ARTIGO ORIGINAL

What is ideology?

O que é a ideologia?

Qu’est-ce que l’idéologie?

¿Qué es la ideología?

 

John Levi Martin*

* Professor at University of Chicago, 1126 E 59th Street, Chicago IL 60637. E-mail: jlmartin@uchicago.edu

 

ABSTRACT

Political ideology has been a confusing topic for social analysts, and those who attempted to eschew judgmental reductions of others’ conceptions and develop a non-polemical political psychology found ideology behaving in ways that defeated their theories of political reasoning.  I argue that political ideology can best be understood as actors’ theorization of their own position, and available strategies, in a political field.  

Keywords ideology, political parties, political psychology.

 

RESUMO

A ideologia política tem sido um tema confuso para os investigadores sociais e para aqueles que tentam evitar julgar as limitações de outras conceções e desenvolver uma psicologia política não polémica que procure encontrar um comportamento ideológico que ultrapasse as teorias do raciocínio político. Defendo que a ideologia política pode ser melhor entendida como uma teorização da posição dos próprios atores e de estratégias disponíveis no campo político. 

Palavras-chave ideologia, partidos políticos, psicologia política.

 

RÉSUMÉ

Le thème de l'idéologie politique suscite la confusion chez les chercheurs sociaux et chez ceux qui s'efforcent d'éviter de juger les limites d'autres conceptions et de développer une psychologie politique non polémique, afin de trouver un comportement idéologique qui dépasse leurs théories du raisonnement politique. Je soutiens que l'idéologie politique peut être mieux comprise en tant que théorisation de la position des acteurs eux-mêmes et de stratégies disponibles, dans le champ politique 

Mots-clés idéologie, partis politiques, psychologie politique.

 

RESUMEN

La ideología política ha sido un tema confuso para los investigadores sociales y para aquellos que intentan evitar juzgar las limitaciones de otras concepciones y desarrollar una psicología política no polémica que busque encontrar un comportamiento ideológico que trascienda las teorías del raciocinio político. Defiendo que la ideología política puede ser mejor entendida como una teorización de la posición de los propios actores y de estrategias disponibles en el campo político.  

Palabras-clave ideología , partidos políticos, psicología política.

 

Different ways of conceptualizing ideology[1]

It is common for sociological discussions of ideology to begin by acknowledging, if not bemoaning, the plurality of different ways of using the term “ideology” (Eagleton 1991).  Marx and Engels used it to denote the most abstract conceptions that populate an imaginary world of ideas independent of material life; later Marxists often used it to denote a conspiratorial ideational wool pulled over the eyes of the masses; political scientists use it to denote packages of positions, often believed to be unifiable in a single preferred optimal state, and, of course, many of us use it to denote the beliefs, attitudes and opinions of those with whom we disagree.

A conventional solution in sociology to these problems comes from our nominalist epistemology—that is, we tend to assume that general theoretical terms must be created by the analyst and are heuristic devices used to greater or lesser success in particular analyses.  Thus we assume that each investigator is basically free to choose how to define his or her terms, and the worst that we can say regarding a particular case is that the definitions didn’t help much.  

Now there are some good reasons for accepting such a nominalist position, but it is far from obviously the best one for the social sciences, and there is much to recommend a quasi “realist” position instead.  That is, we assume that the generalities that we talk about are not open to definition at the whim of the investigator, but are treated as largely pre-given.  It is worth emphasizing that this sort of realism (as opposed to nominalism) is separable from the issue of realism as opposed to idealism (for more on this distinction, see Martin, 2014).  For example, many sociologists are being realists in this (anti-nominalist) sense when they argue that sociologists should focus on the categories (“emic”) that are used by actors.  Even though some of these sociologists may be more like “idealists” in that they argue that all the categories that actors use are (potentially) disconnected from material reality, the investigator is not free to define categories for his or her particular analytic purposes, but must be guided by the externally created ones.  Thus if a certain group has a definition of “witch,” the investigator must attempt to grasp this, as opposed to defining what counts as a witch for her purposes of investigation.

Most field theorists, following Bourdieu (e.g., 1984 [1979]), have such a realist position regarding the nature of some of the key constructs they use to understand social action (though some, like Wacquant, 2002, and Bourdieu himself, will criticize certain other definitions used by the actors they study, specifically, those in which some sort of “bad faith” is inherent; Wacquant, 1999: 276, has a clear defense of a rationalist interpretation of Bourdieu—that is, one that privileges a coherent and defensibly true vision of social conflicts).  In particular, the endogenous definition of the “stakes” of any field, and what capital can be legitimately (if questionably) used to pursue these, leads the investigator to need to have her concepts guided by those of actors.  The sociologist who “defines” what “art” is is not studying the field but playing a role in it.

