Loving and Hating Straw Dogs: the Meanings of Audience Responses to a Controversial Film
Summary: This essay evaluates a recent BBFC-sponsored study of responses to the film Straw Dogs (among others), and compares its methods and findings with a research project conducted in Aberystwyth. Through the presentation of the Aberystwyth research, a model of audience responses is proposed, and the film itself reconsidered. This essay is presented in two parts. The second part will be published in the next issue of Participations.
Keywords: audience research; Straw Dogs; research methods; controversial films.
In late 2002, Sam Peckinpah’s (1971) Straw Dogs was finally released on DVD, after a long period of unavailability. The film was one among a number which aroused considerable controversy in the late 1960s to early 1970s, over their presentation of (in particular sexual) violence. Often discussed in books and magazines as an exemplar of a troublesome film, Straw Dogs has mainly been remembered for one scene, in which Amy, a central character in the film, is raped by two local men. It was not merely that the scene is long that made it such an issue, it was that Amy apparently eventually derives pleasure from the first (although definitely not the second) rape. Given the news that a new film of the book on which it was based, Gordon Williams’ The Siege of Trencher’s Farm, is soon to be released, this is perhaps an opportune moment to report on some research into actual audience responses.1
The British Board of Film Classification, having originally permitted its cinema release uncut, long withheld a video classification, on the grounds that the film might conceivably offer to some viewers a justification of the ‘male myth of rape’ – that women who say ‘no’ secretly mean ‘yes’, and can enjoy rape as rough sex. This judgement derived much of its logic from the BBFC’s attachment to a particular style of media research and theorisation which derives from the United States mass communication tradition. Their argument depended on two key steps. First, an authoritative body must find a potential ‘message’ in the film. Second, they identify an implied audience for that message – in this case, an audience combining two characteristics: they are ‘weak’ in the sense that they are prone to being aroused and persuaded simultaneously by an act of viewing; but they are also strong, in the sense that they are potentially dangerous. The third, implied step – the hunt to locate that audience – is virtually the history of this tradition of research.
Within film studies, the judgements were not much kinder. Straw Dogs suffered on the one hand from being treated as a ‘typical Peckinpah film’. Peckinpah was until recently commonly tagged as ‘bloody Sam’, the choreographer of violence. Here was the man who hymns the decline of the West in The Wild Bunch, and in the course of that celebrates male violence. On this kind of reading, Straw Dogs has been judged as a para-Western, the women within it marginalised and victimised. In the same period, feminist critics have vilified – and then dismissed – Peckinpah’s work as a whole, and Straw Dogs in particular. Straw Dogs exemplifies something worse than the normal sexual objectification perceived in the ‘male gaze’. Here, it is active male revenge. The savage male conquers the initially unwilling, ultimately giving woman, and initiates her into sex. Implicit claims about the audience work as strongly here, albeit with a different speculative-theoretical base, as in the policy positions of the BBFC. Carol Clover’s now-classic book on the role of women in horror films discusses Straw Dogs several times in passing, her longest comment on it being the following: ‘The rape in Straw Dogs is a classic in the “asking for it” tradition: Amy goes bra-less and flaunts her body in front of the local men, and when they undertake to rape her, her “no, no” turns to a “yes, yes” (so during the first man’s turn, in any case). Director Peckinpah is quoted as saying “there are women, and there’s pussy”, and his Amy is “pure pussy”.’2 This highly tendentious dismissive account of the film, attaining rhetorical force from that one-liner from Peckinpah3, typifies one major strand of judgement. More recently, a series of re-evaluations of Peckinpah’s work have sought in different ways to reclaim Straw Dogs from such criticisms. Based on a combination of critical biography and textual criticism, these books are part of a wholesale reconsideration of his work. Although, as will become evident later, I see great virtue in these analyses, I am approaching the film from another direction altogether – from that of that ‘missing’ audience.
