Knowledge, Morality and Hope: The Social Thought of Noam Chomsky
New Left Review, 187, May/June 1991, pp. 5-27
In his first published essay on politics, Noam Chomsky announced his conviction that ‘[i]t is the responsibility of intellectuals to speak the truth and to expose lies.’  Acting on that conviction, Chomsky has long supplemented his work in linguistics with writing on contemporary political affairs, focusing principally on the politics of the Middle East, the immorality of us foreign policy, and the role of American mass media and intellectuals in disguising and rationalizing that policy.  By contrast with his work in linguistics, which is principally theoretical, Chomsky’s political writings in the main address more straightforwardly factual questions. As he emphasizes, these can be settled without special methods or training, and their significance can be appreciated through the application of common-sense norms and beliefs (for example, that aggression is wrong, concentrated power is dangerous, and citizens have greater responsibility for the policies of their own country than for those of other states), as aided by ‘a bit of open-mindedness, normal intelligence, and healthy scepticism.’ 
The characteristic focus, intensity and hopefulness of Chomsky’s political writings, however, reflect a set of more fundamental views about human nature, justice and social order that are not simple matters of fact. This article explores these more fundamental ideas, the central elements in Chomsky’s social thought. We begin (section i) by sketching the relevant features of Chomsky’s conception of human nature. We then examine his libertarian social ideals (section ii), and views on social stability and social evolution (section iii), both of which are animated by this conception of our nature.
To anticipate what follows, we take Chomsky’s social views to be marked by four key claims: (1) human beings have a ‘moral nature’ and a fundamental interest in autonomy; (2) these basic features of our nature support a libertarian socialist social ideal; (3) the interest in autonomy and the moral nature of human beings help to explain certain important features of actual social systems, including for example the use of deception and force to sustain unjust conditions, as well as their historical evolution; and (4) these same features of human nature provide reasons for hope that the terms of social order will improve from a moral point of view. Thus stated, these four claims are clearly neither concrete nor precise. But neither are they vacuous. They provide what we take to be a distinctive, optimistic perspective on human beings and human possibilities. The exposition that follows aims principally at a sympathetic clarification of this perspective. While our discussion is often critical, the criticisms themselves are intended to clarify Chomsky’s views and to underscore deeper points of agreement with them.
Before turning to that discussion, however, a cautionary remark about the character and self-conception of Chomsky’s work in this area is in order. Most important, Chomsky does not have a theory of society or justice, in the sense of a clearly elaborated and defended set of fundamental principles. In fact, he believes that significant progress in ethical and social inquiry requires a systematic theory of human nature, something that does not now (and may never) exist,  and that in the absence of such a theory social and ethical thought must rely on relatively speculative and imprecise ideas (‘guesses, hopes, expectations’  ). Moreover, Chomsky denies any originality for his social and ethical views, identifying himself as a merely ‘derivative fellow traveller’  in the anarchist and libertarian socialist traditions.
Finally, and no doubt in part owing to his conviction that his social and ethical views are neither systematically developed nor original, Chomsky presents those views in an occasional and sketchy fashion. Almost always announced as speculative, and often advanced only in response to promptings from interviewers, their presentation commonly takes the form of quotation from and endorsement of certain views of other thinkers (for example, Rousseau, Kant, Humboldt and Marx).  Apart from creating natural difficulties for any attempt at systematic summary, the character of Chomsky’s presentation underscores the need for caution in reading more into, or expecting more of, his work in this area than he invites. We hope that we have heeded our own warning in what follows.
I What Can We Know? Rationalism Romanticized
As already noted, Chomsky believes that a substantive conception of human nature must play a central role in both the ethical assessment of social arrangements and in the explanation of their operation. By a ‘conception of human nature’ he means an account of the biological endowment of the human species, and in particular the aspects of that endowment that figure in the development of human cognitive systems—aspects that are common to all human beings (excepting those suffering from pathologies) and perhaps unique to the human species. At its core, Chomsky’s own conception of human nature draws together a romantic emphasis on the distinctive human capacity for creative expression and a rationalist contention that there is an intrinsic and determinate structure to the human mind.  In his work, these romantic and rationalist strands are joined through the contention that the intrinsic structure of mind provides a framework of principles that underwrites the possibility of the relevant forms of creative activity, while at the same time limiting the attainable forms of human expression. 
This conception of human nature is most fully developed in Chomsky’s linguistic theory, which emphasizes both the creativity exhibited by normal human language use and the modularity of the human language faculty.  According to Chomsky, the ‘fundamental fact about the normal use of language’  is its ‘creative aspect’. Human beings have the capacity for a potentially unbounded novelty in the production of utterances that are appropriate to their circumstances, but that are not controlled by immediate stimuli (though they are commonly prompted by such stimuli). The linguistic knowledge expressed in such creativity is acquired by virtually all human beings in a relatively short period of time, and in the face of unstructured and impoverished inputs from the environment. Given this ‘poverty of the stimulus’, Chomsky argues that language acquisition can plausibly be explained only on the assumption that human nature includes a language ‘module’—an innate system of language-specific principles, or ‘universal grammar’.
Chomsky’s more general remarks about human nature are fuelled by the speculation that these ideas about creativity and modularity in the domain of language might have more general application, and that the theory of language might suggest a paradigm for a more general account of our nature. Thus he notes that ‘at some very deep and abstract level some sort of common-core conception of human nature and the human drive for freedom and the right to be free of external coercion and control’ of the kind that figures ‘in a relatively clear and precise way in my work on language and thought’ also ‘animates my social and political concerns’. 
Language and Human Nature
To understand how the ideas featured in the linguistic theory might be extended to provide a broader account of human nature, consider first the modularity hypothesis. In areas other than language—for instance, vision, scientific reasoning, aesthetic and moral judgment—human beings might be thought to attain complex and determinate cognitive systems in the face of relatively impoverished data, thus suggesting the presence of modules governing cognitive development in these other domains as well. For example, human beings appear to have a moral nature. That is, there appears to be a natural human tendency to interpret human interaction in moral terms—to display a concern about the justifiability of actions in light of their effects on the well-being of others. That concern is manifest in, among other things, complex systems of moral judgment that are deployed in response both to actual problems and to the hypothetical cases of moral philosophy; furthermore, these systems appear to go well beyond anything that might plausibly be found in the ‘impoverished and indeterminate’ data available to a child who receives moral instruction.  The hypothesis of moral modularity would explain the acquisition of a system of moral understanding in part in terms of a set of intrinsic features of mind that are specific to morality. A characterization of the moral module would of course need to be consistent with the variety of moral systems, but it would also impose limits on possible human moralities. Thus, features of it might help to explain certain formal properties that are allegedly displayed in systems of moral understanding (for example, why moral conceptions are either deontological or teleological, or why moral norms feature elements of symmetry, generality and impartiality) and/or certain substantive features that seem to be common in moral systems (for example, prohibitions against killing innocents, want only imposing pain, or enslaving members of the community). 
