In the first half of this century, cartoonist Harold T. Webster became famous for his sharp depictions of the affective hang-ups of the average American. He drew cartoon series under telling titles such as "The Timid Soul" and "The Thrill That Comes Once in a Lifetime", and he had no mercy for the TV audience. He never owned a TV set himself and was thus no regular practitioner of the art of TV viewing - an art that, when he died in 1952, was rapidly becoming a general domestic activity in the United States. Nevertheless, he felt sufficiently qualified to visualize what he believed happens when people watch TV. In a once-a-week cartoon series for the New York Herald Tribune, he presented taunting portraits of the TV and radio audience. In a relentlessly mocking manner characteristic of cartoons, Webster pictured the broadcasting audience "at its moments of greatest strain: clubbed senseless by commercials, drowned in the soap opera flood, lacerated by thrillers, held slack-jawed and limp before the endless, banal assault on ear and eye and mind", according to Time magazine. 1
Television's entrance into the fabric of everyday life was accompanied with great ambivalence in postwar America. The medium was not simply promoted as a new technological and cultural blessing to the public at large; instead, it provoked deeply contradictory expressions of hopes and fears, excitement and anxiety. 2 Webster, considered by Time to be "one of the nation's best and best-known critics of radio and television", 3 was undoubtedly only one of many cultural critics who contributed to this mood of distrust and caution. What makes his series of cartoons particularly interesting, however, is that he explicitly made the medium's unknown relationship with its audience the central object of concern. In other words, the cartoonist's critical discourse established itself by pictorially fixing the medium's presumably distressing effects on its ordinary consumers. And he was not alone in this. Indeed, as Lynn Spigel has shown, the popular press at that time was full of articles which expressed worry about the problems that might arise with the spread of TV as a mass consumer item: there was widespread concern about TV's possible disturbance of family harmony and children's education, about the medium's power to distract the housewife from her daily domestic tasks, about the problems the TV set would pose for home decorating and architecture, about TV's damaging influence on people's health (eye strain!), and so on. 4 Such worries indicate that the anxiety around TV primarily found expression in a sense of suspense about the impact of the medium in the private sphere. Because TV was set up as a domestic medium, it was hard to assess and control the process and context of its consumption. It is not for nothing that Webster called his cartoon series "The Unseen Audience". This title suggests that the TV audience is an invisible mass, hidden behind the closed doors of private homes, virtually inaccessible to the outsider.
What Webster did then was to take up the outsider's perspective, from which he drew scenes and events that depicted the strainful confrontations of people with TV in their homes. In this process, he did not implicate himself: his own position was that of the knowing subject who stands at a comfortable distance from what he visualised. Thus, through his cartoons, he established a symbolic relationship with the TV audience: a relationship characterised by distance, non-involvement, skepticism. Rather than empathising with those whose activities and experiences in front of the TV set he strived to represent, he constructed images of them as strange and alien human beings with whom no-one would like to identify.
Webster's pictures of the TV audience still feel surprisingly familiar today because we recognise in them the stereotypical type-casting of TV viewers as not-so-smart, naive and pitiful cultural dupes. The popular (self-)derogatory term 'couch potatoes' to designate people who watch (too much) TV in America is only one illustration of the persistent cognitive power of the way in which critics continue to imagine the TV audience.
It is important to emphasise that this image of the TV audience is not just stereotypical because of the negative typifications that emanate from it (stereotypes can also contain positive idealisations), nor simply because it produces extremely partial fantasies of how people watch TV - fantasies that, in their exaggerations and exclusions and their emotionally-charged value judgments, are a combination of distortion and validity. Decisive for the stereotypical quality of this image is the rhetorical tone of certainty by which its crude simplifications are presented, and by extension, its short-circuiting effect: not a trace of doubt that it tells as essential truth about a group of others. 5
In her article "Rethinking Stereotypes", Tessa Perkins has remarked that "[a] stereotype will probably develop about a group because it has, or is presenting, a problem". 6 The core of the 'problem' of the TV audience, I suggest, lies precisely in its invisibility: it is an evasive, intangible reality that is hard to define. We generally refer to the TV audience as if it were a self-evident category, but as with all common sense, its meaning seems utterly obvious until we have to define exactly what we mean by it. Exactly who or what is the TV audience? The difficulty of finding an unambiguous and final answer to this question is a problem for those who want such an answer. In such a context, stereotypes are useful because they reduce the uncertainty, ease the anxiety for the unknown. As Sander Gilman has noted, "the stereotype is a momentary coping mechanism, one that can be used and then discarded once anxiety is overcome". 7 The production of stereotypical representations, then, can be seen as a discursive strategy to come to terms with the invisibility of the TV audience.