Thus if politics is one of those spheres of actions that can be called a field, guided by the reciprocal orientation of actors to one another, we cannot allow ourselves simply to define things as they suit us best.  It is for this reason that we cannot simply solve all our problems by agreeing to disagree, and to disagree by defining our terms differently—at least when it comes to aspects of political life that are within the phenomenological experience of subjects.  And I think there is good reason to think that there is a working consensus as to what we mean by ideology in politics.  That is, actors will tend to agree as to who (other than themselves, of course), “has” an ideology and when they seem to be deploying it.  Thus here I want to attempt to understand the nature of this ideology—that which actors seem to consensually develop and use in ordering their political attachments.

Here I am going to argue that the conceptions that might seem furthest apart—that of Marx and Engels on the one hand, and those of political scientists on the other—must be put together.  That is, political ideology is “ideology” in the sense of Marx and Engels not because it is false or distracting, but because it is the ideational equivalent of actual patterns of relations, in this case, specifically political relations.

 

Political ideology and political reasoning

The Classic Approach

Here we are interested in political ideology, which means that we must distinguish it from (on the one hand) what might be considered ideology more generally, and (on the other) from non-ideological political beliefs.  While some theorists may argue that all ideology is, by its nature, political, there is also a consensually defined more restricted use of the word “political,” especially in democracies.  This is to refer to processes and institutions turning on the quest to control the state machinery (or, analogously, other organizations, but let us put such analogous usages to the side).  In most democracies, this means an orientation to political parties, as these are the organizations that have arisen to pursue such a quest.  I will accept this usage here, and be concerned with beliefs that are understood as relevant to party contestation.  Thus someone may have an opinion about a state policy, but if this is disconnected from partisan struggle, we do not consider it a “political” issue (it may be, for example, a technical problem).

Can we say something about the sorts of cognitive elements that might compose political ideology?  For example, can we list them?  When political and social analysts define ideology, they tend to give extremely broad definitions, usually including beliefs, attitudes and values (e.g., Adorno et al., 1950: 2; Campbell et al., 1964: 111, 192; Jost, 2006: 653; Kerlinger, 1984: 13; Tedin, 1987: 65).  This basically runs the gamut of all possible cognitive elements.  Could it be that we attempt to restrict the class of things included by ideology in some other manner?  Are there specifiable qualities of the elements that constitute ideology?

Most social scientists have assumed that if ideology is separable from some other political beliefs or opinions, it is because ideology is intrinsically normative and generative (see Lane, 1973: 85; for a recent synthesis see Hinich and Munger 1996).  A classic example of an intrinsically normative definition of ideology comes from Downs (1957: 96):  “We define an ideology as a verbal image of the good society and of the chief means of constructing such a society.”  This idea that ideological differences are fundamentally about differences in valuations, both abstract and concrete (that is, “values” and “attitudes”), is widespread (e.g., Billig, 1984: 446; Rokeach, 1968: 123-124; Tedin, 1987: 65; also see Jacoby, 2006; Jacoby and Sniderman 2006; Peffley and Hurwitz, 1987; cf. Minsky, 2006).   

Here I am going to use the United States as my running example, in part for reasons of familiarity, but also because its two-party system highlights some of the more fundamental dynamics involved in political contestation, as this seems to be the form that most politics spontaneously develop, when there is not a well developed system deliberately designed to channel party formation in a particular direction (for an example of such spontaneous dualistic politics, see Barth, 1965).  The US, like the UK, had its governmental system designed before the existence of stable party organizations, while the parliamentary systems that support multi-party systems were designed after the development of mass suffrage and the existence of parties, and these were taken for granted by constitution writers.   Now it is not necessarily the case that a two-party system leads to a division into “liberals” and “conservatives,” though I will argue below that there is actually good reason to expect the development of a “unidimensional” understanding of partisan differences.  However, as this is the case in the United States, I use these terms to describe the consensual self-understanding of actors.  Thus if ideology leads to political choice, it does through “liberalism” and “conservatism.”   But the question is what these terms mean—what the “ideologies” are.  The conventional approach assumed that these were, above all else, oppositions of packages of values.

Thus conservatives are said disproportionately to value self-reliance, limited government, and so on, while liberals are thought disproportionately to value equal opportunity, tolerance, and so on (Klueger and Smith, 1986; Goren, 2004, 2005; Jost et al., 2008).   It is such differences in values that we generally think about when we consider a political “clash of cultures” (see DiMaggio et al., 1996).