Consider next a natural generalization of the creative aspect of language use. Here the thought would be that, by underwriting the acquisition of a variety of complex cognitive systems, our nature enables us to engage in creative activities that deploy those systems. Thus creativity would be a feature not just of language use but of moral judgment, common-sense understanding, and the use of that understanding in productive work, artistic activity and scientific achievement. Of course, a human potential for creativity might be joined with a desire for habit and repetition and a hatred or dread of novelty. But Chomsky’s contention is that, associated with the intrinsic possibilities for human creativity and underscoring their relevance for ethics and social explanation, there exists an innate propensity to pursue the forms of creative expression for which our nature suits us. Borrowing a phrase from Bakunin, Chomsky sometimes refers to this as an ‘instinct for freedom’.  Elsewhere he suggests that we have not merely an instinct towards but also a need for such free activity, and that the failure to accommodate that need results in individual and social pathologies.  Taking the remarks about instinct and need together, Chomsky appears to endorse the view, associated with Aristotle, that human beings enjoy the exercise of their natural powers, with enjoyment perhaps scaled in part to the complexity of the activity in which those powers are engaged, and that we enjoy as well the exercise by others of their powers, at least when we have the opportunity to engage in like activities ourselves. 
Certain features of this view strike us as significant and plausible. Much historical sociology testifies to the claim that there is a fundamental human interest in free activity (alongside other interests in some level of material comfort and in being treated with respect by others), and that human beings have a moral nature in the sense just described.  It is, for example, not implausible to explain the destruction of slavery, the development of religious pluralism, and the evolution of democratic ideals and their (admittedly highly imperfect) expression in social and political arrangements, in part by recourse to these basic features of human beings. It is less clear, however, that these matters are illuminated by considerations of modularity and creativity drawn from the theory of language. It may be the case, for example, that human beings have a moral nature but lack a specifically moral module. What underlies our acquisition of moral systems might instead be, as Kant supposed, intrinsic features of human reason as such. On this account, the construction of moral systems would not involve the deployment of principles specific to a moral faculty, but rather the application of a ‘pure reason’ to matters of human action.  This might be sufficient to account for moral competence—for the complexity and determinateness of human moralities, and the capacity to respond to novel cases in ways that transcend specific instruction—even if the parallel contention about the acquisition of linguistic knowledge is not at all plausible. We do not wish to defend this alternative perspective, but only to stress that the important contention that human beings have a moral nature does not depend for its plausibility on the assumption of a moral module.
Similarly, reflection on the creative aspect of language use does not seem to help in understanding human aspirations to self-directed activity—the fact that ‘people really want to control their own affairs. . .[and] don’t want to be pushed around, ordered, oppressed, etc.’  What is central to linguistic creativity is the capacity for unbounded novelty. But what matters to people in their aspirations to selfdirection—to form conceptions of a decent life and to act on those conceptions—appears not to depend on the prospect of novelty of expression. For example, even if one were persuaded that one’s artistic, craft or intellectual activities would issue in nothing novel, one’s interest in pursuing that activity free of external control, without being ‘pushed around’, would be likely to remain undiminished.
In sum, we agree with Chomsky’s claims about our moral nature and ‘instinct for freedom’, but are less persuaded by the suggestions about how these features of human nature might be understood on the model of the theory of language. Reflecting this conclusion, our references in what follows to Chomsky’s ‘conception of human nature’ will be confined to the fundamental points of agreement.
II What Ought We to Do?
Chomsky’s normative views, and in particular his account of a good society, are set against the background of his conception of human nature, and the conception of an instinct for freedom that lies at its heart. In outline, Chomsky takes freedom to be the supreme human good, and endorses the libertarian principle that the evaluation of social arrangements should proceed by considering whether those arrangements impose tighter limits on human activity than are necessary, given existing material and cultural constraints.  This, of course, is not an especially determinate view. Embracing the supreme value of freedom, and endorsing the principle that unnecessary constraints on it should be eliminated, does not settle issues about the appropriate treatment of time trade-offs (for example, the relative importance of freedom for present and future generations) or distributional issues (in particular, whether it is legitimate to trade the freedom of some for the freedom of others). Still, the conception is not empty. According to Chomsky, indeed, it supports a particular ideal for human beings operating under the material and cultural constraints of modern industrial societies—namely, socialist anarchism. The best known of his ethical-political views, Chomsky’s anarchism is worth exploring at length.
It may be useful here to distinguish two conceptions of anarchism, each of which figures in Chomsky’s work. In the first, anarchism is represented not as a substantive ‘doctrine’ but as a ‘historical tendency’, a ‘permanent strand of human history’.  Reflecting the human aspiration to freedom, this tendency underlies the emphasis on freedom as the supreme value in assessing social forms, and encourages scepticism about familiar contentions of the ‘necessity’ of social arrangements constraining freedom. To endorse anarchism in this sense is straightforward enough; it amounts to endorsing a critical standpoint in social thought and action, rooted in a concern to eliminate unnecessary constraints on human freedom.
The second sense in which Chomsky (more tentatively) invokes anarchism is less straightforward. Here anarchism does appear as a substantive, libertarian socialist, ideal of social order. Chomsky’s anarchism is socialist in that it endorses the social ownership of the means of production, a form of ownership permitting the extension of democratic procedures to economic decisions both in individual workplaces and across the economy as a whole. What that means in detail is not explored at any length in his writings, but the essential principle is clear enough. The content of the more specifically anarchist aspect of his views, however, is less clear.
The basic idea of anarchism is that social cooperation can and should proceed without a state. But the content of that idea depends on the underlying conception of the state and its offending aspects—what is to be eliminated—and this has no single interpretation within the anarchist tradition. To locate Chomsky’s views, it may help to begin by distinguishing three aspects of states that an anarchist might seek to eliminate—that states exercise coercive power at all, that they claim a legitimate monopoly on the exercise of coercive power, and that they specialize in the exercise of coercive power.  Corresponding to these three aspects of states are three distinct visions of social order, each of which has a claim to capture the basic anarchist ideal of state-free cooperation.
Conceptions of the State and Social Order
The first conception is a coercion-free system. In such an order there is no state, in that socially organized coercion does not exist at all and is in fact unnecessary, since the members willingly comply with rules and standards that are publicly announced.  Chomsky’s endorsement of anarchism seems not to depend on the thesis that a coercion-free system is possible. For example, he often cites the 1936 Spanish anarchist experiments as exemplary of anarchist practice, but they were hardly free of coercion.  Furthermore, while he assumes that an anarchist order will reflect and further a ‘spiritual transformation’ of human beings, registered in much greater self-confidence and voluntary cooperation among them,  he seems to agree that it would be unreasonable to expect complete civic consciousness and a fully harmonious coordination of interests in a social order that operates on the scale of a modern society.  Thus he notes the likelihood that (in part because of popular empowerment) ‘factions, conflicts, differences of interest and ideas and opinion’ will be expressed throughout a libertarian socialist society.  Especially given his recognition of such a high degree of disagreement and conflict, it seems safe to assume that Chomsky recognizes as well the need for the continued existence of some agencies with powers of enforcement—if only to assure those who were willing to comply that others would not take advantage of their compliance.
