symbolic degeneracy and stardom

Stardom and symbolic degeneracy: Television and the transformation of the

stars as public symbols

BARRY KING

question of the ideological impact of the star system in its various ions - cinema, television, and video - is arguably a hotter topic

in the area of film studies than it has been in the past. Certainly if activity and publishing are reliable indicators of a deepening of

well-established popular interest, then we are in the midst of a miniin writing on the stars. 1 At the same time, this revived interest needs

reckon with the possibility that the phenomenon of stardom has itself with the advent of television. Certainly there is an argument,

needs empirical demonstration, that popular interest in stars has from a collective level of involvement with a 'powerless' elite toward individualized forms of interest that are at best only 'proto-populist'

1972: 75-98). For a number of writers who see themselves primarily as social critics

than students of the cinema, popular interest in the stars has into a narcissistic and privatized involvement that is more or

pathological (Baudry 1986; Mitroff and Bennis 1989; Postman 1985; ~~.>mckell986). This apparent fulfillment of the Frankfurt School's hypoth

of the link between capitalist development and the progressive desubof culture certainly entails a 'disenchantment' of stardom .

..-.rPr-mPr, it is routinely seen as demonstrating the validity of Andy dictum that fame will become a generalized commodity, promot

a restive form of solidarity between the famous and the obscure that reminiscent of a queue for a tum on a video game. Yet there is a need clarify whether the theorized change is as dramatic or as epochal as

or indeed, whether it is a new phenomenon at all. As those with .knowledge of cinema history will know, the 'dis-enchantment of the

'was a factor noticed in the 1930s (Thorp 1970). It is not my intention to address these matters centrally here, though I

by implication be led to argue that television may be prioritizing an of stardom that was latent in the era of the studio system. Thus my

will lend credence to the view that the way the stars are currently

Semiotica 92-1/2 (1992), 1-47 0037-1998/92/0092-0001 $2.00 © Walter de Gruyter

2 B. King

represented does indeed entail a change, though the question of such a shift parallels a change in audience involvement or promotes a change is not something that current research can answer with a degree of confidence.

If any generalization about the study of the stars is viable, it would that the topic continually threatens to expire from a surfeit of meaning Indeed, for many writers 'excess' is the essence of (and precisely radically exciting about) stardom. The stars refuse to sit comfortably cultural stereotypes, and there is a pleasure to be derived from the tance' this semantic fidgetiness implies. In theoretical terms, the . quality of stardom translates as polysemy. Thus Richard Dyer's ' text judiciously took as its organizing premise the notion that there not be a theory of stardom, and presented a collage of approaches

1979).2

While I would not want to propound the view that there should or could be one theory of stardom, it is another matter to deny that ever could be, a comprehensive theory that would locate (rather than a diversity of approaches. As is well known, there is no theory of though there are a few basic propositions that define via family resem a field of study (Bordwell 1989: 146-148). Is there a unified field of in the case of the stars? To date, and despite many efforts, I answer must be no. Moreover, it is necessary to recognize that paradoxical inversion in the study of the stars - an inversion would suggest is related to the ideological impact of the stardom anyway - whereby the more focused and substantive the of the topic becomes, the more elusive and 'polysemic' the phenor11 appears. Obviously there should be (and invariably is) a movement increasing complexity and nuance in any theoretical study: the pron from which the study of the stars started- for example, that stars consumerism, patriarchy, or mythic aspects of human experience the margin by rationalism- were (even if they weren't recognized at the time) research programs that would engender further scholarly activity went on (Haskell 1961; Lowenthal 1961; Morin At the same time, the dispersal of these primary simplicities that most people to the topic in the first place is an indirect recogni basic fact about the stars: namely, that stardom is a complex located in a multiplicity of determinations that stem from its in the spheres of production, textuality, and consumption.

If we recognize that stardom is more complex than earlier gested, the question of the roots of this complexity still remains. A complexity is no better than a dogmatic simplicity, and it is responsibility of any informed approach to sort out whether the

Stardom and symbolic degeneracy 3

arise from the deficiencies of theory itself or from the basic of the object of study. The recognition that stardom is a complex

process does not mean it is complex in the way a particular proposes. What I will argue is that the tendency to see stars as

of 'undecidability' rests on an endemic failure to integrate an of the nature of the actor as a performance sign into a field of that has so far detailed the function of the actor within the context

and distribution (Clarke 1989; Kindem 1982; King 1985). words, as I will attempt to demonstrate, the tendency to read 'wild' nemeses of ideology and control rests on the failure to the exact nature of the actor as a sign.

other area of determination that is central to stardom is, of course, I want to suggest that the admittedly thin literature on audience

toward stars will provide a prima facie case for a relationship of :aaaaation between the specific character of the actor as a sign and the

characteristics of audience response. That is, the stars are taken Jmu1ogically real individuals rather than as quasi-fictional constructs

by such individuals. Such a denotational foundation - if it is to flirt with architectural analogies in matters of semantics -inscribes a discursive limit on the ways in which stars can be and still be understood as specifically cinematic phenomena.

of the defect of earlier studies is that they assume that the discussion social significance of stardom can take place without discussing how

are constructed as signs in the first place. There is no question stardom is about social meaning, but there is a prior need to specify

in which, semiotically speaking, they accede to this level of the symbolic. Such an expunging of the specificity of stars as signs is in, to take one example, Orrin Klapp's Heroes, Villains and Fools,

the discussion of stars as social types - as found objects, as it were, horizon of celebrity- is conducted as if the cinematic provenance stars has no bearing on their social impact (Klapp 1962).

therefore, that the manner in which stars are constituted as within the media of film and television must be factored into an

of how they are interpreted. It would be nice to go further and my arguments to a detailed ethnography of audiences. But such with few exceptions, is a very recent development that has yet to considerable results (Seiter 1989). 3

thrust of my argument here is to specify how the stars work , always allowing that denotation and connotation are relative of textual meaning. Thus from the perspective of the audience,

appear as the finished products of a semiotic labor, as givens, they are analyzable as complex signs with their own processes

4 B. King

of connotation and denotation. In this sense, I am arguing for the view that stars constitute a basic ontological level that audiences take as given, prior to the development of secondary connotations (Condit 1989; RossiLandi 1979; Withalm 1988). I would not want to claim that the discussion about audiences offered here is complete; only a detailed ethnography of how audiences concretely interpret films can deliver this kind of knowledge. What I have to argue herein may hopefully delineate the final stage of a production-based theory of stardom on which a more comprehensive study of audiences can selectively draw. But at the same time, it is something of a general embarrassment that after the resurgence of the topic in the late 1970s, even a basic statement such as the following seems dubious and tautological: 'Stars matter because they act out aspects of life that matter to us; and performers get to be stars when what they act out matters to enough people' (Dyer 1986: 19).

In what follows I will move from a general theoretical discussion of the role of reference in the formation of the visual sign, arguing from what can be called, roughly, a N eo-Bazinian approach to the semiotics of cinema. The key differences in my account stem from three lines of argument: First, I approach the issue of reference from a baseline in Peircean semiotics rather than Bazin's generalized, if suggestive, emphasis on an anthropological impulse toward preservation- the Mummy complex (Rosen 1987). Second, I will center my remarks on the actor as performance sign, and indeed argue in general that the film text is not a self-contained semantic system, as the term text would tend to imply, but rather a fractured intertext of performance signs, cinematic signs, and signs that are particular neither to the cinema nor to performance. To these 'significant' departures from Bazin, I must add a third: since my interest in stardom is finally -much like that of other students of the phenomenon - an interest in the role of the media in maintaining or challenging social order and social inequality, it is necessary to recognize that television is now centrally implicated in the production and reproduction of fame and celebrity. Accordingly, I will contrast the role of the actor in the cinema with changes effected in that role by television.

The nature of symbolic degeneracy

It is well known that Saussurean semiotics have a singular disadvantage in relation to the kind of sign that concerns me here: the visual sign. For Saussure, the paradigm sign is the verbal sign wherein the relationship between the signifier and the signified - a significant phonic combination and the concept or mental image this combination evokes -is arbitrary.


The phonic signifier 'cat' does not resemble, is not causally connected to the real creature in the world of things. Neither does 'cat' resemble the signified concept 'felinity'. Therefore, the verbal sign can be said to lack motivation.

For Saussure, what defines a sign qua sign is its conventionally defined connection to its concept and the corresponding referent. Such a quality is most emphatically realized in the case of the verbal sign. As crosscultural comparison reveals, the connection between the phonic materials of the verbal sign and its concept or referent exhibits great variability, even though within the confines of a particular culture such a connection assumes the force of 'nature'. If, with a few debatable exceptions such as onomatopoeia, arbitrariness is seen as the essential element of a sign, then the development of semiology could be free from the burden of elaborating a theory of extensional meaning- how terms relate to external reality. In other words, the process of meaning-making - signification - can be considered in isolation from matters of mimesis or reference. In the case of visual signs, where issues of resemblance are unavoidable, Saussure tended to doubt that such phenomena were signs at all or that they were amenable to a rigorous semiological scrutiny. Accordingly, images were assigned a marginal role as a kind of para-sign (which he termed symbol) that 'mimicked' as it were from a junior partner status the mature capacities of the verbal sign (Eco 1979: 54ff; Hervey 1982: 13-16; Todorov 1982: Chapter 9).

In contrast, the American semiotician C. S. Peirce provides a better point of entry into theorizing images as signs. This is because Peirce's approach allows that, under certain circumstances, the material relationship between the sign and its referent can play a constitutive role in meaning. For Peirce, unlike Saussure, signification is a product of a trichotomous relationship between a sign (representamen), an interpretant (not a person, but a concept or another sign), and a referent or object (CP 2.228; Greenlee 1973: 28-33; Zeman 1977).

A sign mediates between the interpretant and its object- and in this sense is 'something which stands to somebody (or could) for something in some respect or capacity'. Eco adds on the basis of an established social convention (Eco 1979: 15-16). In Peirce's approach it is necessary to regard signs from three different aspects: (1) the relationship to the referent/object; (2) the relationship of the mediating sign to its mode of appearance or being; and (3) the relationship to the interpretant. This relationship defines the kind of concept a particular sign embodies- an instance, a proposition or an argument. This dimension is implicit in the foregoing, but it leads on to questions about the logical status of a sign and will not be explored further here.

6 B. King

In relation to the referent, Peirce classified signs in terms of the familiar schema:

Icon - a sign that denotes by virtue of a real similarity that holds between some physical properties of the sign and some physical properties of the object.

Index- by virtue of a real cause and effect link, or contiguity. Symbol - by virtue of a general association of ideas; a convention or

habit or rule. At the outset it is necessary to forestall the notion that the terms icon,

index, and symbol define different types of signs. Depending on the mode of sign production, a specific sign can be concurrently related to its object through relations of motivation (resemblance, causal connection) and conventionality. Moreover, relations of motivation and conventionality overlap. As St. Augustine pointed out with regard to visual signs, everyone seeks to make such signs resemble their objects. But what constitutes resemblance varies greatly, because things can be similar in a large number of ways. Therefore the choice of the criteria of resemblance is largely a matter of convention, even if such a choice has the force of a law - e.g., the depiction of Christ (Todorov 1982: 49).

A specific medium determines as it were according to its own limitations (e.g., in the case of the cinema, two-dimensionality) the pertinent elements of similarity, even if the universe of features selected from relate, as they usually do, to cultural norms that govern perception in general. Once the medium, so to speak, has selected the features by which it establishes a relationship of reference, these features reciprocally impose a constraint on its operation, since they determine what is or is not a valid usage. Obviously, notions of what is valid usage change, but the context of change is always overdetermined by established usage. Thus in the mainstream cinema, which has embraced the tenets of illusionism, stylistic innovations tend to derive their impact from their perceived flaunting or radical intensification of the power of the image to engage the real.

The triad icon, index, and symbol, therefore, defines not types of signsin practice Peirce speaks of iconic-indices, indexical symbols, and so onbut dimensions of motivation or conventionality observable in a given sign which are more or less marked and variously combined. Further for Peirce, the question of the degree to which a sign is motivated or conventional is related to the mode of being implied by its materiality. This mode of being in turn determines the extent of a particular sign's access to the symbolic. Crudely, verbal signs are sui generis symbolic whereas visual and aural signs are only contingently symbolic (i.e., when they are used or interpreted in a symbolic context, failing which their semantic status remains intrinsic or motivated). To express this even more concisely, we can say that visual


signs are internally motivated, while verbal signs are externally motivated (i.e., derive their connectedness from an established usage within the culture in which they operate). Such a usage or conventional understanding regulates the relationship between a particular signifier and the category to which it belongs. Any particular sign is an admixture of the internal and external in this sense, and its functioning or semantic impact will depend upon the extent to which its materiality determines its access to the level of the symbolic. To make this explicit, it is necessary to relate the triad icon, index, and symbol to Peirce's second triadic distinction: the relationship of a particular sign to its mode of being. The modes of being are as follows:

Firstness - a mode of being which is such as it is, positively, without reference to anything else, mere qualities or percepts - iconic mode.