Of course alongside, and in complex ways connecting with, ‘mainstream’ concerns about rape on screen, is a history of feminist arguments about rape. Taking its distinctively modern shape through the work of writers such as Susan Brownmiller, Catherine McKinnon and Andrea Dworkin, rape became a central motif of the development of the theory of ‘patriarchy’. ‘Patriarchy’ is a notoriously slippery term, denoting at times little more than a catalogue of men’s unfair treatment of women, at other times becoming a full-blown social theory, in which ‘Rape is nothing more or less than a conscious process of intimidation by which all men keep all women in state of fear’ (Brownmiller, p. 14). Although the actions of particular men at the moment it happens, every rape is a re-invigoration of an overall male domination – therefore all men benefit, and are complicit in the act. This radical feminist theory famously extended itself from the act to the representation in Robin Morgan’s ‘Pornography is the theory, rape is the practice’. I do not intend here to try to survey the range of subsequent debates within feminism over these claims. I note one point only before taking as a case-study one very recent and thoughtful exploration of the complexities of the issues in here. The point I raise is that in Brownmiller’s foundational argument is that required move: that from the dominating male sex in general, to the individual man. This feeds into the debates about representations of sex, and perhaps especially rape, in and through an assumed identity: between a ‘male spectatorial position’, and the individual male viewer. The substantial debates within feminist film theory, following Laura Mulvey’s ridiculously over-influential essay, have of course addressed the implications of Mulvey’s claim that women enjoying mainstream films are being ‘masochistic’. It is hard to think of an equivalent argument over actual male viewers.
A very recent and thoughtful book provides an opportunity for revisiting this topic. Tanya Horeck has re-examined the history of feminist debates about rape, and has argued very effectively that within all formulations within the debate is an element which has not been made fully explicit – that rape is always and necessarily hermeneutically defined, and therefore has to involve representation. It involves it in the sense that it always involves an act of story-telling, an account of the relations between the involved males and the females. It involves an imagining, in Brownmiller, of a primordial encounter. It involves real court-room and other battles over whose ‘story’ will be accorded truth-status, and indeed what will count as ‘truth’ in a situation of structured sexual inequality. Most importantly, it increasingly involves multiplying layers of circulating stories. In her book, Horeck explores (among other cases) the complexities of representation involved in the moves from the events in Big Dan’s store in New Bedford in 1984, which led to the world’s first live televised rape trial, and four years later to the release of the film The Accused – all of which turned on the issue of the relations between raping and watching rape: issues of ‘spectatorship’. Was it a crime to watch a woman being raped? Could the men who watched, then be ‘witnesses’ to the truth of the events? If we then watched the film version of it, were we as viewers implicated in their watching? Why else did we need to see the rape at the end of the film? In another rich, complex investigation Horeck examines the 1999 Florida fraternity house case, in which a group of students filmed themselves having sex with a young woman over several hours. Her subsequent charge of rape led to the release and wide circulation of their film, and ultimately to its incorporation within a documentary Raw Deal which was broadcast on British Channel 4.
Without engaging in much theoretical exposition, Horeck’s book is nonetheless heavily dependent on – and pretty much assumes the validity of – psychoanalytic approaches, as of course does so much feminist work on textuality. Here, this shows particularly in two ways: first, in her frequent turn to the concepts of ‘spectatorship’, and ‘voyeurism’. Horeck insistently, and rightly, asks the question: what does it mean to watch a rape? (Actually, she does also consider the issues in reading about rape, but like many other theorists tends to regard to the act of viewing as somehow automatically special, and specially voyeuristic.) What I find interesting and curious in her generally excellent analysis, is that this question remains essentially rhetorical. It is something to be posed, and if answered at all, posed as a textual question, something to be answered by consideration of the positioning devices of the film, or whatever. At the end of her discussion of The Accused, Horeck writes:
Moving away from a generalized consideration of the film’s depiction of rape as either ‘positive’ or ‘negative’, I have examined how its representation of sexual violence, as well as the commentaries on it, bring to light cultural unease concerning gender, race and ethnicity. The representation of rape continues to be one of the most highly charged issues in contemporary cinema. And while popular images of rape will undoubtedly continue to be decried for sensationalism and exploitation, what needs to be explored further is how these images open up wider questions about changing viewpoints on the relationship between audience and film. (p. 115)
These would be wise words if they did at least recognise the importance of doing actual audience research. In fact, Horeck stays firmly on the ‘safe’ terrain of talk about ‘the spectator’ and ‘his’ voyeurism – thus once more eliding the gap between the putative ‘male’ interest in rape, and actual men. I want to operate in this gap – and I want to do so provocatively by, precisely, considering what it means for a man to enjoy watching a film which includes rape.