A second possibility, then, is a dispersed-coercion system. Recognizing the need for powers of enforcement, but concerned about the dangers of a concentration of power,  the anarchist might identify the ideal of statelessness with a condition in which no single institution successfully claims a legitimate monopoly on the use of force. Instead, enforcement powers would be dispersed across a set of institutions, each of which would nevertheless be specifically political—that is, defined chiefly by its administrative and enforcement powers. One example of a dispersed-coercion system might be an order governed by a variety of administrative agencies, each with its own enforcement capability. Another would be a territory featuring a variety of specifically protective associations, each providing security for certain people in the territory, but none with a monopoly on powers of enforcement.  But this view also fails to capture the core of Chomsky’s anarchism. To see why, let us contrast it with a third interpretation of the ideal of state-free cooperation.
This third possibility is a system with dispersed coercive powers, but in which those powers, and all other traditionally political powers of collective decision-making and administration, are dispersed over institutions that do not specialize in the performance of political functions. Rather, on this conception, which we take to be Chomsky’s own, political powers are exercised by institutions that also, and perhaps chiefly, perform other (for example, productive, associative) functions. Here, the ‘statelessness’ of society is achieved neither by the abolition of coercion (the coercion-free system), nor by the multiplication of its authoritative dispensers (the dispersed-coercion system), but by the transcendence of the traditional division of labour in governance between specialized political institutions that rule, and the rest of a society subject to their rule.
Several elements of Chomsky’s view underscore the importance of the elimination of specialized political institutions, the distinguishing feature of this third interpretation of anarchism. Thus, he supports political representation, but imagines representation to be based on such ‘organic groups’ as workplace or community associations, and to feature a ‘rather minimal’ delegation of powers.  He sees the need for administration, but thinks that it ought to be a rotating ‘part-time job’ performed only ‘by people who at all times continue to be participants in their own direct activity’ (that is, who act in other capacities as well).  And while he does not think that political parties could legitimately be banned in an anarchist order, he does think that if parties, or any other specialized organizations with an exclusive devotion to political affairs, were ‘felt to be necessary’, then ‘the anarchist organization of society will have failed’. 
Within this conception, for example, workers’ councils and their representatives might have responsibility for economic planning, and the elaboration and enforcement of rules governing workplace relations (such as rules on occupational safety and health). Tasks traditionally assigned to local governments (the maintenance of local order, sanitation, primary education and the like) would devolve to neighbourhood associations composed of citizens active in other affairs. Regional and national tasks of government would be handled through recallable representatives of such groups. National defence would be assured through citizen militias. And so on. What is essential is that all traditional functions of government would be discharged by groups whose members also engaged in non-governing activities.
How plausible is this conception? Anarchist views are typically criticized for resting on implausible accounts of human motivation, for being inattentive to the ways that decentralization can exacerbate political and material inequalities, and for ignoring the attractions and requirements of economic efficiency. Chomsky’s conception avoids at least the most obvious versions of these objections. As noted earlier, for example, he appears to reject accounts of human motivation that would deny the persistence of conflict and disagreement under anarchy. He endorses the idea of a framework of basic rights, and reasonably encompassing arrangements for making collective decisions and administering those decisions, and in these ways addresses some concerns about the pathologies of decentralization. And, responding to concerns about economic efficiency, he stresses that his view does not imply primitive communitarian economic production (the denial of all economies of scale and specialization), and indeed welcomes technological development and increased productivity, which contributes to free social cooperation by limiting the need for human toil. In fact, he thinks that anarchism really comes into its own as a social ideal only with a high level of development of the productive forces of society.
The Weaknesses of Chomsky’s Anarchism
We, however, are less convinced than Chomsky of the attractions of anarchism, for three reasons. First, we are simply unpersuaded that policies of, for example, economic coordination, environmental protection and public health are most efficiently made in the absence of specialized bodies, devoted to formulating policy alternatives and to assessing the likely consequences of their implementation. Assuming conditions of large-scale social interdependence, there are likely to be problematic third-party effects or ‘externalities’ in each of these areas (for example, a significant degree of pollution, affecting large populations, resulting from the actions of individuals and enterprises).
Such externalities, and the complex problems of policy planning and evaluation they pose, are not solely attributable to ‘the irrational nature of [present] institutions’  and the incentives to selfish behaviour that they generate. In some important measure, they are intrinsic to social interdependence itself. If only because of problems of imperfect information, they can be expected to appear in an interdependent and technologically advanced society even under the most favourable of motivational conditions (for example, perfect altruism). Given the persistence of externalities and their attendant complexities, however, we doubt that any reordering of social institutions, however welcome, will simplify public policies to the point that reasonably efficient governance can become simply a ‘part-time job’. 
Second, we would expect tensions between the proposed dispersion of political responsibility and the effective exercise of that responsibility. Underlying this concern is our assumption that actions needed to enforce the terms of order would be costly; that is, both detecting violations and sanctioning violators would generate costs for enforcers. For reasons familiar from the theory of public goods, the coexistence of dispersed benefits (those accruing to citizens in general from enforcement of the terms of the order) and concentrated costs (incurred by those who engage in enforcement) presents a situation ripe for ‘free-riding’, and a concomitant failure to provide appropriate levels of enforcement.
Such problems are, it should be said, much less pervasive in smaller-scale associations. Reduction of scale almost certainly makes violations more apparent, thus reducing the cost of their detection. And since association norms can be invoked in a process of informal sanctioning, this is also likely to reduce the cost of imposing sanctions, including sanction of those who decline to sanction others. But we are assuming a fairly large and complex association. Here, we believe, the problem of providing incentives to enforce the terms of the order has real bite. And here, a natural solution to the enforcement incentive problem is to establish specialized agencies for administration and enforcement. The specialization of agencies eases the task of providing incentives to enforcement, since those who work in them can be held accountable for their failures (for instance, by being removed from their positions). And it greatly simplifies the problem of monitoring enforcement performance, since the division of political labour enables citizens and their representatives to focus their inquiry on particular institutions and individuals.
Third, we doubt whether the proposed anarchist order would encourage the motivations necessary to its stability, in particular whether it would encourage the formation of a sense of justice comprehensive enough to include all members of the order. We noted earlier that Chomsky supposes that an anarchist order will encourage a ‘spiritual transformation’ of human beings. But it remains unclear how the sort of anarchism he endorses would elicit motivations of the required kind. In particular, we assume that a stable democratic socialist scheme must ensure continuing relative equality among the participants in the order. It might be the case, for example, that considerable portions of the surplus generated in some enterprises would be used to benefit those in others. But how is a concern to preserve such equality of condition to be encouraged among citizens, assuming (what seems obvious) that it could not be counted on to emerge spontaneously?
Presumably the idea is that citizens acquire a sense of justice and a willingness to act on that sense through the normal course of their maturation. But if organic groups, based in particular workplaces or neighbourhoods, provide the basis of political cooperation, then it would seem natural that the principal allegiance of citizens would be to those organic groups. And if that were true, then a sense of justice supporting the distributive measures required to maintain the order would not be likely to form. One of the virtues of less organic and more ‘alienated’ political forms that are abstracted from everyday life—political parties, territorially defined representative bodies, and specialized organizations for making and enforcing collective decisions—is that they plausibly encourage the members of society to regard one another as equal citizens, deserving of justice whatever the particulars of their aspirations, class situation, or group affiliations. It seems likely that some such more cosmopolitan sense of citizenship, encouraged by less organic forms of political association, is necessary to provide the motivation needed to sustain the egalitarian background required of a genuinely democratic society.