Secondness - a mode of being formed by a relationship or interaction with some other thing- this is also in·a sense a moment of mutual or dialectical determination, as for example an existent person in the capacity of buyer implies another person in the capacity of seller and so on -indexical mode.

Thirdness - the mode of mediating between two terms by means of a third term which produces recognition and understanding of a general relationship- symbolic mode (CP 2. 233-271; Hervey 1982: 29ff).

Iconic signs are Firsts - an icon is actually a quality or percept, hence Peirce's term qualisign - and strictly speaking not signs at all but units of experience (e.g., redness). Since qualities are not encountered outside of a particular embodiment in an event or an object (e.g., as a red ball), qualisigns are principally encountered as indices or sinsigns. (In this connection Peirce distinguishes icon from its embodiment, which he terms a hypoicon.) Indices are sinsigns or unique particulars without meaning outside the fact of their occurrence or existence. In order to acquire the status of signs (as opposed to mere existents), sinsigns must be subsumed under a general class or type that situates them as a token of a type. In other words, sinsigns only obtain generality if they are placed in a symbolic relationship. This subsumption of a particular (index) under a sign with generality, a symbol or legisign, in effect identifies the sign as belonging to a class of objects or events which have some general meaning and participate in a larger semiotic process: a red ball of a certain material, size, and density which is part of a game (e.g., billiards), which in turn expresses a particular culture (Greenlee 1973: 48).

In an echo of Saussure's equivocation between sign and symbol, Peirce's classification of signs in terms of modes of being undermines to some extent his classification of signs in terms of their relation to their objects. If taken literally, qualisigns and sinsigns (i.e., icons and indices) are not

8 B. King

signs at all. A sign is always a Third, or a Iegisign, a term which mediates between two other terms. Thirdness is a relationship of mediation which by definition entails a particular in a relationship of generality, and the mode of being of icons and indices means that unlike verbally based signs, they can never in and of themselves be symbols (Hervey 1982: 23). They only acquire a symbolic status by being inscribed within a specific semantic field. Such fields can range from a cognitive schema, an exposition, to a variety of narrative forms or an entire iconographic tradition. On the other hand, Thirdness- the defining property of symbols- cannot exist outside of particular instances as indices, which in turn are only instances of particular qualities or icons, so that there is an immanent possibility that the semantic status of a symbol can 'implode' from generality toward particularity. Since the term implosion is taken from Baudrillard, let me point out my polemical intent here. For Baudrillard implosion is the erasure of reference in favor of the signifier. My usage implies that the opposite process is just as likely- signifiers degenerate into particulars (Baudrillard 1983; Kellner 1989).

To simplify even further, we can say that for Peirce there are only two classes of signs - indices and symbols - since percepts or icons are not encountered outside of singular embodiments as indices or exemplary embodiments as symbols. These two kinds of sign have a potential for contradiction and contrariety that any symbolic process must 'solve' and domesticate - i.e., raise to the level of its own inherence.

This latter process of slippage from the symbolic is defined by Peirce as symbolic degeneracy (Peirce 1955: 101ft} As Peirce defines it, a genuine symbol is a sign which by means of a general set of rules or conventions refers to a general object. Genuine symbols are universal and collective rather than particularistic and idiosyncratic. To anticipate, character actors tend to be genuine symbols.

By contrast, a degenerate symbol is a sign which, although governed by some degree of conventionality, nonetheless lacks generality. Symbolic degeneracy is defined by two kinds of symbols:

(a) The singular symbol, which refers to an existent, apparently unique individual, pointing or referring to qualities that only one individual may realize- for example, a proper name or other forms of indexical markers. Film stars approximate to singular symbols.

(b) The abstract symbol, which refers to highly generalized properties such as blondeness, lovability, generosity, etc. that apply to a very wide range of objects and entities that resist placement in any particular class or category, seeming to have a trans-categorical reference. Celebrities approximate to abstract symbols.

Stars, celebrities, and actors (as character actors) constitute representa-


tiona! positions that circulate around the notion of the person as a coherent entity. Thus genuine, abstract, and singular symbols are moments in the construction of stardom and celebrity, not distinct states of being instanced by one individual - though some individuals seem to lay claim to one status rather than another. Compare the persona of Robert De Niro to that of Vanna White.

To sum up: in any sign or sign complex, levels of particularity and generality interact in a variety of ways to determine the status of a given sign. This status depends on the sign's characteristic mode of accession or articulation, given its determinate basis in Firstness, Secondness, or Thirdness, into the level of the symbolic and how contingent this relationship is. The status of a particular sign, therefore, is a function of the interaction between (1) its materiality and the extent to which this materiality places a limit on its autonomy vis-a-vis a referent or the referent world; and (2) the extent to which a sign event or object can be understood as an example or token of a general class of events or objects or as a selfsufficient or unique particular. In the first case, a sign pertains to the status of a symbol; in the second, to an index. The situation of symbolic degeneracy is the situation in which a sign located in a general field of meaning implodes into a particular.

The mediating instance between the materiality of a sign and its categorical status is the code (a set of rules for construing meaning). Symbols tend to be governed by strong codes even in those cases where resemblance is a powerful factor in the formation of the signifier. On the other hand, the weaker the code, the more resemblance derives from the materiality of the object - this is particularly true where it is assumed that signs should reproduce natural perception. Conventionality and motivation are not, therefore, polar states, but rather modes of representation that can substitute or compensate for each other in the making of meaning.

In a strict logical sense the relationship between a sign and its referentor in the Saussurean schema, between a signifier and its object - cannot determine the corresponding signified directly; a relation of reference cannot determine a relationship of meaning, it can only realize this meaning as an empirical fact. But the fact that signifieds have no existence, not to mention social effectivity, outside of a particular embodiment or signifier means that the realization of the signified is conditioned by the nature of the signifier, its materiality and relationship to a code. This relationship applies particularly to visual signs, where the relationship of motivation is internal; but even in the case of linguistic signifiers, where the relationship of motivation is external, the case of neologisms points to a similiar situation (Ducrot and Todorov 1979: 100-1 02).

In sum, visual signs tend, unless textually based mechanisms of connota-

10 B. King

tion intervene, toward particularity, whereas verbal signs tend toward generality, even in those cases (e.g., personal pronouns) where particularity seems evident. Thus the 'I' of verbal discourse- which, as Benveniste has argued, is an example of an indexical 'sign', since it refers to the one who makes use of the language system on a particular occasion - is at the same time transpersonal and general: namely, as an example of a general class of objects, human beings in the capacity of the speaking subject. Likewise the language of intimacy ('I love you') is at the same time a transpersonal anonymous expression in the way the image of the loved one cannot be (Benveniste 1971: 217-222).

!conicity and materiality

With these general remarks and terms in place, I want to deal with the process of symbolic degeneracy in relation to so-called iconic signs, specifically photographically based signs. To approach this effectively I need to clear away one last possible major source of confusion: is the distinction between the linguistic and visual sign the same as the distinction between a symbol and an index? In principle the answer must be no: the distinction between tokens and types, particulars and generals is pertinent to visual signs for the obvious reason that any image can be deployed symbolically: e.g., identifying mugshots of a felon can be used to signify the category depravity, a child's finger painting can connote innocence, and so on. But in the case of photographic signs, the centering of the mode of sign production on indexicality means that the referent is the determining instance of the connotative action of the medium (Santaella Braga 1988). An exception might seem to be the fashion photographer, where style and formal treatment function as indices; but would Helmut Newton be Helmut Newton if he did not record the sexual fantasies of the rich and famous? Likewise, the differential symbolic weight of two identically posed photographs, even if we assume a maximum of formal manipulation in each case, of the family Trubble and the First Family would derive from their respective referents. Comparatively speaking, one would refer to the Bush portrait as symbolic and the Trubbles' as evidentiaL But equally the Trubble family album would render up many symbolic relationships - an example of a petit bourgeois WASP lifestyle, the social eye of the amateur photographer, the power of the patriarchal Trubble to compel underlings to do symbolic labor, etc. - but always from the baseline of evincing, rather than summating as in the case of the First Family, a categorical quality: Nationhood, Leadership, and so on.

What these considerations suggest is that if the semiotic status of the

UNiVERSITY OF CANTERSUiZ•

Stardom and sy7nfblzt ~&em~JY 11

UBRAR'G sign (as an index or a symbol) is determined by the materiality of the medium in which it is produced, the symbolic content of the sign is determined by the evidential status of the referent. Should the status of the referent change, so does the meaning of the photograph. One might reasonably object that such a condition is true of verbal signs; e.g., the term 'crack' no longer refers to what it referred to in 1960. But the essential difference lies in the gearing of the relationship between the sign and its referent. Any sign can acquire a new signified, but in the case of a photographic sign, this acquisition is more evidentially constrained than in the case of verbal signs. In the case of photographic signs there is a particularly strong triangular relationship of reference. The referent, if it is to serve a semiotic purpose, compels the sign to refer to what it refers to at any particular point in its semantic history. 4 In cinematic terms, one might consider how Eisenstein, arguably a formative director par excellence, engaged in frenzied searches for concrete visual particulars or indices to underwrite abstract concepts - going to extraordinary lengths to find the actor Mgebov in order to deploy a shot of his 'insane eyes'. This fetishism of visual detail expresses the 'non-indifferent' nature of things, given that the director or scriptwriter has committed the signifying process to visualize a particular abstract quality (Bellert 1984: 39-52; Eisenstein 1987: 3-4, 11, 181; Lotman 1976: 86).

But if the argument deployed here is not necessarily verbocentric, there is a sense in which in Western culture at least the symbolic is practically equated with the verbal or the linguistic sign. The sovereignty of the verbal over the visual in our culture - a sovereignty that can be regarded as species-specific - is to some extent an expression of functional power. Verbal languages operate as a universal commentator or meta-code - the code into which all other codes (visual, aural) are habitually translated, and indeed, the only code in which such a translation can economically be effected. Verbal languages can say what other codes can say, and can also say things that other codes can never say. Thus, for example, there is no simple visual equivalent of the statement: 'the boat docking at Pier Six is three and a half minutes late'. Because of this combination of concise detail and flexibility, the verbal is often regarded as the ground of meaning per se (Metz 1980: 56-68).

A more accurate statement would be - and this is a truism given the basis of the cinema in the screenplay and shooting script - that visual signs are constituted to conform to a linguistically based meaning, and where they are not they remain subject to verbal forms of interpretation and appraisaL The appropriation or decoupage of the visual continuum is a process of inscribing the semantic content of the linguistic signified into a semantically dense visual object or event strip. Such an inscription is

intrinsically reductive, but it does not follow that it necessarily occurs in one direction socially speaking. Practically, as feminist critics such as Teresa de Lauretis have argued, this verbally based confinement falls most emphatically and selectively on women. Women are not only more likely to be read as corporeal rather than subjective entities, but this objecthood is centered, in the service of verbally based prescriptions, on highly selective aspects of the female figure - though in recent times Stallone and Schwarzenegger may have accomplished a similiar reduction in masculine identity.

The tentative conclusion on the question of the relationship between the verbal and the visual is, therefore, that because of the dominance of the verbal over the visual (and, a fortiori, the aural), images are usually rendered as symbols through inscription in a linguistically dominated semantic field. At the same time - if one exempts metaphysically postulated unobservables or protocols of logic - visual signs and the visual per se constitute the ratifying instance of linguistic propositions. There is a sense, of course, in which every thing becomes a sign when it is spoken of, written, or even thought about from inside the 'prison house of language'. But this absolute foundational condition should not obscure the fact that signs 'impose' characteristic constraints on the manner of their subordination to the word.

With this hierarchical bias in view, we can now return to the matter of visual signs. The general relationships that govern the production of iconic signs have been spelled out by Eco (1979: 178-220, passim). Such phenomena, which in fact embrace a very large category, are not best viewed as signs at all, but as sign-functions articulated around two broad relationships.

(1) On the plane of expression, whether or not a particular expression can be referred unambiguously to a code or code-system to determine its meaning or settle its status as a sign. In other words, is the particular expression a clearly defined token (e.g., a footprint) of a clearly understood class or cultural unit of meaning (e.g., domestic cat)? Or is the token ambiguous in relation to existing codes- a footprint unlike any knownso that the token becomes a sign of itself or a class of one (i.e., has meaning on account of its self-sufficient particularity - the footprint of 'IT')? (Another possibility is that there exists no clearly defined code.)

In such circumstances, Eco speaks of two kinds of type/token ratios: First, where an image can be readily produced or understood according to agreed conventions or a code - a situation of ratio facilis. Second, where there is an expression content which is new or uncoded, or -particularly in the case of visual signs - represents a semantic nebula rather than a unit, so that the production of the image or its interpretation


is not subsumable under an agreed code. In such circumstances, meaning must be derived from the token itself, so that the distinction between type and token is effectively (if temporarily, since a new token tends to create a class) collapsed- a situation of ratio difficilis.