This essay attempts a wide reconsideration of Straw Dogs, through a number of distinct but interrelated stages. I first contrast two recent pieces of research into audience responses to Straw Dogs: the first of which was carried out at the behest of the British Board of Film Classification; the second conducted by myself at the University of Wales, Aberystwyth. The latter deployed a methodology for investigating audience responses which, I argue, can be the basis of a distinctive re-examination of the film itself. From the findings of this audience research, I look at the film itself, drawing first on an approach proposed some years ago, in a different context, by John O Thompson.
How this study came to be.
In late 2002, an opportunity arose to conduct a piece of research on audience responses to Straw Dogs, using concepts and methods deriving from an approach which I have been developing and testing for a number of years. This opportunity arose entirely by accident. Early on in our first year film course at Aberystwyth (which has just over 300 students) we screened A Clockwork Orange. This was partly in order to provoke a discussion around issues of morality on screen, and censorship, and partly to lay the basis for a year-long research project that we had proposed to our field’s national Learning and Teaching Subject Centre. The aim of this was to carry out a piece of action research on how students cope with the relations between vernacular and academic understandings of film. The intention had been to ask them to watch Kubrick’s film twice – once just to garner their personal reactions, a second time guiding them into how academics have analysed and debated the film.4 However, we discovered that Film Four were due to screen it the same week as our intended second screening. Because of this, I asked the students to vote whether they would rather watch it the second time (knowing the brief) or if they would like to watch instead another film which has often been discussed in connection with A Clockwork Orange. Straw Dogs, I explained, had just been re-released after a long period in which it had been banned, and could provide an interesting further case for us to discuss. A class vote showed an overwhelming preference for watching Straw Dogs, with under 10 preferring to watch Kubrick’s film again. I closed this session by warning that some people might find Straw Dogs difficult to watch (without specifying any particular reason), therefore making clear that attendance was entirely optional.
Although no count was taken, a very large number – probably in excess of 250 – attended the screening. I said very little by way of introduction, except to give a date to the film. I started the film, then left, as normal. However, it happened that I returned to the screening theatre some twenty minutes before the film’s end, to find the room in uproar, with large numbers of people shouting out at the screen, occasionally laughing, sometimes yelling in shock. The reactions so fascinated me, that I decided to take a chance on telling them, as soon as the film ended, that I would like to hear their reactions to the film. I had realised, as I stood watching their reactions to the last few minutes of the film, that it would take very little to turn the questionnaire which we had just used with A Clockwork Orange to the purposes of Straw Dogs. However, having thought about the intensity of the reactions I observed, I chose to alter some questions to emphasise their potentially difficult and strictly personal character. Within a few hours I had emailed the group, with the questionnaire attached, asking if they would be willing to complete and return it via email, making clear that unlike the earlier questionnaire this was not part of their course, and guaranteeing that all responses would be immediately anonymised. Ultimately, over a period of just over a week, exactly 60 responses came in5 – not a huge number, but enough to enable an exploration using a research methodology which I have been developing for some time.
It is not often that a chance arises to compare directly what kinds of knowledge and understanding two contrasting methodologies can deliver. In this case, it has happily arisen from a serendipitous coincidence. At the exact point when I was analysing the outcomes of my opportunistic research, I received a copy of the findings of a BBFC-funded project which also explores audience responses to Straw Dogs. From time to time, the BBFC commissions research to provide a basis for its judgements. On this occasion, it asked the Communications Research Group (headed by psychologist Guy Cumberbatch) to research the ways in which people wish the ‘line drawn’ for sexual violence on screen. Cumberbatch is a significant figure in this kind of research. A major critic-from-within of psychological work on ‘media effects’; he nonetheless shares a good deal of their methodological and epistemological assumptions. Cumberbatch’s research is important enough, in my view, to warrant serious attention. Because of the comparison with my own study of audience responses, and because it can be a basis for observing a series of both substantive methodological issues, I have to examine it at considerable length.