Of course it may be that the establishment of specialized political bodies to address these problems would engender a concentration of power, and that such a concentration would produce greater threats to human freedom than those resulting from the absence of specialized political arrangements. This is, clearly, an empirical issue, and one about which we can only hope someday to have data. In the absence of more compelling evidence than we now have, however, Chomsky’s anarchism seems to stand on relatively shaky grounds. This limits its appeal even to those who share his commitment to eliminating unnecessary constraints on human beings, and in particular constraints deriving from economic inequality.
III What, If Anything, May We Hope For?
Setting these particular criticisms aside, Chomsky’s account of human nature and the social conditions appropriate to its full expression naturally suggest two questions about contemporary societies. First, why are present social arrangements and the distribution of basic material resources and political power they provide so distant from the arrangements and distribution appropriate to human nature?  Second, given the distance between the actual and ideal, what reason is there for maintaining even the hope that current arrangements will come to approximate this ideal more closely? In this final section, we consider these questions in turn.
At least at an abstract level, Chomsky’s answer to the first question is clear. Once an unjust order exists, those benefiting from it have both an interest in maintaining it and, by virtue of their social advantages, the power to do so. Maintenance is of course not in the interest of those who do not benefit from the order. So, dominant groups must use their power against subordinate ones to ensure the latter’s consent or acquiescence to the unjust scheme. The basic mechanisms for achieving this are force and fraud. Either those who do not benefit must, in effect, be frightened and beaten into submission, or they must be distracted from their real interests and deliberately confused about the way the world works.
More particularly, Chomsky believes that the relative importance of force and fraud to social reproduction depends on the specific scheme of unjust distribution in place. Fascist orders and Stalinist ‘socialism’ are marked by the denial of liberties of expression, association and participation. Thus, while they feature well-developed propaganda systems, they rely primarily on force to suppress the natural aspiration to freedom. Capitalist democracies, by contrast, provide at least formal rights of expression, association and participation. While they feature considerable use of force, the availability of channels of expression and association increases the importance of what people think, and fraud accordingly plays a more central role in preserving order.  In capitalist democracy, the real ‘enemy’ of governing elites, their ‘ultimate target’, is the human mind itself.  The preservation of unjust advantage requires thought control, the deliberate ‘manufacture of consent’. 
In elaborating these themes, Chomsky has focused principally on the role of the mass media in capitalist democracies, and then almost exclusively on the case of the media in the United States—a case marked, in comparative terms, by an apparently paradoxical combination of extreme media servility and minimal state control of those media.  These writings, some authored jointly with Edward Herman, present a ‘propaganda model’ of the media’s operation—not, it may be stressed, of its effects on consciousness or behaviour (see discussion below, p. 20)—with two component parts. 
The Propaganda Model
The first component is the contention that the principal propagators of ideas in capitalist democracies (again, and throughout, particularly in the us) advance ideas that conform to elite interests. More specifically, the range of positions featured in the media, the issues that receive emphasis, the timing of stories, the sources that are treated as respectable, and the interpretation of the role of the media itself, are all ‘highly functional for established power and responsive to the needs of the government and major power groups’ [emphasis added]. 
The second component of the model is an account of the mechanisms of social power, and the particular organization of the media, that explain this ‘highly functional’ pattern. Here, in an account of why the media operates as it does, general propositions about the organization of capitalist democracies are fused with specific contentions about the role of four broad groups of social actors within such systems.
At the top are various interconnected elites, predominantly business elites (including those who own the media) and government elites (often the same as the former).  These actors, or at least important segments of them, appear to be relatively ‘free of illusion’, a freedom explained by the fact that a clear understanding of the world is essential to maintaining their privileged positions within it. (‘The propaganda may be what it is, but dominant elites must have a clearer understanding among themselves.’  ) Thus, capitalists need to know how the world works in order to compete successfully, and state managers need to know in order to serve the interests of business. Understanding their own interests and what is needed to advance them, and recognizing that their interests are commonly opposed to the interests of the rest of the population, the elites sow illusion outside their ranks.  In this connection, it should be noted that Chomsky thinks that ‘[m]ost people are not liars’,  and that a low tolerance for cognitive dissonance leads most propagators of falsehood to selfdeception; they tend to say what they believe, having first come to believe what they say. But Chomsky seems to except at least certain crucial elements of elites from this generalization, since he holds that they both know the truth and regularly deceive others about it.  Given his assumptions about normal people, such behaviour, which on his view was, for example, characteristic of the ‘stunning lunatics and liars’ who prosecuted the us effort in the Vietnam War, must be termed ‘pathological.’ 
A second group is composed of journalists and the ‘secular priesthood’ of (primarily) academic experts.  Members of this class typically attain their positions by propagating views that serve elite interests. But, displaying the normal human intolerance of conscious deception, they commonly come to believe what they say. 
A third group of actors is composed of the educated and politically active middle classes. They provide the ‘primary targets’  for propaganda, since elites recognize that they could and would do enormous harm to existing arrangements of authority if told the truth. That they could do such harm is a function of their resources and political activism. That they would is a function of the fact that, as Chomsky puts it, ‘most people are not gangsters’.  He believes, and believes clear-minded elites believe, that the middle class would not support immoral us policies if it knew the truth about them, which is why elites seek ‘to prevent any knowledge or understanding’ of those policies. 
A fourth group is composed of the politically immobilized lower classes. Members of this group benefit least from the operation of the system, and thus might be thought to pose the greatest threat to its continuance. But their lack of resources and difficulties with collective action (‘They are not part of the system; they just watch.’  ) in fact make them less threatening than the middle classes; they are as a consequence only a secondary target of the propaganda machine. Moreover, while the fact that members of this group are less highly ‘educated’, and thus less indoctrinated,  than other elements in the population, they are effectively discouraged from political activity by such distractions as spectator sports, lifestyle preoccupations, and ‘religious fanaticism of an almost Khomeinist variety’.  Often profoundly alienated from the operation of the system, and thus susceptible to the attractions of ‘charismatic figures who promise to lead them out of their problems and to attack either the powerful or some other bogeyman, the Jews or the homosexuals, or the communists, or whoever is identified as responsible for their troubles,’  they typically ‘can be satisfied, it is hoped, with diversions and a regular dose of patriotic propaganda, and fulminations against assorted enemies.’ 
As these last observations may suggest, Chomsky does not study the secondary target or popular culture in any detail. His work concentrates instead on the ‘dominant intellectual culture and the values that guide it’  —the interactions among the first three groups just noted. To summarize, his claim is that the functional pattern of media propagation of ideas conforming to elite interests (the first component of the propaganda model) reflects a relatively conscious policy of deception pursued by elites, who act with the willing if unwitting support of intellectuals and journalists, and who are concerned to forestall the emergence of opposition to their power that would likely arise from a (middle class) population that knew the truth about their immoral aims and actions. Before assessing the force of this conception, two clarifications of its aims may be helpful.
