rita award winners may also have enhanced the comparability of books across time.
Single-title contemporary novels (i.e. released individually and taking place after 1945) were chosen because these books were thought to be the most likely to reflect the social mores regarding sex and sexuality at the time of their publication.
Clearly Ménard and Cabrera had some valid reasons for choosing these romances but it might have been wise for them to take a look at at least some erotic romances, lesbian romances, m/m romances, African-American romances and interracial romances. These are subgenres which have not tended to be recognised by the RWA's Rita awards but are nonetheless important parts of the genre.

As it was, their sample was "100% heterosexual," "95.0% non-discrepant (both Caucasian)" and "deviant sexual behaviours (e.g. use of lubrication, masturbation, anal stimulation, BDSM-inspired behaviours) were rarely depicted."


Had Ménard and Cabrera included erotic romances in their sample it seems highly unlikely that they would have concluded that
The total number of sex scenes was surprisingly low, given the lay reputation of romance novels; several books published recently included no sex scenes at all. These results may indicate an increasing trend towards less explicit sexual content in romance novels.
Indeed, that last statement makes me wonder if they were even aware of the existence of erotic romances.

It's important to note that Ménard and Cabrera do show some awareness of the limitations of their sample:
Given the small number of books, and the fact that all were Rita award winners, the generalizability of findings from this study is limited to English-language, North American, single-title contemporary romances. [...] It is possible that there may be more diversity in other romance sub-genres; future investigators may wish to compare and contrast sex, sexuality and gender roles across a variety of romance sub-genres (e.g. historical, paranormal, suspense).
The trouble is, there are plenty of erotic romances, lesbian romances, m/m romances, African-American romances and interracial romances which can also be classified as "English language, North American, single-title contemporary romances" and those sub-genres are not even mentioned here. I would urge "future investigators" working in this area to be aware of these sub-genres (and other relevant sub-genres which may emerge in the future).

--
  • Ménard, A. Dana and Christine Cabrera. "‘Whatever the Approach, Tab B Still Fits into Slot A’: Twenty Years of Sex Scripts in Romance Novels." Sexuality & Culture, Online First™, 3 April 2011.
The photo of the 21 ice-cream cones containing white ice-cream was taken by Thomas Hawk and is available at Flikr under an Attribution-NonCommercial 2.0 Generic licence.

43 comments:

  1. what the heck? since when is using lube a deviant sexual behavior?

    ReplyDelete
  2. Interesting. I agree that the sample size was insufficient to make any realistic conclusions, and the generalizability is weak. Very weak. Why would they not at least use the finalists from each year--the reasoning would be the same in terms of cream of the crop for the time.

    Overall, while it's fairly common for psychological research to conclude something opposite my experiences, this conclusion really seemed off. In my reading, the books have only gotten hotter! The number of scenes may be fewer, but if so, I hypothesize a lot of writers shy away from gratuitous sex, even in erotic romance, to make it count for something more than just titillation.

    (Oh, and I'd love to know how they quantified the content--will be heading to my EBSCO soon).

    ReplyDelete
  3. Nekobawt, I suspect that they might mean "deviant in the sense that it deviates from the sexual practices outlined in Abramson and Mechanic's study." Here's the full sentence, in context:

    Our findings, like those of previous investigations on representations of sex in the media (i.e. Abramson and Mechanic 1983; Duran and Prusank 1997; Ménard and Kleinplatz 2008), show that romance novel characters tend to be young, attractive, heterosexual, able-bodied and childless. Most sex scenes depicted what Castleman refers to as ‘a one-way drive downfield to the end zone of intercourse’ (2004, p. 203), confirming narrow, heterosexist notions of what constitutes ‘sex’. As in fictional bestsellers, deviant sexual behaviours (e.g. use of lubrication, masturbation, anal stimulation, BDSM-inspired behaviours) were rarely depicted (Abramson and Mechanic 1983).

    and here's what they'd written about Abramson and Mechanic earlier in the essay:

    A study by Abramson and Mechanic (1983) examined depictions of sexuality in mainstream best-sellers (i.e. nonromance novels) from 1959, 1969 and 1979. The characters were usually described as attractive, single, healthy and young (19–35 years of age). Sex was typically initiated by the man or by mutual consent, with a trend towards more female initiation in later books. As in television and movies, non-heterosexual sex, masturbation, lubricant or contraceptives were rarely depicted. The results from existing studies on sex and sexuality in romance novels tend to confirm the findings from research on other media forms: the content of romance novels tends to conform to the usual sexual and gender role stereotypes.

    ReplyDelete
  4. "I'd love to know how they quantified the content"

    They write that

    The unit of analysis for this study was scenes of sexual activity in which the characters were engaged in penetration and/or behaviour that was likely to lead to orgasm for one or more of the participants. The gender of the character being stimulated was also coded [...]. In addition, the coders took note of each incidence of touching or sexual behaviour between the main characters to determine the sequence of behaviours across the novel. Instances of solo sexual activity (i.e. masturbation) were also noted and coded. In books that featured more than one couple, the sexual behaviours of each couple were noted on separate coding forms.