The BBFC/CRG Research
The CRG’s research6 is a very good example of the kind of research most commonly undertaken into media audiences. It is good, in the sense that it is very well done, but also in the sense that it thereby displays the inherent problems in this kind of work.7 Shaped by the need of the BBFC to be able to say how far it is ‘mandated’ by public opinion to continue regulating the availability of sexually explicit content on screen, the research sets out to discover the public’s view on ‘acceptable limits’ [2002: 1]. We will see in a moment that there is a strange paradox within this goal, which infects the research in significant ways. The conclusions it offers fit with striking convenience with the general tendency of current BBFC thinking. The report concludes that people generally tolerate screened sexual violence ‘so long as it was justified in the story-line and it was “in context”’ [2002: 2], but at the same time there was widespread concern over who might get to see such videos, and a belief – which they report to have been strengthened as a result of taking part in the research – in the importance of restricting under-18s from seeing them. I am not interested in doing some ‘hatchet job’ on the research – the opposite. I hope to show that we can learn a great deal more by taking it seriously and seeing its qualities as research. That way, a number of issues emerge into clear view which allow us to see why this kind of research is doomed to missing the point.
Cumberbatch reports first on a survey of video renters, undertaken at 15 outlets and using a structured questionnaire. A total of 277 people (55:45 men to women) were classified by age, sex, frequency of video viewing, and by their attitudes to the current levels of control over film content either on video or in broadcast media. From this, a smaller sample of 50 was selected, with a deliberate slight skew in order to achieve a better gender balance, whom they interviewed by telephone. This was followed by a less-clearly explained selection of a few people to discuss the issues more fully in focus groups. Cumberbatch stresses that the people they studied are not entirely ‘representative’ of national opinion on issues such as sexual violence on screen. Wider research has shown that those who go to the cinema and rent videos tend to the ‘liberal’ end of the public opinion spectrum. From a quantitative analysis of the survey, some preliminary findings and relationships emerge – for instance, the assertion that ‘the overwhelming support (by 75%) to protect children is impressive’ [2002: 8].
The section on the survey closes by pointing to a series of ‘promising attitude predictors regarding regulation and the right to see graphic material’ [2002: 18]. These cluster around people’s strength of interest in seeing the film-types for which they indicate a preference. With the one exception of ‘Story-led’ films (which reversed the trend in all other categories, and looks like it holds a high proportion of older respondents who tended to be more conservative), those with high interest in their preferred category of films, or indeed in knowing about the research’s target range of films, veered quite strongly towards the ‘right to see’ position. What exactly this means, and how we might draw any conclusions from these results, is not at all clear.
Other interpretative work on the survey bears the same difficulties, for all that the findings look promising. For instance, the survey showed that ‘heavy’ film viewers and video renters were more likely to agree that adults have the right to see graphic portrayals of sexual violence; while in the other direction, viewers with strong religious views tended to assert that controls are currently too weak. These are hardly surprising findings, and of course beg interpretation. Are ‘heavy viewers’ numbed, or do they just know their films better? Are religious people saved by their moral frameworks, or are they more prone to want to determine the moral choices of others? On these, of course, such research cannot help. But there are dangers when the categories used seem to imply judgements. For instance, what is the difference between being a ‘heavy’, and being a ‘frequent’, or an ‘experienced’, film viewer? And what sly implications may accompany calling people ‘risky’ (as against, say, ‘curious’ or ‘experimental’) viewers because they will sometimes rent videos about whose content they know nothing? Albeit in small and deniable ways, judgements are implied by these category-names. The fact that these are inherited categories, widely used in for instance George Gerbner’s ‘cultivation analysis’ research, does not relieve the problem; it merely indicates its depth.8 A small related problem arises from their account of the ethical dimensions of the research. They announce, with apparent pride, that they closed the whole research process by using professional counsellors to check whether any involved in the latter stages had been at all ‘scarred’ by participating. This is interesting – not wrong in itself, but it still presumes that the expected mode of effect is one of harm, as against, for instance, feeling better able to perform as citizens, or having enjoyed the opportunity to participate, or having discovered dimensions of film which had previously been closed to them, or etc. In such small pre-categorisations are the first indications of bias to be found.