First, and as emphasized above, the propaganda model is offered as an account of the operation of the major media, not of its effects. The model in itself neither states nor implies that ideological conformity in the target group(s) is actually produced by propaganda, or that ideological conformity is the principal cause of obedient behaviour, or even that ideological conformity exists. So the propaganda model would not be falsified if it were established that propaganda was unsuccessful in generating false beliefs, or that it was irrelevant to the production of consent, or that consent was produced by means other than illusion (such as self-interest or cynicism).  Nevertheless, the propaganda model is advanced in aid of understanding the actual manufacture of consent in capitalist democracies, and not simply as an account of the functioning of major media in them. Its interest and importance thus does depend on there being some real and significant effects of propaganda on social action,  and Chomsky in fact holds the view that propaganda efforts are successful in generating both illusion and consent. 
Second, it is no objection to the propaganda model to observe that the distortions featured in the major media are not complete fairy tales, utterly at odds with the facts. On the contrary, attention to the model’s underlying mechanisms would suggest that distortions will be more often a matter of framing and emphasis than simple fabrication.  As already noted, for example, governing elites need to understand the world. But since they in part rely on the major media for information about the world, this imposes some truth constraints on reporting in those media (constraints more severe than those exerted on more ‘yellow’ or ‘tabloid’ journalism). 
An Overstated Case?
Even with these qualifications noted, however, our assessment of the propaganda model is mixed. Chomsky presents reams of evidence for the model, most of it addressed to the first of its two components—that media representations are highly functional for elite interests. With copious documentation, he effectively makes the case that the bulk of information provided by the major media is extremely and systematically biased toward the maintenance of existing arrangements of power and advantage; that departures from orthodoxy, particularly among those who threaten to reach a more than miniscule audience, are deliberately sanctioned; and, above all, that debate about us foreign policy commonly proceeds within a set of presuppositions about the role of the United States in the world that are quite distorted but rarely even noted, much less disputed. With all this we are in agreement.
The recent Gulf war, for example, provided an advanced display of all these phenomena. Information supplied to the press was sharply restricted by American diplomatic and military personnel, while the restrictions themselves were barely noted or challenged. Accordingly, press ‘coverage’ of events, particularly in the us, consisted largely of canned human-interest stories, military briefings, and pool reports from sites selected by military authorities. Information unfavourable to the official us position—on, for example, the details of Iraqi peace initiatives before the war, or the destruction of Iraq during the war—was generally dismissed or not reported at all. Even the slightest departures from orthodoxy—for example, the fact that the Cable News Network actually offered reports from inside Iraq on war damage—were objects of high-ranking political attack. And the most preposterous and cynical statements of public officials—for example, George Bush’s repeated declaration that ‘America stands where it always has, against aggression, against those who would use force to replace the rule of law’—were repeated and amplified without critical comment. 
All in all, then, this was an almost laboratory-perfect demonstration of manipulation of and by the media.
Nonetheless, Chomsky’s view of the media and the manufacture of consent seems overstated in three ways. First, the claim that business people and state managers are in the main relatively ‘free of illusion’ seems overdrawn, at least when that claim is offered (as Chomsky usually offers it) without substantial qualification. There is of course ample reason to believe that business and state elites are on the whole better informed about their interests than ordinary citizens, since they have more resources to acquire information, and, as a rule, greater incentives to ensure its accuracy. But they are not immune to ‘ordinary’ failures of human understanding (such as shortsightedness, excessive attention to the status quo), evidence of which is legion in the ranks of business and the state. Nor, critically, are they immune to the distortions of ideology, which not uncommonly grips elites with at least the force that it grips other citizens.
On the latter point, Chomsky himself provides evidence for a more complex picture in his essay on the ‘Backroom Boys’ who administered the us war on Vietnam.  The history he discusses there suggests that during and immediately after World War II, us policymakers did operate with a relatively clear understanding of their interests, the requisite elements of a world order that would conform to those interests, and the content of the ideology needed to provide popular support for their imperial designs. Once the terms of the postwar world were set in place, however, political and economic elites themselves accepted the terms of postwar ideology, and clung to that ideology even when it no longer served their material interests. In particular, we find that by the 1960s, ‘other and more irrational considerations (than the economic interests of us capital) may have come to predominate’ in the prosecution of us policy in Southeast Asia,  as United States policy-makers were ‘caught up’ and ‘trapped’ by the ‘fantasies’ they had earlier devised as illusions for the public.  The lesson to be drawn from this example is, we think, straightforward. During some periods—for example, the ascendant period of a new world power—elites may be relatively free of illusion about their interests. During others, however, they may be much less so. But at no time does it seem warranted to assume that elite understanding, however clear about the short term, extends much beyond that.
Second, and closely related, the model’s claim that elite-generated ideologies are always ‘highly functional’ for elite interests seems exaggerated. The ‘Backroom Boys’ example just given indicates otherwise; there, elites were not only trapped by illusions dysfunctional for their interests, but were as a consequence propagating such illusions. How often and how significantly dysfunctionality is produced by such illusion (or by other mechanisms) is, of course, an empirical issue beyond the scope of this discussion. It seems plausible, however, to think that such investigation would yield other examples of elite propagation of ideologies that are not ‘highly functional’. In the area of foreign policy, for example, the conviction that world ‘order’ must be supplied chiefly by the us—a notion that postwar elites did their best to encourage among the general population—now arguably inhibits rational elite response to a range of military and economic concerns occasioned by political transformation in Eastern Europe, turmoil in the Soviet Union, and abiding economic challenge from Western Europe and Japan.
Consider, for example, the Gulf war’s effective definition of ‘collective security’ in the New World Order as the raising of foreign subsidies to pay for American military force. This might seem a perfectly rational us elite response to problems of military management after the decline of American hegemony. Crudely put, if a nation is less competitive in machine tools than in bombers, it makes sense for it to ‘trade’ in bombers. Collectively underwritten but American-led, such an arrangement might help stabilize the us comparative advantage in military power by defraying its costs, while offering continued us influence in the world disproportionate to its economic power.
It might. Or it might not. The underwriting project could get bogged down in ways that military threats themselves could not overcome. Celebration of the comparative advantages of American military power might lead to the neglect of other sources of power. First World needs for the submission of the Third World (against which force is most readily applied) might simply diminish, leading to a decline in the demand for American Hessians. And even if none of these things happened, the choice of the bomb trade may simply not be the best, or even close to the second-best, strategy for maintaining us dominance. It might after all be the case that more constructive uses of us resources—to educate its children or train its workforce, clean its environment or nationalize its health-care system, rebuild its material infrastructure or invest more in basic research—would benefit American elites substantially more than adoption of the mercenary route. At the very least, this question seems open. 
Third, we think Chomsky exaggerates the importance of the media’s informational bias in explaining consent and stability in capitalist democracies. In entering this objection, a word of caution is in order. As noted earlier, commitment to the propaganda model itself does not imply commitment to the view that the media’s informational bias is the fundamental source of ideological conformity, nor that ideological conformity is the fundamental source of consent (at least for the target groups). Nevertheless, Chomsky’s writings do suggest, at least as a general matter, commitment to both of these claims.  So it seems worthwhile to register our objection to the claim about the importance of ideology in producing consent.  Since, however, this is not a claim to which Chomsky is committed by the propaganda model, and is not a claim he makes explicitly in his work, we use the name ‘Chomsky*’ in the following paragraph to underscore our uncertainty about attributing these views to him.
Ideology and Consent
The source of our objection to Chomsky*’s emphasis on the importance of ideology in producing consent is simple. Chomsky uses the term ‘ideology’ in a pejorative sense; the term denotes a system of false beliefs about the world, popular action in accord with which is favourable to the realization of dominant interests.  In our view, however, such false beliefs (‘false consciousness’) play a less central role in explaining consent than Chomsky* suggests. Even individuals who know the ugly truth may consent for reasons of, for example, material self-interest, cynicism, fatigue, or simple lack of concern, and much evidence suggests that many do consent for some combination of these reasons. Survey data in the us (the country that has been the primary focus for Chomsky’s political writings) regularly confirm a very widespread (exceeding the bounds of the ‘secondary target’) public conviction that public officials are corrupt, that the country is run in an undemocratic fashion, and that many public policies are immoral.  But this confirmation is provided in a context of profound political stability. This suggests that something other than illusion and ignorance are producing that stability.
An expansion on this suggestion may be used to conclude our discussion of the propaganda model. Consider the case of capitalist democracy, the social system in which, on Chomsky’s view, propaganda plays the greatest role in producing stability. There, instead of explaining the generation of consent chiefly by reference to ideological mechanisms, one might rely on two, arguably more central, features of that system. First, the private control of investment featured in a capitalist democracy subordinates the interests of workers to those of capitalists (without profits there is no investment, and without investment no jobs), and thus leads workers to restrain their demands on employers and the state. It also tends to focus those demands on material gain, a concern that can be in some measure satisfied within the system. Second, the characteristic inequalities of resources between capitalists and workers systematically favour the former as collective actors, providing further obstacles to organized opposition and further bases for worker consent to the system. 
Assuming that such an account seems generally plausible (an assumption that we do not propose to defend here), it has an additional attraction in this discussion, namely that one can accept it as consistent with endorsing our first two criticisms of the propaganda model and embracing that model’s most central claims about media bias and the narrowness of debate. That is, once explanatory emphasis has been shifted onto non-ideological sources of consent, there is no difficulty in granting that elites are often confused, and that ideologies are not always highly functional for elite interests. At the same time, the view does not exclude ideology as a possible source of consent, let alone exclude the media as a source of its propagation. Indeed, by softening some of the claims made for the importance of propaganda, while admitting the central insights of the propaganda model, such an account of consent seems to us a natural way to highlight the force of Chomsky’s work in this area.
Expanding the Domain of Freedom
We come, finally, to Chomsky’s views on social evolution, and the second question posed at the outset of this section. Why should one think that existing societies, structured in ways so hostile to the exercise of human freedom, might change in a direction more in keeping with that essential human capacity? Our discussion of Chomsky’s view of social order (even admitting the amendments to that view we have suggested) underscores the force of this question, for it may suggest that he sees dominant groups as in some sense invincible. In light of that discussion, it is easy to see why he believes that the ‘struggle for freedom and social justice’ is ‘unending, often grim’,  and why political actors face ‘temptations of disillusionment. . .many failures and only limited successes’.  But it may be more difficult to understand why he also thinks the struggle against oppression is ‘never hopeless’,  or to understand the grounds for his ‘hope that our world can be transformed to “a world in which the creative spirit is alive, in which life is an adventure full of hope and joy, based rather upon the impulse to construct than upon the desire to retain what we possess or to seize what is possessed by others.”’ 
In Chomsky’s view, the source of hope lies in human nature itself. He speculates that constraints on human freedom that are not ‘required for survival in the particular state of history’ will tend to be sloughed off, as a result of the moral nature of human beings, the ‘instinct for freedom’, and the ‘continual efforts to overcome authoritarian structures and to expand the domain of freedom’ that result from that instinct.  Put otherwise, systems that impose unnecessary constraints on natural tendencies to human expression will, by that very fact, face intrinsic sources of instability. The significance of these destabilizing pressures for reform will of course depend on a range of factors relevant to political mobilization—including the number of people who feel the constraints of existing order, the capacities for and willingness to pay the costs of collective action on the part of those who do, the willingness of dominant groups to repress or murder their own populations, the power of foreign states to thwart mass action, and the strategic choices of opponents of oppressive regimes. Still, such pressures are present, and their presence suggests a weak evolutionary tendency toward societies more accommodating of human freedom.
Illustrating these claims about basic human nature and the pressures it exerts on unjust arrangements, Chomsky suggests, for example, that the ‘propaganda system’ in the United States—its tremendous power and durability notwithstanding—is ‘extremely unstable because of the reliance on lies. Any system that’s based on lying and deceit is inherently unstable.’  The thought that underpins this view is that lying about us policies (or disguising them in other ways, for example through covert action) is necessitated by the decency of the population, who if confronted with the truth would resist those policies. And elsewhere, in arguing (correctly, we believe) against the familiar ‘expert’ view that the American public has recently shifted profoundly to the right on social and foreign-policy questions, he notes the continuing dissent of the population from many of the more brutal aspects of recent us policy, ‘despite all the brainwashing and indoctrination and so on’.  Here again, such resilient decency, and the threat it poses to indecent institutions and policies, derive from our moral nature and our fundamental aspiration to freedom.
More broadly, Chomsky rejects the currently fashionable neoNietzschean view that the history of the world is merely a history of change, in which old forms of domination are simply replaced by new ones, without significant progress in meeting such fundamental human interests as the interest in freedom.  A ‘child of the Enlightenment’,  he finds instead that human history, at least at some moments, exhibits ‘detectable progress in the guarantee of fundamental human rights, difficult as it may be to pronounce such words in the century that has given us Hitler and Stalin, agonizingly slow as the process may be.’  Such progress reflects ‘continual efforts to overcome authoritarian structures and to expand the domain of freedom.’ And those efforts, in turn, ‘probably (reflect) instinctual patterns that are just part of our moral nature.’ 
To us, Chomsky’s ‘optimistic view’  seems highly plausible. As noted earlier, a fundamental human interest in autonomy and capacity for moral judgment appear to have played a significant role in many historical achievements in the cause of human freedom (for example, the abolition of slavery, the extension of religious and political toleration). Despite the murderousness of the twentieth century, such aspirations and capacities have clearly been operative in many of the great political struggles of the recent past—from Third World efforts to break free of colonial bondage, to the civil-rights movement in the us, the worldwide movement for women’s liberation, or the revolution against Stalinism in Eastern Europe. And whatever the political doldrums of the present, there is every reason to think they are operative now.
Chomsky has disavowed a ‘faith’ in any project or tradition, including a faith in reason itself.  But his weak evolutionary theory suggests a ‘reasonable faith’  in human beings that works in support of hope about social advance.  While the evidence about people is not decisive, nothing that we know about human nature is inconsistent with the contention that aspirations to freedom and decency are fundamental features of that nature; and nothing that we know about social order defeats the hope that the pursuit of these aspirations will produce significant improvement in human circumstances. The fact that such hopefulness is consistent with the evidence enables children of the Enlightenment to be optimists of the will, without condemning themselves to being irrationalists of the intellect. It is Chomsky’s insistence on this point, his commitment to both reason and moral hope, that we take to be his signal contribution to social thought. 
 ‘The Responsibility of Intellectuals’ (1967), in N. Chomsky, American Power and the New Mandarins: Historical and Political Essays, New York 1969, p. 325. References to Chomsky’s work that appear in the footnotes are intended only to support the claims we make about his views, not to provide an exhaustive inventory of relevant passages.
 Chomsky also, of course, has long enaged in more direct forms of political activism, including civil disobedience.
 Language and Responsibility, New York 1977, p. 3.
 Chomsky’s own work in linguistics may be understood as a contribution to the development of such a theory, but he would be the first to assert its distance from a complete and systematic account.
 Language and Politics, edited by Carlos P. Otero, Montreal 1988, p. 756.
 ‘The Relevance of Anarcho-Syndicalism’, in Radical Priorities, edited by Carlos P. Otero, Montreal 1981, p. 247.
 See, for example, ‘Language and Freedom’ (1970) and ‘Notes on Anarchism’ (1973), in For Reasons of State, New York 1973; ‘The Relevance of Anarcho-Syndicalism’.
 On the background in rationalism and romanticism, see Cartesian Linguistics, New York 1966; ‘Language and Freedom’, pp. 402–403; Language and Mind, New York 1972, pp. 76–7.
 In the case of language, Chomsky’s idea was to secure the connection between creativity and rules—to show what enables us, as Humboldt put it, to ‘make infinite use of finite means’—by incorporating recursive rules into a representation of the grammatical knowledge of ordinary speakers. For discussion of the role of recursive rules in solving Humboldt’s problem, see Aspects of the Theory of Syntax, Cambridge, Mass. 1965, p. 8; Language and Politics, p. 146, and the striking remark: ‘I think that the ideal situation would have been to have someone in 1940 who was steeped in rationalist and romantic literary and aesthetic theory and also happened to know modern mathematics.’
 The discussion that follows draws on various remarks that Chomsky makes about the connections between his views about language, human nature and politics. For a representative sample, see Cartesian Linguistics, pp. 24–6, 91–3; ‘Language and Freedom’; ‘Equality: Language Development (1976), Human Intelligence, and Social Organization’, in The Chomsky Reader, edited by James Peck, New York 1987, pp. 195–9; Language and Problems of Knowledge: The Managua Lectures, Cambridge, Mass. 1988, ch. 5; Language and Politics, pp. 143–8, 240–46, 318, 385–7, 402–3, 468–9, 566–7, 593–4, 696–7, 755–6. In these passages Chomsky does not separate, as we do, issues of modularity and creativity in considering the relevance of the understanding of human language for a more general theory of human nature. But the passages listed above suggest the relevance of both, and they are, so far as we understand, independent. Thus, the existence of a visual module employing a rigidity principle in the interpretation of visual experience does not imply anything about the creative use of visual perception (whatever that might mean). And the absence of a module in a particular domain is consistent, at least in principle, with creativity since the latter might reflect general features of human reason (as Descartes seems to have thought). Finally, the relevance (such as it is) of the study of human language to a more general theory of human nature is not affected, so far as we understand, by the shift from a conception of language in terms of rule systems to Chomsky’s more recent ‘principle-and-parameters’ view. For discussion of the shift in the account of language, see Knowledge of Language: Its Nature, Origin, and Use, New York 1986; Language and Problems of Knowledge.
 Aspects of the Theory of Syntax, p. 57.
 Language and Politics, pp. 696–7.
 Language and Problems of Knowledge, p. 153.
 Orlando Patterson emphasizes that slaves are characteristically represented as outsiders to the community that enslaves them, as ‘natally alienated’. See Slavery and Social Death, Cambridge, Mass. 1982, ch. 2.
 Language and Problems of Knowledge, p. 155.
 For example, ‘Language and Freedom’, p. 405.
 On enjoying the activities of others, see ‘Equality: Language Development, Human Intelligence, and Social Organization’, pp. 198–9. For discussions of the ‘Aristotelian principle’ and its connection with the value of community, see John Rawls, A Theory of Justice, Cambridge 1971, pp. 424–33, 523–5 (the background in Humboldt is noted on p. 525, n. 4).
 For discussion of these features and their relevance to historical sociology, see Barrington Moore, Injustice: The Social Bases of Obedience and Revolt, White Plains 1978; Roberto Unger, Politics: A Work in Constructive Social Theory, Part i False Necessity, Cambridge 1987; Joshua Cohen, ‘The Moral Arc of the Universe: The Case of Slavery’, n.p. 1989.
 Immanuel Kant, Critique of Practical Reason, Part i, Book i, trans. Lewis White Beck, Indianapolis 1956.
 Language and Politics, p. 756.
 See ‘Notes on Anarchism’, p. 404; ‘Equality: Language Development, Human Intelligence, and Social Organization’, p. 195; Language and Politics, p. 147.
 ‘Interview’ (with James Peck), in The Chomsky Reader, p. 29; also ‘Notes on Anarchism’, p. 371.
 Our discussion of Chomsky’s anarchism has been greatly aided by Michael Taylor, Community, Anarchy, and Liberty, Cambridge 1982.
 Alexander Berkman, for example, asserts that ‘Anarchism teaches that we can live in a society where there is no compulsion of any kind.’ See What is Communist Anarchism?, New York 1972, p. 182.
 ‘Objectivity and Liberal Scholarship’, in American Power and the New Mandarins, pp. 72–124.
 ‘The Relevance of Anarcho-Syndicalism’, p. 260.
 Drawing on a range of anthropological and historical studies, Michael Taylor argues that virtually all anarchist communities have relied on some scheme of social controls (threats and offers of sanction) to ensure compliance, and not simply on socialization and education. See Taylor, Community, Anarchy, and Liberty, pp. 39, 76ff.
 ‘The Relevance of Anarcho-Syndicalism’, p. 250.
 Chomsky commonly emphasizes a general concern about concentrated power, suggesting that the nineteenth-century libertarians focused on concentrations of power in the State and Church, while twentieth-century libertarians (especially socialist anarchists) have extended such concerns to the concentration of economic power. See, for example, ibid., p. 248; Language and Politics, pp. 301, 744.
 See, for example, the discussion of the possibilities of providing protection within a state of nature in Robert Nozick, Anarchy, State, and Utopia, New York 1974, pp. 12–17.
 ‘The Relevance of Anarcho-Syndicalism’, p. 249.
 Ibid., p. 251.
 Ibid., p. 250.
 Ibid., p. 250.
 Ibid., p. 251.
 Chomsky indicates that this question arises naturally, given his views on human nature; see Deterring Democracy, London 1991, p. 397.
 See ‘Manufacturing Consent’ (1984) in The Chomsky Reader, pp. 131–2; Deterring Democracy, ch. 12.