    ReplyDelete
  5. I think one of the fascinating things about the article is that none of the normal critics were present: Janice Radway, Ann Snitow, Kay Mussell, Tania Modleski, Lynne Pearce, Pam Regis. The authors of the article completely ignore the fact that there is a growing body of scholarship on popular romance -- I'm not certain if this helps their argument or not. They hardly entertain some of the standard debates in popular romance studies.

    ReplyDelete
  6. I don't think it's surprising that social scientists wouldn't cite literary critics in their work, but this study does show some of the perils of cross-disciplinary research (as does Radway's not very good social science). For instance, the assumption that historical romance is less likely than contemporary to reflect the mores of the time in which it was written seems problematic to me. A lot of hisorical romance published barely nods at the mores of, say, the Regency period.

    As Laura points out, the fact that there's a RITA for Inspirational romance but not one for erotic romance or GLBT romance might indicate to anyone with a knowledge of the genre that they tend to the conservative. Not only is it small numerically, but it represents a very narrow segment of the genre.

    That they can conclude, even tentatively, that there might be a trend toward less explicit sex scenes in romance novels shows the limits of their sample. A lot of the things they cite as "deviant" (LOL) are showing up more in mainstream contemporary romances, let alone subgenres, just as they are, apparantly, in North Americans' sex lives: witness the recent Smart Bitches review of "the first Blaze with anal."

    ReplyDelete
  7. "The authors of the article completely ignore the fact that there is a growing body of scholarship on popular romance"

    They do cite a number of previous works on romance: Assiter; Clawson; Diekman, McDonald,& Gardner; Ruggiero & Weston; Thurston. These are all works of romance scholarship which are focused on the genre's depiction of sexuality, and I imagine that could be why Ménard and Cabrera thought they were of particular relevance to this study. Also, as Liz said, "I don't think it's surprising that social scientists wouldn't cite literary critics in their work."

    ReplyDelete
  8. For instance, the assumption that historical romance is less likely than contemporary to reflect the mores of the time in which it was written seems problematic to me. A lot of historical romance published barely nods at the mores of, say, the Regency period.

    Some authors of historicals do try very hard to make their characters seem "of their time," though, even if inevitably some aspects of the depiction reflect modern tastes and beliefs.

    It strikes me that one of the aspects of the novels that they were interested in studying was contraceptive use, and although contraceptives are sometimes used by characters in historicals, the methods available and their efficacy aren't going to be identical to those in contemporaries.

    ReplyDelete
  9. It seems then that there is a necessary methodological question which we, as scholars of romance, must attend to -- not only as "literary scholars" but also as "interdisciplinary scholars." Of course, this is a problem that isn't likely to go away, likely because genuine "interdisciplinary" research is, as noted above, quite difficult

    ReplyDelete
  10. Hmmm ... clearly they haven't read either (80's) Joanna Lindsey or Beatrice Small or Johanna Anderson. I remember quite a bit of BDSM in those books. In fact, I have read a lot of romances where the sex is kinky enough to qualify as a corkscrew.

    ReplyDelete
  11. Ack! "Corkscrew" was NOT a pun.

    ReplyDelete
  12. "It seems then that there is a necessary methodological question which we, as scholars of romance, must attend to"

    What's the question, Jonathan? And do you think you could possibly attend to it in a blog post?

    ReplyDelete
  13. Kyra, were their novels all historicals, and did historicals from that period tend to be more kinky than the contemporaries? If so, I wonder if it was because the past was seen as a safer place to explore kink than contemporaries because if the novel was a historical it might not be considered "to reflect the social mores regarding sex and sexuality at the time of their publication."

    Actually, that raises the question of how often the novels depict "sexual fantasies" (i.e. acts which the reader might never wish to incorporate into their actual sexual activity and/or which might not even be likely/possible to happen in real life) and how often they're an accurate reflection of contemporary sexual practices and experiences.

    ReplyDelete
  14. The question, I think, is: when is enough, enough? For instance, how large a sample size is needed before one could make a "generalized" comment about, for instance, sexuality in romance? One of the fascinating things about romance is the number of romances produced each and every year. For instance, in attending to subgenres of romance, it is quite likely that the authors of the study would find more examples of sexuality. This is likely very true, but what would happen, if, for instance, they included evangelical romances in their research? I'm thinking here particularly of Steeple Hill Romances. The Writing Guidelines read: "Any physical interactions (i.e. kissing, hugging) should emphasize emotional tenderness rather than sexual desire or sensuality. Please avoid any mention of nudity" and that the stories may not include "hero and heroine remaining overnight together alone." Thus, were these romances to be studied, one would find even less sex and sexuality; while were the authors to study erotic romance, m/m, lesbian, menage, etc., they'd find more examples of sexuality. I am not trying to be polemical for the sake of the polemic, but at what point has one read "enough"?