The key research stage comes after the survey, when those 50 people representing a range from ‘liberal’ to ‘conservative’ were asked to view, on their own, three from among a series of films which the BBFC have found problematic: A Clockwork Orange (1971), Straw Dogs (1971), Last House on the Left (1972), I Spit On Your Grave (1979), Death Wish II (1982), and Baise Moi (2000). Once the films had been viewed, people were telephoned and asked to give their responses to a checklist of questions about these films. The report presents the percentages of participants who take up different positions (eg, of Straw Dogs the proportions who believed that ‘the film gives the message that women might enjoy being raped’), and in the subsequent analysis cross-matches these in a variety of ways, often pointing up puzzles and apparent paradoxes among these results
It is important to point to some of the striking findings offered. The first half of the report offers a series of quantitative cross-tabulations of survey responses (for instance, the relations between film preferences, and attitudes to the right to see sexually violent materials). And some interesting findings do without question emerge – for instance, the strong ‘predictive’ relationship between high interest in seeing films within one’s preference categories, and belief in adults’ right to watch sexually violent materials if they choose to. Again, with the telephone interviews, the report presents the results of simple statistical analysis of responses. These are coupled with sample quotes illustrating the various positions adopted. What is striking is the report’s attitude to some of these findings. While the support for protection of children is called ‘impressive’, as we have seen, the cases where results point in other, more complicated, directions are repeatedly called ‘intriguing’ or more often ‘puzzling’.9 These ‘puzzles’ arise when the research appears to display inconsistencies or contradictions within people’s thinking. For instance, despite strong support for the propositions that Straw Dogs gives messages ‘that women might enjoy being raped’, ‘that when a woman says no to sex she might really mean yes’, and ‘that women like being knocked around a bit during sex’, still, 77% believed that the film ought to be released uncut as an ‘18’ on video (a figure curiously not labelled ‘impressive’ by Cumberbatch).
Let us reflect on what is happening here. The research is evidently working with a model which presumes that its respondents hold a series of equally-grounded ‘opinions’, and whose responses therefore ought to be consistent. This is the only ground on which it makes sense even to note apparent inconsistencies. The fact that at key points they are not so consistent poses a problem. Where to go next? One way would be to deny there is a problem in the first place, and see people’s proffering of ‘opinions’ as essentially context-dependent. But that would undo the entire research programme to which CRG belongs, for it would see ‘opinions’ as research artefacts. A related response would be to enquire into the status of the several inconsistent assertions, to find out what they mean to people, and whether they seem them as inconsistent. Again, however, that way lies a different research regime. It is hardly surprising that Cumberbatch takes neither of these routes. In fact, overtly, he takes no route at all – appearing simply to note the inconsistencies, and leave them ‘dangling’. But right at the end of the report comes an unargued ‘solution’, which turns apparent contradictions into marks of audience ‘maturity’ on the issues. Offering the overall conclusions to the research, Cumberbatch writes: ‘A number of participants had said that they did not think there were any general features (such as famous director, art house) which would normally influence their judgements about a film’s acceptability. Their sound advice was that decisions could only be made on a film-by-film basis. At first sight, this seemed to be so true of most members of the viewing sample that the variety of classification recommendations seemed to swamp the expected individual differences. Fortunately, the macro analysis of all of the viewing judgements revealed a far more consistent pattern where many of the expected differences between groups emerged quite clearly. However, equally important was the flexibility shown by individuals in their judgements about individual films. These demonstrate that few participants could be accused of following their beliefs to the point of prejudice’ [2002: 64 – my emphasis, MB]. Suddenly, what had begun as puzzling inconsistencies have become celebrated as flexibility and lack of dogmatism.