Now this approach to reducing political ideology to a collection of “typically conservative” or “typically liberal” values runs into the problems that most value- or norm-based explanations have, namely that our key explanatory elements are very proximate to that which is to be explained—sometimes crashing into tautology.  Explaining citizens’ preference for, say, a war effort or for welfare benefits by pointing to their supposedly distinct values (militarism or equality)—that is, their political ideology—is somewhat akin to explaining that the reason opium induces sleep is its “soporific quality” (cf. Lau et al., 1991).  Of course, if it turns out that it is indeed values that separates conservatives from liberals, one cannot complain that these are not the analytic elements we wished for, but, given the proximity of such values to the opinions they are to explain, we must be somewhat cautious of the initial appeal of the approach to ideology that treats it as fundamentally about valuation.

The second common understanding of ideology is that it is, as Downs (1957: 96) stressed, generative: it facilitates our taking a stand on a particular issue (Higgs, 1987: 37-38; also Lau et al., 1991; Zaller, 1992: 26).  In particular, most analysts of public opinion have embraced what Goren (2004) calls the “political sophistication” model.  Ideological values are then combined with political information to produce non-random opinions on specific matters.  

For example, consider persons in the United States attempting to decide whether they will support a policy, say, one that gives benefits to out of work persons in American inner cities (who are likely to be of Afro-American descent).  Our imaginary citizen first draws on his ideological values—let us say equality and fairness—and then combines these with what he knows about the world—that there is a great deal of unemployment, and that the changing economic structure and persistent racism make it hard for American blacks to get jobs no matter how hard they try—and produces an opinion, in this case, to favor the policy.  In sum, according to this conception, values + beliefs = opinion; attitudes are a fusion of otherwise separable prescriptive and descriptive cognitive elements.

This suggests that ideologues should be those who have clear value commitments, and mutually supporting value commitments.  Thus one would be hampered as an ideologue were one to emphasize both individual freedom and state regulation, as increasing one seems to logically imply decreasing the other.  Further, even in the absence of such logical contradiction, the nature of the world may be understood to be such that other sorts of valuations are incompatible—for example, valuing equality of opportunity and equality of outcome may be understood as incompatible given the existence of good and bad luck distributed across persons, whether randomly or not.  Finally, this conception suggests that ideologues without sufficient information about the world would be unable to form opinions, as they would only have the “ought” part of their cognitive orientation, and not the “is” part.

Problems with the Classic Approach

However, there have been a few recurrent anomalies for this approach.  The first problem is that ideology seems to have a direct effect on many policy preferences that cannot be explained according to a chain of reasoning whereby the abstract principles of the ideology imply more proximate principles that, when combined with political information, lead to the preference.  For example, we might imagine that (A) a liberal ideology leads people to favor, in principle, (B) racial equality, which in turn might influence (C) a particular policy choice such as one involving regulation of housing law.  However, well informed ideologues choose the “correct” side of some issue even when they do not hold the beliefs that should mediate between ideology and choice (Federico and Sidanius 2002; Sniderman et al. 1991: 65- 67, 81-84).  That is, A seems linked directly to C, without mediation by B.  Political psychologists have generally assumed that just as you can never be too smart or too rich, you can never be too ideologically consistent:  indeed, they have tended to assume that such consistency (in the sense of the work of Festinger, 1957, Feldman, 1966, and Abelson, et al., 1968, is a prerequisite for good political participation.  For this reason, the “hyper-consistency” of well-informed ideologues has not been treated as problematic, even though it forces us to reevaluate our assumptions about how ideologues reason.

The second problem is that it turned out that this sort of hyper-consistency wasn’t quite matched by a similarly high degree of consistency regarding the fundamental values.  This in no way implies that any conviction is wanting among ideologues—however, this conviction appears to be turned on and off selectively.  Those who argue against the separation of church and state when it is a matter of their religion (usually Christians in the US), making recourse to very abstract values, had no problem arguing for this same separation when it came to others‘ religion.  And similarly, those who were used to arguing for the separation of church and state when it came to battling the conservative Christians, switched over to arguing against too stringent a separation when this became linked to intolerance of Muslims.  Most wonderfully, Jarret Crawford and Eneda Xhambazi (2013) studied how Americans evaluated two different recent populist movements, the “Tea Party,” which became associated with right wing causes, and the “Occupy Wall Street” movement, which became associated with the left wing.  They show that Tea Party supporters tend to appeal to values of the right of protest when asked about the Tea Party, but appeal to the importance of social order when asked about Occupy Wall Street; and Occupy Wall Street supporters tend to appeal to values of the right of protest when asked about Occupy Wall Street, but appeal to the importance of social order when asked about the Tea Party.