    ReplyDelete
  15. I am not trying to be polemical for the sake of the polemic, but at what point has one read "enough"?

    Well, it's always good to (a) be widely read, (b) draw on the knowledge of other readers and critics and (c) always be aware that however many books you read, there will still be thousands and thousands of books you haven't read.

    I think that the minimum number of books you need to have studied will vary depending on the type of conclusions you want to be able to draw. If you just want to give examples of a particular character type (e.g. heroines with a Glittery HooHa) and demonstrate that they turn up in a range of sub-genres and over a long period of time, I don't think you need to analyse as many novels as you would if you wanted to carry out a statistical analysis designed to determine what proportion of romances contain heroines with Glittery Hoo Has. It also depends on how large your area of enquiry is: for example, are you planning to draw conclusions about the entire genre, the genre during a specific period, or a specific sub-genre during a specific period?

    There's likely to be a trade-off between depth (how many books one can analyse in detail) and width. And like Kyra with her corkscrew, I'm not intending that to be taken as a double entendre.

    I have the impression that in statistical analyses, more data is generally better, whereas in literary analyses, long lists of examples could get rather boring. It's tricky, though, and certainly where literary criticism is involved I don't think there's any formula that you can use to be certain that you've included just the right number of examples.

    Oh, and to state the obvious, people who have not carried out analysis of all the texts under discussion (whether it's all the novels in a particular sub-genre, or all the novels in the entire genre) should avoid drawing sweeping conclusions (e.g. "all romance heroes are older, taller and richer than their heroines"). In a genre of this size there's always going to be at least one (and probably far more than one) exception. That shouldn't, though, stop anyone from writing something along the lines of "a great many romance heroes are older, taller, and richer than their heroines" or "in the romance genre as a whole, virgin heroes appear to be relatively rare."

    ReplyDelete
  16. Wow! You guys have left some very helpful comments here. While I can't address everything that got said, I thought I'd chip in my two cents here and there:
    - Use of lubricant in these books was classified as "deviant" only because this is the approach that was used in previous studies. By "deviant", we were referring more to statistical deviance than any moral judgment on our part.
    - We did consult some of the authors that were mentioned (Radway, Snitow). Discussion of their work was omitted from the paper largely for reasons of space and relevance. As we were intending to publish in a psychology journal, we thought it was more important to focus on authors in psychology rather than literary criticism.
    - Generalizeability is a problem, no question. If we had had more time and money, we would definitely have wanted to sample more novels. Unfortunately, it's just the two of us grad students and no funding!
    - We are definitely aware of the existence of romance sub-genres (e.g., erotic, paranormal). We didn't include any other type of books in order to enhance our ability to make comparisons across books (comparing single-title contemporaries to single-title contemporaries seems more fair than comparing Regencies to paranormal romances in terms of sexual content). It's a sad fact that if you want any hope of successful statistical analyses, you need to keep your sample relatively homogenous. It was not our intention to ignore the diversity of romance novels out there. Again, it was a question of time, manpower and money.

    Thank you all for taking the time and energy to comment on this post. Christine and I will definitely take these comments into consideration as we continue our work in this area.

    ReplyDelete
  17. Thanks for posting on the blog. It is a fascinating article and one which I shall likely refer to often in my own work.

    ReplyDelete
  18. Thanks very much for your comment and clarifications, Dana.

    "We are definitely aware of the existence of romance sub-genres (e.g., erotic, paranormal). [...] It's a sad fact that if you want any hope of successful statistical analyses, you need to keep your sample relatively homogenous. It was not our intention to ignore the diversity of romance novels out there. Again, it was a question of time, manpower and money."

    I fully appreciate that there are time and other constraints, and I also very much appreciated that you included the statement to the effect that "the generalizability of findings from this study is limited to English-language, North American, single-title contemporary romances."

    I was a little concerned, though, that people reading this article who are less knowledgeable about the genre might not fully understand the implications of that caveat. They might also not understand the internal policies and politics of the RWA which might have affected the Rita selection process. I don't know a great deal about this myself, but as far as I'm aware, all books entered for the Ritas are entered by the authors themselves, and it costs them money to do so.

    * If authors who know that the kind of books they write rarely, if ever, win a Rita, therefore decide not to bother entering, this could affect the outcomes of the contest.

    * Some authors have found their books barred from participation due to rules about ebook and epublisher eligibility. In more recent years this may have affected the pool from which the final sample was drawn.

    * Books have been known to be given low ratings because of sexual content. This is particularly relevant to the subject of the paper. Here's an example from 2006:

    The most recent online RITA/RWA furor concerns Jamie Sobrato, whose entry was judged to be a story with "no strong romantic elements". In other words, her book, As Hot As It Gets, wasn’t considered a romance. I think this shocked her somewhat, seeing as she is a romance author after all. But on reflection, she does write for the Blaze line, so this may have been the deal breaker. You know what I’m saying here, don’t you?