The invisibility of the TV audience is closely related to the already mentioned domestic context of TV consumption. And precisely the fact that people generally watch TV at home, is felt as problematic and threatening because it renders the TV audience out of control - in a double sense: both chaotic and beyond control. Consider, for example, the following quote from a horrified cultural commentator, written up in 1958 in the New York Times:
The scene in front of the living-room TV screen is apt to be much different [from that of the public space of the cinema]. Casualness of dress is apparent, for one thing, and the master of the house may be unshaven and the mistress of the house in curlers or facial cream. Food snacks and liquid refreshments are often on hand. When the commercial appears, the husband may rise to put another log on the fire and the wife to attend to something in the kitchen that needs attending to. During the performance itself silence is not at all rigorously observed. Remarks of displeasure at a hammy line or gesture come forth uninhibitedly from those gathered in front of the TV set. 8
This fantasy displays an acute awareness of the lack of order and discipline that characterises the everyday setting in which people watch TV. From this commentator's perspective, watching TV is a 'wild' practice as it were: the living room is imagined as a barbaric space, full of 'wild', uncivilised viewers.
It must be noted, however, that this problem of the 'wild viewer' is not just a cultural question, the focus of moral concern over the presumably debasing effect of TV. More fundamentally, it poses a very material, practical problem for the TV producing institutions. This is because, in the words of John Hartley, TV institutions "are obliged not only to speak about an audience but - crucially, for them - to talk to one as well: they need not only to represent audiences but to enter into relations with them". 9 To put it simply, the 'wild viewer' is a problem for the TV industry because the TV must 'catch' viewers and have an audience in order to survive.
In his article "The Imaginary Signifier", film theorist Christian Metz has identified a similar problem for the film industry. Here is how he has characterised the problem:
In a social system in which the spectator is not forced physically to go to the cinema but in which it is still important that he [sic] should go so that the money he pays for his admission makes it possible to shoot other films and thus ensures the auto-reproduction of the institution - and it is the specific characteristic of every true institution that it takes charge of the mechanisms of its own perpetuation - there is no other solution than to set up arrangements whose aim and effect is to give the spectator the 'spontaneous' desire to visit the cinema and pay for his ticket. 10
Conjured up here is the problem of institutional reproduction. The cinema can only continue to exist if and when enough people are willing and prepared to be regular members of the film audience, but the film industry does not have the means to provide itself with a guarantee that people will not stop going to the movies. The problem seems to be rather far-fetched one, because since the turn of the century, when the cinema first entered our cultural life, the world has obviously turned into a place full of filmgoing women, men and children. However, the principle of the problem is undeniable, and that it is not entirely hypothetical, is easily exemplified by the sharp and steady decline in cinema-going since the fifties, when TV made its entrance in people's homes. 11 At stake, then, is the institution's control - or better, lack of control - over the conditions of its own reproduction.
Broadcast TV faces the same institutional problem: it too cannot take its audience for granted. Contrary to other social institutions such as the school or the family, TV (as well as all other mass media) does not have the means to coerce people into becoming members of its audience. Television audience membership is not a matter of compulsion or necessity, but is principally voluntary and optional. Therefore, the TV institution is ultimately dependent upon people's unforced appetite to continue watching day after day. Again, the problem seems far-fetched given TV's manifest success in securing huge audiences for its transmissions, but this still does not mean that success comes naturally and effortlessly. On the contrary, numerous institutionally orchestrated activities such as the publication of TV guides, advertisements, and press interviews with TV personalities, as well as previews of forthcoming programs during an evening's flow, the use of teasing jingles, logos, and so on, testify to the enormous amount of money and energy being spent to reinforce and update people's desire to watch TV. 12 The very fact that these strategic institutional activities are of a continuous, never-ending character indicates that TV networks and broadcasting organisations know that they must continually find ways to attract the attention of potential audience members, because they cannot control them in any direct manner.
The audience, sine qua non for both TV's economic viability and cultural legitimacy, forms its ultimate insecurity factor because in principle there is no way to know in advance whether the audience will tune in and stay tuned. Audiences must constantly be seduced, attracted, lured. How-to-get-an-audience is, willy-nilly, the TV industry's key predicament, even though this not always acknowledged as such.