(2) On the plane of content, whether or not the particular sign qua sign shares material properties with its object, or more precisely, whether the sign is constructed out of the same materials as might be the object in the external world. Once again the notion of similar properties cannot be taken in the literal substantial sense - i.e., the jdentical substance. Since the character of a sign is to operate in absentia, the literal presence of an object does not occur in a relationship of signification. Rather, it is a matter of the articulation of conventions that govern sharing - what elements of the real object (size, shape, texture) enter into the formation of the signifier, what is the degree of sharing, and to what extent does the use of the medium accede to the parameters of natural perception in shaping its own signifier?

According to Eco, it is necessary to distinguish between a formal and a substantive aspect of material sharing, even though these are phenomena of degrees of sharing rather than exclusive modes. This distinction mirrors a similar one made by Hjelmslev between the substance of form and the substance of content. 5

The formal aspect of material sharing, or what Eco calls heteromateriality, indicates a circumstance in which a sign is produced from a material which is radically different from the material of the referent, but shares certain formal characteristics - relative size, color, shape, etc. Since the selection of the formal markers of similarity is by definition guided by conventional criteria, these criteria may be more or less explicit, as in the case of strongly coded symbolic systems - e.g., Kabuki theater or Classicism in the plastic arts. Alternatively, this relationship may be more or less implicit in the case of the application of procedures that mirror natural perception, such as naturalism in the theater or 'objectivism' in forensic photography. But whether explicit or implicit, the relationship between the sign and its referent is evidenced by shared formal properties that are held to centrally define the category. In this sense, heteromateriality works via the construction of prototypes of resemblance as defined by Rosch (Rosch 1976; Thompson 1985). The paradigm case of heteromateriality is the replica - e.g., a matchstick model of Chartres Cathedral.

The substantive aspect of material sharing, or what Eco terms homomateriality, indicates a circumstance in which the sign vehicle is not merely prototypically related to the referent through formal markers, but also literally has the same material as the referent. This material can be 'stuff' in the case of objects or 'behavior' in the case of performance. In this

14 B. King

circumstance, relations of resemblance between sign and referent do not confine themselves to the level ofform, but work through common substantive properties that imply identity. To repeat, such an identity is not, of course, complete; if it were, one would be outside of a semiotic relationship, since signs 'stand in' for absent entities. It is a semantic rather than an existential identity. This semantic identity is not any less mandatory, however, since the extended degree of penetration of the qualities of the referent into the sign-vehicle or signifier means that the sign refers to the same meaning as its extra-discursive 'real' world object. Obviously, the character of Abraham Lincoln in the theater or cinema is not the real historical personage, nor is this representation merely an expression of the actor's own personality. But these evidences of fabrication do not demonstrate the arbitrariness of the sign. Since in both the construction of the character and the selection of the actor to play the role the intention is to make narrative use of a public symbol, this usage is limited by the fact that in the culture in question the meaning of Abraham Lincoln is given (or at least limited to a few semantic options). Even if the real Lincoln were available, he would be 'used' in the same way- i.e. he would play 'himself' (cf. the use of Henry Kissinger in 'Dynasty').

A parallel example from the object world, which actually is closer to the phenomenon ofheteromateriality, would be the use of the Statue of Liberty as a sign of location at the end of The Planet of the Apes. Such a 'quotation' only works narratively if the semantic integrity of the object is reproduced in the sign. Conversely, if the referent is not conspicuously a symbol within the narrative, but merely embodies the 'zero-degree' coding that underwrites the given and the 'doxic' in a particular culture, then the signifier will be constructed to correspond to how the referent would look if seen from the angle of view or visual 'take' implied by a particular framing. In the case of the mainstream cinema this is likely to imply forming the signifier according to the parameters of natural perception, suggesting a passive recording or unmarked transcoding (Lotman 1976: 32ft} In this sense, homomateriality is a more fully realized 'zero-degree' indexical mode that 'presents' rather than markedly represents existential objects. The paradigm case of homomateriality is the process of ostension: the 'picking up' of an object which is then shown as an expression of the class of which it is a member (Eco 1979: 225-227). To indicate the relevance of ostension to acting, let me anticipate once again and say that the process of type-casting is precisely 'picking up' and showing the actor's social type as an expression of this or that class - a semantic operation which rests finally on social discriminations.

The distinction between the formal and substantive aspects of sharing is, as Eco makes clear, a matter of degrees of indexical determination in


the formation of the sign. But, as is probably apparent from the foregoing discussion, the explication of this process has clear limits if one conducts the discussion at the level of the individual sign. Generally speaking, the status of a particular sign is determined by a specific textual use, so that it is more appropriate to speak of a process of signification and to see the question of the sign functioning as a symbol or index as determined by larger textual processes of denotation and connotation.

In the case of :visual signs the relationship between the signifier and signified is articulated within a concrete image or visualization of a meaning. The derivation of the signified or meaning entails on the part of the reader an inferential scanning of the body of the signifier in order to determine those aspects which are meant to be taken as cues to meaning and those which ground the meaning in a cultural context. The latter, denotative level functions to provide an intertextual support to the specific work of the text, a backdrop of natural or purely evidential signs that permit the subcode of connotation to do its work. The implication of this argument is that the functioning of a sign as an index or a symbol is always relative to a specific deployment or usage - indeed, if no usage were identified, then the very discrimination between index and symbol would be inappropriate. But signs are already given as symbols or indices: the space of culture is by definition already semanticized, and aesthetic texts work in the space of re-signment and recoding (Rossi-Landi 1983: Chapter 6; Eco 1979: 261ft} Moreover, any sign placed within the kind of text considered here- the published text - is proto-symbolic, since it participates in a socially symbolic act. As such all signs within the field of the text can be construed as standing for or summating a collective reality, unless textual mechanisms negate this possibility.

This framing of signs at the level of the collective is complicated by the problematic phenomenon of 'hyper-semiotization' identified by Bogatyrev (quoted in Elam 1980: 8ft} All the elements within a text have a potential relevance to its total semantic impact. Therefore, how can elements that have a specific narrative function be distinguished from those which merely provide supports for that function as, in Barthes's terms, 'effects of the real' (Barthes 1968)? Is the vaccination scar on the arm of the actress narratively motivated, or not? In a sixteenth-century costume drama, probably not; but in a sci-fi thriller about a time-travelling murderess, possibly. To give a more ambiguous example, is an all-black casting of Romeo and Juliet a statement about the irrelevance of color or a claim that such a casting will revitalize a 'tired' classic? Indeed, should the casting of a person of color in one or another of the leading roles be taken as meaningful at all?

As these few examples suggest, indexical qualities can acquire a symbolic

16 B. King

status simply by being textually embodied, in much the same way that an actor's personal qualities can enter into the symbolic field of a film. The issue is, therefore, will a particular sign be treated textually as a selfsufficient particular that fills out the text but is not instrinsic to its structuring principles? Or does the sign enter into the connotative process as a metonymic surrogate of narrative or, at least, expository purpose? Further, the intratextual usage of a particular sign is constrained by its established usage. What functions indexically, as an existent, within a specific text may be intertextually a symbol which a particular narrative re-signs in order to make a particular representational politics possible - e.g., the casting of a 'dumb' blonde as a rocket scientist. Conversely, any particular text may elevate what is usually treated as an existential given to the center of the connotative process - e.g., the issue of the casting of ethnic types. In other words, any particular narrative may semantically reorder the status of a sign in order to promote a different politics of representation. So from the perspective of any particular deployment, symbols may be required to function as indices of a narrative which undercuts or elides their sign status in other discourses.

The process of symbolic degeneracy is therefore a phenomenon of intertextuality, in respect of horizontal semantic relations with other texts of the same kind and vertical semantic relations with different kinds of texts (Fiske 1987: 108ff). Such a process is therefore dialectical; moreover, it is emphatically situational. The question of whether a sign functions as a symbol depends on its articulation within a specific textual field, and this placement is over-determined by the status of the sign in other textual fields and proto-textual social practices- experiences and events not yet given a social symbolization. Accordingly, from this perspective, texts must be viewed as differentially articulating some meanings over others, in a semantic space in which the suppressed and the exaggerated meanings inhere as immanent potential for contradiction and interdiction. The process whereby the meaning of a sign 'slips' out from under its envelope of textual meaning is symbolic degeneracy.

While this process is possible in verbal signs - though there it would seem to be a process the reader accomplishes inferentially rather than a process enacted in the body of the signifier- it seems particularly relevant to visual signs, and especially photographically based signs. Such signs tend to effect an erosion of the distinction, strongly marked in symbolic based systems, between the signifier and the referent, especially in those circumstances in which the substance of the signifier seems to be conspicuously the same as that of the referent (i.e., without an obvious or strongly marked mediation by a code or system of correlation) - the paradigm case being the photographic sign, which according to Barthes constitutes,


at the level of denotation, a message without a code. 6 A message, in other words, that is only made available or foregrounded by the erasure of coding procedures that enter into the formation of denotational aspects of the image (Barthes 1982: 196~ 197).

Photographically based signs function chiefly as indices, as self-sufficient particulars that convey a denotational charge that any particular 'text' or image complex can deploy in its own connotational work. Such deployment confers a symboli<; status on particulars or tokens, though the manner in which a particular sign enters into symbolic relationships will depend upon its materiality. Photographic signs enter into symbolic relationships through a reduction of their semantic complexity. This reduction is endemically unstable, since part of the meaningfulness of a photographic sign rests on its power to bring denotational weight to the process of connotation, producing evidence of the existence of the signified in the body of the signifier. Thus each sign has a double semantic status.

In particular, as argued above, connotational processes can be subject to implosion: an interdiscursive based erosion whereby the referential center of the sign expands outward from its encasement in a particular semantic field toward becoming a self-sufficient particular or token (Barthes 1982: 317~ 333). 7 The implication of this approach is, therefore, that the mainstream cinema is committed, by the parameters of illusionism, to a seamless investment of its imaginary in the 'real', which given its mode of operation at the level of collective representations is the real as reproduced in public and 'official' discourses. In other words, the effect of illusionism is to commit the cinema to concretize aspects of the collective process of reality definition. But because this process of definition occurs by means of an effacement (never absolute, of course) of the cinema's formative characteristics there is an endemic and chronic potential for exposing the lack of fit between such collective definitions, which are mainly verbal or verbocentric, and the visual concreteness of its own signifiers, which show a complexity that exceeds symbolization. The regime of illusionism, in other words, by prioritizing the indexicality of the cinematic sign, tends to undercut the power of the connotational process to subordinate its referents.

The reader, cognitively speaking

The thrust of my argument so far is clearly and deliberately in the direction of a textual determinism. Following Peirce's pragmatic emphasis, meaning by this the manner in which signs inflect and determine the development of human understanding and symbolic interaction, I locate the potential for symbolic degeneracy within the realm of sign production. Clearly, an

18 B. King

argument could be made that the process of discrimination between the denotative and connotative aspects of a 'text' rests within the power of the spectator. In general terms, Bordwell has been advancing the argument, by no means unchallenged, that texts are semantically thin grids that spectators 'fill in' by undertaking a variety of cognitive operations (King 1986). If such a claim is not true by definition, then the operations undertaken must be shown to be as Bordwell specifies them. While this issue cannot be broached here, it can be noted that the effect of Bordwell's argument is to locate the semantic impact of film more or less exclusively within the cognitive operations of the beholder, who sits like a latter-day Cartesian genie peering out as the film unfolds. Accordingly, the nature of a particular 'reading' and even the perception that films are amenable to a variety of interpretations depends critically on the resources the spectator brings to the act of consumption. While it would be hard to quarrel with the view that some readers are more skilled than others, it is less apparent that textual complexity rests on reader-centered processes alone. Obviously, the argument proposed here suggests that films are literally complex, and not just because a number of readings can be applied to them. In particular, I have argued that films are semantically dense enactments or strips of experience that are already structured, both referentially and narratively, albeit that this structuring process is volatile. Clearly the notion that films 'cue' interpretations, rather than embody interpretations of their own which the reader may reject, but not ignore, is too 'weak' to account for the process of spectator engagement. The notion that spectators inferentially plod through films shuffling schemata clearly ignores the extent to which spectators are positioned already - by biography, by the pre-release publicity, and by a host of other sociological factors. Such a notion also ignores the extent to which films provide a direct vicarious engagement with visualized meaning as opposed to a semantic 'algebra' that spectators use to 'make meaning'. There is a sense in which Bordwell treats film as though it were a signal, a message traveling from a sender to a receiver that only gives up its meaning if subjected to a code - a conception that sits oddly with Bordwell's collateral call for a 'historical poetics of cinema' (Bordwell 1989: 2-11 ).

From a different theoretical tradition entirely, the argument has been that how texts mean is a matter of determining how they are actually read rather than a matter of conformity to a model of a textual hierarchy, of preferred, negotiated, and oppositional readings vis-a-vis a dominant ideology (Wren-Lewis 1983: 179-197).

For Wren-Lewis, visual texts are encoded and ordered hierarchially, but the actual variation in readings- as opposed to the hypothesized variation expected by the theorist using an encoding model based on the dominant


ideology thesis- is a complex question. In this approach, then, a measure of textual determination is retained which the methodological individualism of the 'cognitive' approach tends to elide even as it renders the reader an operator of a series of decontextualized, supracultural inferences. For the purposes of my argument here, I would suggest that the position of WrenLewis provides a way of preserving a dialectical relationship between the text as a structuring agent and the reader in the process of 'reading' that Bordwell's model does not.

For a theory of the reader's activity that is on the same level of generality as my arguments here, it is useful to examine briefly the theory of symbolic strategies. As Worth and Gross have argued, the observer in confronting a particular visual text - or for that matter, a real occurrence - is required to adjudicate between two strategies of reading or interpretation, which they term symbolic strategies (Gross 1985: 2-11). Which of the aspects of the sign complex are formed with the intention to directly express a message, and which merely provide a supportive background? This primary decision, taken at the level of the ontological status of the sign, determines (until overturned by a similar decision which likewise determines, and so on) all the subsequent epistemological decisions that lead to a specific reading. In other words, what we can infer depends on what we decide warrants an inference and what does not - the latter merely conveying an existential meaning.

Such a distinction between existential signs and communicative symbols is certainly reminiscent of or coincident with Peirce's distinction between sinsigns and legisigns or indices and symbols. But it differs in that the distinction proposed is at the level of reading rather than at the level of textual articulation. That such a discrimination is central to the process of interpretation is not in doubt. The attribution of intention is a necessary methodological assumption even for a deconstructionist reading (Ellis 1989: 13-14). What is more problematic is the tendency to locate this process of discrimination more or less exclusively in the reader's inferential processes, and the link that is established between a literal notion of intentionality and the semantic status of signs. Dealing with the latter linkage first, obviously the attribution of intention is not the same thing as the fact of intention. Signs can be read as intentional which from the point of view of the empirical author are entirely gratuitous or accidental (Eco 1984: 4-7). Intentionality is not so much a property that is uniformly related to the author's psychology as a composite sense of purpose established by the text - the intention to show or to tell - and the intertexts of motivation that surround it. One of these would be the author's conscious motives, but others would include intertexts derived from traditions of narration and so on (Burgoyne 1990: 3-16). Likewise, the notion of com-

20 B. King

municative intent is no less problematic. If communication can only be said to occur if A has constructed a message X with the intention to influence B and B has recognized X as having this intention behind it and has altered his or her behavior accordingly, then such a draconian requirement would eliminate most aesthetic and cultural texts and performances from consideration. Indeed, it is possible to argue that even a weak notion of communication as the orderly transfer of meaning entails a view of textual effectivity as an impediment to meaning, rather than a place where different forms of intentionality converge (Clarke 1987: Chapter 4).

Even if one accepts that the symbolic strategies approach is really concerned with the perception of communicative intent rather than the empirical fact of intention, the fact remains that such an approach fails to emphasize the extent to which a text is formed out of the articulation of such a distinction- the distinction between denotation and connotationprior to any engagement with an empirical reader. So while it is correct to see the reader's attribution of symbolic status as determining the way in which a text is read or interpreted, this process should not be understood as resting primarily, or even solely, on the fulcrum of the reader's inferential processes. While such an assertion clearly warrants further investigation, there is evidence to suggest that the variability of readings or polysemy is a third-order phenomenon of connotation that rests on a textually encased distinction between a primary 'existential' and a secondary 'symbolic' level of meaning: a distinction which a subsequent reading assumes (even erroneously) and uses as an inferential base. This primary level is determined by the interaction between textually encased denotational and connotational markers. Another problem with this approach is that like Bordwell's 'cognitivism', it tends to propound an abstract rather than a concrete reader - a cultureless, ahistorical creature who nonetheless, in an exact inversion of the text-subject couplet, subordinates the text to its cognitive operations.

Notwithstanding these difficulties, the symbolic strategies approach, if given a Peircean twist, supports the arguments above: First, that there are in fact only two orders of signs - particulars or tokens existentially or causally connected to their referents, and symbols or generals. Second, that the semantic status of a particular sign is decided textually, and that this process of mediation has two dimensions: denotation, a first level of meaning (codes of recognition, conventions specific to a culture), which is deployed to generate a second level or subcode - connotation (Eco 1979: 54-57). Third, that the status of a particular sign as connotative or denotative is relational: does it refer to codes that govern perception in general (i.e., as realized in objects outside a particular, e.g. visual text), or does it participate centrally in the meaning construction of the text itself? Fourth,


as regards actually functioning signs deployed in texts (as opposed to signs theorized taxonomically), the property of Thirdness is common to all signs within the text - which is a socially symbolic act - so that the question of whether a sign is an index or a symbol is really a question of whether the system of the text deploys the sign in the status of an index or a symbol. Since this status is a matter of framing, offoregrounding or backgrounding, of subordination and superordination, it has no essential basis: it is given not by the nature of the sign, but by the manner in which the sign is used. At the same time the material basis of the sign, or for that matter a particular mode of signification, renders the process of internal discrimination more or less open to the slippage from a symbolic to an existential status.

As pointed out above, in the case of photographic signs, the process of discrimination between evidential (or natural) signs and intentionally created or 'communicational' signs 'within an array is complicated by the actual interpenetration of these aspects of the sign within the body of the sign itself. In part this problem arises because photographs show aspects of the object world with a minimum of semantic alteration, such metonymic aspects inhering in the image itself irrespective of photographic manipulation, or at least minimally inflected by it. Even in those cases where individuals are influenced by the presence of the camera- 'camera faces' and 'acting up' - such behavior is a meaning which the camera records prosthetically rather than constitutes externally. Once again, I should stress that this should not be read as identity: that the process is triangulation. The photographic signifier looks the 'same' as the object would look if so photographed, and this is true even if the object so photographed is a 'fake' or impossible event. 8

As was argued above, the process of cinematic signification is located within an indexical matrix. This means that the relationship between a visual sign (token) and its corresponding type is worked out in terms of codes of resemblance. These codes have two broad dimensions: extracinematic codes and cinematic codes, which in the mainstream cinema are mainly (though not exclusively) narrative. I see this relationship as hierarchically ordered: denotational signs; percepts, icons such as shape, color, and size; locations or sets; and actors as passive objects, walking scenery, must be rendered as indices if the film is to build narrative connotation.

To put it more exactly, the denotational signs of. 'illusionist' cinema are mainly the signs of other discourses and other referents which the cinema needs to re-sign in order to inaugurate its own discourse, its own version of the real. In illusionistic cinema, the tendency is to take the indexical level of sign formation as given - natural perception gets transcoded in filmic terms, with a minimum of marking in the transition. The 'world' we

22 B. King

see in the cinema is an assemblage, under rules of visual conformity to natural perception, of things and actions we see in the real world and these things and actions as indices of the narrative. The dark street of the film noir is both how a street could look if viewed from the appropriate position on an appropriate night and an evidential trace of the world evoked by the narrative, which is always a narrative within a genre. In other words, we are dealing with an indexical mode of sign production subsumed under a symbolic system.

Thus, the process of connotation in the cinema partakes of the features of connotative processes in general: the deployment of a reference to a meaning located outside the text in order to naturalize the meaning within. The process of naturalization is in general terms one of articulation - the selection of signs, constituted at the first-order, referential level, which are then construed, through a process of suppression of 'surplus' meanings in terms of a second order - connotation. The particular character of this process in the cinema is that the connotative aspects of the image are presented in a context in which negated semantic materials are present and retain in potentia their semantic effectivity as structures or systems of meaning. In other words, the relative status of the denotational and connotational aspects of the image remains open. This openness is compounded by the factor of movement through time and space. The cinematic image, of course, is drawn because of its mobility more intensively into a discursive context than is the still photograph, which maintains its integrity as an existentially fixed semantic particular. Nonetheless, the cinematic image has a parallel potential to 'slip out' from its connotational encasement precisely because of its implication in more complex and permeable relations - one symptom of which is the 'use' of stills to capture and by implication summate the effect of narrative flow. Likewise, the activity of the spectator in 'reading' the film is forced back in the absence of strongly foregrounded connotative markers to pattern recognition based on extracinematic elements in order to cope with the complexity of the image track. Such is the implication of Eco's view that the cinema has a triple articulation: percepts or qualisigns are organized into sinsigns, which express global cultural units (or sememes). The level of the photogram or frame is further articulated out of figurae of movement that are not units of meaning at the level of gestures and behavior. Equally from a cognitive viewpoint, Jaglom and Gardner found that the acquisition of image comprehension among young children is based on a constant comparison with the same objects in the real world, and that this basic level of discrimination is developmentally replaced by an emphasis on the behavior of characters (Jaglom and Gardner 1981: 33-47).

This textually induced 'regression' to a referentially based reading strat-


egy is most marked, of course, in the case of the naive or non-specialist spectator- though it seems likely that as first-time viewers even specialists are subject to it. It is in all likelihood compounded by the fact that given verisimilitude, the connotative process of film favors a predominantly unmarked deployment of cinematic technique. The characteristic modes of connotation in the mainstream cinema, not to mention photography in general, tend toward a cognitive framing of the object world rather than the radical constitution of that world. This complex process can be summarized as entailing excerption - the singling out and foregrounding of elements of the event or object world- and condensation of those events and objects given a symbolic weight as 'traces' of a film's narrative teleology (Roskill and Carrier 1983: 96-98r These narratively based emphases, I should add, also provide the ground for a 'fixation' of the object since its spectacular quality is conjointly brought into view. Such connotative framing, even if occurring within a generic context that lays no obvious claim to contemporary 'reality' (e.g., historical romances or science fiction), nevertheless relies on the inherence of the denotative level to render its own semantic particularity as a parallel reality.

I take from Stephen Heath the notion that narration, in the expanded sense of this term as the cinematic inscription of the diegesis, has a particular flaw: efforts to establish the integrity of the narrative rest on combining signs that are potentially disruptive, that inhere as negativities, because they point to alternative meanings and alternative (i.e., textually non-specific) modalities of the 'real' (Heath 1981: 19-75, esp. 62ff; Rodowick 1988: Chapter 7).

The actor as a sign

For my purposes here, I would emphasize the centrality of the presence of 'people' in film - actors, extras, and stars - in providing a central mechanism through which the various codes of denotation and connotation are integrated.

An actor's performance in cinema is composed of three levels of coding: (a) directorial, (b) everyday life, and (c) acting (Lotman 1976: 85). While these distinctions seem basic, if one asks where these three codes are fundamentally located it is clear that the last two are mainly provided through performance. Hence the distinction is probably more exactly rendered as between signs produced by the cinematic apparatus, which are in part directorial, but in their total effect may be summarized as narrative and performance signs. Obviously the cinema makes use of everyday

24 B. King

behavior; but it is less clear, given the emphasis on illusionism, that such everyday behavior before the camera should be distinguished from acting.

More to the point, I want to suggest that the centrality of the actor in the process of cinema signification means that 'directorial' effects are formative only if the actor can ground them in his or her performance. Clearly, there are many examples of the camera substituting for the actorthus Morin argued that the cinema atrophies the actor's performancebut such 'apparatus' effects rest on the actor's capacity to render these effects as credible in terms of human motivation and behavior.

The place of the actor in the process of signification is important here: the actor mediates between the narration in the capacity of character and the diegesis in the capacity of a really existent, already constituted sign. As such he or she concretizes as a subject and an object the narrative process as a whole, replicating in his or her self the various modalities implied by the film's textual processing. Thus the actor can be the object of narration which the camera scans and at the same time its subject -telling a story to other characters, to the camera, or as a voice-over (Kozloff 1988: 48ff).

This centering of the narrative process on the figure of the actor seems to have its correlate in reception because of the person-centered nature of everyday perception and because the subsumption of the various signs deployed under narrative is an emergent process. We have access to elements of plot chiefly through a focus on the concrete behavior of characters -whether from a character's point of view or from the omniscient viewpoint propounded by the camera as it scans their behavior in various ways or gives access to details which fill out the meaning of their behavior (e.g., the sight of an object hidden from the characters- Bordwell 1989: 15lff; Forgas 1981: 26lff).

Correspondingly in mainstream cinema, which rarely foregrounds its own devices, all other elements of the narrative process tend to be confined to performing as signs of the object world or of the process of telling, securing the illusion of the real. It is true of course that the use of location - cosmomorphic signs - can combine both a denotative and a connotative function (e.g., Wim Wenders's Paris, Texas. But this is rare, and in any case is used most often as a metaphor (a pathetic fallacy) of central character mood or situation. Likewise Heath points out that film as a total system entails a circulation between the actor as a person, image, agent of narrative and character, which creates a figure - a semantic field that always exceeds these various elements (Heath 1981: 180-182). This associational excess, while grounded in the total semantic field generated by narration and extratextual discourses such as star biography, is nonetheless, as Heath's examples show, internally focused on the star, the process


of performance forming what I term the persona - characterization plus figure.

The double aspect of the actor - as a really existent person whose personal characteristics, selectively utilized, provide the stuff of a narnttively existent person - indicates a relationship of signification wherein the stuff of the object/referent enters into the materiality of the signifier. The actor as it were uses himself as an ostensive instance of what the character is supposed to signify, unifying the signifier and the signified in his person as evidence of the reality evoked by the text. This relationship is narratively unstable if the actor begins, so to speak, to play himselfthe denotative qualities of the actor, his or her private person, begin to constitute or declare themselves as extracinematic realities which the cinema merely 'presents'. When this occurs at the level of the frame of reference established by the text - that what is happening is happening to 'real' or plausibly real people - the actor ceases to operate as an intertextually constituted sign (a symbol or an index) and begins to refer outward to his or her 'real' self. Such an external reference may be symbolic - the actor seems to be the social type he or she plays - or it may be indexical -the actor is a self-sufficient particular, an individual paragon who summates a class and is apparently the only member of a class so summated.

With these specifications in hand, I can now address directly the question of the actor as sign. The obvious point to be made about the actor as a sign is that the impression of a 'really existent' person is a textual or discursively realized phenomenon. To put this succinctly, the actor is an example of anaphor or coreferentiality. The identity postulated by theatrical discourse is not the 'I' of the actor who speaks and enacts, but rather an 'I' in a quotational mode. The distinction between the quotational 'I' and the 'I' of the actor parallels the distinction between character (a sign complex representing a really existent person utilized in a discourse for dramatic purposes) and personality (the idiosyncratic and existentially given, concrete realization of attributes that are recognized as humanly valid within a host culture). The process of symbolic degeneracy in the case of the actor takes the form of anaphoric implosion, the process whereby the distinction between character as a textually inscribed 'I' and the 'I' of the actor as a supratextual biographical entity is eroded. Obviously the process of anaphoric implosion is not the only form of symbolic degeneracy that occurs in film; auteurism in its most literal sense is another variant. Again, various approaches that view film as a record - e.g., of stunts, erotic moments, evidences of social mores - provide other examples, but for my purposes anaphoric implosion is particularly important (Urban 1989: 27-52).

When such an implosion occurs, the text or narrational process, as it

26 B. King

were, relinquishes its capacity to vary meaning attributed to the actor in favor of what the actor already means. If one focuses on the actor per se, rather than on guest appearances, this means in the average circumstance that the actor is a pre-given type. The typing of actors, given naturalistic assumptions, is not usually perceived as problematic, but clearly in relation to minorities - or for that matter, symbolically marginalized majorities, e.g. women - such a zero degree of coding can be seen as an imposition of an arbitrary symbolic labor in the service of oppressive or at least restrictive stereotypy.

The critique of stereotypes notwithstanding, for reasons of financial and semantic economy in casting, actors are generally cast as types so that character is generally stereotypical - it conforms to the naturalistic assumption that visible human attributes define an individual's membership in a social category. In response to typecasting and the competition for parts, actors selectively develop the stereotypical aspects of their appearance and behavior and project these aspects as coherent, high-resolution personality. Accordingly, the casting process 'finds', as it were, readymade 'characters'.

The normative or 'professional' view of acting advances the proposition that the relationship between character and personality should be flexible. Basically, the actor should be able to mold his personality to project a wide range of characters so that the actor's self is systematically distinguished from character, always allowing that some characters come closer to the actor's sense of his or her self than others. From the point of view of a particular role, therefore, there is a potential for the personality of the actor and character to emerge.

The process of merging occurs in two directions:

(a) Through typecasting - a character is played repeatedly with a minimum of variation, and hence becomes, from the point of view of any particular role, an index of a transtextual personality. The actor appears to play him- or herself. This phenomenon is observed in the cinema, but it is most fully developed in television.

(b) A leading player has a favorable screen image whose semantic resonance exceeds any particular role and in fact operates as a transcendent purport which each narrative instance must accommodate as a referential given. The actor again appears to play him or herself, but the self assumes a prototypical rather than stereotypical resonance. This phenomenon is a product of cinema - though it is also observed in the rock music business, where its material preconditions warrant a separate analysis.

Both typecasting and stardom represent examples of anaphoric implo-


sion, though in one case the character absorbs the actor and in the other the actor absorbs the character, which is in any case positioned as the center of the narrative.

In order to effect the incorporation of cinematic and extra-cinematic codes into the flow of narrative established by a particular film, it is necessary for the conventional aspects of acting to disappear into the character as an apparent indexical type (sinsign) at one with the fictional world. In other words, character acting is the process whereby a particular- an individual- accommodates his or her self to a social type.

There are three strategies that position the actor in relation to character: (a) impersonation, in which the actor disappears into the character by modifying his or her behavior and appearance; (b) typecasting, wherein the actor is selected by type in advance and fits the character through a minimum of self-manipulation, being available as it were as a natural sign that fits the narrative purposes embodied in the screenplay; and (c) personification, where the actor in effect types himself in real life according to his image on screen, so that he or she seems to be the embodiment or the evidential guarantee of what is seen - like a worker precisely adapted through emotional labor for an occupational role. Where the actor is successful at the box office and the efforts to become 'the stuff of the signifier' reach a level of intensity that exceeds the conventions governing typecasting and characterization.

Since the term 'personification' can be held to indicate the process whereby characters represent spiritual or moral truths - a convention derived from popular Christian doctrine- I should note that in the period since the sixteenth century, impersonation has become the dominant ontological premise of theatrical characterization. By impersonation I mean the convention, saturated with the assumptions of naturalism, that characters refer to types in the known and observable social world. The appearance of the actor is assumed to be a reliable, if not apodictic, clue to his or her character, and character in tum indicates social position and deserved social destiny (Bums 1964: 162ff).

Personification as I use it assumes the historic fact of impersonation and naturalism as a conventional field in theater and cinema, and is therefore closer to the contemporary usage: the attribution of human qualities to a thing. In this case, a persona is a cluster of textual effects that are reductively inscribed onto one semantic core: the professional personality of the actor given a figural enhancement through cinema. The strategy of personification should be viewed as a process of personal entailment whereby the symbolic and figural potentialities of the cinema are inscribed onto an exclusive purport: the star or would-be star. Persona building is the imposition of unitary meaning on a diversity of textual effects - not merely film

28 B. King

and TV appearances, but the subliterary outlets of celebrity - to create a personal monopoly, a semantic category with a membership of one. In this process the dimension of critical negativity implicit in the fictional projection of another reality - in this case, a utopian version of the power of culture to realize what is latent in the individual - becomes an affirmative example of bourgeois culture (Marcuse 1964: 60).

The transformation of symbolic leadership

As stated at the outset, I want to relate the process of symbolic degeneracy to the social role of the star. As is well known, the tradition of criticism of stars has rested either explicitly or implicitly on Boorstin's contrast between a hero, a doer of great deeds whose fame rests on real accomplishments, and the pseudo-hero or celebrity who is 'well-known for being wellknown' (Boorstin 1963: 55-85). In the case of actors, heroism in Boorstin's sense could only apply to great feats of performance- the 'test' of Learthough clearly the public may often confuse the portrayal of heroism on screen as being warranted by the real-life qualities of the actor. In this sense, Audie Murphy was a hero who was also a star, whereas John Wayne, who never saw real action, was a star who played heroes. I do not want to enter the fray of moral evaluation here, but I do want to identify the mechanisms by which the 'category mistake' identified by Boorstin comes about. In the light of the foregoing, I am in a position to do this now, but let me first state the issue as I see it: I am particularly interested in exploring symbolic degeneracy in relation to stars and celebrities, not the least because popular understanding of large-scale issues is often mediated through such biographical surrogates.

I take the view that the conversion of stardom to celebrity represents a further step in the erosion of the capacity of texts to represent collective issues and concerns- converting the 'occasion' of the text into an occasion for personal visibility. To make these points, let me begin by defining some basic terms, some of which I have argued for elsewhere (King 1985) and some which I now introduce by way of modification.

Personality: the transformation of the private self of the actor via professional training in which the private self is rendered pliant and manipulable in order to cope with the vicissitudes of acting life - self-presentation for rehearsals, determination, general practical and intellectual routines for presentation of self as a viable candidate for casting. Personality is the projection of those aspects of the self that have a market value.

Character: the scripted Personality, or socially symbolic self- the bare notation of a narrative agent which the actor fills out deploying pre-


performance skills of conceptualization- performance skills. Associated with or constituting the norm of impersonation: that the skill of acting rests on the ability to assume any character by the application of technique to the actor's body and behavior as the raw materials of character portrayal.

Persona: the personality implications of a successful portrayal- whether as a result of 'deep' character playing or personality playing in charactercollapsed into the 'private' person of the actor as a prototypical example of his or her character type - subject to circulation in the media under the rubric of a personal success (and failure) story. Whereas the 'truth' of character refers to a place in narrative, the 'truth' of persona refers to the 'presence' of the actor, a 'presence' which in reality is generated by the figural power of cinema. The process of personification is therefore an example of symbolic degeneration. Ultimately, the space of the persona can shrink to the space of celebrity, another level of degeneration. The relation of these public identities to the question of social type can be briefly summarized by defining another basic term:

Stereotype: a representation of a person based on inferences about his or her membership in a social group - in other words, as a social type or character, as a token of a given social type (class, ethnicity, race, role relations). Character acting is primarily operating as a stereotype.

I define prototype as the apparent (i.e., discursively articulated) realization by the individual of the maximum number of traits held to be central to a social type. While prototypes are primarily cognitive phenomena, the action of personification as a discursive strategy is to inscribe the superrealized attributes of a stereotype into the persona.

The star operates under the rules governing stereotyping, but does so to some superlative degree and via personification claims the status of a prototype. The type-token ratio- the relationship between a given individual and a class - still pertains, but is 'straining' toward individuation. The star seems to be the essence or platonic embodiment of the type: e.g., Marilyn Monroe as the definitive dumb blonde (cf. Lakoff 1987: 84-85; Forgas 1981: 216ff; Givon 1989: 35-67).8

Hypertypes: (a) the formation of a generalized celebrity via identification of the star with what appears to be fundamental, qualitative parameters that go beyond or above type; (b) the need to be loved, to have security, to realize oneself, etc. In another terminology, this mode of transcendence is the level of mass values, myth, etc. 9

Symbolic degeneracy, therefore, is the process whereby the collective implications built into stereotyping are subject to a reductive inscription onto the individual qua individual. In the case of the prototype, which equates with Peirce's singular symbol, the typical implications of the actor are repolarized so that the token (the 'personality') begins (via an act of

30 B. King

symbolic violence) to exhaust his or her type. This metonymic inversion is accomplished within the confines of performance and the occupational field of acting by the process of anaphoric implosion. The actor becomes · the existential prop or node of the process of personification. In the case of the hypertype, the process of individuation and metonymic inversion moves beyond the confines of occupational reference- the star, in contrast to the celebrity, still remains within the terrain of character portrayal and stereotyping. The hypertype moves the referential center of 'personality' into, so to speak, semantic 'hyperspace'. The celebrity, as an abstract symbol, lays claim to representing needs and desires that are common to all individuals, regardless of their typical social positioning. It is in the nature of celebrity to assume a referential positioning that sheds the grounding of the individual in a social type, whether this is related to occupation, race, class, gender, etc., in favor of more comprehensive articulations of social typing. Thus, for example, while gender plays a strong part in the projection and reception of personality, it is framed as a more fundamental and undifferentiated set of experiences and needs that transcend the fact of being a white as opposed to black, middle-class as opposed to working-class, man or woman. Such a minimum of differentiation can be summarized as a species-specific variant of populism- what we all feel 'deep down' in the radical individualist sense, as beings outside of social position and social destiny. In other words, the celebrity is the concrete individual positioned as the subject of the discourse of possessive individualism in a situation of high social visibility.

Let me stress once again that stars and celebrities do not represent exclusive modes of 'symbolic' existence; rather, what is under consideration here is the positioning of the individual as the node or referential fulcrum of discursive processes. Any individual- in this case, any actor- may, depending on the context of reference, function as a stereotype, prototype, or hypertype. What is missing from my account so far is how the process of attenuation is effected, and this is intimately tied up with the place of the actor in the discursive practices of television.

Television as text and supertext

The first question television raises is: does the process of symbolic degeneracy apply at all to its textual processing? One answer might be that such a process is paradoxically not a factor at all because of the poor resolution - this side of High Definition Television - of the image track. Television tends to function as a minimalist recording of performance, with a predominating emphasis on soundtrack, especially dialogue, in order to


compensate for a loss of semantic richness. As is well known, theatrical films on TV suffer from a flattening of depth, and loss of textural detail even in close-ups. In this sense, television is illustrated radio.

Such an attenuation of the narrative process would imply in cinema that the elements of spectacle and paradigmatic 'richness' would be highlighted in a sort of backhanded fulfillment of the Bazinian project. But this process is combined with a collateral attenuation of the image, a diminution in its indexical charge with- a corresponding loss of semantic texture and complexity.

A more exact specification - especially as it applies to the fictional forms, such as sitcoms and soaps, etc., that television has specifically developed- is that the narrational process is attenuated as well. A blatant example of this attenuation is the cropping of the image that occurs when theatrical films are shown on TV - so that in a medium shot of a conversation scene, one or the other half of the interaction is lost. Such an exogenous example only serves to highlight the sorts of accommodations that made-for-TV films take as given. Other externally derived erosions can be cited: the post hoc or prior editing of films to accommodate ads which are in themselves strongly articulated narrative skits and semantically dense images. The segmented flow of television operates as a supernarrational context that weakens the impact of internal closure when it occurs and provides an overall context of loose boundedness (Merelman 1984: 70-115). In general terms, then, compared to film, television tends to record rather than constitute its symbols.

This becomes apparent if we note that in the case of the cinematic textbarring the case of the B movie, which television resembles- denotational and connotational relationships are articulated and played out over three broad codal dimensions: (a) the relation between a text and the codes of natural perception; (b) the relation between a text and the codes of performance; and (c) the system of the text and its codes of narration.

In comparison to cinema, the semantic layering of the television text tends to reduce internally to the coding of performance, and very light coding at that. The prim~ry referent of the television text is the interaction between actors, an interaction that necessarily occurs in character because the other markers of narration (e.g., the aspects of setting and framing) are relatively depleted. Unlike cinematic narration, which deploys three general layers of coding to add propulsion and articulate difference and semantic depth, television narration tends to deploy one - dialogue and gesture - to ground its internal meaning.

This process of attenuation is apparent in what may be the most characteristic form of televisual fiction: the soap opera. Soaps show a weakening of stylistic markers that in cinematic practices encode the diegesis as a

32 B. King

quasi-autonomous 'reality' - haphazard and 'loose' correlation between the framing of shots and narrative propulsion, unmarked passages of time within segments, abrupt and discontinuous compression of 'natural' time in between, multiple plot lines chronically disaggregating from the level of overall narrative cohesion, lack of closure, limited variation in sets and locations.

In addition, it is the behavior of actors that tends to control framing, with camera deployment tending to record rather than construct the action. Likewise, the marked use of medium to close shots (two and four groupings) which settle the dramatic landscape on the poorly resolved human figure, the tendency to hold the camera in close-up on the actor at moments of 'high drama' longer than is narratively necessary, the expressive style of acting which emphasizes the dynamics rather than the texture of character reaction, the siting of complexity in multiple plotlines that cohere if at all at the micro-level of character motivation- all point in the same direction: performance assumes the burden of inscribing the narrative on television, albeit in an qverall situation of depletion of the image track. If television constitutes itself in a relation of reference, whether to the external world or to its own particular conventions, it is through the mediation of performance (Butler 1986: 53-70; Hindman 1982: Part One).

Compared to film, the visual (and relatedly aural) aspects of the image track are filtered through and referentially embodied in the process of performance, which in itself is a secondary or metalinguistic framing of the 'real'. In mainstream cinema, 'everyday life' and directorial and performance signs circulate in and out of the denotative and connotative levels of the narrative in an ongoing process of combination and recombination, super- and subordination. Such a circulation generates both the textual complexity and the semantic density of the cinematic image, which strengthens and at times destabilizes through 'excess' the narrative process. 10 In this sense, cinema is further from the textual and closer to the complexity of 'natural perception' than television, which is more decisively centered in a second-order textual process: the regulated behavior of actors.

Indeed, the entire impression of 'liveness' on television is less a matter of utilizing the medium's capacity for simultaneous transcription and transmission than the orchestration of reference through the presence of the anchor person, newscaster, and chatshow host. This basic thrust locates television's ontological framework as subjective realism: ranging from the authentic voice of the expert to the sincerity of the chatshow and fictional portrayal. Certainly the perception of greater 'realism' in television drama (e.g., 'Hill Street Blues') seems to combine the parameters of'live' reportage with strongly foregrounded emotional 'realism' (Feuer 1983: 12-22; Barker 1988: 42-56).


In sum, narration in television fiction tends to eschew a constitutive role for the camera in favor of a minimalist recording of an ongoing reality. In the case of fiction this ongoing reality is the performance of leading actors qua characters who are required to provide more or less exclusively 'mood' and emotional resonance that in cinema is produced by a broader texture of coding - sometimes, indeed, in excess of what is dramatically necessary. This intensification of the actor's contribution as a subject and object of narration may explain the observation that acting on television is often perceived as excessive.

A more exact specification is that character playing on television is endemically caricatural - albeit to varying degrees, depending on the player- in respect of a given occasion of performance. This seems true, notwithstanding the fact that ensemble playing necessarily prioritizes acting in character if the performance is to work at all. Overall, the projection of character on television is subject to time constraints which push in a caricatural direction. The lack of rehearsal time, the one-shot pattern of production, all require the actor to manifest character with very little preparation and in a state of preparedness that resembles improvisation rather than character research in depth. Where the role is significant -and in the case of television stars, this usually means where it is part of a continuous portrayal running over a number of seasons - then character acting develops complexity in breadth. But in each unit of performancein each show - character playing is minimalist, even though a character accumulates resonance over many performances, with various aspects of character being shown, often in a way that does not strive toward consistency overall. Thus the character of the star is a thin semantic or 'manneristic' core through which the actor is shown as responding to various human situations with a variety of mood states and personal qualities and attributes (Hindman 1982: Part Two).

In other words, television stardom is derived from character playing, but draws on the semantic equivalent of personality. This means that the television text tends to mimic at the level of performance the dynamics of parasocial intimacy, the play of personalities that marks television overall.

TV stardom: Characters as stars on the edge of celebrity

The analysis above would suggest that internal textual mechanisms provide the basis of the observed difference between stardom on television and in cinema: TV appears to make stars out of character players, in accordance with the truism that TV stars encountered in real life are encountered in character. Obviously, the phenomenology of the viewing situation must be

34 B. King

factored into this account. Certainly factors such as the viewing situationrelatively small-scale screen, the fact that the size of the actor's face in close-up is roughly natural size; a viewing context - in the home, in a mundane and familiar setting that unlike cinema does not encourage projective involvement; a programming mix which makes parasocial address a routine matter- tend to promote an observational set rather than the projective involvement which seems to underlie the more extreme examples of star-fan identification.

But television, as a discourse that exceeds any particular text it embodies, adds synergy to these reception phenomena. More importantly, television also imposes a framework to which stars seeking television exposure (which is a routine matter nowadays) are required to conform: the discourse of managed sincerity.

Let me now spell out the textual differences between stardom on televi-sion and film stardom.

Identification with actors is a common point of entry into the narrative working of fi~m and television texts. In film, the actor playing in character, given a privileged and intimate interaction with the camera, becomes meaningful in his or her own right - as a persona, or an evidential trace of a glamorous and attractive being who seemingly predates and is the exclusive pretext of a particular character. In such a circumstance the relationship between the stereotypical aspects of character- a person with such and such characteristics - and the material stuff- the behavioral and psychological cues of personality that the actor differentially deploysis hierarchically reversed, so that a further differentiation of personality, persona, starts to overdetermine casting. In other words, the need to reproduce the persona, a prototypical 'essence' of the actors' social type, becomes paramount. Discourse about the star tends to circulate around the question of ratification or authentication: does the character allow the star to realize what he or she already is?

In the case of the television star, the opportunity to implode character into persona is not provided textually, though it can be sought via intertextual framing (e.g., guest appearances). In terms of quality of image and camera deployment, television does not generate the figural excess that makes a particular player stand out as a spectacle and a self-referential entity, and the facts of reception and exigencies of performance do not correct this neutralization of the figure. Accordingly, audience interest in television stars relates less to the notion of stellar attractiveness than to everyday phenomena of collective and group life, the selection of particular players from the ensemble of characters on the grounds of their individual qualities- i.e., as personalities in character (Langer 1981: 351-365).

Other contextual factors, such as the fact that television stars are invited


into the home as 'intimate strangers', or that television stars do not benefit from the transcendent style of hype that Hollywood routinely deployed in the studio period, are obviously important here as well.

For these reasons, and others mentioned earlier, television stardom is a phenomenon of character playing and impersonation with the question of the relationship between the personality of the player and the characteris he or she like the character? - replacing the cinematic question: does the character adequately reveal the essence of the star? In other words, audience interest in television stars tends to reflect the discourse of sincerity on television in general - does the 'I' of the character represent what the 'I' of personality feels? Is the actor's personality visible through his or her casting?

It is worth emphasizing that the career path of the character actor in general dictates staying in character and maintaining a separation between character and personality in order to broaden role range. Television transforms this because the series format entails many performances of the same character, and this character in itself becomes a professional 'personality' asset usable in self-promotion and in product advertising.

The fact that television acting is character playing does not mean, of course, that television actors do not seek to be stars -witness the various attempts to make it into films. There are obvious financial incentives for this move, not to mention a longing for recognition and prestige that film stardom alone can satisfy.

But the route through television for getting outside character is problematic. The entire use of the actor on television is likely to reinforce the emphasis on playing in character, and media interest in the actor is in terms of overlaps between character and the actors' personality. One prominent heavy-footed patrolee on the boards of fame puts the problem for the character actor very clearly:

A lot of actors don't like to give interviews, on the theory that you only want to sell the public the character you play .... You don't want them to know the real you.

But I want people to know the difference between the characters I play and the real me. Because the reality is that I don't take a gun in my hand to solve the situation, that I'm not the Terminator, that I don't wield the sword around like Conan, that I'm not a stem character who doesn't smile, but that I'm a peaceful guy who has a great time. (Schwarzenegger 1990)

But this desire to be known for oneself is of course premissed on a certain level of success that less fortunate actors, and even stars, cannot take for granted. The notion that a star who plays 'heavies' is a nice guy is really only a tired twist in a pre-established persona. For the majority

36 B. King

of actors, factors of marketability and the need to promote themselves or product pose the problem of reconciling the need to maintain a line of unambiguous existential connection to character for employment purposes with the desire, if it arises, to do different work. One way to negotiate this is to be in and out of character and personality at the same time, signalling that one is a character actor. Thus, for example, the beer-guzzling character Norm of 'Cheers' and the actor who plays him, George Wendt, are presented in tandem in a TV ad for Meisterbrau beer by having a voice address the actor, in character as Norm, as Mr. Wendt.

On the other hand, on the chatshow circuit, where the actor can appear voluntarily (and sometimes involuntarily) outside of character, the format imposes its own conditions of generalized celebrity. If we define personality as a preferred image of the self and we recognize that personality is a selection of the best aspects of one's appearance and behavior, chatshow formats currently tend to want to go beyond personality in a breathless revelatory rush for greater 'realism' and sincerity.

As has been observed, the discourse of celebrity today has its own terms: the protagonists may be invited on because they are social types- banker, sportsperson, celebrity, actor, rescued hostage, etc. and even a found type, the audience member rapping with Oprah Winfrey. But they are required to function as 'interesting people', sincere specimens of the human comedy, who provide evidence of the gritty reality of a problem, or manifest personal concern over a burning social theme of the moment as evidenced in personal experiences, in the latest novel or film or some other promotable item.

In such appearances, the emphasis is decisively on speaking from one's personal (i.e., intimate rather than professional) experience. Such populist ventriloquism means manifesting an ability to speak with conviction even if lacking insight on experiences or feelings relevant to a theme. It means revealing oneself- saying what one feels and how one relates to fundamental questions: are human beings evil by nature, are men and women necessarily incompatible, what is the nature of fatherhood, etc.

Such a compendium of 'progressive' topics -which in many ways seem to be a recuperation of such notions as 'the personal is political ergo there is no need to talk politics' or 'sexual repression is bad ergo let's talk about sexual encounters'- seem to lead to an intensification of the paradox of massively overlooked intimacy: that the celebrity becomes more available to the anonymous viewer, more candidly on display, as the process of candor becomes more managed and ritualized. I believe that this process may be bringing about a shift in the relative position of the celebrity-fan, and I will briefly touch on this matter below.

But clearly part of the reason for this shift must be attributed to changes


in the nature of celebrity journalism per se. The chatshow has had a rapid diffusion over the last decade- a fundamental recognition of its cheapness as a source of programming and personal promotion - and this expansion has been marked by changes in the nature of celebrity journalism. It is not at all clear that this expansion has been significant in quantitative terms; the era of radio (e.g., Walter Winchell in the 1930s) and the subsequent careers of figures like Louella Parsons arguably represented the heyday of mass gossip. But whatever the nature of the quantitative expansion, there has been a qualitative change which may be more telling. Compared to the era of Hollywood, with its controlled publicity, there is currently a trend, however patchy, towards the professionalization of the genre of celebrity gossip. This trend is evident in a greater degree of frankness -this genre's version of objectivity - if not accuracy in the detailing of peccadillos. On the other hand, compared to earlier times, there is an apparent relaxation in public attitudes toward sexual morality and personal weaknesses (Henry 1990).

The implications of this change for public morality require examination. Prime facie, if there is a cardinal sin recognized by the celebrity system, outside of those forms of deviant behavior still defined or currently redefined as unacceptable, it is a false confession. Indeed, openness about one's failings, if convincing, seems to work as a moral cauterization of blame for lapses from past and currently held standards.

Of course, to some extent this resort to the public confessional is a ploy for reviving a flagging career. Yet if this is so, the notion of a failing or a weakness becomes a positive attribute: woe betide the celebrity who is unable to fill out his or her official biography with interesting or acceptable foibles.

Yet there seems to be more to the phenomenon than career opportunism. Is there not in all this a suggestion of a public impatience with hypocrisy, not to mention a widespread cynicism about the standards of public life? Again, it is difficult not to suspect a certain complicity on the part of the audience, since all of us manipulate the 'self' in the 'great drama' of public life - the recent vogue for home videos and the staging of such videos, not to mention the staging of family photographs, for public consumption serving as one example (Goffman 1973). Obviously these matters need further investigation. But in terms of moral discourse it seems that the long history of sincerity- which before it began to be perceived as a front manufactured for the exigencies of public life meant unabashed purity -is entering a new stage of contrivance (Trilling 1971). The chatshow and the discourse of celebrity have as their human reality term the authentic self, even as they admit that public life is a sham. This reflexive critique of their own artificiality is of course undertaken on our behalf: celebrities

38 B. King

are real people whom the little people demanded be better than they were or even could be. What is paradoxical in all this is that it is yesterday's stage-managed sincerity that is exposed in a public ritual of confession which is, of course, no less managed. In this scenario the notion of truth as correspondence to an independent state of affairs outside the process of observation becomes an acknowledgment that all realities are manipulated.

Among professional symbol bearers the operation of the norms and agendas of celebrity gossip serves as a form of performance that turns on the 'drama' of the constant signing and re-signing of personal reality. Individuals such as actors are compelled to obtain publicity by assuming the role of a 'private' person for public consumption alongside a bewildering array of other 'interesting' people. So even those actors who wish to emphasize their professional status can be drawn into talking about the subjective non-professional and social encounter side of film-making: how was it to work with X; were Z and G romantically entangled? It is this movement, all expressive of the fact that 'celebrity' per se now has a market value, that completes and reactivates the process of symbolic degeneracy initiated by employment relationships in the film industry.

Film stars and TV stars - those who were once divine and those who were once mundane - now face in the dynamics of publicity and selfpromotion a further reduction. They shift from being prototypes and stereotypes to becoming sometimes willing hypertypes- individuals whose point of interest is not what they are, let alone what they do, since they have the emotions and foibles we have, but where they do it: in public and at the center of the social imaginary, television. The final step in the process of symbolic degeneracy is a high media profile: the semantic vacuum of the hypertype, which places a possessive biographical lien on the media evocation of everyday existence. Thus emerges a populism without an identifiable collective subject, a type without a class.

Conclusion

In general terms, I argue that the role of reference in the formation of the film text is both ingredient to the construction of meaning and at the same time potentially disruptive of this process, since it leads to the fragmentation of the textual process into a series of intertextual mentions. Obviously, if the narrative process is seen as propounding an ideological problematic the reader does not like, then this a good thing. However, if this fragmentation is a generalized process, it undercuts the possibility of narration per se - eroding both 'progressive' and 'reactionary' narratives. In such


a circumstance, what ~ccurs is an evacuation of the socially symbolic aspects of narrative which in the final analysis favors the status quo.

Clearly there are some forceful reminders here of the old debates about the desirability of 'materialistic' or non-representational film as exemplified by the work of Peter Gidal. But I would take the view that such strategies are in fact self-defeating as interventions - a point that barely needs reiterating in relation to popular cultural interventions- though they may have had an indirect influence on mainstream practices by their radical critique of illusionism, not to mention a service rendered as research and development satellites for mainstream production. Equally, the welldeveloped debate around spectacularity and woman as bearer of the 'look' is an obvious antecedent of some of my points here, and one would have to recognize that the process of textual erosion can have a positive function in respect to gender and race. But to repeat, as a generalized process, such an erosion undermines the possibility of a cultural politics per se.

As indicated at the outset, I want to conclude with some speculations on the likely social consequences of the foregoing. Since such speculations mean second-guessing audience behavior, it is necessary to underscore the provisional nature of what follows. Probably the following should be viewed as queries for further research. The process of symbolic degeneracy exemplified by the television celebrity can be seen overall as an another element in universal commodification in which the personality becomes subject to the pressure of the market, becoming the simulacrum of a person reductively adapted for economic exchange (Hochschild 1983).

Such a process seems to be connected to the following general social consequences:

(1) Dispersal of social functions of narration. The phenomenon of anaphoric implosion means that the boundary between the socially symbolic field of narrative and the private biography of the actor- a personal narrative that is effectively subject to amendment and, like the soap opera, without closure - is eroded. This fusing of a bounded narrative with an unbounded 'life story' tends to convert the former into a documentary resource for writing and reading the latter. Stars become swamped with so many meanings that they disperse narratives, even if one discounts the impact of vehicles and deliberately constructed spectacles. In the case of a scandalous development - e.g., Rob Lowe's unintentional contribution to the world of celebrity pornography - the narratives in which a star is involved can become a documentary resource for a papparazzi style approach to celebrity-watching. Media attention genres such as 'Whatever Happened to ... ' and the informal activity of poring over the effects of age on a former idol represent another aspect of the erosion that operates here.

(2) Erosion of the function of the actor as a professional symbol maker,

40 B. King

ritual agent, etc. The development of a persona and its parallel attenuation through the pursuit of celebrity promotes an erosion parallel to that occurring at the level of the narrative. But there seems to be an effect stemming from celebrity journalism itself: character portrayals with strong ideological charge get neutralized by gossip.

To take an obvious case, celebrity gossip journalism about Roseanne Barr, which suggests that in her private life she is actually a mess, with no cause to criticize anyone, undercuts the political and social weight of her on-screen character as an assertive working-class woman. The claim of having seen through patriarchy clearly relies on a notion of experience and competence that images of emotional erraticism undercut. In this case, the category confusion between the character and the person works productively to recuperate criticism. Indeed, since the actual behavior of Ms. Barr is strictly irrelevant to the character of Roseanne and yet damages the credibility of both, one could claim that character assassination is occurring on two levels. Arguably, the emphasis on the 'personal' undercuts the socially symbolic role of public figures in general.

Similiarly, Bill Cosby's portrayal of Heathcliff Huxtal:Jle, while criticized as a sell-out to white petit bourgeois aspirations, can be seen as an attempt to erode the discourse of racism by emphasizing the educational achievements and the aspirations of the black American middle class. Generally, it seems that the importance of race as a discriminatory factor declines with the achievement of middle-class status, and Cosby's persona more or less reflects that fact, arguing that the positive evaluation of respectability - and all that entails in terms of education, personal culture, and profession - is an aspiration common to blacks and whites. Given the fact that the basic consequence of racism is to deny humanity to people of color, this assertion is not without its positive side (Gray 1986, 1989).

What is more problematic is Cosby's own relentless exploitation of the Huxtable persona to sell commodities such as advice books on marriage and fatherhood, toys, film, sweets, and so on. This promotional activity is evidently lucrative and may be justified on the grounds of Cosby's undeniable generosity to educational and charitable causes. But semiotically speaking, this projection of an abstract persona (in fact a hypertype) that serves as an existential guarantee for the claimed qualities of a diverse range of products clearly undercuts the more community-centered aspect of Cosby's persona as a prototype of the black middle class achiever, and by implication a positive model for the community. This latter deployment clearly entails a notion of a centered commitment or, to use an old word in an old way, sincerity that functioning as a referential passe-partout cannot. More than that, since after all cynicism about the role of the shill is commonplace, the emergence of the hypertype tends to cause a referential


hemorrhage in the anchoring of Cosby as a prototype. Thus part of the cost of deploying a character over semantically disparate contexts is not merely that the public sees too much of Cosby, but that what the public sees becomes devoid of personal or collective meaning. In the end Cosby just stands for Cosby.

(3) Repolarized parasociality. In the original formulation, parasocial intimacy was the illusion of face-to-face interaction initiated and controlled by the television performer (Horton and Wohl 1956). The chatshow repre·sents a variant of this phenomenon of intimacy at a distance. For the spectator, it offers the opportunity to combine voyeurism (always a feature of cinema) with the simulation of direct address and conversational involvement as the celebrity and the host alternately address the camera and the studio audience (per proxy the home audience). In short, the spectator is alternately cast as an onlooker or a party to the interactions on screen in a manner that mimics dyadic and triadie face-to-face interaction.

In the literature it is a common finding that where the illusion of intimacy is sustained in the mind of the spectator, in a minority of cases this can lead to strong identification. It is not inappropriate, given a certain context of consumption and a certain kind of spectator, to speak of 'capture' (Rosengren 1976). But the essence of this phenomenon is that the star operates as the primary definer of this mass-mediated relationship. The managed sincerity of the chatshow may be repolarizing this interaction, shifting, through its dynamic of revelation and interrogation, the celebrity into the position of junior partner or even patient of the public interest. After all, revealing oneself in public is a form of voluntary relinquishment of rights to the self.

Such a shift may be evaluated positively as representing the moment of the undoing of the elitism that was always implicit in traditional stardom. But there is some evidence that in cases of extreme star-fan identification this shift is driven by an impatience with the celebrity's existence as an independent being, not to mention a jealous sense of possession. Rather than the fan worshipping the star, the star, unwittingly, is being measured against the demand that he or she provide a suitable foil for the fan's fantasies, as though the apparent frankness of the star or celebrity marks a personal contract to supply the 'dreamwork' of the worshipper. Some fans, at least, seem to feel that stars no longer represent models to be emulated, but potential co-partners in a fantasy which they, the fans, initiate and direct.

Clearly, star-worship has always involved the use of stars as projective vehicles for personal fantasies, but what seems new is the literalness of the phenomenon. If, as I have argued, the construction of celebrity represents the detachment of the persona of the star from a collective process of

42 B. King

typification, then another impediment (in an interaction where, after all, impediments are only textual and semantic) to a narcissistic appropriation of the star image is removed. The hypertype enters with ease into what Schickel aptly calls the 'private cinema' of the fan's mind - in the most extreme cases, connecting with an individual whose practical detachment from collective processes seems to mirror the textually generated detachment of the celebrity (Schickel 1986).

There are, of course, well-documented extreme cases - e.g., John Hinckley's plagiaristic use of the scenario of Scorsese's Taxi Driver and the symbolic capital of the Presidency in order to impress Jodie Foster. But it is more interesting to speculate as to what these extremes might suggest about the attitude of the average fan.

What is new is that the discourse of celebrity now encompasses a reflexive admission that a public image is often, if not always, manipulated. The public has little reason not to know that showbiz is 'glitz', since the celebrity system flaunts its own artificiality. Why, therefore, does the phenomenon of capture, surely more comprehensible in the heyday of studio-based publicity which carefully disguised its own artfulness, persist with an increased literalness? Obviously, sociological factors would need to be brought into the account - is the lonely crowd getting more lonely or feeling a greater sense of insignificance? Let me focus here on the question of meaning and meaningfulness.

Everything seems to operate as if the managed sincerity of the star is read as signalling an exceptional commitment to truthfulness, so that he or she alone, in an insincere business, is sincere. Such a reflexive gearing of the truth seems to offer the fan in tum a guarantee that his or her involvement with the star is premissed on an intimate access to the self that is behind the front that the star 'puts' on for the general public. Thus the fan can conceive of his or herself as someone who has seen through the 'glitz' and is perspicacious enough to 'know' who the star really 'is', and to know it even better than the star.

The problem with this demotic variant of the knowing bystander- a perspective mass gossip clearly promotes- is that it is not really as cynical or as knowing as it would seem. Rather, it seems to suggest, if I can coin such an oxymoron, the naivete of the casual cynic: the process whereby believing the worst about people makes one defenseless against someone who seemingly defeats one's weary expectations. In other words, the recognition that show biz is fake seems to leave some individuals especially susceptible to someone who, by a curious variant of the liar's paradox, proves their sincerity by admitting that they are in a business that normally and routinely requires that they be insincere. Thus fans seem to develop an obsession on the basis of conceiving the star as someone like themselves


who is seeking an intimate, meaningful experience. Accordingly, the star no longer models publicly approved behavior, but rather the various strategies that individuals develop for deriving self-satisfaction and personal meaning in an increasingly meaningless or hostile world. In short, an elitism of soul searchers is replacing the old elitism of superconformity to social types. Hence, the scene is set for a relationship of attraction based on apparently fundamental needs and drives, and the fan is positioned as someone not paid to pretend, the 'genuine' other ofthe public confessional.

This particular trope will in tum explain both the fan's sense of 'knowing better' what celebrities are really seeking and the potential for violence and range when it becomes apparent that the celebrity does not know (much less want) what his or her co-seeker on the dark side of celebrity is offering (Caughey 1984; Dietz et. al. forthcoming; Vermorel and Vermorel 1985).

In terms of a situational logic, at least, it would seem that the chatshow, by extending the parameters of confessional literature to cover the moment of the authentic revelation before a mass audience, may be encouraging its own 'pathological' version of intimacy at a distance.

Once again, how widespread this phenomenon is - and whether it is really a new as opposed to a newly observed 'dark side' of celebrity -remains an unanswered question. Yet the fit between a culture of narcissism (Lasch 1978) and this apparent development in star-fan identification suggests that a larger constellation of attitudes to the self in public has come home, courtesy of the chatshow - that spectacle of exhibitionistic confession - to stardom.

Notes

I. For example, two collections are currently in press, edited by Jeremy Butler and Christine Gledhill. Richard de Cordova is about to publish a book on the early roots of the star system, and the current writer is preparing a book for publication by Polity Press.

2. Though this pluralism probably reflects the purpose of providing a wide coverage of the field. Subsequently Dyer has emphasized the star as a problematic embodiment of individualism (Dyer 1986).

3. The besetting problem with the 'natural' study of audiences is that the culture of everyday life is already extensively permeated with media-based representations of what the social relations of media consumption are. To some (possibly indeterminable) extent audiences are already 'auto-ethnographers' editing their responses to the media.

4. Such a history is irreversible. Consider the impact on the meaning of the family album if Pa Trubble was discovered to be a serial killer of young women.

5. For purposes of argument, I take the other aspect of the Hjelmslevian distinction, the substance and form of expression, to be virtually (i.e., semantically) transparent in the case of the mainstream cinema (Milo 1986).

6. Strictly speaking, no message is without a code; Barthes means weakly coded or overdetermined by codes that organize the 'natural' perception of the referent.

44 B. King

7. Barthes's notion of obtuse meaning is close to the notion of figural excess. Since I am making a case for reference, I do not explore this dimension herein, arguing that it is essentially contained by the strategy of personification.

8. The contention that visual signs are not signs at all but concrete images would be relevant here.

9. In Lakoff's terms, personality, character, and persona are metonymic models. Such models can be categorical and individual. My usage here cuts across this distinction. I have hesitated over the term hypertype, not least because of the rather weak pun it implies- types produced by 'hype'. Yet other candidates such as archetype are too loaded with metapsychological baggage to be of much use.

10. The notion of the classical realist text seriously underestimates the fact that even in the mainstream cinema, the 'membrane' between the metalinguistic and object levels of cinematic 'language' is highly permeable and variable, as shifts in sightlines, framing, and narratively marked and unmarked camera deployment would suggest (McCabe 1974).

References

Alberoni, Francesco (1972). The powerless elite: Theory and sociological research on the phenomenon of the stars. In Sociology of Mass Communications, Denis McQuail (ed.), 75-98. Harmondsworth: Penguin Books.

Barker, David (1988). 'It's been real': Forms of television representation. Critical Studies in Mass Communication 5, 42-56.

Barthes, Roland (1968). L'effet de reel. Communications 11, 84-89. -(1982). A Barthes Reader, Susan Sontag (ed.). New York: Hill and Wang. Baudrillard, Jean (1983). Simulations. New York: Semiotext(e). Braudy, Leo (1986). The Frenzy of Renown. New York: Oxford University Press. Bellert, Irena (1984). The sign and its object in everyday language, literature and theatre. In

Semiosis: Semiotics and the History of Culture, Morris Halle et. al. (eds.), 39-52. Ann Arbor: University of Michigan Press.

Benveniste, Emile (1971). Problems in General Linguistics, trans. by M. Meek. Miami: Univer-sity of Miami Press.

Boorstin, Daniel (1963). The Image. Harmondsworth: Penguin Books. Bordwell, David (1989). Making Meaning. Cambridge: Harvard University Press. Burgoyne, Robert (1990). The cinematic narrator. Journal of Film and Video 42 (1), 3-16. Burns, Elizabeth (1964). Theatricality. Boston: Harper and Row. Butler, Jeremy (1986). Notes on the soap opera apparatus: Televisual style and 'As The World

Turns'. Cinema Journal25 (3), 53-70. Caughey, John (1984). Imaginary Social Worlds. Lincoln: University of Nebraska Press. Clark, Danae (1989). Actors' labor and the politics of subjectivity: Hollywood in the 30's.

Unpublished Ph.D. dissertation, University of Iowa. Clarke, D. S. (1987). Principles of Semiotic. London: Routledge and Kegan Paul. Condit, Celeste Michelle (1989). The rhetorical limits of polysemy. Critical Studies in Mass

Communications 6 (2), 103-122. Dietz, Park et al. (forthcoming). Threatening and otherwise inappropriate letters to Holly

wood celebrities. Journal of Forensic Sciences. Ducrot, Oliver and Todorov, Tzvetan (1979). Encyclopedic Dictionary of the Sciences of

Languages. Baltimore: Johns Hopkins University Press.


Dyer, Richard (1979). Stars. London: British Film Institute. -(1986). Heavenly Bodies. New York: St. Martin's Press. Eco, Umberto (1970). Articulations of the cinematic code. Cinematics 1, 3-9. -(1979). A Theory of Semiotics. Bloomington: Indiana University Press. -(1984). Postscript to The Name of the Rose, trans. by William Weaver. San Diego, CA:

Harcourt, Brace and Jovanovich. Eisenstein, Sergei (1987). Non-indifferent Nature, trans. by Herbert Marshall. New York:

Cambridge University Press. Elam, Kier (1980). The Semiotics of Theatre and Drama. London: Methuen. Ellis, John M. (1989). Against Deconstruction. Princeton: Princeton University Press. Feuer, Jane (1983). The concept of live television: Ontology as ideology. In Regarding Televi

sion ( = AFI Monograph 2), E. Ann Kaplan (ed.). Fredrick, MD: University Publications of America.

Fiske, John (1987). Television Culture. New York: Methuen. Forgas, Joseph (1981). Social Cognition: Perspectives on Everyday Understanding. London:

Academic Press. Givan, T. (1989). Mind, Code and Context: Essays in Pragmatics. Hillsdale, NJ: Lawrence

Erlbaum. Goffman, Erving (1973). The Presentation of Self in Everyday Life. Woodstock: Overlook

Press. Gray, Herman (1986). Television and the new black man. Media, Culture and Society 8 (2),

223-242. -(1989). Television, black Americans and the American dream. Critical Studies in Mass

Communications 6 (4), 376-386. Greenlee, Douglas (1973). Peirce's Concept of.the Sign. The Hague: Mouton. Gross, Larry (1985). Life vs art: The interpretation of visual narratives. Studies in Visual

Communication 11 ( 4), 2-11. Haskell, Molly (1974). From Reverence to Rape. Harmondsworth: Penguin Books. Heath, Stephen (1981). Questions of Cinema. Bloomington: Indiana University Press. Henry, William A. (1990). Psst...Did you hear about? Time (March 5), 46-51. Hervey, Sandor (1982). Semiotic Perspectives. London: George Allen and Unwin. Hindman, John et al. (1982). TV Acting: A Manual for Camera Performance. New York:

Hastings House. Hochschild, Arlie Russell (1983). The Managed Heart: Commercialization of Human Feeling.

Berkeley: University of California Press. Horton, Daniel and Wohl, Richard (1956). Communication and parasocial intimacy. Psychia

try 19, 215-228. Jaglom, Leona and Gardner, Howard (1981). Decoding the worlds of television. Studies in

Visual Communication 7 (1), 33-47. Kellner, Douglas (1989). Jean Baudrillard: From Marxism to Post-Structuralism and Beyond.

Stanford, CA: Stanford University Press. Kindem, Goreham (1982). The American Movie Industry. Carbondale: Southern Illinois Uni-

versity Press. King, Barry (1985). Articulating stardom. Screen 26 (5), 27-50. -(1986). The classical Hollywood cinema. Screen 27 (6), 74-88. Klapp, Orrin (1962). Heroes, Villains and Fools. Englewood Cliffs, NJ: Prentice-Hall. Kozloff, Sarah (1988). Invisible Storytellers: Voice-over Narration in American Fiction Film.

Berkeley and Los Angeles: University of California Press. Lakoff, George (1987). Women, Fire, and Dangerous Things. Chicago: University of Chicago

Press.

46 B. King

Langer, John (1981). Television's personality system. Media, Culture and Society 4, 351-365. Lasch, Christopher (1978). The Culture of Narcissism. New York: Norton. Lotman, Yuri (1976). Semiotics of Cinema ( = Michigan Slavic Contributions 5), trans. by

Mark E. Suino. Ann Arbor: University of Michigan Press. Lowenthal, Leo (1961). The triumph of the mass idols. In Literature, Popular Culture and

Society, 109-140. Englewood Cliffs, NJ: Prentice-Hall. Marcuse, Herbert (1964). One Dimensional Man. Boston: Beacon Press. McCabe, Colin (1974). Realism and cinema. Screen 15 (2), 7-27. Merelman, Richard (1984). Making Something of Ourselves: On Culture and Politics in the

United States. Berkeley: University of California Press. Metz, Christian (1980). The perceived and the named. Studies in Visual Communication 6 (3),

56-68. Milo, Daniel (1986). The culinary character of cinematic language. Semiotica 58 (1/2), 83-99. Mitroff, Ian and Bennis, Warren (1989). The Unreality Industry. Secaucus, NJ: Birch Lane

Press. Morin, Edgar (1960). The Stars. New York: Grove Press. Peirce, Charles Sanders (1931-1935). Collected Papers of Charles Sanders Peirce, vol. 2,

C. Hartshorne and P. Weiss (eds.). Cambridge: Harvard University Press. -(1955). Philosophical Writings, Justus Buchler (ed.). London: Routledge and Kegan Paul. Postman, Neil (1985). Amusing Ourselves to Death. New York: Viking Press. Rath, Claus-Dieter (1988). Your life please: Autobiographical moments in television pro

grammes. Paper presented at International Television Studies Conference, London. Rodowick, David N. (1988). The Crisis of Political Modernism. Urbana: University of Illinois

Press. Rosch, Eleanor et al. (1976). Basic objects in natural categories. Cognitive Psychology 8,

382-439. Rosen, Philip (1987). History of image, image of history: Subject and ontology in Bazin. Wide

Angle 9 (4), 7-33. Rosengren, K. E. et al. (1976). Adolescents' TV relations: Three scales. Communication

Research 3, 347-366. Roskill, Mark and Carrier, David (1983). Truth and Falsity in Visual Images. Amherst, MA:

University of Amherst Press. Rossi-Landi, Ferruccio (1979). A theory of sign residues. Versus (Quaderni di Studi Semiotici)

23, 15-31. -(1983). Language as Work and Trade: A Semiotic Homology for Linguistics and Economics.

South Hadley, MA: Bergin and Garvey. Santaella Braga, Maria Lucia (1988). For a classification of visual signs. Semiotica 70 (1/2),

59-78. Schwarzenegger, Arnold (1990). Interview. New York Times Magazine (June 10), HIS. Schickel, Richard (1986). Intimate Strangers: The Cult of Celebrity. New York: Fromm. Seiter, Ellen et al. (1989). Remote Control: Television, Audiences and Cultural Power. London:

Routledge. Thompson, John 0. (1985). Beyond commutation: A reconsideration of screen acting. Screen

26 (5), 64-76. Thorp, Margaret Farrand (1970). America at the Movies, reprint of 1939 edition. New York:

Arno Press. Todorov, Tzvetan (1982). Theories of the Symbol, trans. by Christine Porter. Ithaca, NY:

Cornell University Press. Trilling, Lionel (1971). Sincerity and Authenticity. Cambridge: Harvard University Press.


Urban, Greg (1989). The 'I' of discourse. In Semiotics, Self and Society, Benjamin Lee and Greg Urban (eds.), 27-52. Berlin: Mouton de Gruyter.

Vermorel, Fred and Vermorel, Judy (1985). Star Lust: The Secret Fantasies of Fans. London: Comet Books.

Withalm, Gloria (1988). Self-referential semiosis in TV: Discursive strategies and hidden ideologies. Paper presented at the International Television Studies Conference, London.

Wren-Lewis, Justin (1983). The encoding/decoding model: Criticisms and redevelopments for research on decoding. Media, Culture and Society 5, 179-197.

Zeman, J. Jay (1977). Peirce's theory of signs. In A Perfusion of Signs, T. A. Sebeok (ed.), 22-39. Bloomington: Indiana University Press.

Barry King (b. 1944) is Assistant Professor of Media Studies at Widener University, in Chester, Pennsylvania; he was formerly a member of the Editorial Board of Screen. His principal research interests include semiotics of performance, acting as a labor process, and political economy of culture. Among his publications are 'Articulating stardom' (1985), 'The classical Hollywood cinema' (1986), 'The star and the commodity: Notes towards a performance theory of stardom' (1987), and Understanding Stardom (forthcoming).

symbolic degeneracy and stardom

Documents