    It always seems to come back to S.E.X, doesn’t it? As an author, if you’re judged to have too much sex in your book, you’ve got as much chance as winning a RITA, as Tom Cruise has of regaining his sanity. Nope, I don’t fancy your chances at all.

    Apparently a significant number of Blaze authors didn’t enter this year, due to their belief that they had no chance of winning, because of the overtly sexual nature of their books.


    That quote's from a post by Karen Scott at All About Romance.

    ReplyDelete
  19. Also in 2006, the RWA's magazine published a letter which reopened a controversy from the previous year, raising:

    concerns that the RWA has exhibited a growing conservative streak in recent years.

    The issues raised in Butler's letter stem from a poll conducted in 2005 by RWA leadership on the definition of “romance” as a genre. According to lesbian romance writer Radclyffe, who is also a member of the RWA, “A poll was taken by the Romance Writers of America asking the members to vote on the definition of ‘romance,' specifically distinguishing between a love relationship between a man and a woman versus a love relationship between ‘two people.'”
    (post by Malinda Lo at AfterEllen).

    I don't know if the way the Ritas are judged tends to favour more conservative depictions of sexuality, but I did notice that the one novel by Jennifer Crusie which won in this category is her most conservative. That may just be a coincidence, but in the novel that won (Bet Me) there's no extra-marital sex (as in Tell Me Lies), sex in public places (as in TML, Crazy for You, Welcome to Temptation), intercourse between the heroine and someone other than the hero (as in Fast Women), kissing between the heroine and another woman (also Fast Women), or depictions of less than satisfactory sex (Faking It). In addition, Bet Me only has one sex scene, which occurs at the very end of the novel, whereas Crusie's other single-title contemporary romances all have more than one sex scene.

    Unfortunately I don't know enough about the other winning novels to be able to comment on them and how they compare to other works in the same authors' oeuvres.

    ReplyDelete
  20. And how about romances published in other countries? Jilly Cooper, anyone? What she describes could be considered kinky in other societies, but not in Britain. They're mainstream. Even if the study is confined to North America, this isn't a typical sample of romances sold.
    There are no generalisations to be made, and an academic looking for generalisations is likely to be disappointed. Any genre that sells books in the kind of volume that romances do is hard to generalise. In a way, "romance" is an umbrella to a huge diversity of stories taking a variety of themes and worlds. Dividing books that have one thing in common (the central relationship and the happy ending) is largely a marketing ploy, because they could equally be divided into historical, contemporary, fantasy, etc etc and sit on the shelves in those sectors, as they do elsewhere. The genres are separated into what sells best and they are a modern construct.
    The Ritas are not "typical" romances. In many cases, they're not even the best selling ones. They are self selected, in that you have to enter to win. So already the sample is biased.

    ReplyDelete
  21. "And how about romances published in other countries?"

    That would be a fascinating subject to study, but also a vast and very complex one. Since one of Ménard and Cabrera's primary interests was in trying to detect whether there had been changes over time, it makes sense that they would choose a sample in which the main variable was the date of publication. Having made that choice, though, they acknowledge that "the generalizability of findings from this study is limited to English-language, North American, single-title contemporary romances."

    Given the way the Ritas work, I think one might want to state that the generalizability is even more limited than that because it seems that an English-language, North American, single-title contemporary which is also one or more of the following stands less chance of winning and/or is perceived to stand less chance of winning (which may well affect the submissions process, and therefore the pool from which the winners are drawn): erotic, AA, inter-racial, LGBT, BDSM, epublished-first.

    "There are no generalisations to be made, and an academic looking for generalisations is likely to be disappointed."

    As I said to Jonathan, I think one can make generalisations, but one has to do so very cautiously, in the knowledge that there are likely to be exceptions (and in some cases very significant numbers of exceptions). I'll put myself on the spot and quote from some of my own work: I think it would be fair to say that "many a romance hero" possesses a Mighty Wang, and that the MW can be "thought of as a symbol of the male sexual drive discourse" (Vivanco and Kramer) and I also think that "beliefs about gendered sexuality frequently appear to underlie the sexual behaviour of characters in the romance genre" (Vivanco and Kramer).

    Unlike Ménard and Cabrera, we weren't able to give precise figures, and we were trying to generalise about the whole genre (using the RWA definition, and referring only to English language novels) across an even longer period of time. Given the dangers of generalising, we were therefore careful to use words such as "many" and "frequently" rather than words such as "every" and "always."

    ReplyDelete
  22. In a way, "romance" is an umbrella to a huge diversity of stories taking a variety of themes and worlds. Dividing books that have one thing in common (the central relationship and the happy ending) is largely a marketing ploy, because they could equally be divided into historical, contemporary, fantasy, etc etc and sit on the shelves in those sectors, as they do elsewhere. The genres are separated into what sells best and they are a modern construct.