The seriousness of this predicament is deeply ingrained in the domestic content of TV consumption. Television's dominant institutional arrangement is embodied within the framework of broadcasting, a framework whose basic configuration has been extended, without any radical changes, to cable and satellite TV. According to Raymond Williams, the broad-casting framework is characterised by a "deep contradiction between centralised transmission and privatised reception". 13 This 'deep contradiction' refers to the circumstance that while TV is generally seen as a form of 'mass communication', no true communication - in the 'ritual' sense of that word: an exchange of meanings which is both intentional and interactive 14 - between the TV institution and the TV audience generally takes place. Broadcast TV transmission is both adamantly intentional and resolutely non-interactive: the diffuse and dispersed TV audience, locked in its condition of privatised reception, is an invisible and mysterious interlocutor. 15
Over the years, a range of risk-reducing techniques and strategies of regulating TV programming such as serial production, usage of fixed formats and genres, spin-offs, horizontal scheduling, and so on, have been developed. These strategies do not only serve as a way to facilitate the organisation and co-ordination of the industry's production practices, but are also aimed at the codification, routinisation and synchronisation of the audience's viewing practices, so as to make them less capricious and more predictable. In this sense, the iron repetitiveness of TV programming, especially in the most commercialised systems, can be seen as an instance of orchestrating and disciplining the people's leisure time - the time people are supposed to be 'free'. 16 But all these strategies can only help to manage, not remove the basic uncertainty with which the TV institution has to live. There are no guarantees that actual viewers will comply to the codes, routines and synchronicities of viewing behaviour as designed by the institutions. Ultimately, then, the problem of (lack of) control amounts to one thing: the impossibility of knowing the audience - in the sense of knowing ahead of time exactly how to 'get' it, and tame the 'wild viewer'.
This does not mean that no knowledge about the audience is produced in the multi-layered organisational process of TV broadcasting. On the contrary, both formal and informal knowledge about the audience is constantly operative in the complex decision making procedures which determine the shape and content of TV's daily output of programs. 'Know the audience' is the first basic principle every handbook for commercial broadcasting teaches the would-be TV programmer. And in this respect, one type of knowledge about the audience has over the years gained uncontested prestige and truth value in industry circles: the knowledge emanating from audience measurement.
Consider, for example, the language used by network executive Brandon Tartikoff, president of NBC Entertainment and responsible for airing much-acclaimed series such as Hill Street Blues, Cheers and Miami Vice. But this does not stop him from setting his standards within a discourse which problematises - and, as a result, stereotypes - 'the audience' as object of concern and control. He asserts:
I probably have more esoteric tastes than the average television viewer. I'll go to see Amadeus and pay my five dollars and fifty cents, but when the salesman from Orion [a production company] comes and asks me to buy it for the networks, I'll say no, because it's going to get a twenty-two share,
which is, he implies, not enough because 'as a programmer I had to ask myself if it was something that would get a thirty share or better'. 17
Thus separating personal taste and market judgment (and consequently creating an antagonism between 'us' [the industry] and 'them' [the audience]), Tartikoff self-evidently adheres to the principle of 'audience maximisation', which reigns to supreme in the operations of American commercial TV. The language in which he articulates this principle, is the quantitative one of 'shares' and 'ratings' - a language that is only made possible by audience measurement.
As a form of systematic research in which empirical information is gathered through scientific methods, audience measurement supplies a technical, formal and seemingly objective and factual kind of knowledge. The conventional assumption is that ratings figures indicate the popularity of TV programs, pure and simple. Apparently, this is a straightforward and unambiguous fact. Can we say then that audience measurement is the perfect solution for the TV industry to deal with the invisibility of the audience? And does the quantitative discourse of ratings escape the stereo-typing process we have discussed earlier?
A look at the recent history of how ratings are produced, tells us that this is not the case. It is a fascinating history - a history about the contradictory intertwinings of knowledge, technology and subjectivity - which makes clear that no matter how successful TV is in attracting viewers, it will never be able to completely domesticate the wild viewer. It is this history to which I will now, necessarily in an abridged way, turn.
Let us begin with a short exposition of how ratings data are produced. Historically, two major methods have been used: the diary and the setmeter. In the diary method, a sample of households is selected whose members are requested to keep a (generally, weekly) diary of their viewing behaviour. At the end of the week the diaries must be mailed to the ratings firm. In the second case, an electronic meter is attached to the TV sets of sample households. The meter gives a minute-by-minute automatic registration of the times that the TV set is on or off, and on which channel it is turned on. The data are transmitted to a home storage unit, where they are stored until they are accessed by the central office computer during the night.