These points do matter. At several points during the presentation of the research, Cumberbatch offers readily quotable ‘conclusions’. For instance, early on, he cites people’s responses on the amount of regulation there is, and should be, for sex, violence, and sexual violence on screen. Cumberbatch comments on the finding that the last was seen differently by many: ‘Evidently the mandate for more liberal policies over the portrayal of sex is not sustained for sexual violence’. The problems with the lurch to a mini-conclusion ought to be obvious: what counts as ‘sexual violence’ has not been addressed. It has yet to be established that any of these are coherent categories. And of course a great deal of the remainder of the research, if taken seriously, proceeds to undo just those assumptions. For if anything is meant by the finding that people tend to measure the appropriateness of acts against their context of showing, at the very least what we have here is an undoing of the cohesion of that category ‘sexual violence’. What that high figure for regulation of ‘sexual violence’ almost certainly shows, is that more people don’t like the idea of appearing to consent to unbridled scenes of assault and rape on screen. Therefore, almost certainly with fewer empirical examples to hand, many have played safe and criticised it. But even to admit that possibility is to undo the security of such ‘mandates’.
Let’s return to that notion that people are simply being asked their ‘opinions’. In the telephone interviews, those who had viewed Straw Dogs were asked their reactions to the following propositions: ‘The film gives the message that women might enjoy being raped’; ‘The film gives the message that when a woman says no to sex she might really mean yes’; and ‘The film gives the message that women like being knocked around a bit during sex’. My problem here is not that these are leading questions. It is that they can only be answered at all if one has taken up a very peculiar, indeed very ‘British’ orientation to this or any other film. To watch a film for its ‘messages’ is to watch in a very peculiar way. It is to the credit of the research that it does note the tensions that result. On each of the three questions, responses were sharply divided, with 60% agreeing or strongly agreeing with the second. Yet, as they themselves have noted, 77% felt it should be released on video, either uncut or with minor cuts.
In fact, the report does present some further materials which, looked at closely, run sharply contrary to the emphasis on ‘messages’. Right at the end of the report, Cumberbatch reproduces a substantial chunk of the discussion from one of the two focus groups, which they held with the intention of gathering people of opposite tendencies in the hope that this might bring into the open and accentuate the positions which are generating disagreements. Cumberbatch reports that almost the opposite happened – discussants veered towards a point of agreement. The section quoted centred on a discussion of Straw Dogs, which interested the research just because responses were so paradoxically opposite to what they might have expected. Here is a film which, on the BBFC’s account, is particularly problematic, because of the scene in which Amy is raped and appears to respond with pleasure. The BBFC, operating within this ‘find-the-message’ framework, found plenty. But people discussing it didn’t seem to want at all to talk in these terms. What interested them was to try to make sense of Amy’s motivation, her relationship with Charlie, the man who rapes her, and how this scene contributes to our understanding of them.
What the CRG’s methodology prevents them from seeing, is that here in this conversation is a practical working example of the ways in which people work in and through a ‘context’; and that this very process cuts against looking for and finding ‘messages’. These two belong to different discursive worlds. The first is an illustration of an everyday process of making meaning from a film. The second is an application of an external way of worrying about films, which frequently collides with people’s vernacular talents for making sense.
As I’ve already said, I believe this research remains valuable, and am not wanting to dismiss it. But I am concerned at a number of levels with what it doesn’t and cannot do – because the way its questions are posed, and the way its methodology is framed, actively discount other possibilities. And these possibilities are at the heart of the research which I have undertaken. I clearly have, therefore, to mount a critique of the self-imposed limits of this research.