    Genre labels as such aren't a modern construct, of course: in the classical period there were separate muses for comedy, tragedy, epic poetry and lyric poetry (and for a few other genres too).

    I'd agree with you that the use of the term "romance" to describe works of prose fiction with "a central love story and an emotionally-satisfying and optimistic ending" (RWA) is new, and the word doesn't even have that very precise meaning in all the markets where romances are sold.

    Like you, Lynne, I live in the UK and in my local library the "romance" shelves include family sagas and a number of other types of novel which would not be considered "romances" according to the RWA definition.

    That difference between the US and UK markets is also reflected in the name of the Romantic Novelists' Association, and in their definition of what constitutes "romantic fiction":

    try giving each thread of your story a title—love story, family relationship, experienced tragedy, ambition etcetera. Next, picture a mountain range, the Cairngorms, the Himalayas, whatever, and give each mountain peak the title of a strand of your novel, with the most important thread going on the highest peak, the least important, the smallest. If when you’re done, the love story is stuck on K2 you’ve written a romantic novel.

    A "romantic novel" is not necessarily a "romance" as defined by the RWA.

    ReplyDelete
  23. One of my MA students took an interesting approach, I thought, to the sampling issue.

    She let her initial selection pool be this year's nominees for the DABWAHA competition (http://dabwaha.com). She then used a random number generator to pick a number, which turned out to be 6, if memory serves, so the sixth listed entry in each category of the competition was the book she investigated.

    She's being very careful in how she frames her conclusions, of course. But her sense was that there were too many limiting factors in the RITA competition for that to serve as her selection pool, and that the DABWAHA categories were a better representation of American romance today.

    ReplyDelete
  24. As far as I can remember, in one of the early years of the DABWAHA I noticed that they had no AA romances. That may have changed in more recent years, but I don't know. And do they include inspirationals?

    ReplyDelete
  25. Good points, Laura. There's no separate category for inspirationals or for AA romances; I'd have to look to see whether there were any authors of color.

    Maybe we scholars need to come up with our own competition (with our own sense of the widest range of categories), so that future scholars can sample it!

    ReplyDelete
  26. Eric, Jonathan had a fascinating idea in an email he sent me: that we collect from scholars syllabi for a hypothetical "Intro to Popular Romance Fiction" course, with explanation of the texts chosen, so we can see if there's some sort of consensus about a romance "canon". That doesn't deal with corpus issues for articles like this, but it's potentially an interesting topic for a Pedagogy article/section for JPRS, if anyone wants to take it on.

    ReplyDelete
  27. I've just seen an article written by Eloisa James/Mary Bly which addresses the issue raised by Ménard and Cabrera's choice to study contemporary rather than historical romances on the grounds that the former "were thought to be the most likely to reflect the social mores regarding sex and sexuality at the time of their publication."

    James writes that:

    sex -- its practices, its customs and conventions, and prevailing attitudes toward it -- is a function of the historical, cultural and social conditions of a given time and place.

    What was considered fun to read in the 1980s isn't necessarily considered fun to read now, and thus the bodice-rippers of the '80s went the way of that decade's aggressive shoulder pads and crumpled leg warmers.

    The United States of the 21st century is no longer in the same place when it comes to desire, women and the need to wreak havoc on apparel: In keeping with the times, my heroines tend to do their own button-scattering.

    What I'm saying is that although eroticism is culturally, geographically and historically specific, we writers of historical romance sexualize history without regard for the specific epoch in which we set a novel.

    No matter how historically accurate the details and language in our novels might be (and mine, in case you're wondering, are pretty accurate), we write sex from the point of view of our own contemporary attitudes and mores.


    ---
    James, Eloisa. "Bringing past sex to life is complicated." CNN. April 25, 2011.

    ReplyDelete
  28. I skimmed this paper, and I agree with the criticisms you all have raised. I don't see how these findings could be generalized. The sample just suffers from too much selection bias. The books nominated for RITAs are already a non-random subset of all books eligible, since nominees must be entered and nominations cost money. Then, of course, you have the problem of all the books that aren't eligible or don't fit the RITA requirements.

    It's better to think about these results as generative of questions to ask in the next round of studies rather than anything that sheds light on the larger genre.

    It would certainly be possible to create a statistically valid study of romance novels, given the large number published over the years. But I have yet to see such a study.

    ReplyDelete
  29. Sunita, I was wondering if something like the Google Ngram Viewer would be helpful (if it could be applied to a corpus of romance novels rather than to the existing corpus, which I assume contains a wide range of texts).

    I suppose it would depend on the kind of analysis one was wanting to carry out: that only looks at the prevalence of key words, and many of the features Ménard and Cabrera were looking for could only be identified via careful reading of the texts.