The data so provided offer more than just factual information about who's watching what. They bring about a more general sense of knowing, and the reassuring feeling of certainly that goes with it. Consequently, their function is not merely instrumental, but ritualistic: one of the most important achievements of ratings is that they help establish a relationship between the industry and the audience, a relationship however that is not 'real', but symbolic. Through the discourse of ratings the 'TV audience' is represented as a unified totality, a thing than can be known in terms of size, profile, and demographic composition. An object the industry can define its relationship to by cutting it into segments or using it as a target, a commodity that can be bought and sold. It is thanks to ratings that TV companies can 'deliver audiences to advertisers', as the slogan goes. 18
But this audience object/commodity is a fictive entity. This does not mean, of course, that ratings dream the audience into existence. They are based on actual information on what real people do. The knowledge produced by ratings is therefore neither 'false' nor 'untrue'. On the contrary, ratings are so powerful precisely because of their ability to define a certain field of empirical truth. That regime of truth is fictive, however, in the sense that its very terms to describe TV viewers inevitably foreground very specific conceptualisations of the audience. For one thing, it is already a rather peculiar move to perceive the audience as something that can be measured - an assumption originating in the general idea of 'measurability of markets' quintessential for the parameters of marketing thought as it began to be developed in the United States in the early twenties. 19
However, 'TV audience' as represented by audience measurement is not a static, stone-like object whose characteristics can be described once and for all. It is continually changing, dynamic object that always seems to elude definitive descriptions. This slipperiness is already obvious enough in the fact that the production of ratings is an ongoing practice, repeated day after day. Ratings are very fleeting products: they become obsolete almost instantly.
A crucial factor in the dynamism of the construction of the 'TV audience' is induced by the subjective nature of TV viewing as such. However, inclusion of the singular concreteness of the heterogeneous practices and experiences of TV viewing would make the production of ratings utterly unmanageable. Therefore, audience measurement, as is the general rule in quantifying social science, is designed to abstract from the detailed singularities in experience and practice. In other words, in order to construct an objective 'TV audience', it has to squeeze the 'wild viewer' into a manageable and measurable mould.
The technical instruments used for audience measurement testify to this reductionist tendency. Nielsen's setmeter, for example, can only register whether the TV set is on or off. Here, then, 'audience behaviour' is implicitly defined as a simple, one-dimensional, and purely mechanical act. As Todd Gitlin has rightly remarked, "The numbers only sample sets tuned in, not necessarily shows watched, let alone grasped, remembered, loved, learned from, deeply anticipated, or mildly tolerated". 20 In other words, what the discourse of ratings erases from its field of discernment is any specific consideration of 'the lived reality behind the ratings'. TV viewers here are thus merely relevant for their bodies: strictly speaking, they appear in the logic of ratings discourse only in so far as they are agents of the physical act of tuning-in.
In short, what the discursive strategies of audience measurement succeed in constructing is a streamlined image of the 'TV audience'. This imagined streamlined audience is stereotyped as a 'disciplined' audience. It is constructed by ratings discourse through a smoothing out of problematic subjectivity and translating it into ordered and regular instances of viewing behaviour. In the streamlined audience, each viewer will ideally find her or his exact place in a comprehensive table of knowledge, formed by the central components of size and composition. After all, the quantifying perspective of audience measurement inevitably leads to emphasising averages, regularities and generalisable patterns rather than idiosyncrasies and surprising exceptions. What all this amounts to is the construction of a kind of map of the 'TV audience', on which individual viewers are readable in terms of their resemblance to a 'typical' audience member whose 'viewing behaviour' can be objectively and unambiguously classified. In other words, in foregrounding the stable over the erratic, the likely over the fickle, and the consistent over the inconsistent, ratings discourse symbolically turns 'wild viewers' into well-organised, serialised viewers, displaying dependable viewing habits and routines.
Imagining viewers in this way is very handy for the industry indeed: it supplies both networks and advertisers with neatly arranged and easily manageable information, a form of knowledge which provides not only a vision of predictability, but also controllability, of the audience. Empirically found variations within the streamlined audience are conveniently contained and fixed in a limited number of 'types' and 'patterns', developments over time are straightened out in terms of convenient 'trends'.
But TV viewers are by definition more than just 'typical' audience members. When people watch TV they are of course inevitably positioned as members of the audience, but they also always simultaneously inhabit a myriad of other subject positions such as parent, critic, fan, democrat, Southerner, or whatever - subject positions which elude the symbolic world of audience measurement. As a consequence the map of the streamlined audience as provided by audience measurement inevitably stands in a strained relationship with what actual audiences are up to. The map never quite fits, as it were.