The problems start at the outset. There is a singular oddity in the way in which the research question is posed, indeed in its very conception. What exactly do the BBFC hope to learn through this? The grounds on which they may make their judgements are two: taste and decency; and probable harm. These must be held separate, because they are entirely different kinds of proposition. The first depends upon good accounting with ‘public opinion’. The second depends upon expert evidence.10 It is absolutely of no relevance if a large number of people believe, or indeed don’t believe, that there is harm. Yet the research is framed around propositions that wholly conflates the two. And those propositions take us back to the problem we have already encountered – of seeing films as ‘vehicles for messages’. Framing the report at the beginning, Cumberbatch reports that the BBFC ‘has always maintained a strict policy on the portrayal of rape and sexual violence, most often insisting on cuts particularly to acts considered by their treatment to eroticise or endorse sexual assault’ [2002: 3]. What would it mean to do research to check whether they had a mandate to continue this policy? This is to reach to the core of the issues here. Are the BBFC’s methods for determining such eroticisation or endorsement relevant to the ways in which lay viewers of the films make meaning, gain enjoyment and understanding from such films? To find that out, the research would have to put at risk the very notion of ‘messages’. Instead, this research spins uneasily between being a public opinion poll (do large numbers agree generally with the policy of the BBFC? If they do, OK) and research into how people perceive and understand films containing sexual violence.
There is no way out of this conundrum without radically altering the nature of the research. It is to mapping and then illustrating an alternative that I now turn.
Cumberbatch emphasises that while this is qualitative research, it has virtues that much of such research lacks: ‘Most qualitative research does not show how representative the quoted views are and, indeed, may give undue weight to the most articulate voices’ [2002: 23]. This is an important criticism, and one that others have made in different ways (see for instance Deacon et al.’s criticism that qualitative media researchers often embed para-quantitative claims in their accounts (‘some’, ‘a good many’, ‘very few’, etc), without going through the tests normally applied to quantitative research). But the criticism turns on what kinds of question we may want to answer, and what kinds of knowledge we may want to gain. The problem, I would argue, is not that qualitative researchers do not provide proper quantitative validation of their materials, but that they have not in general developed an alternative approach within which it becomes proper to choose for particular attention certain responses, and where there are checkable procedures for so choosing. In the absence of that, Cumberbatch’s criticism can look strong.
But if that is so, then in the reverse direction it is appropriate to comment on problems in the analytic framework which Cumberbatch himself uses. Or rather, the absence of a clear analytic framework for listening to people’s talk. He does, it is true, helpfully group all the elements of some responses – notably, all the occasions on which people uses one cluster of terms: ‘gratuitous’, ‘pointless’, and ‘unnecessary’. But while this is interesting, it takes us very little further in understanding what people are doing when they use such words, or what is the status of such uses. A discourse analytic approach would have taken note of the ways in which terms like ‘gratuitous’ operate as conversation-closures. They do not attribute any particular quality to a film – they remove it from analytic attention.
I’ve said that Cumberbatch has no declared analytic framework for listening to how people talk. That is only partly true. There is one small, semi-detached indication of a methodology which is in fact quite troublesome. In his initial quotes of people’s comments on the listed films, he introduces a distinction between ‘restricted’ and ‘elaborated codes’. Here is his explanation of this:
Many of the film descriptions were quite clinical, with little evidence of abstraction or affective responses. In order to summarise the patterns, these accounts were classified in terms of the kind of language used: as essentially restricted code or essentially elaborated code. Those judged to involve essentially restricted code language (37% of all descriptions) used basic vocabulary and mainly concrete descriptions (such as, ‘and they went out, and they killed her’). Those considered to show distinctively elaborated code (12% of all descriptions) tended to use more complex vocabulary, conceptual synthesis and evidence of abstraction. The remainder (50% of all descriptions) were not easily categorised and are referred to as ‘average’. [2002: 24]
Why is this troublesome? In the actual research, little further use than this simple classification is made of these terms. Its effects, then, if any, are primarily negative – using this system forestalls others being used. But actually I think a little more is at work here than is at first evident. For a start, this distinction has a clear history, deriving from the highly controversial work of Basil Bernstein. Bernstein uses these terms to mark what he claimed was a distinction between two class-based styles of speech: a working class mode of speech which was essentially descriptive, and which limited the thought of its users to the concrete situation; and a middle class mode of speech which transcended the concrete and sustained abstract thought.