    ReplyDelete
  30. Laura, I hadn't thought of the Ngram but you're right, that would be very interesting. I have a colleague in the humanities who's pretty savvy about using the Ngram as well as more sophisticated measures in his research.

    If the authors had taken a random sample of books over the same time period (say, HPs), their results might have been generalizeable. At least the sample wouldn't have been as corrupted. And with the various lists of Harlequin lines available, it wouldn't be that hard to do.

    [Apologies if this posts twice; Blogger & Chrome are not cooperating on my Mac at the moment]

    ReplyDelete
  31. "If the authors had taken a random sample of books over the same time period (say, HPs), their results might have been generalizeable.[...] And with the various lists of Harlequin lines available, it wouldn't be that hard to do."

    Yes, because although some lines of Harlequins have come and gone relatively fast, there are others which have been around for a very long time. The Harlequin Romance line and Presents have been around a very long time. There's also the Temptations, which mutated into Blaze.

    ReplyDelete
  32. Can an amateur/non-academic jump into this discussion?

    Linda Hilton chiming in here. I used to subscribe/be a member of this list several years ago but life intervened. However, as romance author, as a reader, and as a college student back in 2000, I wrote an honors thesis on romance novels, "Half Heaven, Healf Heartache: Discovering the Transformative Potential in Women's Popular Fiction." (Arizona State University - West Campus, Barrett Honors College, 2000). After graduating with a BA in Women's Studies (at age 51), I hung around ASU long enough to pick up a Master of Arts in Interdisciplinary Studies in 2003. Kind of design your own MA, but what the heck. And my emphasis was on sociology. So my academic background is as non-traditional as I was a student, and I have both the literary and social science background. Sort of.

    When I wrote the thesis, one of my sources was George Paizis, "Love and the Novel: The Poetics and Politics of Romantic Fiction" (Macmillan 1998). In it he estimated "a selection of about thirty novels . . . seemed an adequate number" on which to base his study, and he limited his sample to short novels with contemporary settings -- essentially Harlequin/Mills & Boon. My thought on reading that was ARE YOU KIDDING ME???

    With all due respect to Ménard and Cabrera, PUH-LEEZE. At the time I wrote my thesis in 2000, there had been approximately 20,000 paperback romance novels published since 1972. TWENTY THOUSAND. How many since then? I have no idea, but I'd be willing to bet it's more than 500. :sarcasm: And out of that they could only find time to analyse 20???!!!!???!!!

    And on this kind of sample they're going to draw conclusions about how depictions of sex and sexuality have changed in romance novels over a period of 20 years? They're assuming their 1989 "sample" of ONE BOOK, ONE FREAKIN' BOOK, is their baseline from which change will take place??? Am I nuts here, or is this poor scholarship?

    Maybe my outrage is out of place here, but as someone who has read and written and written about and defended romance fiction for at least 50 of my 62 years, I'm pretty d**ned appalled.


    Linda Ann Wheeler Hilton

    ReplyDelete
  33. "Can an amateur/non-academic jump into this discussion?"

    Yes, of course!

    When I wrote the thesis, one of my sources was George Paizis, "Love and the Novel: The Poetics and Politics of Romantic Fiction" (Macmillan 1998). In it he estimated "a selection of about thirty novels . . . seemed an adequate number" on which to base his study

    I've found Paizis's analysis useful and he's one of a relatively small number of romance critics to have analysed "The Poetics [...] of Romance Fiction." Some of his ideas about the uses of time and place in romance fiction are interesting, for example. I'd tend to see his findings as suggestions for further research, though, because as you say, the genre is vast, varied and constantly changing.

    I also think Ménard and Cabrera's findings are interesting and what they say about increased condom use fits with what other readers have mentioned in discussions online, but I wonder if, in the end, this study reveals more about the preferences of the RWA judges and attitudes within the RWA than it does about the genre as a whole.

    ReplyDelete
  34. Thank you, Laura, for the welcome. And my apologies for being so angry last night -- so angry I didn't even go back and fix my typos -- but there is a reason for it. And I'm still pretty upset.

    When I was writing romance novels, my projects were 130,000 historicals. Whenever I told someone I wrote romance novels, the stock reply was, "Oh, you mean Harlequins." And then off I went on the explanation of the differences. (Most people who didn't already know the differences didn't really care, but I explained it anyway.) I can't imagine anyone who has been a romance reader for more than a few years that would accept a sample of 20 books, which is ONE PER YEAR, as representative of changes over two decades.

    You wrote, opening your commentary after posting the excerpt from their paper, I'm a literary critic, not a statistician, but it seems to me that sample size does matter if one is going to try to draw conclusions about an extremely large genre. Your first observation was about the small size of the sampling, too. You recognized, maybe just on a gut rather than academic level, how serious this situation was. Serious enough to be your very first observation.

    So, maybe what the study says more about is the continuing dismissal of romance fiction by academia -- not only did Ménard and Cabrera's advisors approve of their "sampling," but the editorial board of Sexuality and Culture didn't seem to see anything wrong with this either. I have to wonder, therefore, what would have happened if the genre being studied were westerns or mysteries or spy thrillers -- you know, something more "manly." Would the advisors and the editorial board have sent the study back for a more representative sample?

    I don't have access to the whole article, only the abstract, so I can't really comment on what they wrote or how they wrote it. But one of the aspects of the "change" they saw over this 20 years has something to do with condom use, according to your comment. (In 20 books, out of shall we say at least 25,000, maybe more?) They started with 1989, with one book, and then claimed condom use "increased" over the 20 years. I guess that means in the ONE BOOK they sampled from 1989, he didn't, and by 2009 he did. How can they make that claim based on such a puny sample? Hell, Alison Hart was advocating authors have their heroes "wrap that puppy" at the 1989 RWA conference in Boston; both she and Jayne Ann Krentz, among others, featured "safe sex" in books written before Ménard and Cabrera's 1989 starting point.

    Who knows? Maybe if M & C had used a really representative sample they'd have found out condom use in contemporary romance novels DECREASED!!!! (Readers perceptions are hardly a scientific study.)

    That's why I remain appalled at the minuscule size of their sample. After all, they're doing at least something of a quantitative analysis, aren't they? How valid can that analysis be on the basis of that kind of sample?

    Whether RWA or the RITA judges have a conservative bias is, I think, almost irrelevant. (I happen to think they do, and they have since the beginning, and as an RWA member for 15 years I tried to fight it, but you see where it got me.) That bias would have been somewhat neutralized had M & C chosen a larger sample. They didn't study "Twenty years of sex scripts in romance novels" at all; they studied "Sex scripts in 20 single-title contemporary romance novels chosen for us by someone else whose motives and biases we don't know and don't care about."

    The thing is, this will enter the body of research available for future academics. It will held up there along with Carol Thurston and Tania Modelski and John Cawelty (ugh, don't remind me) and Pamela Regis as a valid, scientific, academic analysis of romance fiction. IMHO, it's not. Not even close.


    Linda Ann Wheeler Hilton

    ReplyDelete
  35. My apologies to Tania Modleski for misspelling her name. I guess I'm still angry.

    ReplyDelete
  36. @Linda, we're doing our best to change this, you know. I mean, we're not on the editorial board of Sex and Sexuality, of course, but JPRS is doing all we can to change the scholarship of popular romance fiction. Thanks for your passion--that's what we need to keep going, because we've got it too. :)

    ReplyDelete
  37. I am perhaps an optimist in a sense here in that I saw the article by Menard and Cabrera as an opening from which they (or other critics) can depart. A study of the scale being proposed by some would be fascinating and undoubtedly useful -- but it would be a project that would seem to be a never ending project. Even if one were to survey 500 novels written between 1989-2009 and even if there are a total possible sample size of 25000, that is still, in reality, a rather small portion of romance novels. I imagine that such a project would need an incredibly generous funding agency to allow for that project to unfold. I admit wholeheartedly that the question of sample size is one of the most pressing concerns for romance studies: what is an appropriate sample size?

    I think Sarah is right -- the scholarship of popular romance fiction is growing and as it grows it will continue to affect other areas of scholarship. JPRS may very well be an ideal place for Menard and Cabrera to publish another paper, or for another critic to draw on their paper to produce another set of conclusions.

    ReplyDelete
  38. "So, maybe what the study says more about is the continuing dismissal of romance fiction by academia -- not only did Ménard and Cabrera's advisors approve of their "sampling," but the editorial board of Sexuality and Culture didn't seem to see anything wrong with this either."

    As Sarah and Jonathan have said, I think there's a growing academic interest in romance fiction, and more academics are taking it seriously. That said, of course, plenty of misconceptions about the genre and its readers persist.

    I have the impression that, whatever the flaws of their study, Ménard and Cabrera do take the genre seriously and when I read their article I was very pleased to see that they were willing to reassess their preconceptions on the basis of what they found. In addition, I was impressed that Dana Ménard engaged with me on this thread and responded to my criticism. So, like Jonathan, I hope Ménard and Cabrera carry on with their research on the genre.

    I don't want to minimise or dismiss your anger, but I'm reserving my ire for academics like Ogi Ogas. Sunita outlines the problems with his research at her blog. He seems to have analysed vast amounts of data but

    This is not academic research. This is not even academic research results popularized for a general audience. This is just someone with a half-baked theory and an unrelated Ph.D. trolling through mounds of data which may or may not support the theory. We can’t tell because we don’t know anything about the shape and scope of the data or how he analyzes it.

    And he gets to publicise this "research" at Psychology Today and in The Wall Street Journal.

    ReplyDelete
  39. Thank you for the link to Sunita's blog and the commentary on Ogas. I agree: Whatever Ogas is doing, it's not "research," and it's certainly not academic.