But this epistemological gap does not necessarily constitute a problem for the industry. So long as the map works, the industry will not bother to look for more 'realistic' knowledge. The gap will only be problematised when the streamlining process tends to slacken; that is, when it seems no longer possible to establish fixed viewing habits, behavioural categories, and so on, by which viewers can be unproblematically typified and classified. At such moments, consensus over the map of the 'streamlined audience' provided by ratings discourse breaks down, and audience measurement is thrown into crisis. In fact, the late eighties are witness of such a crisis, and it is to this that I will now turn.
The rise of the so-called new technologies has provided people with new options and choices, especially the VCR and cable. An entirely different TV landscape unfolds before the viewer's eyes these days, one characterised by abundance rather than scarcity. And viewers seem to have responded by eagerly multiplying their viewing behaviours: they zip through ads when playing back their taped shows, they zap through channels with their remote controls, they time shift - in short, they have become less and less 'loyal' to the carefully composed schedules of the mighty networks. "After years of submitting passively to the tyranny of [network] television programmers, viewers are taking charge", comments Bedell Smith. 21
Indeed, from the industry's perspective, some sort of 'revolt of the viewer' seems to have erupted: they feel that they are losing their grip on the audience. And in this chaotic situation, audience measurement is a central focus of concern, the site on which the uncertainties and worries are expressed and articulated. What we witness here then is the industry's anxiety and uneasiness about the growing unpredictability of audience behaviour, or viewing habits, as a consequence of the fragmentation of the electronic media landscape. In short, what is at stake is a disruption of the streamlined 'TV audience'. Consequently, diverse branches of the industry began to call for more finely tuned knowledge of the audience, to be acquired through better measurement.
This call for better measurement was articulated by criticising the prevailing techniques and methods of measuring the TV audience: the diary and the setmeter. For example, the proliferation of channels through cable has acutely dramatised the problems inherent to the diary technique. All of a sudden, the built-in subjective (and thus 'unreliable') element of the diary technique was perceived as an unacceptable deficiency. Thus, officials of the pop music channel MTV complained that their target audience, young people between 12 and 24, consistently comes off badly in the demographic data produced through diaries, because "younger viewers (...) tend not to be as diligent in filling our diaries as older household members". 22
The video cassette recorder has also played a major destabilising role in the measurability of the TV audience. 'Timeshifting' and 'zipping' (fastforwarding commercials when playing back a taped program) effectively deregulate the carefully composed TV schedules of the networks. This phenomenon has come to be called 'schedule cannibalization' in industry circles, 23 a voracious metaphor that furtively indicates the apprehension, if not implicit regret, felt in network circles about the new freedom viewers have acquired through the VCR. Through the VCR, the wild viewer clearly manifests her/himself!
Wiped out then are the good old days when TV was a relatively neat and simply-structured business, with a limited number of channels and viewing styles for viewers to choose from, and in which the diverse branches of the industry could complacently believe in and rely on the premise of a streamlined 'TV audience'. That premise can no longer be taken for granted. Consequently, the one-dimensional assumptions about TV viewers that has been the traditional basis for audience measurement can no longer be held. Thus, behind the controversy over the diary methodology lurks a great suspicion of the capriciousness of the new viewing behaviours brought about by the emergence of the new TV landscape. Viewers just can no longer be trusted to report their viewing accurately: they lack perfect memory, they may be too careless. In short, they behave in too subjective a manner! In this situation, the sentiment thrives that there should come a better method to obtain ratings data. And better means: more 'objective' that is, less dependent on the fallibilities of viewers in the sample. A method that erases all traces of wild subjectivity.
It is in this situation that the ratings business has now come up with the people meter, a new audience measurement instrument that was introduced in the US in 1987, and that has already been in use in several European countries and in Australia as well. 24 The people meter is supposed to combine the virtues of the traditional setmeter and the paper-and-pencil diary: it is an electronic monitoring device that can record individual viewing rather than just sets tuned in, as the traditional setmeter does. It works like this. When a viewer begins to watch a program, she must press a numbered button on a port-able keypad, which looks like the well-known remote control device. When the viewer stops watching, the button must be pressed again. A monitor attached to the TV set lights up regularly to remind the viewer to the button pushing task. Every member of a sample family has her or his own individual button, while there are also some extra buttons for guests. Linked to the home by telephone lines, the system's central computer correlates each viewer's number with demographic data about them stored in its memory, such as age, gender, income, ethnicity and education.
The intricate measurement technique is attractive for the industry because it holds the promise of providing more detailed and accurate data on exactly who is watching what. The people meter boosts the hope for better measurement of the wider specter of cable and VCR viewing (including zipping). Smaller audience segments may now be detected and described, allowing advertisers and broadcasters to create more precise target groups. New sorts of information are made available, hitherto hidden and unknown minutiae of 'audience behaviour' are now becoming visible. The people meter, then, opens up the potential of drawing a new, more detailed map of the audience. In line with the cartographic metaphor, it could be said that the old map was no longer adequate as a guide in the rocky terrain of the TV business, because more traffic is now on the road, making careful driving more necessary. With the people meter, the industry sees the promise of a new map that should enable finding the correct signposts and avoiding the danger zones more efficiently.
Still, the existing versions of the people meter are by no means considered a perfect measurement instrument, as they still involve too much subjectivity; after all, they require viewer co-operation in the form of pushing buttons. A professional observer wonders: "Will the families in the sample really take the trouble? Will they always press the buttons as they begin watching? Will they always remember to press their buttons when they leave the room - as when the telephone rings, or the baby cries?" 25 It should come as no surprise, then, that furious attempts are being undertaken to develop what is called a passive people meter - one with no buttons at all - that senses automatically who and how many viewers are in front of the screen ... For example, Nielsen has recently disclosed a plan for a rather sophisticated passive people meter system, consisting of an image-recognition technique capable of identifying the faces of those in the room. The system then decides first if it is a face it recognises, and then if that face is directed toward the set (unfamiliar faces and even possibly the dog in the house will be recorded as 'visitors'). If tested successfully, Neilsen executives expect this system could replace the imperfect, push-button people meter by the mid-nineties. 26
These developments illustrate the attractiveness of the idea of a 'perfect' people meter in industry circles: so pressing is the need felt for exactly knowing who is watching what and how. But they are also reveal the more general 'political' issue involved in the audience measurement project as such. At stake in the turmoil around people meters, is more than just a methodological 'improvement' of measuring the TV audience: the significance of the people meter is not just a technical matter, but a matter of control.
Now that the industry as a whole is confronted with multiplying ways of practising the act of watching TV, a loss of manageability of the audience - a streamlined audience, one than can be objectified and acted upon - threatens to take place. The solution, of which the people meter is a historical embodiment, is sought in more and faster information, allowing for more microscopic differentiations and characterisations. A more detailed categorisation of the peculiarities of audience behaviour, so the implicit philosophy goes, will supply the industry with new symbolic means to sharpen its power to define the audience, and to specify its relation to it in more concrete ways. The people meter, as an idea, symbolises the wish for provision of a continuous stream of precise data on who's watching what, everyday, all year long. Its projected task is to put the streamline back in the profile of the 'TV audience'.
In trying to make sense of these developments in audience measurement, I am, in fact, reminded of Michel Foucault's considerations of "panopticism". 27 The principles of panoptic discipline are central to the technological operation of audience measurement; its core mechanism, and ultimate ambition, is control through visibility. Audience measurement is a form of hierarchical examination: its aim is to put TV viewers under constant scrutiny, to describe their behaviour so as to turn them into suitable objects in and for industry practices, to judge their viewing habits in terms of their 'productivity' for advertisers and networks alike.
But, unlike prisoners or schoolchildren, TV viewers cannot be subjected to officially sanctioned disciplinary control. In the school or the prison, disciplinary techniques are aimed at transforming people's behaviour through punishment, through training and correction. The living room however, is emphatically not a classroom or a prison cell, nor is TV a carceral institution. Therefore the commercial TV industry does not have the power to impose the conversion of viewers into what Foucault has termed "docile bodies" - that is, to force them to adopt the 'ideal' viewing habits (for example, watch all the commercials attentively) - to stop being 'wild viewers'.
