Continuum: The Australian Journal of Media & Culture
Vol. 7, No. 2, 1994
Screening Cultural Studies
Edited by Tom O'Regan & Toby Miller

Two or Three Things I Know about Meaning

Tom O'Regan

For Bill

Of all affairs, communication is the most wonderful....When communication occurs, all natural events are subject to reconsideration and revision; they are re-adapted to meet the requirements of conversation, whether it be public discourse or that preliminary discourse termed thinking. Events turn into objects, things with a meaning. They may be referred to when they do not exist, and thus be operative among things distant in space and time, through vicarious presence in a new medium. (John Dewey 166)


In 1966 Jean-Luc Godard made Two or Three Things I Know about Her. It is an essay film about a host of different things: city buildings, public spaces, sex, love, prostitution, capitalism, the Viet Nam war, Paris...filmmaking. A quite opinionated commentary on these different things is sometimes spoken, sometimes whispered over appropriate and inappropriate images alike. This chapter also voices an immodest position - on nothing less than the politics of meaning. Grouped here are those matters at stake in: the instability of meaning, the distribution of texts and their arrival and non-arrival at their intended destinations (their adestination), the designing of texts for readers and viewers, reading strategies, and questions of truth and relativism. Like Godard's film it finds connections between things, sites and textual technologies normally kept separate. 1 It fabricates a continuity between: film and television production, film and cultural criticism, policy development, science, and the popular comic's concerns to create polysemous resources for future uptake. All are creative responses to meaning indeterminacy and textual adestination. The relations between meaning stability/instability, uptake and dissemination are critical to social practices which handle inscriptions, whether the concern is to maximise or minimise subsequent interpretative play.

The textual technologies of television, policy, science and popular entertainment all entail a number of reading and writing practices (or inscriptive practices) in addition to those with which they are most familiarly associated with. The policy technology, for example, includes not only policy documents but also journalism, public commentary, politician's speeches, press releases, manuals, meeting minutes, submissions, research (commissioned and otherwise), and the interpersonal communication commenting on, interrupting or advancing policy documents in formal and informal lobbying, meetings and other networks of actors. There is not so much a "policy discourse" as a wide range of inscriptive and reading practices, social actions, different dispositions of actors and the like. This makes it difficult to establish where policy leaves off and representative politics and cultural criticism begin. Such mutual interpenetration not only suggests that the boundaries between the different textual technologies are fuzzy, it also suggests that the analysis of particular textual technologies needs to take into account the heterogeneous parcel of inscriptive and non-inscriptive resources involved in each. The general nature of the relations between these resources can be listed as involving the relation between words and people, texts and other texts, reading and writing strategies, discursive and non-discursive resources, language and institutions.

Four quite different books, drawn from outside contemporary cultural studies, guide my examination of the politics of meaning. Horst Ruthrof's Pandora and Occam, a work of literary philosophy, advances a semiotic rather than language-based understanding of meaning ('language cannot mean by itself but can do so only semiotically, i.e., in relation to and through corroboration by non-verbal systems' (6)). Additionally, Ruthrof theorises constraints upon meaning as enabling conditions for the production of meaning as inscriptions, texts, are made meaningful through (potentially multiple) instantiations. Ruthrof's insistence upon 'the semiotic' (12) as the limit to what we can say and think, his emphasis upon constraints, and the instantiation of meaning in time and space all provide a way of enclosing concerns of textual theory and philosophy within the 'theories' of scientific discourse developed by Bruno Latour in Science in Action.

Latour brings together phrases and actors - or, in his terms, verbal associations and actor networks and stresses the importance of the linkages between non-human resources and human actors, their necessary provisionality, and the rhetorical nature of scientific writing. Here, texts lead away from themselves. Consequently the emphasis is placed upon the provisional networks that are created around texts as they are handled by actors. As in Ruthrof the "fate" of a text is seen to be attached to the ongoing transformative work of later statements, later users, and later texts. Latour extends his claims about the text-handling characteristics of scientific discourses to include economics, sociology and governmental techniques (e.g., statistics) through which the economy is managed and social welfare delivered. But I want to go further and claim that such text-handling characteristics are present, in a transformed way, in a range of public texts ranging from the more esoteric domains of policy development to more popular texts. Like the activities and texts of science, these other activities and texts also entail heterogeneous materials. The film producer needs human allies (publicists, distributors, exhibitors, financiers) and machines; and, even Batman comics - contain their own user legends so as to exclude unwanted readers (McHoul and O'Regan).

For his part Michel Foucault theorises the character of verbal associations in The Archaeology of Knowledge in terms of their regularities, dispersion, extension, transformation, heterogeneity and maintenance in space and time. These verbal associations (Foucault labels them discursive formations) are unstable assemblages held together through multiple meaning instantiations in space and time; and they are inextricably wedded to non-discursive elements. Foucault extends Latour's sociology and Ruthrof's philosophy of meaning by enabling us to theorise the nature of the verbal associations while retaining close attention to the instantiation of meaning.

The perspectives developed from the work of these three writers are sharpened through a dialogue with the pragmatist philosopher, Richard Rorty (especially his arguments in Contingency, Irony and Solidarity). There are the points of agreement: Rorty's discussion of vocabularies as central to the generation of new and unpredictable verbal associations links his work to Ruthrof's semiotic dynamic and Foucault's discursive formation. His call for a shift from epistemology to politics and his understanding of truth as a politics of competing vocabularies and agents not only links but usefully extends Latour, Foucault and Ruthrof's work. But there are the points of disagreement: Rorty overvalues language and in so doing undervalues the 'broader semiotic dynamic' (Ruthrof) in which verbal language is simply one among a number of semiotic systems. He undervalues the importance of networks of actors and institutions in creating meaning. He valorises new vocabularies, self-creating artists and 'free and open encounters' substantially downplaying the effectivity of older vocabularies, the "ordinary" artist and the many different conditions under which meaning is produced ranging from the coercive to the free and open.

This collocation of vocabularies drawn from literary theory, the sociology of science, Foucauldian archaeology and philosophy leads me to propose a redescription of cultural studies' project. If, in an ideal sense, we can say that textual criticism pursues vocabularies and phrases, and sociology social agency and structure, cultural studies' project - like the 'science in action' of Latour - is necessarily caught between doing both and trying to reconcile both. Indeed the emergence of cultural studies reflect the increasingly messy boundaries separating the traditional domains of textual analysis and the study of social agency and structure. 2 Cultural studies practice does generally insist upon the importance of language and verbal associations but, by the same token, does not collapse language into social practice, discourse into institution. So too cultural studies' work starts from the premise that to focus simply upon "discourse", upon "texts", and upon "inscriptions" in themselves will tell us little about different forms of life, and the ways in which particular actors have become mobilised and enrolled in space and time, using such discourses and texts. Similarly to focus exclusively on actors and their mobilisations will tell us little about the verbal associations and logics that sustain such actor networks. Furthermore, cultural studies' work often understands discursive formations as so many heuristic devices to describe those resources we utilize, plunder and mistake when we produce meaning (rather than as so many structures speaking through the subject - in doing so it stresses "using" agents; not used-by subjects). Here the work of Ruthrof, Latour, Foucault and Rorty enable a focus on something cultural studies' practice does not often do - a systematic conceptual reflection upon the theoretical bases of its insistence upon the social embeddedness of meaning. But this redescription of cultural studies' practice is not only designed to reflect upon some of the field's basic assumptions: it has a practical intent of generating a productive perspective on a fundamental problem for cultural studies. And this is how to productively conceptualise and act upon the relation between the top-down practices of the cultural industries, government and oligopolist capitalism and the bottom-up practices of "users" - the viewers and readers of popular entertainment, the users of street and housing scapes, and others that are the "targets" and "subjects" of government policy, real estate development and commodity capitalism. Ruthrof, Latour, Foucault and Rorty's work provide one means of showing how these are brought together in a politics of meaning.

Meaning Instability as Constraint

It is now taken for granted within cultural studies that the meaning of the cultural industries' popular artifacts are necessarily unstable despite their relatively secure settings and delivery systems. Another way of putting this is that texts cannot in themselves convince their audiences to attend to them, or dictate how those audiences take them up. This makes it difficult to secure "identical" meanings even within the same community of viewers and readers. (At this point Derrida (72-197) would note the endless iterability of inscriptions, Peirce the unlimited semiosis of an 'endless series of representations' (339), and Dewey (166) the marvelousness of conversation in its unlimited capacity for 'reconsideration', 'revision' and 'readaptation'.) As a corollary, inscriptions (or textual technologies) are generally held to be 'adestined' in the sense that their arrival and non-arrival alike are structural to their existence as texts. Meaning instability and adestination are then fundamental not only to fictional and aesthetic texts but also to administrative, policy, informational, scientific, and interpersonal texts such as letters.

It is possible, having come to this point, to construct a hierarchy of textual forms and critical practices. We could identify those texts and critical practices which acknowledge this discovery and those texts and practices which seem to want to hide or constrain its application. We would then be on the comfortable terrain of open versus closed film and television texts, the polyvocal versus the mono-vocal, deconstructionist tactics and those of traditional textual exegesis, the humanities and the arts versus the sciences. In each case we would be attaching our discovery of the necessary interception of texts (in Ross Chambers' (214) elegant formulation) to some rather than other texts, to some rather than other textual technologies, and we would be associating our own critical and ethical practice as the most proper response to this discovery.

However, it would be more appropriate to explore the different kinds of interception available in social practices and therefore the different character of interception available in different textual technologies. For his part Ruthrof notes that the discovery of meaning instability, the polyvocality of expressions, and the indefiniteness of semiosis, encourages the conclusion that there are, in principle, no constraints upon meaning and reading. But in actual practice, there are certain social constraints upon semiosis, certain conditions which ensure that meaning is socially negotiated (see discussion immediately below and Ruthrof 15-51). Similarly, Eco (24) counsels us that '(t)o say that a text has no end does not mean that every act of interpretation can have a happy end'. Latour goes even further. He claims that scientific writers are successful through being able to channel the interpretation of readers along particular pathways. Scientists are able to direct this later uptake within reasonable limits because they have created one of the most powerful rhetorics invented (it enlists on its behalf heterogeneous human and non-human resources including machines (Latour 57-61)). Latour contends that meaning instability and the non-arrival of texts to their hoped for destinations and/or their arrival at other destinations (their adestination) is fundamental to processes of meaning determination in science (rather than an alternative to it).

Applying Latour's position more generally, we can say that meaning indeterminacy and the adestination of texts generally - when instantiated in social practices - simultaneously create meaning (providing pathways for meaning to develop along) and constrain the range of other possible meanings (by closing off other pathways as a condition of developing a particular pathway). All writers and producers are concerned to channel and direct the character and direction of meaning instability and control; within limits, the adestination of texts.

The differences between the various textual technologies encountered here turn on how users (writer/producers and readers) deal with and turn meaning instability and adestination to their advantage. The historical constraints upon meaning are then tied to the 'semiotic potential' of utterances and texts. To say this is not to endorse constraints in principle but to note the existence of such constraints as a necessary aspect to the functioning, sustenance and creation of social meaning. The term "constraints" is being used here in a general descriptive sense of a framing that necessarily confines and restricts (in Ruthrof's terms 'the larger semiotic constraints of an utterance' (182), it is not being used to give some in principle endorsement to constraints which may be repressive and productive only negatively (as when unconvicted prisoners have limits placed upon their time with their lawyers, or governments limit the time for comment on the exposure draft of legislation).

"Constraints" upon meaning give the organization, transformation, translation or (re)iteration of statements a social and negotiated directionality. They are not simply a brook upon thought; they permit it. The social constraint of meaning is an inherently productive constraint (capable of good or ill). It is integral to what we can build and what we can invent just as it inevitably precludes certain buildings, certain theories, certain inventions.

As Donald MacKenzie (248) argues, what is important here are 'circumstances that in practice constrain challenge and that in practice facilitate it'. That one knows that 'what one learns is flexible, and must be affected by the presuppositions and goals of the learners as well as by the situation learnt about' can be a cause for celebration:

I would not merely concede that the specific constraints on the use or testing of nuclear weaponry inhibits certain forms of learning about it that would otherwise be possible: I would celebrate that, and note its usefulness as one means of exerting political control over the development of nuclear weapons. (248)

If meaning instability is a force for stability as much as disorder in a field, there is no purpose to be had from dividing texts into those that liberate versus those that limit meaning potential. Limitation and constraint are basic to any performance of meaning. Rather, the issue is one of the kinds of constraint, or the kind of channels, routes or pathways that develop for the management of communication. Following H.A. Innis we can call this the bias of communication.

Negotiating Meaning Instability

A text that is absolutely comprehensible is at the same time a text that is absolutely useless... An absolutely understandable and understanding partner would be convenient but unnecessary, since he or she would be a mechanical copy of my 'I' and our converse would provide us with no increase in information; just as there is no increase in money if one passes a purse from one pocket to another. A dialogue situation does not blur the distinctions between the partners, but intensifies them and makes them more significant. (Yuri Lotman 80-1)

Meaning instability refers to the fact that there can be no inherent identity of meaning between speakers within a field or discipline employing natural languages. Furthermore, meaning can be neither guaranteed by any inherent qualities of the text (whether these be conceptual qualities or formal textual arrangements) nor can it be guaranteed by its manner of presentation (Ruthrof 99). As Ruthrof puts it - in agreement with Derrida, though he argues the point differently - 'meaning is always already contaminated'. Such an account of meaning precludes us from imagining uninstantiated or disembodied meaning. Meaning is a socially occasioned activity in the sense that it is so many instantiations, so many denunciative events bearing no necessary relationship to each other. This implies that 'meaning conferment is never complete but continuously changing along the irreversible arrow of historical time.' (30) The future deployment of a statement, and/or a text, means extension, enlargement and expansion through subsequent speech occasions. This is transparently a building, not an unraveling process. With Ruthrof we can assert that 'there is no longer any presence of the signified....but only processes of signification...processes which provide meaning with no more than differing degrees of instability' (201).

If meanings are unstable then we always have Lotman's imperfect communication as our "perfect communication". 3 By implication we cannot have identical readings or interpretations of our inscriptions, whether these are ones we produce ourselves or ones that are produced for us to read. We also, therefore, have "consensus" in a loose rather than a strict sense. But this does not prescribe a world in which there is a cacophony of readings, where communication never appears to take place and a chaotic mess of contending "facts", "texts" and "truths" co-exist without any possibility of consensual agreement or "understanding" amongst them. Such a world would be impossible as speech communities - be they of friends, of relatives, of textual critics, of disciplines, of scientists, of government - simply could not function. Mostly we do communicate, we do agree and disagree on things, and there can be a remarkable sameness about our readings and writings. We just have to be clear about what we mean by communication, agreement and similarity.

Our speech communities function 'by negotiation', not by an inherent identity of meaning (Ruthrof 16). We can never seek more than relative corroboration, or in Ruthrof's words 'partial understanding and misunderstanding' (see 119, 185). Communication is thus approximative, it is marked by similarity and the 'partial overlap of semantic fields' (23). As Ruthrof elaborates:

language praxis is always characterized by a high degree of partial understanding as well as a considerable amount of misunderstanding, features not to be eradicated by making speech partners into better logicians....All that is needed is for utterers to be able to decide whether or not they are operating within the same - not the identical - semantic field. Where a higher degree of specificity is called for meanings are negotiated further...(75)

Such patterns of further negotiation are critical to specialised communication such as science and cultural studies and help create a sense of consensus about objects, policies and politics. By contrast, broadcast and entertainment communication situations entail much larger and more loosely connected "communities" who do not require the same kind of further negotiation to be successful. This may not be due to the fact that we understand each other better here but because there need not be any purpose served as audiences or readers in noting our misunderstanding. What we call "consensus" about some notion is different things to different people and may indeed be - as Mulkay and Gilbert (112-40) demonstrate of the scientists they studied - different things to the same person at different times. Nonetheless social, political and scientific consensus can still be confidently talked about. Consensus operates by networks of statements, speakers, and speech occasions bearing a family resemblance to each other as through 'recurrent interpretative methods...variable symbolic products, such as consensus accounts, are contextually generated' (140). This suggests that there must be degrees of approximation, variable corroboration, and varying forms and kinds of cohesion and consensus (e.g., the political consensus that politicians construct about foreign policy is of a different order from scientists' consensus about the nature of DNA). Communication is Dewey's remarkable achievement not only because so much needs to be in place for even relative corroboration to occur, but also because such approximations, such partial overlaps are the conditions under which "fact" building, theory creation, and machine development occur.

Meaning here becomes understood as so many meaning events; so many instantiations taking place in historical time and places. Ruthrof suggests these instatiations are both 'individual' and 'social':

The event of meaning making is both individual, in that it requires an enunciating and re-enunciating medium, a consciousness or a high-powered "machine" of the future, and social in that this very medium is always already woven into the semiotic fabric of a speech community. (27)

Furthermore, the rate and nature of change in meaning making from individual instantiation to instantiation is set not by 'linguistic meanings' per se 'but by a broader semiotic dynamic' (31). 'New expressions' and 'new metaphors' take their cue:

from the larger semiotic from which they emerge: the ghetto, digital technology, experimental writing, the Australian outback, or computer commerce. It is there, in the semiosis which exists between social action and discursive formations, that we must look for meaning. Use is never purely linguistic. (21)

Meanings therefore link linguistic expressions and other sign systems (31). (For Ruthrof these other systems include 'pictorial, kinetic, proxemic, aural, photic, electrical, thermal, and other forms of non-linguistic grasp...[and] our semiotic relations with both the animate and inanimate environment' (12)). Thus the 'semiotic' is given a purposively wide definition here; it encompasses non-verbal and verbal sign systems. Such a 'broader semiotic dynamic' is a resource used by the enunciating medium to constrain the 'semantic drift' that occurs between speakers, and between readers in different acts of meaning instantiation. Because of the heterogeneous cultural resources available to us through sharing social worlds, meaning making need be only approximate, negotiated. This permits us to "muddle through":

On the one hand, all utterances of texts always take place within constraints of speech act formations and the flexible boundaries of language use as circumscribed by a dynamic speech community; on the other hand, uttering is also always an individual historical event, an instantiation....whenever a text is uttered, whenever in other words language takes place or a text finds a tongue (or a membrane), the inscription undergoes a complex process by which its content is profoundly affected....the acts of uttering as they occur in a good deal of everyday speech and in all literary reading result in a significant bending of the propositional contents which can be abstracted from sentences and texts (28-29).

Conferring and creating meaning (e.g. reading, writing, painting, appreciating) is necessarily caught between individual enunciation and its social frame.

Recipient Design

Writers and producers deploy this social frame in particular directions. They seek ways of encouraging the reader/viewer to follow suit - the subsequent reader/viewer may 'approximately' follow but s/he may merely intersect and 'escape'. As Latour aptly describes this process for scientific writing '(t)he writer draws so many pathways going from one place to another and asks the reader to follow them; the readers may cross these paths and then escape' (57). Such channelling says to the reader: "read me in this fashion", "take me up in this direction", "build me this way", "explore and construct within these parameters".

The "reading" situation is one in which there is 'a race between authors and readers to control each other's moves' (Latour 58). The writer projects an "ideal reader" - and so "negotiates" the reader's response in advance of the reader's later reading. In this way the writer feigns being a reader him/herself, projecting a context and motivation for his or her reader to read. For scientific literature, this projected reader is according to Latour a skeptic; by contrast, the projected reader of broadcast television entertainment is, for John Hartley (108) a much more distracted subject - a viewer with 'childlike qualities and attributes. Television discourse addresses its viewers as children'.

The reader, in his or her turn, projects an "ideal writer" and this is itself a renegotiation of the writer's anticipation of the reader's response. In such a context the reader is always dealing with him or herself. In Ruthrof's words, the reader is constructing 'propositional meanings against the backdrop of the parallel construction of a speech stance and an utterance motivation' (51). That is, s/he is not merely developing propositional meanings but is doing so in a particular way by inferring a speech stance and a motivation for the text before him or her. S/he is imaginatively projecting a speaker, an author (178). To do so s/he makes use of:

typical ways of speaking, typical postures of speech, typical semiotic traits which accompany and modify acts of speaking, typical social roles which inform a speech event, typical philosophical, moral, religious, aesthetic and political convictions as well as a typical ideological blindness. In short, a range of 'recipes for social action,' as Alfred Schutz called them, and not just discursive formations are at my disposal as I respeak an inscription. (178)

Authors appear able to control readers only insofar as readers share similar cultural resources and recipes for social action as the author. Their projections of the "author" will match the author's projection of the "reader" to the extent that this applies. It is through such "idealisations", such necessary "abstractions" that understanding is approximated and consensus produced.

Scientists, bureaucrats and academics are drilled in such a way as to maximise shared cultural resources and recipes for social action among a discrete community or group of writers and readers. These actively control the kind and nature of cultural resources used by readers. They turn reading into a public matter. It is no accident that these practices achieve this control by limiting and excluding readers, including other scientists, bureaucrats and academics. Such texts cannot, by definition, be as universal as their popular counterparts. For Latour, they protect themselves from 'malevolent readers' by carrying their own 'legend' or 'user's guide' explaining how and by whom they should be read (52). This makes them difficult to read and analyse. Expertise is required to read the literature, participate in scholarly debate, or be effective in the political process. This presents the writer with a quandary. As Latour describes the position for the scientific writer:

Remember that the authors need the readers' willingness to have their own claims turned into facts. If the readers are put off, they are not going to take up the claim; but if they are left free to discuss the claim, it will be deeply altered. The writer of a scientific text is then in a quandary; how to leave someone completely free and have them at the same time completely obedient. What is the best way to solve this paradox? To lay out the text so that wherever the reader is there is only one way to go. (57)

Latour is not denying that readers may be disputatious and go in different directions. The normative power of scientific discourse stems from the fact that it is a writing practice which makes it more difficult for the reader to go in other directions than do other discourse genres. Some writing practices make it easier for readers to do this. It is easier for the film and cultural critic, the historian and the sociologist to go in other directions than it is for the scientist, the analytical philosopher and the mathematician. 4 And it is still easier for the 'wild viewer' (in Ien Ang's words (see "Stalking")) of popular entertainment.

C.S. Peirce (176, 177) once attempted to protect 'pragmatism' from my sort - malevolent literary and cultural critics - by using a word for his philosophy, 'pragmaticist'; a word which he considered so 'ugly' that it would discourage theft. 5 I do not cite Peirce's admonishment of cultural criticism to indict cultural studies for a lack of rigour. Rather I want to use it to point to the greater degree of bending of propositional contents and mis-readings available in more publicly subscribed enterprises such as film and cultural criticism than in the more esoteric and exclusive frameworks developed in scientific activity, policy development, Peirce's pragmaticism, and specialist branches of sociology such as ethnomethodology. Of course, public texts - such as journalism, politicians' speeches, and the "reading of policy" by those whose business it is to deal with these texts - bend propositional contents to an incredible extent. "Consensus" here is often a matter of the same words, phrases and sentences being put to vastly different purposes and programs of action.

For their part television and popular entertainment is even more 'dirty' in Hartley's sense. Hartley (23-4) describes television as 'a dirty category' because 'the interface between texts and readers is capable of producing both meanings and 'relations of power and domination' precisely because it is not a clean opposition and necessarily ambiguous'. Both the Hollywood movie and the television quiz show have "writers" and "readers" who only weakly share cultural resources. Producers and viewers alike are required to negotiate cultural cleavages in their viewing audience. Michaels (19) suggests that this encourages producers to adopt a 'highly complex rhetorical stance which makes it quite difficult to say what the intended meaning of many programs might be'. In other words the "conversation" between producers and audiences is designed to minimise obstacles to participation on the part of potential audiences, but this strategy of incorporation is achieved through a communicative inefficiency (which is exploited most efficiently) as propositional contents are bent further, opportunities for partial misunderstanding are increased and even encouraged. And this is not a problem. Consuming such texts along with, in Dewey's (203) unfussy list, 'letters, poetry, song, the drama, fiction, history, biography' involves 'engaging in rites and ceremonies hallowed by time and rich with the sense of the countless multitudes that share in them'. These are 'modes of discourse' that are 'ends for most persons' in that they are 'detached from immediate instrumental consequences of assistance and cooperative action'. In them 'discourse is both instrumental and final'. For Hollywood it does not particularly matter that wildly divergent or astonishingly convergent interpretations are routinely accomplished by audiences through Hollywood's global circulation so long as tickets are sold and videos rented.

It would make no sense to criticise one of these textual technologies for not being the other (which is what we often do when we counterpose art to science; policy to cultural criticism). Each of these technologies requires different kinds of critical protocols and political action. Nonetheless, all of these domains share to varying degrees the characteristics of scientific discourse elaborated by Latour. Each attempts to lay out the text for the reader. This is so whether it is the Hollywood story-telling which puts the labor into script, plot and image - providing a digestible, acceptable and attractive story which belies its engineering feat and encourages diverse uptake; the cultural critic or historian mobilising her evidence and rhetoric trying to "shape" the reader to his or her argument; or Latour's scientist driving readers out to bring in fewer, more powerful readers.

For Latour the task of the writer is to 'dam up' (58) the alternatives the reader has:

No matter where the reader is in the text, he or she is confronted with instruments harder to discuss, figures more difficult to doubt, references that are harder to dispute, arrays of stacked black boxes. He or she flows from the introduction to the conclusion like a river flowing between artificial banks.(58)

This "damming" process is successful where the reader feels that s/he is herself coming to that conclusion; that, in some way, s/he is shaping the object. Latour says that scientific literature does this by incorporating the reader's response in the text. (Film and cultural critics as diverse as Hugo Munsterberg (41), Sergei Eisenstein (34, passim) and William Routt (chapter 6) note that cinema and popular art respectively - popular music, television, popular literature - also involve producers getting the reader, the listener, or the viewer to feel that he or s/he is shaping the object. These producers also incorporate the reader's response into the text.) Such control over the reader can be thought of as being achieved by on the one hand, the mobilising of the 'semiotic dynamic' of other sign systems (Ruthrof's position), or, on the other hand, the mobilising of 'heterogeneous resources' such as machines, animals, and instruments (which in Latour become the writer's external allies (Latour 61)). 6

Latour calls scientific writing rhetorical because it mobilizes so many external allies to win the argument. He further claims that the only difference between scientific rhetoric and ordinary rhetoric is that scientists 'invoke so many more external allies besides passion and style in order to reverse the path of common reasoning'. Scientific writing is thus that rhetoric which is 'able to mobilise on one spot more resources than older ones' (16). But scientific writing is not alone in this. Consider the extraordinary array of external allies including the projection, distribution, reproductive and marketing technologies available to the popular filmmaker to - in his or her own way - reverse common sense. What distinguishes, on the one hand, the rhetoric of common sense and, on the other hand, scientific rhetoric and the rhetoric of the popular is the resources available to each.

For Latour the 'fact builder' is thus someone in the business of attempting to construct particular pathways, to create that channelling in a particular direction. As Latour puts it, the fact builder needs to do particular things:

there is a set of strategies to enlist and interest the human actors, and a second to enlist and interest the non-human actors so as to hold the first. When these strategies are successful the fact which has been built becomes indispensable; it is an obligatory passage point for everyone if they want to pursue their interests. From a few helpless people occupying a few weak points they end up controlling strongholds. Everyone happily borrows the claims or the prototypes from the successful contenders' hands. As a result, claims become well-established, facts and prototypes are turned into routinely used pieces of equipment. Since the claim is believed by one more person, the product bought by one more customer, the argument incorporated in one more article or textbook, the black box encapsulated in one more engine, they spread in time and space. (132)

These 'black boxes' - whether a statement, a network of associations, an interpretative method or machines - are inherently contingent. A 'black box' becomes 'durable in time only through the actions of many people; if there is no one to take it up, it stops and falls apart however many people have taken it up for however long before'. (137) We may formulate this as the revenge of adestination inasmuch as inattention is fatal to any 'black box'.

Control over the reader here means the reader's enlistment to think along certain corridors, to adopt and then use particular instruments and machines. This is control through seduction. Readers have to be interested in the claim; there must be something in it for them. Constraints, therefore, are not so much things to be overcome as the abstract potentiality of meaning becomes instantiated in use, but are rather fundamental to dismantling old - and building new - objects and machines.

This detour through Latour's account of scientific meaning discloses meaning instability and adestination as powerful tools for explaining how meaning becomes provisionally stabilised and texts relatively destined. Semiotic potential is a creative edifice built upon constraint; readers/audiences are held in ways that preclude their significant altering of the utterance and enhance the prospect of their finding inventive ways of extending the scope and range of the statement or "text" in time and space.

The Recipient's Activity: Going Along, Working Through and Giving Up

For Latour's readers to accept a claim without discussion and significant transformation 'the number of associations, linkages, resources and allies locally available' (62) has to be increased to such an extent that the cost of disputation in human and monetary terms becomes extremely high. By incorporating within the scientific text so many heterogeneous elements - by mobilizing so many allies - the reader is left little negotiating space. He or she has limited options. He or she can go along with the text turning its claims into facts (9%), give up (90%), or work through it (1%) (Latour 60). Importantly each of these options leads the reader away from the text. Scientific literature limits the kinds of possible readings and so sacrifices numbers of readers (including those who might be interested) in order to achieve closer approximations of meaning between the meaning instantiations of writer and reader; it also discourages 'working through' (60). All these options share this in common:

the only possible readings all lead to the demise of the text. If you give up, the text does not count and might as well not have been written at all. If you go along, you believe it so much that it is quickly abstracted, abridged, stylised and sinks into tacit practice. Lastly, if you work through the authors' trials, you quit the text and enter the Laboratory. Thus the scientific text is chasing its readers away whether or not it is successful. Made for attack and defence, it is no more a place for a leisurely stay than a bastion or a bunker. (Latour 61)

Whether we 'go along' or 'work through', this reading and writing practice incessantly leads us away from the text into an engagement with other texts, other sign systems, applications, labs, and instruments. Each of these options are so many social transformations of the text. Reading, like writing, is channelled in certain directions. Within the tight frames that such channelling provides, contingency and instability are a matter of course. Meaning instability, textual adestination and polysemic potential are present in all these alternatives. We cannot usefully confine it to those in a position - i.e., who have the resources, the allies etc. - to 'work through'.

Readers who could be expected to read and simply do not are showing us how texts routinely fail to arrive at their intended destination. Latour does not see this as the "system" failing. He sees such 'failure' as being built into the system from the start. Only in these sorts of competitive conditions in which inattention is the norm can we imagine the need for one of the most powerful 'rhetorics' invented in the case of science and the massive distribution and exhibitionary machinery of popular texts. Readers may give up for reasons apparently unrelated to the text. Giving up to do something else is not "slackness" but integral to scientific, academic and bureaucratic rigour and to the popular appreciation of entertainment. Competing rhetorics may be more attractive; activities other than reading journal articles may need to be done, and these will, in turn, produce new inscriptions and form the basis of new articles which others can then ignore, and so on. Giving up is then not just the property of popular art's supposedly slack viewers, readers and listeners who switch off or switch over.

By the same token those who 'go along' do not simply repeat the claim. They turn the statement into a fact, they abstract it, condense it, think with it, and create further uses for it. In this way the provisional consensus about facts and concepts turns them into, in Latour's terms, 'black boxes' (29-30). Going along is thus characterised by statements, networks of associations and interpretative methods being turned into black boxes which are not understood as being opened for investigation and dispute, but turned into so many facts from which to work (60). By doing so those who go along enlarge the black box's scope and extend it (just as this chapter accepts and modifies the statements of Latour, Ruthrof and Rorty). In this fashion later literature expands possible applications. Polysemic potential is exploited in these later uses. Even when we may be most directly 'going along' with a statement and understanding it, the apparent consensus between ourselves and the author is itself made up of variable understandings and interpretations of the statement, comment or text (thus it is not surprising that Mulkay and Gilbert find scientific researchers in an area having different accounts of the meaning of a particular scientific statement /fact even when they are 'going along' with it (130)).

It is remarkable ('wonderful' in Dewey's words) that in these conditions the semblance of consensus is maintained. Just because there is such variability to the consensus involved in 'going along', we have to consider 'consensus' in ways which take this into account. Too much time and energy gets spent disclosing the variability in scientific and other kinds of consensus making for the purpose of dismissing it as an ideological construct or as an artificially imposed ritual performance. This does not allow for variability in consensus to be neither chaotic nor a lie but intrinsic to building, creating new machinery and technologies, to discoveries, to cohering a community and creating channelled communication. The difference between disputation and consensus might consist in disputation making a topic of variability of meanings and interpretative methods; while consensus creates unacknowledged variability through broadly similar interpretative methods. The former elides important commonalities between the disputants; while the latter elides contradictions and disjunctions between fellow travelers. Such variability can become a topic under certain conditions and then 'going along' may become 'working through'.

By contrast, working through is explicitly about taking apart the black box - disassembling and deconstructing it. But to deconstruct also requires equipment, resources and powerful allies. For those of us trained in aesthetic forms of life this working through appeals as the critical moment par excellence; it appears as the moment of innovation and creativity equivalent to the actions of an artist, a poet, a filmmaker, a revolutionary. This is the moment where stability can be rendered in all its instability and contingency; where the blockage in the chain of signification represented by the black box is done away with, where the critic becomes the "equal" of the aesthetic text. In principle 'working through' does create new machines, new objects, new kinds of associations, new theories, and it remakes our world but no more than 'going along' does. To 'work through' in science means not simply to dispute and to critique. It must in turn lead away from the text to the formation of new objects, new instruments, or even a return to an older set of conceptions. Those who 'work through' are engaged in the same channelling, damming and controlling activities as those who 'go along'. Indeed, in some sense, they must work more closely on the text and in the laboratory than those who 'go along'. Paradoxically those who work through may be more constrained by the text, by the "fact" in dispute than their colleagues who work from the text, not acknowledging their "bending" of the facts. 'Working through' supports science and extends its instruments. Like 'going along' it does so by closing off as much as opening out. It too "arrests" the chain of signification and makes certain claims more credible and others less so.

'Working through' is thus as highly channelled and controlled activity as is 'going along'. Both are critical to the development of new machines, technologies and instruments, one to create new facts, new association of words, phrases and instruments, the other to build upon the claims made, to find applications, to widen scope. If we assume that meaning is not implicit but always explicit and developed through use, then (scientific) reading can be understood as consisting of the polysemic potential inherent in meaning instability being channelled in certain directions, and moved along certain pathways.

Latour's discussion of the meaning of a text, fact, instrument or machine gives critical centrality to 'later users' and to 'later statements' (27-29) to the extent that the fate of a scientific text is in those later users' hands, in later scientific texts. Additionally, the scientific text appears to be shaped from the start for later uses (and constructed with that purpose in mind). We are dealing with a retrospectivity to texts and to facts. Such retrospectivity is a fundamental characteristic of public texts, whether they be bureaucratic, scientific, or popular.

The very provisionality of the associations brought together to form 'black boxes' - whether scientific facts, or the taken for granted ideas which become taken to be facts by film criticism - requires that scientific and critical facts are creations of specific historical usages and users. Latour insists upon the paradox that whilst scientific writing limits readers' options, what those same readers' use or do not use, support or reject, retrospectively determines the fate of a scientific statement, text and scientist.

Because so much is known to be in the hands of later users and later statements, scientists and cultural critics must attempt to influence those later users, those later statements, in some way. This may be done through 'damming up' the readers' options. Alternatively it may be done through 'interesting' and 'enlisting' them among the allies to the project (Latour 173). This is a process of enlisting actors so as to form networks interested in extending the claim made in the research through space and time. For science these other actors might include the funding bodies, politicians and benefactors who have underwritten the laboratory work in the first place. (For cultural criticism it is the 'invisible colleges' the networks of people in informal and formal associations, publicity, publishers and journals, conferences, public forums, willing research "subjects", funding and employer organisations.) This is what Latour means by being led away from the text. The reader may be skeptical but s/he also needs to be interested and shown how s/he can be interested; s/he needs to be shown how s/he might extend the claim without fundamentally altering it. So too the scientist needs those other allies - the labs, the funding agencies, the patents - in order to produce the research to begin with. S/he needs external allies who can have the research translated for them in order to be persuaded to work together on it.

If the different facets of Latour's description are general features of social practice shared to varying degrees by popular and elite art, academic, policy and journalism, 7 there are, however, differences between scientific activity and texts and other social activities and social texts. A sense of some of these differences will become evident by considering 'working through' and 'going along' in film criticism.

Working Through and Going Along in Film Criticism

In film criticism working through and going along are more closely aligned without the sharp distinction between them characteristic of science and policy making. David Bordwell shows in Making Meaning that going along has been critical to the project of criticism in film studies as successive "new" readings are shaped by the twin demands of conformity and limited innovation. As in science, film criticism creates its own black boxes as, over a period of years, provisional statements about films get turned into "artifacts" about films and film meaning which require no explication; 'ideas had become facts' (Bordwell Making Meaning 241). So too these 'facts' can be rendered as 'questionable' in competing elaborations striving to work through and achieve exemplar status (Noel Carroll attempts just this in Mystifying Movies with respect of psychoanalytic and Marxist readings of the cinema). 8 Further, only a handful of "readings" of films can function as exemplars: Laura Mulvey on visual pleasure, Stephen Heath on Touch of Evil, Thomas Elsaesser on 'New German Cinema'. Like scientific activity, the public achievement of working through is only available to a handful. Most film critics are like Thomas Kuhn's 'normal scientist'. The gap between underlabouring and achieving exemplar status is narrower here than it is in science, but just as contingent.

This normative reality to film criticism sits side by side with a critical rhetoric embracing innovation, renovation, and textual productivity in the object being discussed. 9 Assisting this is the different objects of scientific practice and film criticism. While science and film criticism have an interest in stabilising and renovating their respective black boxes, there are critical differences between the black boxes and the kind of consensus about the object that is managed to different ends in them. Unlike the scientific black box, which can be a diesel engine or an equation, film criticism's black box is not so much the "film" before it as the rhetorical strategies and interpretative techniques brought to bear upon it, only some of which may be actively supported by the film. Scientists created the diesel engine and the equations which subsequent users build upon; film critics deal with found objects in films whose construction is not of their making (although there may be much sharing of sense-making procedures, and directors from Jerry Lewis to Sally Potter may actively create texts designed to enlist critics). Bordwell is therefore right to insist upon the object that criticism constructs and to distinguish it from the material text of the film:

What permits the endless variety of meanings to be generated from a film are in large part the critical practices themselves, particularly the indefinitely large variety of semantic fields and salient cues that can be "processed" by a set of schemata and heuristics in force. The ambiguity sought by the New Critic, the polysemy praised by the structuralist, and the indeterminacy posited by the post-structuralist are largely the product of the institution's [film criticism] interpretative habits. Our ability to recognize, however tacitly, these habits in action emerges in our praise for the text's "richness"; it must be polysemous if we can imagine using different, but equally permissible, procedures to makes sense of it, and to make cases for its discrete meanings. (245)

Unlike the scientific community, the film criticism community does not make consensus such an explicit topic and end point of analysis. Yet consensus is nonetheless critical to stabilising interpretative methods and ensuring the continuity of the institution. It is simply a differently ordered consensus based more on 'familiarity' in which 'interpretation' is 'part of the communal effort' reaffirming the 'social cohesion - of critic and reader, of critic and critic' (215).

From the rhetorical standpoint, the interpreter's basic task - building a novel and plausible interpretation of one or more appropriate films - becomes a matter of negotiating with the audience's institutionally grounded assumptions. There is a trade-off. Risk a more novel interpretation, and you may produce an exemplar; fail, and you will seem merely odd. Stick closely to the limits of plausibility, and you will pass muster, but you may seem routine. In general, the best preparation is to study exemplars. This teaches the critic what will go down with an audience and what degrees of originality are encouraged by particular institutional circumstances. (206)

Science, by contrast, makes an explicit topic of consensus and has evolved particular techniques to ensure that polysemy is controlled and channelled such that, in their passage through many hands, scientific facts do not become so distorted. Scientific inscriptive practices therefore control readings more tightly. But equally, the scientist - like its counterpart the film critic - aims quite explicitly to ensure that copying and extending innovations takes place with attribution, that many will speak the terms turning them into facts; in short, that the innovation becomes a necessary point through which others must pass through. In this way their innovation is rewarded.

Against this background the institution of film criticism favours novelty. In practice it constrains the form, direction and nature of that novelty. Bordwell shows how novelty becomes the basis for staging consensus of interpretation, rhetorical procedure and viewpoint. Yet there are also significant differences between film criticisms. Wide Angle and The Journal of Popular Film and Television are not Positif or Film Comment; Wide Angle is relatively autonomous with respect to contemporary cinema, while Positif and Film Comment have an obligation to cover the ephemeral flow of cinema output and "taste" and consequently maintain important connections to film reviewing. Additionally, film output goes on in spite of criticism and can be hardly affected by it. The same could not be said for innovations in science and technology.

But criticism is not as autonomous from the "film world" as Bordwell suggests. Whereas the text's polysemy may be a product of critical institutions, television's ethnographic audience research suggests that diverse audiences' varied readings are the product of polysemous audience interpretative habits (see Ang Desparately Seeking 153-65). There is thus an homology between criticism and the nexus of normal film and television viewing, production and consumption. Screen criticism conforms to the characteristics of cinema and television in similar structures of novelty and similarity, ephemerality and remakes, centralised and peripheral production sources.

While new readings of standard films are the explicit business of criticism, and standard readings of new films the explicit task of film reviewing, these readings are constrained and limited by contemporary interpretative methods and critical institutions, just as audiences are constrained by their interpretative and sense-making habits. Thus film criticism - like popular films and literature - can be best thought of as a machine exploiting polysemy within particular and delimited confines. Popular texts tolerate simultaneously divergent reading possibilities as a means of maximising audience reach on an international basis. Film, television and literary criticism constrain divergent readings to a more limited frame as a means of similarly expanding, though in more rarified academic and intellectual spaces.

Little wonder then that those associated with literary and screen criticism celebrate the productivity of readings and the bending of propositional contents rather than lament them. Going along means producing new interpretations; the experience of "consensus" is obtained through applying accepted interpretative methods and procedures to "new" objects and "new" methods to "old" objects. The very polysemic potential of the aesthetic text enables it to function as the same cultural resource but to generate different meanings. This plurality helps us understand the obligation on the part of film criticism to develop them further. If film critics are opposed, in principle, to "fixing" and "controlling" film meaning, seeing in it the arresting of the endless significatory chain, the scientist, by contrast, sees such control and fixing as a precondition for the creation and extension through time and space of new machines, new technologies, new ways of behaving and new forms of social and cultural organisation.

Literature and screen media should not be posed as inherently more flexible, more polysemic, more available to unexpected and powerfully diverse interpretative acts than fields such as science and policy. Since the range of resources we bring to cinema, television and literary criticism are likely to be fewer such criticism is likely to be more stable, less dynamic and less transformative than fields which routinely deal with more heterogeneous elements.

In Pandora and Occam Ruthrof argues for a 'ladder of discourse' in which the literary text, television and the cinema are all on a higher rung than 'the discourse of history', 'juridical discourse', the 'discourse of the everyday', 'technical discourse', 'scientific discourse' and at the bottom 'formal logic' (153-164). Rather than celebrating this heterogeneity of the aesthetic and the literary which Ruthrof does at this point, it might be better to note that such texts have to be more embedded, more 'polysemous', 'more fecund' in a ladder of discourse precisely because there are fewer, and less powerful options and actions available to later users. Textual heterogeneity is all they have to exploit because the options for the viewer and critic alike are, mostly, simply more talk.

Centring Inscriptions

In the approach developed here inscriptions can be thought of as linkages, as facilitations, as agencies permitting readers to be simultaneously drawn in and driven away. Following Latour's injunction, we would also follow the actors involved in circulating texts. By following them we would be concerned with their practices of "text handling", or "information handling". Such a concern would illuminate how an inscription's verbal associations become currencies and items of value to particular networks of users.

"Inscriptive" practices need to be inspected for their incorporation into particular "forms of life". It would be misleading to develop a general argument about "text handling" on the basis of a particular form of text handling. For example, if textual and inscriptive matters are approached through literary and cinematic metaphors there will be a tendency to project the inscriptive text as a site of pro-literary hermeneutic interpretability; if it is approached through the policy and political text, the emphasis will be on moving away from the text towards social effectivity. In the case of the literary or filmic text the social membranes extending out from filmic and literary texts can be understood by immediate reference to other texts. The pathways have become so routine, the system of verbal associations so routinised and the attention so focussed on the text in its contours and its appearance, that we cannot easily see that it is precisely through these analytic techniques that we are drawn away from the text to create an ideal text which may bear little relation to the audience making it over, ignoring it, going along with it, working through it. The aesthetic metaphor is one which prioritises the inscriptive text and its relations to other like texts. Similarly, the policy text requires us to understand not only its written language but also the semiotic richness of the bureaucratic office, the meetings, the memos, the discussions, the bets and the gambles as a discourse is tamed and normalised. Policy texts always imply more than the text, more than the semantics of this or that inscription. It implies a definite circuit of which this inscription forms but a mark and a trace of a larger system. Thus policy metaphors encourage the de-textualising of inscriptions and the elaboration in its stead of social actor worlds - thereby losing a sense of the importance of language use to the technologies and rationalities of governance.

Inscriptions represent varying opportunities for meaning conferment, development and opposition which are the consequence of the kinds of circulation these inscriptions have in later users' hands. They are, therefore, an important component in the story of textual technologies to be told in so far as their handling - their destination and insertion within a field of forces - is a critical component in generating meaning.

Inscriptive practices may be varied, but each acts at a distance. The policy text is used by networks of actors in far flung places to guide their funding applications; the exemplary film analysis serves as a template for film criticism across continents and languages; government economic information underwrites market calculations such as share values; the scientific journal article (itself often based on machine generated inscriptions) is disseminated in a form accessible to an international community of scientists; the popular film is screened in so many different locations using more or less identically struck prints released in a coordinated fashion; and the newspaper is simultaneously delivered and sold within a definite time frame to many readers across the geography of a city, a state or a nation. All are characteristically texts "acting at a distance" in the sense that they are purposefully designed to have long distance effects even if it is the minimal one of going through the turnstyle to buy a ticket for a show we end up sleeping through. All may be durable, comparatively light of weight, all may have their transmission interrupted, corrupted, or bypassed.

But, just because these are "long distance" texts in which the identification by the "author" of the reader is necessarily ideal and often unknown to the "author", we should not assume that such texts are decontextualised entities capable of arbitrary re-invention and re-description along the way. As we have seen, texts - like meanings - may be in principle 'adestined'. The same words may be capable of multiple and divergent programs of an antagonistic nature, and thus play multiple roles in competing discourses. But as we have also seen, texts may be channelled to greater or lesser extents. Texts designed to "act at a distance" depend on the network of actors; the degree and kind of the enrollment of those actors and the "logic" by which the text drives its readers away.

Popular entertainment texts represent one version of this. These are designed, by and large, to go no further than their ephemeral, though intense, entertainment. They do not require the tight damming of the reader managed by the scientific and academic article (such as this one). They are - as texts - rhetorically weak, yet in their mobilisation for popular consumption in distribution and exhibition structures, they are rhetorically powerful, achieving an extraordinary reach; being "too popular by far". The outcome of "going along" and "working through" will, as a consequence, be qualitatively different. For the popular film, as with aesthetic texts generally, ranges of interpretations are actively solicited and even invited. It is not a problem, for instance, that the Australian director George Miller's Witches of Eastwick can be equally seen as 'our film' by feminists, fundamentalist christians, sci-fi buffs etc. (Durgnat 268). Indeed such pluralising of meaning is an important component of the "demand management" of Hollywood. By contrast, the corruption over distance and through social space of an academic argument is a point of disputation, boundary policing, and may even lead to the destruction of academic reputations in literary and cultural studies.

Our public inscriptions are mobile, durable, and capable of being worked upon many kilometres away from their source. The character, nature and form of this "action at a distance" varies with each. Policy and scientific texts have certain delimited spheres of influence and readership, usually marked nationally, bureaucratically, and politically (and sometimes internationally). In these spheres a text is assessed, ignored, abused, endorsed, and/or modified. Like the work of popular entertainment, policy and science also have their own particular locations where active and possibly intense engagement takes place.

But this is not an argument for a necessary incommensurability of "forms of life" and forms of inscriptive text. The same ideas, phrases and sentences may be present in a journalistic text, a policy text, a politician's speech in the legislature, an aesthetic text, and an academic text. The distance between the interpretative resources mobilised to read and position policy, political speech, legal, aesthetic, entertainment and journalistic texts may be themselves slight but significant. Such interconnection, repetition and redundancy is socially typical of inscriptive practices, just as there is a sharing of many mundane resources to understand and interpret inscriptive texts. So too, a work of translation between the different "spheres" is routinely accomplished not only by individuals but also by authors of specific general inscriptive texts such as journalist texts which are designed to provide "translation" services, offering a broader and intersubjectively shared "language game" in which social interconnectedness is routinely staged, as too are readers' relations with other readers, authors and institutions.

Another function these different inscriptive practices share in common is that each is engaged (in Callon, Law and Rip's words) in 'building a world and then attempting to convert that world into reality by enrolling others in it...actors attempt to impose worlds upon one another' and with this comes all the 'uncertainties involved in maintaining power and size' ("Putting Texts" 227-8) .11 We therefore need to consider 'how texts were able, historically, to become emissaries that are durable, transportable and forceful, and therefore crucial agents of social control' ("Putting Texts" 229). Following this line of reasoning Callon, Law and Rip argue that it may be possible to pose the 'classical functioning primarily by a combination of drilled people and texts...' ("Putting Texts" 229). I would argue that popular texts constitute another highly particular relationship, that of drilled producers and texts, and distracted audiences and texts. Inscriptions are linkage and relay points of exchange between people, and between people and non-human resources.

Differences between inscriptive practices can then be understood in terms of the different kinds of actor networks created, the differential enlistment of the author and the audience/reader (the nature of their mutual imbrication and intelligibility), the different notions of "character" and "author" under each regimen, the orientation to the materials and the outcomes, the difference between certain "public spheres" in terms of their techniques for the generation and circulation of statements. In short, the inscriptive practices involved in policy, scientific, journalistic, entertainment, and aesthetic texts may be best understood as so many "local" strategies for creating and extending networks, advancing their power, scope and the resources available to them. Dissemination - or distribution - is thus critical to them.

Latour argues that inscriptions do not have any particular destination or directionality by virtue of any intrinsic quality they may possess. They are capable of multiple and unexpected destinations; because of this their fate is in later users' hands (29). This adestined character of inscriptive practices is managed, controlled, and channelled externally (not intrinsically) in particular ways in particular reading practices through particular disseminatory techniques. Fragile networks of actors and verbal associations manage this channelling and use the very instability of meaning and the, in principle adestined, character of inscriptions so as to create socially productive but constrained and channelled processes of enormous social and technological power.

The link between actor networks and the verbal associations in texts for Latour is mutually constitutive. Verbal associations lead us to actor networks and vice versa. New verbal associations help create new networks of actors and vice versa. In practice it is difficult to separate them out, as there is a constant to and fro motion building up over time and through space. New notions do not emerge fully formed to then act upon the world. As Ian Hacking points out for experimental science, it is not a matter of two different scenes, two different sites, a strict process of this then that of theory and application. 12 Rather, science emerges through the inter-relation and mutual interpenetration of numbers of sites and heterogeneous resources (verbal and non-verbal, discursive and social). Latour insists that we look to phrases and actors, verbal associations and actor networks; that we pay careful attention to the text construction and, just as importantly, the text-handling characteristics of different inscriptive texts (where text construction does not simply involve "writing practices"; and text-handling "reading" practices - both also imply "social practices", organising institutional frameworks, in Latour's terms 'actor networks' and 'actor worlds'). Latour describes these collocations of actors and phrases as 'socio-logics'.

Language is a central resource of these socio-logics. Nikolas Rose and Peter Miller in their study of governmental programs (prima facie socio-logics) note that language 'provides a mechanism for rendering reality amenable to certain kinds of action' (77). Combining insights drawn from Foucault and Latour they argue that certain uses of language are rendered operable in 'intellectual technologies' with these intellectual technologies rendering 'aspects of existence amenable to inscription and calculation' (1). Thus 'theories...constitute new sectors of reality and make new fields of existence practicable' (7). These different 'technologies' exclude and eliminate, redispose and channel certain items; they work to 'formalize' and 'abstract' (Latour's terms) and so narrow the environment of available speech and action. These constraints are not so much institutions, power, and vested interests but rather particular regimes of calculation and action enrolling and enlisting particular persons and entities to the program.

Such intellectual technologies (socio-logics for Latour) are critically reliant upon the view of language which is well expressed in the oft-repeated formula of contemporary literary theory and critical linguistics: 'meaning is not simply something "expressed" or "reflected" in language: it is actually produced by it' (Eagleton 60). But this commonplace is given a particular inflection. Meanings may well derive from language - but only in so far as we mean language in social use, language as intellectual technologies. The meaning of a text may well be indeterminate but this is no more than an in principle possibility. As the meaning of a text is actualized through the many and successive determinations of meaning given to it, these determinations of meaning are not arbitrary, they possess a particular logic which is both socially and linguistically determined. Latour purposively brings together vocabularies associated with 'language' and 'social action'. He calls this 'science in action' (or knowledge in action). "Discourse" and larger regularities such as "discursive formations" are key resources in shaping social and institutional formations, but they are seen to gain their meaning, extension and social force through functioning as intellectual technologies or socio-logics.

The term socio-logic is appropriate in that it does not diminish the role of "language" and substitute for attention to it the study of social circumstances. Rather, it binds both together in the same way they occur in social practices. Even formal logic has an eminently social character (think of the drilling, the training and the educative apparatus required to produce competent logicians). Our logics are always and already social paths, social channels of a particular localised nature. Facts and truth are not then abstract logical schemata, they are social matters. Conversely, the concrete social circumstances which give "facts" their force include clusters of verbal associations which help give a "fact" or a "truth" its shape. If we are necessarily constrained by our available socio-logics, this does not mean that we cannot tinker with, change, transform, or create new meanings from them (but such tinkering will necessarily be done through rather than in spite of them).

Language Games, Discursive Formations & Actor Networks

The terms - language game, forms of life, discourse, discursive formation, actor network - help clarify socio-logics. They provide a means of linking language (concepts, ideas, words, expressions) to social action and things (persons, institutions, technologies, machines) and further specifying what is meant by each. "Language game" gestures to processes of text and information handling in the familiar verbal pathways such handling takes, the routines through which ideas are habitually arranged and associated, and the virtual transformative power of the language game to reproduce, extend, and maintain itself in space and time. For Wittgenstein 'the speaking of language is part of an activity, or of a form of life' (Part I, n. 23, 11). Language game becomes here substitutable with 'forms of life'. In Wittgenstein's conjunction - 'or' - lies the link between verbal and non-verbal, language and social organisation. It is not just a matter of seeing language as a social practice and then going about our business as before, as do much linguistics and social semiotics do. Rather language is always embedded in social practices and is inseparable from them. It can only be known through those practices, through the uses to which it is put. The instability of language lies in the ways in which verbal expressions can be manipulated and combined so as to form new associations enlisting (existing and new) forms of life. This implies a specific understanding of pragmatic use. As Ruthrof advances it, 'use is not simply an independent linguistic practice, it is a practice tied to Lebensformen [forms of life]' (111). For Ruthrof this is a two-way process:

one could say that a social organisation exists by virtue of a specific dynamic of 'forms of life'. And, we must add, it is the specificity of organized formed life which gives rise to (and not merely goes along with) concept formation and language. (111)

'Forms of life' is understood as an assemblage of fundamentally heterogeneous elements. Ruthrof quotes Schatzki who specifies form of life as 'the totality of practices, activities, and ways of being in which a person, or a group of people, is able to participate in by virtue of becoming a member of a particular social organisation' (Schatzki in Ruthrof 111). Meaning instability is therefore negotiated through numerous language games, in numerous forms of life. By implication then our reading of texts would be best understood as channelling processes which simultaneously create and constrain meaning - i.e., as so many language games, so many forms of life.

But this only emphasises the process, rather than the nature of the language used. To get at this nature it is necessary to consider the regularity and dispersion of verbal associations, the extension and maintenance of vocabularies. This is where Foucault's notion of discursive formation is needed. But our way of talking about such regularities must recognise that meaning can be no more than its (in principle limitless and in practice limited and contingent) instantiations. A discursive formation can only be unstable assemblages of verbal associations held together in multiple meaning instantiations through time and space.

Foucault provides just such an understanding of discursive formations. Rather than seeing a discursive formation as a 'chain of inferences', he critically insists on it being a 'systems of dispersion' through space and time (37). His invitation to follow the dispersion and the historical instantiations of meaning, projects discursive formations as intrinsically unstable verbal associations forming fragile networks of varying durability. Consequently a discursive formation cannot be specified as if it was 'a great uninterrupted text', as if it was 'tightly packed', 'continuous', or 'geographically defined'. Nor should it be conceived as a particularly 'well defined' set of notions (37).

The discursive formations that Foucault describes are made up of a 'series of gaps, intertwined with one another, interplays of differences, distances, substitutions, transformation' (37). Far from being comprised of normative statements, they are 'formulations that [are] much too different and functions that [are] much too heterogeneous to be linked together and arranged in a single figure'. Instead of well-defined conceptual architectures, discursive formations are characterised by 'concepts that differ in structure and in the rules governing their use, which ignore or exclude one another, and which cannot enter the unity of a logical architecture'. Rather than finding a discursive formation characterised by a permanent set of themes and preoccupations, they are characterised by 'various strategic possibilities that permit the activation of incompatible themes...or the establishment of the same theme in different groups of statement' (37).

For Foucault the description of a discursive formation is necessarily a description of these 'dispersions'. It is in this dispersal over time and space that the 'archaeologist' finds 'regularities' and practitioners in a field are held together by a loose collections of statements, themes, functions and so on. A discursive formation is therefore the discernment of a regularity:

an order in their [the dispersions'] successive appearance, correlations in their simultaneity, assignable positions in a common space, a reciprocal functioning, linked and hierarchized transformations. (37)

A discursive formation is characterised by commonality and contiguity; association and adjacency; stronger and weaker linkages. Thus we are dealing with a discursive formation:

Whenever one can describe, between a number of statements, such a system of dispersion, whenever, between objects, types of statement, concepts or thematic choices, one can define a regularity (an order, correlations, positions and functionings, transformations). (38)

Discursive formations are a means of describing regularity. Given that our understandings are partial and inherently localised to this or that meaning instantiation, we cannot speak of users (including researchers) fully knowing a discursive formation. Discursive formations do not in themselves produce regularity. Rather, they are a means for researchers and members to describe aspects of the "social frames" within which meanings are instantiated.

Foucault is careful here to distinguish between discursive formations and 'non-discursive domains' ('institutions, political events, economic practices and processes' (162)). Since we cannot know discursive formations except through the many meaning instantiations, the many statements made, the many connections established in them, the many readings, and the many interpretative methods that appear to be used, our understanding of discursive formations cannot be separated from their operationalising in those instatiations. We must therefore consider 'the specific forms of articulation' between 'discursive formations and non-discursive domains' (162). From the start we must seek 'how the rules of formation that govern it [the discursive formation]...may be linked to non-discursive systems' (162). Discursive formations consist of connections made in space and time. It is thus necessary to look to who makes those connections under what conditions; what governs the points of articulation between verbal associations and institutions, machinery, social practices; under what conditions did such statements emerge and are sustained in time and space. To examine a discursive formation is simultaneously to examine a form of life; a set of social practices which are not reducible to language or to discourse - although they may be translated into verbal language, particularly written language.

There is thus a need to hold together regularity and contingency, repetition and transformation in any concept of discursive formation. This is, after Wittgenstein, to see such formations as language games made up of assemblages of things (as so many tool boxes); it is, after Rorty, to see discursive formations as vocabularies (as so many collections of sentences and statements); it is, after Foucault, to focus on the operationalising of these vocabularies in social practice. It is to seek their points of articulation with what Foucault calls non-discursive domains and what Ruthrof calls other sign systems. This implies a textual vocabulary and focus upon the kinds and nature of the verbal associations made; and it implies a sociological vocabulary and focus upon who is making the associations, what networks are being formed through such associations, who is taking up or disputing the claims being made and how the discursive formation is itself a product of the systems of enrollment and enlistment of actors, of the mobilization and extension of non-human resources including machines and so on.

Latour argues that attention to the verbal associations leads inexorably to the actor networks and disseminatory structures, just as attention to actor-networks leads to the verbal associations. This does not imply that one can be substituted for another, only that both are always co-present with no necessary hierarchy or point of origin between each. 14 We may, of course, use the analytical formulations developed in text analysis to go through some characteristics of these inscriptions but any results obtained must be tempered with analyses of how these texts are physically handled, where they go, and what happens to them when they get there. This is because meaning instantiation cannot be reduced to texts - even though language and texts play an important a role in meaning. Discursive and non-discursive elements, linguistic and non-linguistic sign systems need to be held together. Without doing this we could not adequately establish meaning as a linking process connecting "social action" to "discursive formation" nor would terms such as "socio-logics" and "intellectual technologies" make much sense.

I don't want to be particularly legislative here. Of course, we may want to do something else with cultural studies than describe our objects in ways that are appropriate to them. We may want to intervene in them because with regard to other criteria these objects are inappropriate. Some of the writings on the semiotics of culture and film criticism lead to productive and deliberate misreadings of enormous potential, for all their ludic dimension. Manny Farber's practice of 'termite criticism' is a time honoured critical response of inventive film and cultural critics (see Thompson, Farber). Cultural criticism can be, after all, an important political act. I want to neither overvalue nor undervalue it. I simply want to say that such analyses are also engaging in cultural politics and are creating in the process one more local knowledge, one more way of handling this information and extending the range, nature and circulation of those statements.

Truth as Politics: The Contingency of Vocabularies

The view advanced here of language and its consequences, discursive formations and forms of life, holds out particular consequences for how we understand changes in the vocabularies we use, the nature of contingency and the instability of truth. In Contingency, Irony, and Solidarity Rorty understands change as the abandonment of one vocabulary and its substitution by another. He redescribes truth, changes in theoretical approaches, and the necessary reliance upon politics as a zone of competition between knowledges. For my argument Rorty - like Ruthrof - provides an important bridge between, on the one hand, rhetoric and text and, on the other hand, social practices, networks, institutions and non-linguistic behaviours. He shows the centrality of language to the creation and transformation of our world. He suggests with Ruthrof that the productivity of language - and therefore the contingency of the languages we happen currently to use - gives a tremendous semiotic and transformative capacity to our social world.

For Rorty 'a talent for speaking differently, rather than for arguing well, is the chief instrument of cultural change' (7). Diffusion and dissemination of such vocabularies are therefore a matter of others operationalising the vocabulary and thereby adopting it:

The method is to redescribe lots and lots of things in new ways, until you have created a pattern of linguistic behaviour which will tempt the rising generation to adopt it, thereby causing them to look for appropriate new forms of nonlinguistic behavior, for example, the adoption of new scientific equipment or new social institutions. (9)

Intervention capable of producing social transformations involves the development and social extension of new vocabularies in which the world is redescribed. The history of politics, philosophy or science therefore consists in the substitution of one vocabulary for another, one redescription for another. The succession of such vocabularies is seen to be a matter of just stopping doing one and doing another ('we might want to stop doing these things and do something else' (9)). Thus an actor would seek the extension of a vocabulary through making it 'look attractive by showing how it may be used to describe a variety of topics'. S/he says: 'try thinking of it this way'.

For Rorty new vocabularies do not come to us complete with their implications and consequences available to us in advance. They are made in history. In this he is following Dewey who wrote: 'meanings as objects of thought are entitled to be called complete and ultimate only because they are not original but are a happy outcome of a complex history' (171). 15 Rorty emphasises the piecemeal and necessarily accretive process out of which a new vocabulary takes shape and is enlarged upon. In this Rorty pictures "reformism" as much as "revolution", the piecemeal application of a metaphor alongside an emerging sense of gestalt shift. This is a gradual process of enlargement, the literalisation of metaphor, producing a sense of a new vocabulary, rather than some fully formed awareness of discontinuity and newness.

Rorty also notes that recognition that one is even using a new vocabulary may take time. The exposition of a vocabulary is largely a matter of the accumulation of redescriptions through 'enlarging the scope of one's favourite metaphors' (44). There will, therefore, be varying degrees of discontinuity and continuity between vocabularies. New and old vocabularies alike do not present themselves as already formed or hermetically sealed. Rorty sees such vocabularies as formed out of existing linguistic patterns in a piecemeal way through processes of redescription. If these vocabularies are to be successful they need to be in a constant process of development and discursive enlargement. (And, at this point, Rorty is close to Latour who would say that it is not a matter of the intrinsic qualities of a vocabulary but of all the transformations it undergoes later in the hands of others,16 and to Ruthrof who would stress the multiple instantiations of meaning in time and space.) It follows that the enlargement, transformation and adoption of a vocabulary by particular agents must provide a point of focus in the determination of the nature of the new vocabulary. Vocabularies have careers; these careers shape what we do with, and can know about them.

Although Rorty emphasises the intrinsic productivity of vocabularies (and therefore the semiotic potential of language) more than he does actors and their promotional work, his concept of 'use' does allow new vocabularies to emerge, be developed, and be sustained because of heterogeneous conditions: verbal as well as non-verbal, social institutions, machines, promotional activities. By shifting Rorty's conceptions a little, Rorty these can be seen to be close to Foucault's conception of the conditions under which discourse is produced. As Colin Gordon sees it, Foucault regards these conditions as neither unified nor coexisting in a hierarchy. Rather, they are:

concepts and objects, institutions and practices...different in kind among themselves and bear[ing] no resemblance to the form of the discourse they determine, yet they are all equally contingent historical realities. (Gordon 35)

The dynamic of vocabulary production is not one of discourse begetting discourse, but socially produced discourse formed out of the interaction of diverse and heterogeneous elements. In this situation the application of, say, a particular metaphor is part of a process not a point of origin.

Such a view has important implications for our understanding of the contingent nature of the vocabularies we use. As Rorty puts it

Our language and our culture are as much a contingency, are as much a result of thousands of small mutations finding a niche (and millions of others finding no niches), as are the orchids and the anthropoids. (16)

If language and culture are contingencies, then so must be truth. We cannot insist that "truth" is "out there in the world" because the world is simply not 'personlike in that it has a preferred description of itself' (21). Disputes between, say, competing analytical enterprises - e.g. between policy analysis and cultural critique in Australian cultural studies - are incapable of adjudication by some independent party. The "resolution" will come through the relative capacity of each to build up, maintain and expand networks of users of the vocabulary. And because meaning is not monolithic there would be some common vocabulary between them, and each will change over time, permitting permeation and incorporation of elements as well as affording new kinds of opportunities for disagreement. Therefore, for Rorty: 'every specific theoretic view comes to be seen as one more vocabulary, one more description, one more way of speaking' (57).

Given that contending theoretical views are in competition with each other, the acceptance, establishment and maintenance of some rather than other views will not have to do with their intrinsic rationality or superiority over others. Rather, it will have to do with the networks created, the actors that have become enlisted, and the attraction to such actors of the descriptions and machines such theoretical views make possible. For Rorty this entails a shift in our thinking 'from epistemology to politics' (68).

We can go so far with Rorty. But Rorty wants to go further. He specifies that this succession of vocabularies, this succession of truths, has a certain legitimacy in that truth in communication is understood 'as what comes to be believed in the course of free and open encounters' (67). By specifying the liberal-democratic condition of 'free and open encounters' Rorty is unnecessarily narrowing the conditions under which "truths" (or 'self-evidentiality' in Foucault's words) can be arrived at. Rorty seeks - where there can be no intrinsic guarantees - for the succession of vocabularies to result in some intrinsic good. Such a presumption leads him to curtail the shift from epistemology to politics, 'to an explanation of how political freedom has changed our sense of what human inquiry is good for' (68). But it is politics in all its forms that has changed our sense of what human inquiry is good for, not just the exercise of a specific form of politics, free and open encounters. Dewey, with his enthusiasm for science, talked of communication less restrictively describing it as 'consummatory as well as instrumental', seeing in it a 'means of establishing cooperation, domination, and order' (Dewey 202). (Harold Innis repudiates similar thinking ensconced in the judicial work of Oliver Wendell Holmes, providing a powerful argument against such a narrow specification:

Justice Holmes...stated that 'the best of truth is the power of thought to get itself accepted in the competition of the market' without appreciating that monopoly and oligopoly appear in this as in other markets. (32) 18 )

Rorty's description of meaning, truth and cultural and social change also prioritizes language, and more particularly the creation or invention and then subsequent transformation and extension through space and time of new vocabularies. For Rorty the foundational injunction becomes 'metaphor and self-creation rather than truth, rationality and moral obligation' (44). In this context Rorty axiomatically valorises new vocabularies, including those of the self-creating artist (Proust, Nabokov) and, to a lesser extent, philosopher (Nietzsche, Heidegger). These become the agency by which the system renews itself as new descriptions, and new metaphors come about. But, following Latour, we would rather say that new vocabularies enjoy no such intrinsic advantage over old ones just as old vocabularies enjoy no intrinsic advantage over new ones. I do not deny, in principle, the important transformative work of utopian thinkers, self-creating artists, or new vocabularies, but neither do I deny the stabilising, extending and reformist work undertaken by agents of the political status quo, ordinary artists, or older vocabularies. Change and transformation are not properties unique to new vocabularies; they are also available to older vocabularies. 17 Whereas the transformation and enlargement of vocabularies may entail both new and old vocabularies, these can be continuous and discontinuous with each other - and anywhere in between. As Rorty himself would argue decisions about vocabularies are made pragmatically based upon the attractions, entities, machines and corridors of thought that are made possible by each.

Additionally, there are problems with Rorty's over-reliance upon vocabularies and the under-utilisation of concepts of social action (such as forms of life, socio-logics, actor networks) in his account of social meaning. To be sure, as a term, "vocabulary" usefully marks out a stock of words capable of an indefinite variety of uses (and is, as such, close to Foucault's conception of a discursive formation), yet in Rorty's hands the term veers towards a position in which language is superordinate rather than 'subordinate, however efficient' (Ruthrof 12, my emphasis) to the broader social semiotic; and therefore superintending rather than corroborating with other sign systems. Whereas Rorty stresses the primacy of the "vocabulary" to the adoption of new equipment and social institutions, Latour and Hacking argue for an understanding of the important role of equipment, machines, experiment and observation in fashioning scientific vocabularies and theories. Hacking, for example, shows that experimental science 'has a life more independent of theorising than is usually allowed' (viii). Indeed Latour and Hacking would say that the linguistic pattern - in this case the scientific theory - is not itself fully formed prior to its enlistment of actors, its impact upon behaviours, social institutions and the creation of machines. It is also formed in the encounter with actors, behaviours, machines and institutions. Although more open than most, Rorty's theory of meaning suffers from the same problem as do all linguistic theories of meaning: it has difficulty accommodating Hacking's point that observations may sometimes precede 'any formulation of theory' (156). Verbal and written languages do not form an origin and ultimate destination for our social visions and practices. They supply our visions and practices with a potential.

It is surely ironic that Rorty is prepared to criticise Foucault for 'longing for total revolution' (65), while, at the same time, investing 'new vocabularies' and their preferred and typical agencies - utopian thinkers, artists, etc. - with the kind of socially normative power Foucault would have found romantic. In this context Rorty unnecessarily asserts that ' a matter of using new words as well as arguing from premises phrased in old words' (48), thereby making progress and new vocabularies synonymous. There are no grounds upon which such actors can be granted any privileged or necessarily pre-emptive power; nor are there grounds for new vocabularies to be so privileged as agents of progress; nor is it useful to limit 'conversation' to free and open encounters.

If Rorty's work does usefully connect with the understanding of science developed by the "science in action" school (of Latour, Callon, Law and Rip), the important difference lies in that Rorty, with his interest in the aesthetic text and conversation ('free and open encounters'), tends to stress the enlistment of vocabularies as assemblages of language as the principal resource, thereby sustaining a linguistic bias he elsewhere eschews. By contrast the science in action group sees the 'association' and 'juxtaposition' of verbal elements in an enlistment strategy entailing human and non-human resources. They recognise language in action rather than language as the (primary) social action. From the "science in action" perspective the actor, the reader, the writer of inscriptions are given shaping power. This implies a focussed attention (lacking in Rorty) onto the activities of writers, actors and readers and thereby seeing what kind of non-human, non-discursive and discursive resources are mobilised and enrolled. This also turns on a particular understanding of actors (the writers, readers, those who 'go along', 'work through', and 'give up') in these processes. As it stands Rorty is weakest when developing a relationship between his work and the how and why of 'new vocabularies' in social action in time and space. If Rorty had chosen other inscriptive practices to shape his model of conversation and language than verbal art, he would have - like Ruthrof - given more weighting to the polysemy of social practice and not just to polysemic texts. If he had done this he would still celebrate with Ruthrof 'the complex, material heterogeneity and relative fuzziness of language as its fundamental potential for new social visions and practices' (57).

Beyond Relativism

How then do we distinguish between vocabularies? Must we be relativists? Our use of vocabularies is determined by what kinds of things they permit us to produce and who they permit us to enlist. Our choice of vocabularies always has a direction to it - a particular pragmatism. Latour maintains that this selection in fact depends upon the resources and allies that it can recruit. Our vocabularies are shaped by those networks of enlisted actors for whom this vocabulary is translated and whose interests it may serve. This is far from a relativist's statement. It is a political statement in the sense of taking a side in relation to an actor's, an organisation's or a vocabulary's status or influence. This is why Latour cutely describes science as 'politics by other means' (cited in Callon, Law and Rip "Putting Texts" 222). Our use of a vocabulary therefore depends on the forcefulness of the words and the durability of the verbal and social linkages created around them.

The formulation of such linguistic patterns is accomplished through a process of promotion of the 'linguistic behaviour' leading to the enlistment of actors and the establishment of networks expanding through space and time. Users of 'new vocabularies' are thus engaged in Callon, Law and Rip's words in 'building a world and then attempting to convert that world into reality by enrolling others in it' ("Putting Texts" 227). In this sense the competition between theoretical views can be understood as 'actors attempt[ing] to impose worlds upon one another'. We can also understand meaning instability as the 'uncertainties involved in maintaining [the] power and size' (228) of those 'worlds'. For example, the outcome of this promotional work can be unexpected as users may find 'unexpected allies and resources being made available to them which can then be used to tip the balance of force their way' (Latour 259).

Callon, Law and Rip suggest that those involved in such enrolling and enlisting activities - such as scientists - may be best described as 'entrepreneurs'. This is because

They attempt to obtain the use of various kinds of resources, to fit them together and to profit from the results. In so doing they display a general lack of concern with the distinction between what is "internal" to science and technology , and what is "external"....Building a world in which experimental rats in one laboratory have a story to tell to readers of scientific journals all over the world requires the management of heterogeneous resources. The rats have to be bought and paid for, kept free from infection, enclosed in cages, prepared for drug injection and subjected to measurement. But the work of technicians must also be organised, laboratory resources must be secured in the face of competition from colleagues, research applications must be filed and papers must be written and presented. ("How to Study" 9)

The varied people encountered so far - film makers, policy developers, cultural critics - are also, in these terms, entrepreneurs. Consider the case of the film maker. Like the scientist, the film maker is constantly engaged in juggling heterogeneous elements. Film makers mobilize diverse resources such as funding agencies, producers, exhibitors, distributors, actors, instruments (cameras), film stock etc. They need to enrol as many actors as possible so the future of those actors is tied to the production and its subsequent circulation. Bits and pieces are assembled in an attempt to impose coherence and ensure that the film does not go off course, that the film is made and that the reputations of the film crew and the film director are maintained. Once completed and its textual form is stabilized the filmic text becomes a black box in Latour's sense. At this point a different and wider group of allies need to take it up: the public relations experts, the trailer makers, the distributors, and exhibitors. These create in turn another black box. And at this point another group of actors need enlisting - the audience. In turn these create another black box, giving their own particular shape to 'going along', 'working through' and 'giving up'. This reality forms the basis of the Truffaut saying that a film maker needs to be a business person in the morning and an artist in the afternoon. And just like the scientist's article which is mostly not attended to - the resulting film is mostly not watched, not taken up, people may go to the cinema or get it off the video shelf but not stay till the end. Video rental stores advertise movies for a week at $1 and get few takers. Such titles are exhausted, tired, used up, simply worthless. In this way the film production process is made up of different sorts of actors and resources being involved at particular stages of the production and circulation cycle. The film maker starts off with few resources, with few people acting on her or his behalf, s/he must in turn interest actors - often through the attraction of non-human resources. What started as isolated may now involve many different people - from screen actors, producers, funding agencies, studios and so on. Once completed a different network of actors are enrolled on its behalf and use it as an obligatory passage point - among these actors will be the paying audience. At each point the whole machinery may break down; and finally the audience itself may fail to be enlisted at all.

Through their control of resources of the environment and control over the world that is being built, scientific entrepreneurs are also 'practicing politics, economics and sociology' ("How to Study" 9). This has particular implications for those wishing to describe the activities and nature of inscriptive practices of film, policy and the like. It means that we do not have to create linking categories - we can rather follow the ones that they themselves make. As Callon, Law and Rip assert:

the analyst who follows scientists into their laboratories has no need to create his or her own categories and linkages. Since scientists are also practical politicians and sociologists, they are able to supply them in profusion themselves. The job of the analyst is rather to study the creation of such categories and linkages [my emphasis], and examine the way in which some are successfully imposed while others are not. ("How to Study" 10)

This is also the case for the popular art of cinema and television, and the spheres of government and policy making. Cultural studies is faced, as is the sociology of science, not so much with an object to apply a method to, but an object already with methods in place to describe its own activities and practices. Its task is one of adjusting such rationalities and strategies in relation to the analytical techniques and protocols of cultural studies. In these encounters cultural studies is and will be necessarily transformed by what it encounters as much as by what it brings to these encounters.


My argument is paradoxical in that "adestination" and "indeterminacy" are associated with "constraint" and "limits". Consequently the plurality of meaning in social use is understood as indefinite (not infinite). No in-principle political or ethical value is given to ludic misreadings or to near "perfect" communication; no superordinate value is given to working through or to going along. Sometimes the bending of propositional contents can be celebrated. At other moments, the play such constraints enable will be important. If the evidence of meaning instability and the artifice of social constraints upon meaning naturalise transgression and interdisciplinarity by encouraging us to produce instability, to work to mess up, recombine, redescribe, to reform and to revolutionise (indeed this is precisely what Rorty celebrates in his utopia of artists and revolutionaries), equally, that same evidence also naturalises work to protect, preserve, maintain, and extend the pathways and passageways of meaning. To propose one at the expense of the other is to overlook the fact that typically we do both - whether we do one or the other depends on the political and social context.

Intersubjective communication - such as reading, translating, describing and redescribing - is based not upon the identity of meaning, but its necessary approximation. Such an approach suggests the mutual interpenetrability and commonality of interpretive resources and acts shared in the different forms of meaning construction, within the different forms of life. This opens out onto the different kinds of consensus (or black boxes) and disputation that are socially available in our textual technologies.

Differences between and within forms of life - public, policy, aesthetic and academic - are both mobile and fixed. Each work with specific "angles of incidence". These are the typical forms of life, interventionary positions, protocols and strategies encountered in the matters before screen and cultural studies. Since the "styles" of intervention will vary depending on the conditions/forms of life specified, it is necessary to recognise available styles of argumentation and zones for their application. There are no a priori grounds for choosing one rather than the other. We simply do them at different times for particular social reasons.

"Meaning instability" simply exists, there is no purpose served in continually arguing for meaning instability and indefinite semiosis. Let's get on with attending to its consequences in diverse inscriptive practices and their attendant socio-logics. And this means finding out how meaning instability, negotiation and approximation can become the basis of varying kinds of social, intellectual, political and aesthetic program.


1. Godard's is a more polyvocal text, mine more mono-generic. But Godard's is less polyvocal than might be expected - his is, after all, a film essay which today reminds of Michel de Certeau's work.

2. Indeed some of the most rigorous textual and institutional analysis is now being done in sociology and literary and film studies respectively. Textual analysis is the meat of ethnomethodology; for an exemplary analysis within this field of sociology see Jayyusi, "The Equivocal Text and the Objective World". Ian Hunter's discussion of the emergence of English is an exemplary institutional analysis as is also Bordwell et al.'s analysis of classical Hollwyood cinema. See Hunter, Culture and Government and David Bordwell, Kristin Thompson and Janet Staiger, The Classical Hollywood Cinema.

3. Dewey's metaphor of 'conversation' that begins this chapter is apt as long as we maintain his enlarged sense of conversation as negotiation which would include reading, writing, the lover's discourse and the acrimonious interchange over the divorce settlement, the board room meeting and management/shop floor relations, not just pleasant conversations among colleagues and friends.

4. In a sense this is because the 'cost' of falsification in terms of the resources and the degree of 'working through' required is much higher in scientific activity than it is for film and cultural criticism. But, by the same token, film and cultural criticism cannot physically change their "object" as much as can the scientist. But film and cultural criticism can change the interpretation of this or that program and indirectly influence the creation of new programs.

5. Peirce writes: 'at present the word [pragmatism] begins to be met with occasionally in literary journals, where it gets abused in the merciless way that words have to expect when they fall into literary clutches' (176).

6. Like their scientific counterparts, popular authors can also be thought of as using heterogeneous resources such as sign systems, instruments, books, projectors, CD players and distribution and exhibition systems as their allies.

7. There is an interesting comparison to be made between these three options for reading and those advanced by Stuart Hall in his classic article "Encoding/Decoding". Hall's alternatives for the reader/viewer in that article are 'dominant-hegemonic position' (something like 'going along'), 'negotiated code' (a cross between going along and working through in that it is defined as a mix of 'adaptive and oppositional elements'), and 'aberrant decoding' (which does not fit so well with 'giving up' and can almost look like 'working through' in that it involves decoding the message in a 'globally contrary way'). See Hall 136-138. It is beyond the scope of this essay to comment on this comparison except to note that Hall's schema is broadcast media centered and judgemental about the intrinsic value of the different options. Latour's schema has a more immediate applicability to other textual technologies, is quite good at exploring production strategies as strategies for inciting reception, and is purposefully neutral about the different options for readers/viewers and writers/producers alike.

8. Carroll contends that the combination of Marxism and pyschoanalysis that dominated film theory in the 1970s and early 1980s has fundamental methodological problems at a logical and an empirical level (6-7).

9. To be sure Bordwell's purposes are not mine here. He investigates the institution and finds it wanting in part because it is in this shape. He suggests we can go beyond this situation if only there were a differently constructed film theory rather than the existing literary and psychoanalytically dependent film criticism. Hence his support of an historical poetics. For my purposes Film Criticism Inc. is necessarily the way it is and Bordwell's prescriptions for change - which I have much sympathy for - would not fundamentally alter the institutional dynamics of film criticism that he discloses.

10. This is also true for film reviewing which Meaghan Morris aptly describes as a 'bit of a newspaper' rather than as 'parasitic' upon the film industry (121).

11. This point about the commonality of science, policy and popular entertainment is underlined by John Law when he writes 'A run of (scientific) successes does not preclude the possibility that next time another actor-network will array such a powerful structure of forces that one finds oneself localised, unable to publish, perhaps with one's forces so scattered that it will be difficult to publish anything using, for instance, one's favoured method ever again.' ("The Heterogeneity of Texts" 71).

12. See Ian Hacking, Representing and Intervening. See especially the last half of the book dedicated to demonstrating how the socio-logics of science confound philosophical notions of science, knowledge and the mind.

13. For reference to intellectual technologies see Rose and Miller (1-31). In a related fashion Lena Jayyusi, in the context of an analysis of news discourse, argues for the explication of 'the socio-logic of knowledge-in-use, and of the procedures, understandings, and practices, by which a sense of objective reality is constructed. The notion of a socio-logic addresses itself to the ways that the normative, the conceptual and the practical are always concurrent and co-constitutive features of actual real-worldly contexts and of their mundane intelligibility'. See Jayyusi (185-6).

14. Latour likens the situation here to road maps: 'all paths go to some place, no matter if they are trails, tracks, highways or freeways, but they do not all go to the same place, do not carry the same traffic, do not cost the same price to open and to maintain. To call a claim 'absurd' or knowledge 'accurate' has no more meaning than to call a smuggler trail 'illogical' and a freeway 'logical'. The only things we want to know about these socological pathways is where they lead to, how many people go along them with what sort of vehicles, and how easy they are to travel; not if they are wrong or right' (205).

15. As ever, Latour describes the situation more brutally: 'The fate of the statement, that is the decision about whether it is a fact or a fiction, depends on a sequence of debates later on...the status of a statement depends on later statements. It is made more of a certainty or less of a certainty depending on the next sentence that takes it up; this retrospective attribution is repeated for this next new sentence, which in turn might be made more of a fact or more of a fiction by a third, and so on' (27-28).

16. Latour's text actually reads: 'To determine the objectivity or subjectivity of a claim, the efficiency or perfection of a mechanism, we do not look for their intrinsic qualities but at all the transformations they undergo later in the hands of others' (258).

17. The study of society, politics, philosophy and social formations relies significantly on older vocabularies such that claims for new vocabularies can often amount to rediscoveries of older arrangements. In this ideas of 'progress', with its scientific exemplar of a succession of vocabularies, may actually be a misnomer for what happens in the social sciences. Robert Merton has written eloquently about this and Herbert Gans has recently noted the routine reportage as new discoveries of research results already reported by past researchers.

18. Ever since I first read this passage it has been a powerful metaphor guiding my understanding of the career of 'knowledges', their uptake and the investment in them.

Works Cited

Ang, Ien. Desperately Seeking the Audience. London: Routledge, 1991.

_____. "Stalking the Wild Viewer" Continuum: the Australian Journal of Media and Culture 4:2 (1991): 19-35.

Bordwell, David, Kristin Thompson and Janet Staiger. The Classical Hollywood Cinema: Film Style and Mode of Production to 1960. New York: Columbia UP, 1985.

Bordwell, David. Making Meaning: Rhetoric and Inference and Rhetoric in the Interpretation of Cinema. Cambridge, Mass.: Harvard UP, 1990.

Callon, Michel, John Law and Arie Rip, eds. Mapping the Dynamics of Science and Technology. London: Macmillan, 1986.

Callon, Michel, John Law and Arie Rip. "Putting Texts in Their Place". Callon, Law and Rip. 221-230.

Callon, Michel, John Law and Arie Rip, "How to Study the Force of Science". Callon, Law and Rip. 3-15.

Carroll, Noel. Mystifying Movies. New York: Columbia UP, 1988.

Chambers, Ross. "An Address in the Country: Mallarme and the kinds of literary context". French Forum xi:2 (May 1986): 199-215.

de Certeau, Michel. The Practice of Everyday Life. Trans. Steven Rendall. Berkeley: U of California P, 1988.

Derrida, Jacques. "Signature Event Context", Glyph: John Hopkins Textual Studies, 1. Baltimore: The John Hopkins University Press, 1977. 172-97.

Dewey, John. Experience and Nature. New York: Dover, 1958.

Durgnat, Raymond. "Up Jumped the Devil or, The Jack In Pandora's Box", Monthly Film Bulletin 644 (September 1987): 265-8.

Eagleton, Terry. Literary Theory: an Introduction. Oxford: Basil Blackwell, 1983.

Eco, Umberto. Interpretation and Overinterpretation (Cambridge and New York: Cambridge UP, 1992, Stefan Collini ed. with contributions from Richard Rorty, Jonathan Culler, Christine Brooke-Rose.

Eisenstein, Sergei. The Film Sense. Trans. Jay Leyda. London: Faber and Faber, 1970.

Elsaesser, Thomas. New German Cinema: A History. London: BFI and Macmillan, 1989.

Farber, Manny. Negative Space. New York: Praeger, 1971.

Foucault, Michel. The Archaeology of Knowledge, trans. A.M. Sheridan Smith. London: Tavistock, 1974.

Freadman, Anne. "The Vagabond Arts". In the Place of French: Essays in and Around French Studies in Honour of Michael Spencer. Mt Nebo, Brisbane: Boombana Publications, 1992. 257-291.

Gans, Herbert J. "Sociological Amnesia: The Noncumulation of Normal Social Science". Sociological Forum 7:4 (1992): 701-710.

Gilbert, G. Nigel and Michael Mulkay. Opening Pandora's Box: a Sociological Analysis of Scientist's Discourse. London: Cambridge UP, 1984. 112-140.

Gordon, Colin. "The Subtracting Machine". Ideology and Consciousness 8 (Spring 1981): 27-40.

Hacking, Ian. Representing and Intervening: Introductory Topics in the Philosophy of Natural Science. Cambridge: Cambridge UP, 1983.

Hall, Stuart. "Encoding/Decoding". Culture, Media, Language: Working Papers in Cultural Studies, 1972-79. Eds. Stuart Hall, Dorothy Hobson, Andy Lowe and Paul Willis. London: Hutchinson in association with the Centre for Contemporary Cultural Studies, University of Birmingham, 1980. 128-138.

Hartley, John. Tele-ology: Studies in Television London: Routledge, 1992.

Heath, Stephen. "Film and System, Terms of Analysis". Screen 16:1 (1975): 7-77; and Screen 16:2 (1975): 91-113.

Hunter, Ian. Culture and Government. London: Macmillan, 1988.

Innis, H.A. The Bias of Communication (1951). Reprint. Toronto: University of Toronto Press, 1972.

Jayyusi, Lena. "The Equivocal Text and the Objective World: an Ethnomethodological Analysis of a News Report". Continuum: the Australian Journal of Media and Culture 5:1 (1991): 166-190.

Kuhn, Thomas. The Structure of Scientific Revolutions. Chicago: U Chicago P, 1970.

Latour, Bruno. Science in Action. Cambridge, Mass.: Harvard UP, 1987.

Law, John. "The Heterogeneity of Texts". Callon, Law and Rip. 67-83.

Lotman, Yuri M. The Universe of the Mind: A Semiotic Theory of Culture. Bloomington and Indianapolis: Indiana UP, 1990.

MacKenzie, Donald. "The Problem with 'The Facts': Nuclear Weapons Policy and the Social Negotiation of Data". Information and Government: Studies in the Dynamics of Policy Making. Eds. Roger Davidson and Phil White. Edinburgh: Edinburgh University Press, 1988. 232-251.

McHoul, Alec and Tom O'Regan. "Towards a Paralogics of Textual Technologies: Batman, Glasnost and Relativism in Cultural Studies". Southern Review. 25:1 (1992): 5-26.

Merton, Robert. "The Fallacy of the Latest Word: The Case of 'pietism and science'". American Journal of Sociology. 89 (1984): 1091-1121.

Michaels, Eric. "A Model of Teleported Texts (With Reference to Aboriginal Television". Continuum: the Australian Journal of Media and Culture 3:2 (1990): 8-31.

Morris, Meaghan. Pirate's Fiancee. London: Verso, 1989.

Mulvey, Laura. "Visual Pleasure & Narrative Cinema", Screen. 16:3 (Autumn 1975) republished in Popular Television and Film. Eds. Tony Bennett, Susan Boyd-Bowman, Colin Mercer and Janet Woollacott. London: BFI, 1981. 206-215.

Munsterberg, Hugo. The Film: A Psychological Study: The Silent Photoplay in 1916. New York: Dover Publications, 1970.

Peirce, C.S. Collected Papers of Charles Sanders Peirce, v.5. Eds. Charles Hartshorne and Paul Weiss. Cambridge, Mass: Belknap Press, 1974.

Rorty, Richard. Contingency, Irony, and Solidarity. Cambridge: Cambridge UP, 1989.

Rose, Nikolas and Peter Miller. "Governing Economic Life". Economy & Society. 19:1 (1990): 1-31.

Routt, William D. Prolegemena to a Theory of Popular Art, PhD Diss., University of Chicago, 1982.

Ruthrof, Horst. Pandora and Occam: On the Limits of Language and Literature. Bloomington and Indianapolis: Indiana UP, 1992.

Schatzki, Theodor R. "The Prescription Is Description: Wittgenstein's View of the Human Sciences". The Need for Interpretation. Eds. S Mitchell and Michael Rosen. London: Athlone Press, 1983. 118-40.

Thompson, Richard. Interview with Manny Farber and Patricia Patterson. "Termites". Film Comment May-June (1977): 36-45, 54-60

Wittgenstein, Ludwig. Philosophical Investigations. Trans. G.E.M. Anscombe, Oxford: Basil Blackwell, 1974.

New: 2 December, 1995 | Now: 24 March, 2015