 Turning the Tide: us Intervention in Central America and the Struggle for Peace, Boston 1985, pp. 234–6.
 The term, taken from Walter Lippmann, Public Opinion, London 1932, provides the title for E. Herman and N. Chomsky, Manufacturing Consent: The Political Economy of the Mass Media, New York 1988.
 See, for example, On Power and Ideology: The Managua Lectures, Boston 1987, which notes that the us is near the ‘libertarian end in the spectrum of existing societies’ (p. 114) but that debate proceeds ‘within very narrow limits’ (p. 124).
 Herman and Chomsky, in Manufacturing Consent, use the term ‘propaganda model’ to refer just to the first of the two components discussed in the text. For reasons of terminological convenience we use the term to cover a wider range of Chomsky’s views about the media. We do not assume, however, that Herman embraces all aspects of the propaganda model as it is characterized here.
 Herman and Chomsky, p. xv.
 The account of the five filters advanced in Herman and Chomsky is basically a refined version of the conception of elite interests sketched here (though we have abstracted from their fifth filter, namely anti-communism). Thus they discuss the way that news is filtered through (1) the ‘important common interests [that owners of major media share] with other major corporations, banks, and government’ (p. 14); (2) the need of the media to sell an audience to advertisers who are interested in ‘audiences with buying power’ (p. 16); (3) the dependence of media on government and corporate sources of information (p. 19); and (4) the requirement of sensitivity to ‘flak’ produced by ‘individuals or groups with substantial resources’ (p. 26). The basic, ‘guided market’ model of explanation is that the power to fix what gets said is held by individuals and groups who have substantial resources, who have a reasonably good understanding of their interests, and who seek to ensure that what is said conforms to those interests.
 ‘Interview’, p. 45.
 In emphasizing that this policy is deliberately pursued, we do not mean to suggest that there is a conspiracy—that elites explicitly coordinate in the pursuit of a common interest in deception. On the other hand, the explanation is not a classical invisiblehand explanation, since the pattern of distortion is the result of deliberate acts of distortion by individuals.
 Ibid., p. 39.
 For example, in one interview Chomsky claims that ‘The more intelligent people are just lying, but the less intelligent believe it.’ See Language and Politics, p. 713.
 ‘The Backroom Boys’, in For Reasons of State, p. 3. Chomsky borrows the ‘lunatics’ phrase from then New York Times reporter Gloria Emerson.
 While Chomsky’s earliest political essays focused on the ‘secular priesthood’, the emphasis in his writing shifted in the late 1970s to the role of the mass media in the manufacture of consent. At least part of the explanation may lie in the role that academic experts on Latin American politics played in opposing us policy in Central America in the 1980s. See, for example, the discussion of the report by the Latin American Studies Association in the 1984 Nicaraguan election, in Herman and Chomsky, ch. 3. For note of the relative unimportance of such opposition, however, see Deterring Democracy, p. 105, n. 24.
 Chomsky often emphasizes that illusions about the operation of the world are most pronounced among such ‘experts’. See, for example, ‘Interview’, p. 43.
 Necessary Illusions: Thought Control in Democratic Societies, Boston 1989, p. 47.
 Language and Politics, p. 373.
 ‘Interview’, pp. 48–9.
 Language and Politics, p. 685.
 Ibid., p. 765.
 Ibid., p. 602.
 Ibid., p. 765.
 Necessary Illusions, pp. 47–8.
 The Culture of Terrorism, Boston 1988, p. 3, and n. 3.
 Necessary Illusions, pp. 148–9.
 Chomsky himself states that he decided to focus his political writings on the operation of ‘ideological institutions’ for two reasons. The first was a ‘judgment of importance’, the second a matter of personal circumstances and abilities. And he says that considerations of the second kind would have been sufficient to lead him to concentrate on the ways in which schools, universities and media ‘serve to indoctrinate and control’. Language and Politics, p. 372. But saying this is perfectly consistent with what we state in the text.
 See, for example, On Power and Ideology, p. 128; Necessary Illusions, p. 148.
 Herman and Chomsky, pp. xiv–xv.
 Necessary Illusions, p. 151.
 For Chomsky’s own criticisms of media coverage of events leading up to the Gulf war, see Deterring Democracy, ch. 6.
 ‘The Backroom Boys’, in For Reasons of State.
 Ibid., p. 66. See also Chomsky’s remarks about the ‘persistence of. . .astonishing illusions’ among presidential advisors and the existence of ‘historical fantasies at high levels of decision-making’—‘absurdities that must be taken seriously, given the vast resources of terror in the hands of those whose decisions are guided (or justified) by them.’ Ibid., p. 165, n. 193, and the references there to American Power and the New Mandarins.
 ‘The Backroom Boys’, p. 54.
 Adding further openness, or doubt, is the fact that American elites are divided, and what is functional for one elite group may not be functional for others. One consequence of this is that the predictive power of the propaganda model is diminished. Different popular ideologies may be equally functional for elites in general, but not for elites in particular. In these circumstances the model does not tell us which ideology will reign.
 Chomsky does discuss several other sources of consent, agreeing for example with our view (presented in chapter 3 of J. Cohen and J. Rogers, On Democracy, New York 1983) that consent reflects the fact that the operation of capitalist democracies tends to channel political action into the pursuit of interests in short-term material gain and to enable individuals to satisfy those interests. See, for example, Turning the Tide, pp. 233–4. This, however, is consistent with the claims we attribute to him in the text, which concern the fundamental sources of ideological conformity and consent.
 That most people, we think, take Chomsky to be making a claim of this sort provides another motivation for engaging it.
 For a careful discussion of this and other senses of ideology, see Raymond Geuss, The Idea of Critical Theory, New York 1981.
 Chomsky is aware of this. See, for example, Turning the Tide, pp. 240–45. But while he takes it (properly in our view) as a basis for hope in achieving a better political system, he does not explore the limits it suggests to his own account of the stability of that system.
 For such an account, see Cohen and Rogers, ch. 3.
 ‘Language and Freedom’, p. 406.
 Turning the Tide, p. 253.
 ‘Language and Freedom’, p. 406; Deterring Democracy, p. 64.
 Problems of Knowledge and Freedom: The Russell Lectures, New York 1971, pp. 110–11. Chomsky is here quoting Bertrand Russell.
 Language and Politics, p. 469.
 ‘Interview’, p. 49.
 Language and Politics, p. 735; Deterring Democracy, p. 173.
 See, for example, Chomsky’s summary of his disagreements with Foucault, in Language and Responsibility, p. 80.
 Language and Politics, p. 773.
 Necessary Illusions, p. 355.
 Language and Politics, p. 469.
 Language and Problems of Knowledge, p. 154.
 ‘Interview’, p. 48.
 The term comes from Kant, who held that it is reasonable, on moral grounds, to have faith that God exists, that the will is free, and that the soul is immortal. See Kant, Critique of Practical Reason, Part i, Book 2.
 And, recently, he appears to acknowledge as much. See Deterring Democracy, pp. 397–401.
 We would like to thank Robin Blackburn, Robert Brenner, Edward Herman, Paul Horwich and Carlos Otero for their comments on an earlier draft of this paper.