This 'problem' - that is, the 'problem' that viewers are not prisoners but 'free' consumers - accounts for the limits of audience measurement as a technology of power. It is also against this background that the importance of methodological accuracy and objectivity may be understood: emphasising that audience measurement is a matter of research, increases its credibility and legitimacy and reduces distrust against it. All this amounts to the fact that audience measurement can only be an indirect means of disciplining TV audiences; it is through discursive, and not literal objectification and subjection that audience measurement performs its controlling function. It does not effect the actual discipline of TV viewers, it only conjures it up in its imagination. This leads to a fundamental contradiction in the very motif of audience measurement. Just as the disciplinary technologies described by Foucault, it puts viewers under constant examination, but contrary to what happens, for example, in the prison, the visibility of the audience achieved is not linked up with the organisation of direct behavioural control; the observation of bodies and their regulation do not go together here. This does not mean that there is no power and control involved in the set-up of audience measurement (that power is articulated, for example, in what audiences get to see on TV); it does mean, however, that the production of ever more refined knowledge as such becomes a rather autonomous pursuit: audience measurement is~~ carried out in the belief that the production of knowledge per se must somehow automatically lead to actual control over the audience. We will see, however, that the project has quite contradictory effects, not at all uniformly leading to the desired increased control.
No matter how sophisticated the measurement technology, there will always be aspects of the activity of watching TV that will elude the 'gaze' of audience measurement. The subjective moment can never be completely 'domesticated' in ratings discourse, because TV viewing is, despite its habitual character, dynamic rather than static, experiential rather than merely behavioural; it is a complex cultural practice that is more than just an activity that can be broken down into simple and objectively describable 'habits'. In other words, watching TV is an activity replete with significance; in its everyday uses it can take on a myriad of specific and changing meanings, which the sensors of audience measurement technology cannot possibly register fully.
It is true that audience measurement does not need to include all those heterogeneous uses and meanings of watching TV in its discourse. But as we have seen, it is exactly the disruption of the 'streamlined audience' by VCR and cable-related viewer practices that has impelled the development of more sophisticated techniques to put viewers through more meticulous rituals of examination. In other words, the recognition that watching TV is done by people of flesh and blood, and that 'TV audience' is not so easily streamlinable by ratings discourse as had been assumed, has resulted in an attempt to incorporate an ever widening range of components of TV viewing activity in the measurement endeavour.
But paradoxically the increasingly microscopic technological 'gaze' of audience measurement only seems to lead to an ever greater elusiveness of the invisible audience it is presumed to measure. The more it sees, the less it can get to grips with what it sees, as it were. The more visible the audience is made, the less certain it comes to define exactly what takes place in people's homes when they watch TV. No longer can it be conveniently assumed - as traditional ratings discourse does - that having the TV set on equals watching, that watching means paying attention to the screen, that watching a program implies watching the commercials inserted in it, that watching the commercials leads to actually buying the products being advertised ...
Thus, the activity of watching TV loses its imagined one-dimensionality: measuring 'it' can never be the same anymore. The calculable audience member tends to dissipate before the ever more sensitive microscope of audience measurement, and increasingly regains his or her status of active subject - of 'wild viewer'. Audience measurement, in short, is an example of how the practice of panoptic examination, when severed from the attendant power of regulating behaviour, turns out to have a contradictory outcome: rather than facilitating control, it makes it more difficult!
This whole turn of events, and the crisis of audience measurement it has generated, coincides, as we have seen, with the emergence of the new, postmodern TV landscape. The perceived 'return of the wild viewer' is not some sort of romantic eruption of viewers' rebellion on the basis of their 'authentic' needs and desires, but is brought to the surface by the very technological changes introduced by the TV business itself. Perhaps, a permanent disruption of the streamlined audience is bound to take place; perhaps the proliferating range of activities people perform with and around TV will increasingly resist being straightjacketed in a streamlined discursive construct, a fixed stereotype - in the literal sense of that word - of the 'TV audience'.
Brandon Tartikoff, NBC's President of Entertainment, testifies fully to the sense of change within the industry:
Lucille Ball said that TV changed with the invention of the remote control device. As soon as a guy doesn't have to get up from his chair to switch the channel, TV becomes a new ball game. Viewer inertia, which supported many an uninspired show, has given way to viewer impatience. 28
Of course, viewer inertia has never been 'real' in the first place. It was just the fictional stereotype of the streamlined audience member complacently and arrogantly indulged in by the industry. As for viewer impatience, it may be the new stereotype by which the TV industry seeks to combat its uncertainty about the TV audience - a fiction which, if anything, betrays a declining sense of confidence over the power of the medium. It is, in fact, an acknowledgement of the industry's plight always to be stalking the 'wild viewer'.
Large parts of this article are extracted from my forthcoming book Desperately Seeking the Audience: How Television Viewership is Known (London: Routledge).
1. See Time (22nd October 1951), p.83.
2. Lynn Spigel, Installing the Television Set: The Social Construction of Television's Place in the American Home, 1948-1955. Unpublished Phd. Diss., UCLA (Los Angeles, 1988).