But there is a second problem, perhaps more immediately germane to the application of this distinction to responses to films. In calling descriptions of the film ‘restricted’, Cumberbatch is denying the possibility that the way in which a film is described may suggest, presume or directly indicate the nature of character motivations – and that this may be how, for those most involved in a film, meaning is made. By treating the more distanced judgemental responses as ‘elaborated’, with the hint of greater achievement that this carries with it, once again Cumberbatch is unwittingly validating an approach that goes ‘looking for messages’. Simple involvement in the film, with responses that therefore simply adhere to the film’s unfolding narrative, is not for his research worth much consideration.
The Aberystwyth Study
It is, for me. I have become interested precisely in how we may get inside the ways in which people, ordinarily, arrive at understandings and judgements on a film – of whatever kind.11 How they therefore find ‘messages’, if they do. Or indeed other modes of making a film meaningful to themselves.
The methodology I now use calls for a number of distinct stages (although in practice they may not be undertaken absolutely sequentially). The aim ultimately is to answer this set of questions:
to identify the interpretative ‘moves’ that different audience members deploy to generate a working understanding of a film, and how (far) these cohere into an overall account of the film – this is what I mean by the term ‘viewing strategy’;
to identify the range of such viewing strategies, with which different people approach and seek to engage with the film;
to identify the costs and benefits of each, upon encountering the film – what elements of the film become visible and salient, and what pleasures and understandings vs. frustrations and disappointments result from each kind of encounter;
to find out how far, and in what ways, different viewing strategies are mutually aware, and take account of each other;
from these, to what extent is it possible to identify and itemise the conditions necessary for a wholeheartedly positive participation in the film?
A large amount of conceptual work underpins these questions, some of which has been elaborated in previous work.12
The questionnaire (see Appendix 1) first invited people to allocate their responses along two key dimensions (Enjoyment, and Admiration, of the film), but in each case inviting them to explain what they mean and intend by this self-allocation. This was followed by a series of open-ended questions, asking first what they knew about the film, how they might summarise it, and how they had personally felt about two (to me, key) parts of the film – the rape scene, and the ending. There followed an open question inviting them to say anything else that they felt was particularly important about the film that explained their reaction to it. Finally, there was a request for some minimal personal information (sex, age-bracket, the kind of area they had grown up in, and whether they had studied film before now). The self-allocation allowed me to group their responses, and thence to see to what extent common kinds of response unite a category; thence again to compare categories. The aim is to see how far it is possible to glean a sense of the modes of participation and the strategies of viewing that people adopt, and how these contribute to their eventual judgements on the film.
The requested demographic information is identical with what we sought for A Clockwork Orange. It was kept minimal, partly because it is not the centre of interest of this research, partly because I wanted to keep the questionnaire short and light. In one way, I was mainly interested in the sex of respondents, given the nature of the film. But I didn’t want to ask only about that, not least since that might seem to presume that this is seen to be the only ground of discrimination of responses. But actually, given the themes and setting of Straw Dogs, the request for information on where people grew up (city, suburb, small town, or countryside) might have been very interesting. In the event, it hardly was. Finally, students were asked to compare Straw Dogs with A Clockwork Orange, and to say, on a simplified scale, how they had classified their own responses to the latter.
Among the 59 analysable responses, 36 were from men, 23 from women. Overwhelmingly, the respondents were 17-21 (just 5 were in other age groups). The figures for area of origin were: City = 11; Suburb = 13; Small town = 28; Countryside = 7 (a preponderance which fitted with our general sense of our recruitment profile). Previous experience of studying film produced 22 positives, 37 negatives. Asked their self-allocations for A Clockwork Orange, 41 had Enjoyed & Admired; 11 had Not Enjoyed but Admired; 3 had Enjoyed but not Admired; and 4 had neither Enjoyed nor Admired – an overall more positive rating than for Straw Dogs, as we will see.
Table 1 (below) shows the results of looking at self-allocations: