Fuller summary of S Hall 'Culture, the Media and the "Ideological Effect"'

Section one  -- ‘theoretical mapping’ (my title)

This section involves Hall trying to develop a marxist account of culture. We know from the German Ideology that culture is rooted in a double relation to men and to nature. When human beings labour (interact with nature) it very quickly turns into a form of social organisation. For example, human beings collaborate in order to hunt and gather, or to produce even the simplest of things. Soon after this very early stage, in other words, humans (Hall calls them ‘men’) and their relation to nature is always socially mediated. As a result human beings rapidly need to reproduce society as well. That society then needs to reproduce the next generation of human beings as social individuals. However, the material reproduction of existence  (labouring to produce materials needed to survive) is always the first and most important human activity, and is thus 'determinant in the last instance'. All subsequent social complexity arises from this initial process -- complex matters such as the division of labour, different types of society, different forms of association, families, different types of state, different kinds of belief. This account is the basis for any materialist understanding. This is what Marx means by talking of  'the forces and the relations of production' and their interaction.

However, this materialist process is not to be understood as a general abstract process sufficient in itself. Marx urged us to pursue specific historical understandings. It is clear that different types of society can arise from the same form of general material production. Thus, tracing society to its material production in general is not sufficient -- this would be undialectical materialism. We are told in the German Ideology instead that definite individuals enter into definite social relations. It follows that we should understand the production of ideas in the same specific historical way. For example a capitalist mode of production is founded on a specific type of exploitation but different capitalist societies can have different specific superstructural or ideological forms [compare the French, British, German and American social formations, for example]. 

It is worth pointing out that because material production is always contradictory this process of social development will still always show traces of crisis and supercession 

There are definite modes of life as well as the more familiar modes of production. Differences in a mode of life follow from the way labour is organized, the level of technical development, the way institutions are designed to produce and circulate goods, and the way in which they realise value. Different forms of civil association and different forms of a State can also arise. However these different institutions are connected to each other as an ensemble or a series of interconnected levels and they are experienced as a totality

This basic Marxist analysis can be used to grasp the concept of culture (one concept which Marx himself does not use). Cultures can be defined as forms of social existence in definite historical circumstances, whereas notions of the social refer to the content of those forms. Here, Hall is developing an idea first grasped by structuralism [or what some people know as 'semiotics']: what matters is the arrangement of elements and how they produce a social system. Culture can be seen as a design template for social life, as in an anthropological sense, and this also links to the work of Raymond Williams and E P Thompson [two former buddies Hall wants to acknowledge --both have written very important historical works, of course]. 

Culture is also a productive force, a kind of record of our mastery over nature. Culture is therefore materialised, rather than just recorded as knowledge, materialised in institutions and in language. Hall refers back to the German Ideology again here. [and develops an argument referring to "stock of knowledge" -- although not in those terms]. Language is obviously very important in this process -- it objectivates, clarifies, allows abstractions and so on. The use of human language is what distinguishes us from animals, or, in Marx's example, separates the worst of architects from the best of bees. Language permits purposeful design. 

Language is a matter of  'practical consciousness' for Marx, but it can be abstracted and pressed into the service of Capital. What makes this difficult to understand is that our practical consciousness becomes a matter of 'second nature', as our knowledge and language is passed to new generations unreflectively. 

So, ideas and culture are important but they still arise initially from social and material life. In the German Ideology we find a simple reproductive relation between these two levels. Society is thus thought of as 'an expressive totality', a rather simple organism. There are fairly simple forms of determinism and change -- the economy changes, and so does culture and in the same way. These are homologous changes found in each level. At the root of this model is an idea that praxis and its constant alienation is the foundational link [see comment] -- human praxis takes place (but under definite material limits and conditions independent of men's will, Marx reminds us)  and it is this that produces the  'correspondences'  between each level of practice. 

Here, Hall wants to depart from this simple model. Human beings do not always experience their society and culture quite in this way. They can have a  'false consciousness' of the real conditions of their life, Marx insists. In other words, human language enables and distorts, it binds and fetters human culture, it clarifies and conceals. For example, in a famous metaphor Marx tells us that false consciousness represents things as  'upside down, as in a camera obscura'. [a camera obscura was a Victorian optical observation device consisting of a lens mounted on the roof of a building which projected views of the streets outside on to a table. As is often the case with cameras and lenses, the image appeared upside-down]. 

In the middle of this, Hall tells us, there is an argument that human creativity is being limited by the circumstances in which they live. Humans are  'de-centred' [a trendy term at the time] -- affected by definite conditions and unable to fully author their own culture and social life. This is the effect of superstructures  -- they condition practices, they produce ideology. One key function of ideology, therefore, is to de-centre [de-author] and displace the processes that produce culture. 

The simple model founded in the German Ideology needs to be rethought. In Marx's later work, another model seems to appear. This one does not offer a smooth correspondence between different levels but features disjunctures instead between the material relations of production and social practices -- for example between class and social relations -- and between superstructures and ideological forms (forms of consciousness which are appropriate to different social practices).  This new model has three levels  [it is the EPI/C model associated with Althusser] [see comment]. These levels are often apparently autonomous -- as seemed to be the case for ‘ideas’ in Germany in the 1840s and 1850s, which is what Marx and Engels were criticising in the German Ideology. This autonomy and separation arises partly from the split between mental and manual labour, which is institutionalised in modern societies. Moreover, the means of mental labour are also appropriated by capitalists who develop dominant ideology and a set of 'ruling ideas.' 

The tension in the models in classical Marxism mirrors those found in the concept of culture in other work. Culture also operates at two levels. At the social level, tied to the relations of production, we can often find a working class culture, for example, which reflects the specific shape of class relations as they affect the working class. This kind of culture was described in work like Richard Hoggart's book The Uses of Literacy [a classic by Hall's predecessor at the Birmingham Centre]. This working-class culture is lived, or in other words it reflects real conditions, and offers a kind of genuine experience. It is also made sense of, accounts are formed of it, AND ideas are used to bring to it a 'certain imaginary coherence'. (page 322). Only this latter coherence might usefully be seen as  'ideology proper', says Hall. 

These senses of culture, as both lived experience and also as an active account, thus need to be separated, Hall thinks. Marx himself lumped these two senses together in terms like  'superstructures' or  'ideological forms'. They need separation simply because lived experience itself cannot be false [stemming as it does from the real lives of the participants] although ideas about that experience clearly can be. 

We are back to the function of language in providing elements of meaning. In an aside, Hall wants to remind us here of the French work on language (structuralism or 'semiotics'): language comes before individuals, individuals find meaning by locating themselves in given discourses. Certain formal discourses are also called [high] 'culture' [economics, philsosophy, literature etc]. 

Hall says we should expand on this work. He also gives an example that helps us see a link with marxism -- the market. 

In capitalism, labour must be socialised [socially organised], but the market misrepresents this sociality, and makes everything appear as a matter of free individuals negotiating with each other. Of course, this is true at one level. That markets exist and that they operate with individuals is clearly true -- but what is not true is that market forces described the entire social process of production. Here, or actually slightly later, Hall cites the famous example found in Capital volume one. Here is my account of this example: 


Marx invites us to consider a day in the life of a labourer. At the beginning of the day, individual labourers negotiate their day's wages with individual bosses. At that moment, which may take half-an-hour, labourers and bosses interact as free individuals. The labourer is under no compulsion to work for that boss, and may try his luck elsewhere for a higher wage. However, once the contract is agreed, for the rest of that working day, they are extremely unequal. Once the labourer enters the factory, he is under the control of the boss, he labours to meet the boss' s schedule, and he produces value that the boss appropriates -- see discussion. It is therefore extremely misleading to take that first half hour as representative of the whole day. However, that is precisely what liberal theory does -- using that one moment of equality as a description of the whole society. 
Incidentally, and this is another aside of my own, or Marx uses the same kind of argument against parliamentary democracy as a matter of equal votes and equal rights. Again, seeing the business of 'one person one vote' as the heart, guarantee and essence of democracy is true, but misleading -- it is true of just one-day, polling day, and after that the normal inequalities resume and dominate us. 
This kind of idea, whether it is about market forces or parliamentary democracy, acts perfectly to describe the functions of ideology: 
  • in the first place these mechanisms misrepresent the social as a matter of individuals  (which is upside down, as in a camera obscura,); 
  • secondly these mechanisms allow one element to stand for the whole (as in another marxist term 'fetishism' -- see discussion); 
  • and thirdly such mechanisms make the real foundations disappear from view (concealment)  (page 323). 
In all these senses, market relations are both real and ideological: they are important but only phenomenal forms, to take a phrase developed in Marx's later work. 

Hall comments on this discussion using terms developed well in Mepham's work (see comment). The argument assumes the existence of different levels and a dislocation between them rather than an expressive totality. We move up from one level to the other via a transformation. The material forces of production are transformed into relations of production rather than just reflecting those forces. Moreover, this misleading appearance generates another level, as it is itself theorised about -- and this is the error in political economy, which theorises merely about the phenomenal forms [this is where the example of the day in the life of a labourer appears, but Hall also links us to the discussion in Grundrisse]. Ideology, in particular, is not a matter of what is hidden, says Hall, but a matter of how to understand that which appears openly. He uses another fashionable phrase -- conscious activity has an unconscious dimension  'beneath the surface'.  [Here and elsewhere, Hall is using these terms in these little asides in order to try and steal the thunder of rival theoretical perspectives, and reclaim them for marxism: the concept of the unconscious originates in Freud, of course, but was also appropriated and re-worked by Lacan in a way Hall was never really keen upon]. 

Here is another aside, on the notion of  'common sense':

Common sense is used because it works, it is shared, and it appears 'natural'. However, it is clearly constructed and limited by specific circumstances -- Hall cites Robinson Crusoe's common-sense as an example. Crusoe simply behaved ‘naturally’ as he saw it, but in a way which clearly reconstructed 19th century capitalism. Our common sense is a combination of past ideas and traditions too. It does not generate a systematic knowledge of its own origins. Becoming conscious of these origins is a revolutionary act, as Gramsci told us [This is the first mention of Gramsci in the piece]. 
Keeping people unconscious of the origins of their thought is a major aspect of ideology for Althusser, hence his stress on the importance of structures rather than ideas for understanding ideology, on experiences rather than thought, and on how people imagine how they live the relation between themselves and their conditions of existence rather than how they actually know how things are. 

Hall's discussion openly follows Althusser's criticisms of humanism and historicism (see comment for a brief account). These approaches see the social formation as a simple structure (even if a dialectical one), an expressive totality, where the same dynamics are found in all the levels of the social formation, which are expressive in turn of the one fundamental contradiction in the economy [between creativity and alienation].  Praxis is forever objectified and alienated, as a universal underlying process. Instead, following Althusser, we can think of the social formation as a structured complexity lacking a single essence, constructed by its practices, where each level has its own specificity and history. Social formations change as a result of uneven development, because changes can begin in any level, although the effects can spread into other levels. The levels are not fully autonomous and independent, though, but they are articulated, which implies that levels are different rather than expressively similar. We have here a complex determinacy exercised by the economy. Althusser talks of  'over determination' – meaning, in this case, that the contradictions in other levels can accumulate at the economic level. Economic conflicts, therefore, tend to be over-determined by ideological and political elements acting on them as well. Even though the economic level often seems to be the leading site of struggle, this is so only because it is really  'structured in dominance' [i.e. not naturally or inevitably dominant]. 

Section Two -- Language

Language operates at the ideological level and produces various kinds of social knowledge. Ideas are expressed and embodied in it. Language is a system of signs and discourses. Language is social rather than individual, of course. Individuals need access to the codes found in language in order to express themselves. Language operates at all levels of the social formation but it is itself also structured by these other levels.  Hall relies here on the work of the 'Russian formalists', people like Volosinov and Vygotsky, who worked under Stalin in the 1930s and 1940s, and were much discussed as a way of reconciling modern French work on language ('semiotics') with Hall's marxist preferences. Hall argues that language must be open to historical materialist analysis as well -- for example current usage must reflect the class structuring of capitalist social relations [this is where Volosinov comes in as we shall see]. But again language also has its own internal rules and laws. 

This is exposed in the classic work on the sign [semiotics]. Signs refract reality rather than simply standing for an object unambiguously. The internal organisation of signs in systems is equally important in understanding how meaning is conveyed: this internal organisation helps actually to articulate the relations between real objects. This is seen especially when looking at the relation between language and complex social objects. Sometimes, ideological domains describing these objects are fully inscribed, thoroughly articulated, while others are relatively empty, and thus open to signification as a form of labour or practice. 

Reality, especially social reality, can only be understood via the  'relay of language' -- but this is variable, and open to 'ideological displacement or inflection'. This is the origin of 'imaginary lived relations'. Volosinov expresses this best for Hall -- signs can either distort or help an accurate perception according to the point of view being developed [ultimately, a class point of view for Volosinov?]. Ideology is always possible, therefore, and all ideology involves semiotics. [but must the reverse apply too -- must all semiotics be ideological? -- see comment].  We usually get a complex ideological field of discourses, in fact, each discourse offering specifics according to its locality -- for example artistic discourses, religious discourses, political discourses and so on. [It is worth adding here that an argument that gramscians found very exciting is also found in Volosinov's work -- that each discourse bears the traces of ambiguity arising from class struggle, a 'multiaccentuality', and therefore offers a potential for new class struggle within it as well]. 

This connects with the work of Poulantzas on regions [another key author]. Poulantzas argues that there are regions inside a capitalist social formation -- for example a juridico-political region often manages and hides the operation of the economy. This region can often appears dominant and central. [I suppose the political region certainly appears that way in our society on occasion]. That dominance of a particular region is often demonstrated by finding borrowings from it in other ideologies [the example here might be the way in which management-speak finds its way into educational, personal, and religious discourses as well.]  With Poulantzas, we are once more far away from ideology as a matter of false understandings held by individuals. Individuals are always inserted into a social context -- and here Hall just briefly cites the [pretty dated marxist-ish] work of C Wright Mills on 'situated actions' and 'vocabularies of motives'. [There is also a very brief critical aside on recent trends to incorporate Freud into marxism, especially via Lacan -- you can read about this on page 330 if you wish. Basically, Hall thinks it is mistaken]. 

So, to sum up, signs are arranged into codes and sub codes and appear in intertextual applications. These qualities enable a continual effort to go on to signify, to make sense of new circumstances. This can happen through the idea of  'connotations', for example. Codes are organised into domains: they clarify, sediment, and map meanings. This implies a series of corresponding practices. However, there are different degrees of knowledge here, and different decipherings. [Hall refers specifically here to Barthes]. This work clearly links, for Hall, with Marxian analysis on the limits imposed by dominant ideology as in the example of the market. These linguistic limits explain the limits of commonsense, and what Lenin called  'trade union consciousness'. In this way, the limits and qualities of language are also made to perform functions for Capital. 

Section 3 -- 3 related concept of domination 

Firstly, Raymond Williams has recently developed a new idea of dominant ideology, which focuses on the process of incorporation. The education system has an important role here for RW [it is strangely ignored by Hall -- see comment] -- it helps to emphasise certain themes, discard others, offer reinterpretations or dilutions, and to reconcile different concepts, thus producing a dominant system which constantly adjusts and incorporates new elements. This model implies a lot of alternative meanings are in circulation, but it also emphasises how these are controlled or incorporated. The role of  'tradition' is important, for example. There are also two kinds of important elements:  residual elements arise from the past  (e.g. the idea of rural community). These elements can offer critique of the present but from a less effective standpoint rooted in the past. The other kind of elements are emergent ones, which arise from new practices -- [for example from technical change?] -- and these are either partially incorporated or remain as deviant options which vary, but do not threaten the status quo. 

This account owes much to Gramsci and his notion of hegemony [which is the second 'related concept']. This occurs when an ‘historical bloc’ (an alliance of class fractions)  coerces and exerts  'a total social authority' over the social formation. The leading blocs direct, lead, and organise so as to command and win consent. In liberal capitalism, consent dominates rather than force. Hegemony is accomplished on the terrain of the superstructures. It works by ideology -- favourable  'definitions of reality' are institutionalised and come to constitute the 'primary lived reality' (page 333). Hegemony unifies a bloc. It works via the control of frames rather than detailed contents; it sets limits. Consent is won via the existing ideologies, which can have differences --eg some might be residuals. [comment]

The whole mechanism requires a notion of an alliance of class fractions in specific conjunctures [i.e. specific combinations of elements]. The approach stresses the active winning and securing of consent -- it is not a permanent process, there is no total absorption of dissent as in Marcuse' s account in One-Dimensional Man [a famous 1960s account of popular culture, far too 'pessimistic' for the gramscians who missed the upheavals of 1960s, by and large, but yearned for more 'activism' and ddin't like accounts that suggested capitalism was gaining the upper hand]. The subordinate class culture remains but is contained. In some conjunctures, this can be organized to form a counter-hegemonic force, but in others it can be used to prolong subordination -- as in trade unionism. Even when it takes an 'economist'  form it is never totally subsumed, but managed. It remains as a complementary culture rather than existing as a source of radical conflict. The balance between dominant and subordinate class cultures exists as an unstable equilibrium, which is especially apparent in the State, where particular interests have to be managed to appear as general ones. 

This  'immense theoretical Revolution' is attributed to Gramsci, and Hall comments approvingly on the enlarged concept of domination it involves: in modern societies the winning and opposing of consent takes place at all levels, and power clearly includes non-coercive elements as well. Domination is certainly not just a matter of a ruling class following its interests. Further, ideology is both structural and epistemological. 

Althusser offers the third ‘related concept’. In Hall's view this is  'Closely inspired by and elaborated from Gramsci ' anyway (335), but Althusser rejects Gramsci's historicism [whether Hall does is more debateable]. The key is found in the ISAs [ISA = Ideological State Apparatus] essay discussed here. Althusser emphasises that reproduction is the key. This includes the social reproduction of both labour power and the relations of production. Reproduction requires things like a system of wages, a set of skills AND a set of appropriate ideas: these include a submission to ruling ideology for the working class, and the ability to manipulate ruling ideology for ruling class. It requires the intervention of agencies not in production itself -- agencies like the family, education, cultural and political institutions, and the State. These all have a key role to represent the neutrality of society while following the long-term interests of capitalism -- hence Althusser's insistence that these institutions be called State apparatuses. (Hall argues that Althusser has exaggerated the role of the State in this (page 335)). 

The ISAs rule by ideology, indirectly, via displacement. Althusser himself believed that the school-family couplet was an especially important unifier [again, not Hall though?]. 

Ideology exists as system of ideas [and practices] which perpetuate an 'imaginary relation to the real conditions of existence'. This is close to Gramsci, but with more emphasis on the necessary contradictions in ruling class ideology, which becomes a stake and site of class struggle -- [so not the usual functionalist reading then!].  The unity of the social formation is only via a  'teeth-gritting harmony' , not a functional fit. However the overall appearance of Althusser's schema does look  ' more functionalist than he would clearly like' (336). 

Section 4 -- What does Ideology do?

Gramsci says there are two main floors to the superstructure -- civil society and the State. The boundaries shift and blur. Poulantzas helps us make progress via the notion of  'separating and uniting' (336). [ I think here and elsewhere, Hall is close to an excellent piece written by one of his former students -- see comment]. For example, in the market we find separate agents but they are also united, bound together by  'a hidden hand'. In Poulantzas's Politics and the State, people appear as individual subjects rather than classes, but are also bound together as a nation, via the social contract and some idea of a general interest. Hall thinks that a lot of ideological regions display the same mechanism. 

Ideological effects therefore involve: 

  • firstly a masking and displacing of antagonistic foundations; 
  • secondly a fragmentation and separation (e.g. the official  'separation of powers' in liberal democracy [between the judiciary and the polity etc], the splits inside the working class [between skilled and unskilled, for example], the separation of producers from consumers); 
  • thirdly these  'possessive individuals' unite and cohere in the imaginary. 
Hence the appearance of  'imaginary lived relations'. Hall cites ideological totalities like communities, nations, 'the public', the social and so on. 

The State especially does this uniting -- it generalizes and universalises. It helps to secure Capital, but also to extend and innovate as a  'ideal total capitalist', following the long-term interests of the system, rather than the immediate interests of  any section of capitalism [finance capitalists, say, or  industrial capitalists, who are involved in a competititon at the moment, of course]. In this sense, the State is relatively independent of class and of the specific economic level. Indeed, the classes must attempt to control the State and its mechanisms, to get it to represent their interests as a general interest. 

Other institutions in 'civil society’, or ISAs, also do this, such as in the  'structure of political representation' in formal democracy. They also appear as separate agencies and powers, but are also reorganised into a mystical unity of consensus (339). They seem free and emergent, but they are structured by dominant ideology. 

Section 5 -- The mass media

At last we come to the media and their ideological role. Hall has an aside for us before we begin. 

The mass media have other roles, he tells us -- for example they produce cultural objects for exchange and consumption as in the early notions of  'mass culture'. They have helped the transitions from a rural to an industrial urban capitalist society, from laissez-faire to monopoly capitalism. The mass media come into their own in the modern era  [as well as producing this era? -- see page 340]. They become the principal means and channels for the production and distribution of culture, their influence expands into all other channels of public communication [but Hall still wants to warn us against calling this a 'mass society']. In modern societies the mass media have penetrated industrial production, Capital, and the State. However let us focus on their ideological role specifically. 

First the media now lead compared with the traditional culture or channels, and specialise in the production and consumption of social knowledge. They have colonised the cultural and ideological spheres and have a role especially: to construct an image of the lives of other groups and classes; to provide any image, representations and ideas which help us grasp the social totality as a whole. 

Secondly they manage the variety of experiences found in modern societies. They do not offer one great unitary ideological discourse as in the past. The media reflect this variety, they provide an inventory of lexicons, lifestyles and ideologies, and then they map, order, classify and rank these elements [Geertz and Halloran are cited here -- bourgeois sociologists!! Hadn't we clearly departed from them?]. The media accept or disapprove of what they depict, guided by preferred meanings. It is a huge ideological labour to manage this complexity, to map it and make sense of it.  Meanings are preferred and excluded constantly, and negotiated -- a form of struggle. It is a class struggle for Volosinov again, over the terrain of 'sign communities' which are not the same as class communities, which overlap them to some extent, but which are still  'multiaccentual'. This multiaccentuality gives signs their vitality [and holds open the prospect of class struggle in language]. 

Thirdly, the media bring together and orchestrate cultural matters, producing an acknowledged order, and concealing the real structuring effects of class. They manage consensus and consent. Other voices are allowed in the discourses found in the media, but as minorities or acceptable alternatives, and this helps the media appear reasonable and equal. However, beneath the apparent diversity is a dominant structure. 

Fourthly, the media offer particular mechanisms of representation and signification which makes them appear to be semi-independent of the ruling class. They operate in a pluralist framework, not under the direct control of the State. [Hall must be thinking especially of the BBC here -- even that got quite a lot closer to the State in the years immediately following the publication of this article!]. They appear to possess their own set of  'impartial professional-technical' working ideologies, but their configurations are selective and from a limited repertoire. 

Hall offers some examples from [his own earlier work on] television. Events must be made intelligible, real events transformed into a symbolic form, what Hall has called  'encoding' [a reference to the 1974 piece here]. There are fictional codes and codes of actuality and naturalism. There are different ways to encode, especially with difficult and complex events. However, the preferred codes embody the most  'natural' and acceptable explanations, and this places them within the repertoire of dominant ideologies. There is a cluster of these ideologies not a single one. 

There is no deliberate intent to reproduce capitalist social relations. However, dominant ideologies constitute a field of meanings within which coders choose. These choices are then universalised and naturalised to appear as the only possible ones, universally valid. The  'premisses and preconditions' are invisible, ideologically masked  (page 347). The media thus structure and accent events as they signify, but this is an unconscious process, even for the encoders themselves. It looks like they are simply pursuing their professional ideologies -- their news values, their interest in lively coverage, their construction of a good story. This mechanism helps distance the encoders from dominant ideology: as far as the professionals themselves are concerned, they are achieving  'effective communication'.  Really, though, they are skilfully extending the explanatory reach of those ideologies, using other elements in the repertoire. They are reflecting unconscious limits and unconscious connections with dominant ideology, and aiming to win the consent of the audience as subordinates, by affecting their decoding.[ see comment]. 

Of course, audiences do not necessarily decode in this way  (there are references back to the earlier work and to Morley). Audiences reflect their own social and material conditions. However, there is a weight of expertise attempting to make these consonant with dominant ideology, or at least as  'negotiations' within dominant codings -- anything rather than as counter-hegemonic meanings. For example, it is always possible to make an exception of oneself for local reasons while still accepting the wider ideological system. [Perhaps Hall has in mind here the well-known 'levels' within racism? Individuals often exempt themselves or their friends while accepting a general charge against 'the British' On a wider level, no soldier thinks they are going to get killed although they fully accept that in general soldiers are expendable]. The effect is one of    'complementarity', of a provided space for opposition. 

(My) Conclusion 

The media are or identified as ISAs, relatively autonomous ones, of course. Broadcasting must seem separate and tolerant of diversity for the sake of legitimacy: it cannot appear openly as an organ of ruling class power. It must be 'objective', 'neutral', 'impartial' and so on. It must offer 'balance', often as some kind of two-part dialogue, but even this reproduces the dominant discourses. In this way the media 'underwire' the whole system, the whole  'ideological field ' in which different positions contend. [Turning to the ways in which politics is conventionally reported in the UK] the media support a complex unity of political parties, divided by policy, but united in their support for parliamentary democracy. The media construct consensus, but in a field  'structured in dominance', within the limits of what is acceptable, what is included and excluded  (the excluded includes  'extremist, irrational, meaningless, utopian, impractical etc' political alternatives -- page 346). 

So, the media  'ceaselessly perform the critical ideological work of  "classifying out the world" within the discourses of the dominant ideologies' (346). Of course, there are internal contradictions in this work that are also reproduced -- this produces the [predictable] unstable equilibrium. In this sense, reproduction is only a  'systematic tendency' (346). Nevertheless, as a last word, Hall says the media serves 'to reproduce the ideological field in such a way as to produce, also, its structure of dominance'.  (346). 

Well, finally -- what did you make of all this? I think it is a mistake to see this as a piece about the ideological effects of the media as such. Actual conclusions about the media are pretty banal and obvious, although the tweak from Poulantzas about the nation as an imaginary unity etc is nice (and it did lead to some detailed work on 'current affairs' programmes -- well, and ccrime series and televised sport, and lots of other things, which made it all look pretty obvious and repetitive again after a while -- see Harris (1992)). I think most of these pieces could have been written with the briefest of acquaintances with actual TV programmes. 

I think the example about 'the apology' in advertising (in Barthes 1973)  makes the same points, but is much nicer, more amusing, subtler and far briefer (see file). What is even more heretical, I think the analyses in Marcuse's One-Dimensional Man, or some of those in Adorno's work (e g the essay on the 'culture industry') make the same points, often in a much subtler way, even though they use less fashionable theoretical resources to do so, and often after a much more detailed analysis. Hall and the gramscians would be very unhappy at this suggestion, though -- they saw 'critical theory' as outdated and pessimistic.

Really, the main point of this article is to attempt to map theoretically a number of recent academic works, surely, to 'manage' and 'articulate' them. we might say. This is why we have to divert all through the various discussions about ideology and how to read Marx and all that. Hall wants to clarify his own position, perhaps to establish some sort of agreed direction for his Birmingham Centre, perhaps to pursue his own unique 'research programme' which will set him apart from all the other Schools and Centres. This is perfectly acceptable as a goal for a working academic of course, if a bit esoteric -- as long as everyone knows what is going on.