    But it seems to me -- and I'm sure the real academics here will correct me if I'm wrong -- that what Ménard and Cabrera have essentially done is put together a proposal for "real" research. And again, I don't have access to the whole article, so maybe I'm way out in left field here, but my impression is that this isn't in fact a proposal; it's presented as a completed study with a published article to show for it. And a few years from now, undergrads who are assigned a paper in their Women in Popular Culture class will cite Ménard and Cabrera that between 1989 and 2009, references to condom use increased in romance novels. Or whatever.

    When I used the Paizis reference in my thesis, I was able to point out what a horrible excuse for a sampling his was. Regardless what wonderful theories he posits about poetics and politics, how can he -- or anyone else -- draw any conclusions from that paltry a sample? And if academic research on romance fiction is on the increase since then, where's the follow-up analysis of Paizis 13 years on? Where are his theories applied to broader samples? To new subgenres like erotica and steam punk?

    Maybe it's not fair for me to beat this dead horse. Maybe I should be glad there's some, ANY, research being done, even if it's less than stellar. But in the decade-plus since I wrote my honors thesis, I thought the quality and depth of research would have progressed further than using a sampling of one book per year to draw conclusions about changing depictions of sex and sexuality in romance novels. Somehow or other, I suspect that if they'd chosen to analyze, say, depictions of sexual torture in spy thrillers or sexual encounters with non-humans in science fiction, no one would have accepted a one-book-per-year sample. But hey, with romance, they're all alike, right? You read one, you've read 'em all. :sarcasm: :grrrr:

    I just don't think saying Ogi Ogas is ten times worse is a valid justification for NOT criticizing Ménard and Cabrera. Instead of trying to quantify changes in depictions of sex and sexuality, why not do a qualitative analysis? Why does it have to be by the numbers? Why not look at the ways in which condom use is addressed in romance novels? (And this may have been treated better in the "Contextualizing Sex" article, rather than just the "57.9% of the time" according to Ménard and Cabrera.) Is it merely a reference to "a little foil packet" as I believe some of Jayne Ann Krentz's did? How often does the female partner initiate the discussion of whether a condom is used or not? Wouldn't that kind of examination of the literature be more interesting than dull numbers?

    And why is it that research about romance novels always seems to be about sex? Where are the discussions of non-sexual friendships, of sibling relationships (given all the series books detailing the lives and loves of seven brothers, seven sisters, 87 cousins, and all the in-laws), of hero and heroine to their parents, of hero and heroine to their children? How are widows and widowers depicted in romance novels? Ex-spouses? In-laws?

    I just don't think, dare I say it, shoddy research like this does the "academic" romance community much good. Is bad research better than no research at all? That's almost like saying an abusive husband isn't all bad because at least he pays SOME attention to his wife. I think we -- academics and non-academics alike -- deserve better. But that's just me.

    ReplyDelete
  40. "And why is it that research about romance novels always seems to be about sex?"

    I'd agree that quite a lot of the research into romances has focused on sex, but some of the psychologists who've written about the genre have focused on gender roles/relationship dynamics, not just sex. There's also been a fair bit written from a non-psychological perspective about a range of other topics, including capitalism and racism.

    "I just don't think, dare I say it, shoddy research like this does the "academic" romance community much good."

    Not do us much good in what way? Do you mean that it doesn't advance the study of romance?

    I think a lot of the academics who've worked on romance would not consider themselves to be part of an academic romance community. They'd probably see themselves as academics in their own disciplines (psychology, sociology, history, etc) who happen to have explored those subjects through romance.

    Then there are those of us who want to develop "romance scholarship" as a field. I'd like to think that we're growing in number.

    "Why does it have to be by the numbers?"

    Again, I'd go back to noting that academics who study romance come from quite a wide range of different disciplines. Some of those disciplines will use more statistics than others.

    ReplyDelete
  41. Hi Linda,

    I'd be happy to send you a copy of the finished article and of my research proposal explaining why we went about things the way we did, if you'd like. The short answer is that we are two PhD students with no funding at all and very little time on our hands. We assumed (perhaps wrongly) that some information was better than none, hence the small sample.

    Although there may have been more research done on romance novels from different academic perspectives (e.g., literary criticism), there was really very little that had been done from the lens of Psychology. Our goal in doing this project was essentially to start the discussion within our field.

    My e-mail address is dana.menard@gmail.com. Feel free to drop me a line and I'd be happy to continue the discussion with you.

    Dana Menard

    ReplyDelete
  42. Hi Everyone

    Thank you for your feedback on our article. I truly enjoy romance novels and it was a pleasure for me to have the chance to study them from the perspective of academic psychology.

    C. Cabrera

    ReplyDelete
  43. Well said C. Cabrera, Academic psychology is the main focus of this proyect.

    ReplyDelete