3. Spigel,Time, ibid. (footnote 1).
4. Spigel, op.cit.
5. Of course, this is an essential rhetorical strategy by which cartoons in general drum up their message. In this sense, cartoons can be seen as a genre that relies upon, and contributes to the construction of stereotypes.
6. Tessa Perkins, "Rethinking Stereotypes" in Michele Barrett, Philip Corrigan, Annette Kuhn & Janet Wolff eds., Ideology and Cultural Productions (London: Croom Helm, 1979), pp.139-159.
7. Sander L. Gilman, Difference and Pathology (Ithaca and London: Cornell University Press, 1985), p.18. According to Gilman, only pathological personalities hold on to rigid stereotypes about a group they fear on a consistently permanent basis.
8. New York Times (11th February 1958).
9. John Hartley, "Invisible Fictions: Television Audiences, Paedocracy, Pleasure" in Textual Practice, v.1 no.2 (1987), p.127.
10. Christian Metz, "The Imaginary Signifier" in Screen v.16 no.2 (1975), p.19.
11. See Douglas Gomery, "The Coming of Television and the 'Lost' Motion Picture Audience" in Journal of Film and Video v.37 no.3 (1985), pp.5-11; David Docherty, David Morrison & Michael Tracey, The Last Picture Show? (London: BFI, 1987). Docherty et al criticise the technological determinism implied in the popular explanation of the decline of the cinema audience as being directly caused by the rise of TV as the most important mass visual medium. The authors claim that both developments can be explained, in Britain at least, by the same sociological factors, most importantly, the expansion of home-based consumer culture after World War II. Gomer discusses the American context of the same phenomenon.
12. See John Ellis, Visible Fictions (London: Routledge and Kegal Paul, 1982).
13. Raymond Williams, Television: Technology and Cultural Form (London: Fontana, 1974), p.30.
14. On the 'ritual' view of communication, see James Carey, Communications as Culture (Boston: Unwin Hyman, 1989).
15. It is worth noting here that TV was not naturally destined to be a medium for private, domestic consumption. Early experiments with TV technology were set up with several alternative uses in mind. Television's initial entertainment setting was that of public showings on large-screen TV in theatres, while it was also envisioned as a monitoring device for factory production and as a surveillance device in military settings. Furthermore, the possibilities of two-way TV as a replacement for the two-way telephone were explored by AT & T in the twenties, while radio amateurs were also enthusiastic about the interactive potential of TV communication. However, these alternative uses were finally marginalised in favour of a development of TV analogous to that of radio broadcasting. See Jeanne Allen, "The Social Matrix of Television: Invention in the United States" in E. Ann Kaplan, ed., Regarding Television (Frederick, MD: University Publications of America, 1983), pp.109-119.
16. cf. Chris Rojek, Capitalism and Leisure Theory (London and New York: Tavistock, 1985).
17. Quoted in Richard Levinson & William Lin, Off Camera (New York: Plume/New American Library, 1986), pp.256-257.
18. See, for some critical accounts of the functions of ratings in the operations of the (commercial) TV industry, Donald Hurwitz, "Broadcast Ratings: The Missing Dimension" in Critical Studies in Mass Communication v.1 no.2 (1984), pp.205-215; and Eileen Meehan, "Ratings and the Institutional Approach: A Third Answer to the Commodity Question", in Critical Studies in Mass Communication v.1 no.2 (1984), pp.216-225.
19. James Beniger, The Control Revolution (Cambridge, MA: Harvard University Press, 1986).
20. Todd Gitlin, Inside Prime Time (New York: Pantheon, 1983).
21. Sally Bedell Smith, "Who's Watching TV? It's Getting Hard to Tell" in New York Times (6th January 1985) E21.
22. Quoted in Victor Livingston, "Statistical skirmish: Nielsen cable stats vex cable net execs" in Television/Radio Age (17th March 1986), p.130.
23. See Edmond M. Rosentahl, "VCRs having more impact on network viewing, negotiation" in Television/Radio Age (25th May 1987).
24. I extensively discuss the introduction of the people meter in the American TV industry in Part Two of Desperately Seeking the Audience, forthcoming.
25. William F. Baker, "Viewpoints" in Television/Radio Age (10th November 1986), p.95.
26. "New 'People Meter' Device Spies on TV Ratings Families", in San Francisco Chronicle (June 1, 1989).
27. Michel Foucault, Discipline and Punish (Harmondsworth: Penguin, 1974).
28. Quoted in Levinson and Link, op.cit., p.263.
New: 20 December, 1995 | Now: 16 March, 2015