The third problem has to do with the factual basis for opinion formation.  If the “values” part didn’t seem to work as it should, neither did the “knowledge” part.  Since Converse’s, 1964 classic work, political psychologists have been forced to acknowledge that few Americans have enough factual information to allow them to make the sorts of decisions that were assumed by the model of political reasoning.  While it must be acknowledged that there are other polities in which the average citizen has more information than does the average citizen of the US, what is key about the American example is that it demonstrates that lack of factual data hampers opinion formation only slightly.  And this is because the “information” held by an average citizen is, when one considers what would be needed to make a rigorous deduction as to a political choice, necessarily extremely partial.  Consider the question of which candidate to favor in an election.  Presumably, one would need to know what the candidate would actually do when elected, which of course is beyond anyone’s actual knowledge.  Thus even if voters knew what candidates promised to do, they would fall short of a decent model of political reasoning through no fault of their own.  But they would also need to know how the promised actions would affect their own interests, which would require a great deal of knowledge about the world and its causal texture, knowledge that few of us have.

And to top it all off, while the evidence that ideology gives us values gets weaker the closer we look, it becomes more and more plausible that ideology gives us knowledge—which might seem contradictory.  Thus the fourth problem with the conventional view is that ideology gives citizens exactly the wrong cognitive element.  In fact, differences in ideology seems to correlate much more strongly with differences in descriptive statements than they do with differences in purely prescriptive ones (cf. Rumelhart, 1989; Kurtz et al., 1999).  And this is because, as Rokeach, 1968 always held, the thing about values is that they’re all good, considered singly.  It’s only in trade-offs that people begin to distinguish themselves.  So people can agree with one another in their value commitments, while still having diametrically opposed opinions.  

Now to some extent, the way that this happens has long been well understood.  Because there are usually a variety of competing sources of information (such as newspapers) that are more or less strongly associated with different ideologies, ideologues have the capacity to choose the information source that is likely to disproportionately report facts (or would-be facts) that support their previous position.  Further, there is general evidence from psychology that when we come across information that contradicts our strongly held positions, we are less likely to pursue it (e.g., read it), less likely to understand if we do pursue, and more likely to forget if we do understand.

But even more, it seems that ideology provides “knowledge” about the world indirectly (Lau et al., 1991; Dawson, 2001).  Let us return to the example used above, namely  Americans determining whether to support a policy for unemployed blacks.  We walked through the traditional understanding of how an ideologue might be led to support the program (values + beliefs = opinions)—a commitment to fairness, plus a belief that there is discrimination against blacks, leads to favoring the policy.  Yet many conservatives do not favor the policy.  Could this be because they (unlike liberals) value “self-reliance”?  It is certainly true that they do, but as shown by Martin and Desmond (2010), so do liberals—in fact, there are only very small differences between liberals and conservatives here.  Where they do differ greatly is in their belief as to how worthy the recipients are (how likely the poor are to be trying to solve their own problems).

Now this issue refers to a matter of external fact.  We would imagine that at least one of the two positions has to be wrong.  Could we determine this through social science?  The actual wording of the item analyzed by Martin and Desmond is this:  “Most poor people these days would rather take assistance from the government than make it on their own through hard work” (agree or disagree).  Who are “poor people”?  Only adults?  Not on disability?  Under retirement age?  Do we agree that it is an “either/or”?  And, most important, just how hard does someone have to work to “make it,” and how far do they “make it”?  Are we talking about turning down a $30,000 a year union job with medical benefits so as to stay on TANF, or not giving up food stamps when one is working two jobs, each below minimum wage, each with erratic hours?  Taking the question literally, we scratch our heads, and wonder how could anyone answer it with confidence?  The further we pursue the matter, the more implausible the classical understanding seems, and the more difficult it is to salvage it.

 

Political sides and political action

Sides and parties

Given that the classic logic seems implausible, different political psychologists have contributed different possible “heuristics” that citizens can use to construct their ideas and actions (here see recently the work of Baldassarri, 2012).  One popular theory of political action is a “rejectionist” one which directly parallels the falsificationist logic of Karl Popper (see, e.g., Riker, 1982).  Rather than reject hypotheses that fail tests, voters reject candidates who have, in the past, failed their interests.  In the United States, this dynamic is popularly called “throw the bums out”.  The assumption is that members of a party in power are retained until their performances falls below some threshold in a multiparty system, at which point voters will move to replace them, either with their opponents in a two-party system, or with the party that makes the most credible claim to have always argued against the problems that the voters retrospectively identify.[2]

There is a great deal of evidence to suggest that these heuristics are used by voters, and that they can be of fundamental importance in a two party system. However, such a heuristic can only be used to choose whom to vote for (and it does not, in itself, generate an ideology that could inform other choices); further, it really only deals with switching, while we know that most of the time, most citizens stick with their party through thick and thin.

Is there a more general way in which the choice of a side may be understood to be a plausible expression of actors’ interests, one that does not require switching back and forth?  It might be, if sides in terms of political parties correspond to recognized sides of a social cleavage.  In this case, we may not require actors to think through each and every position.  Political reasoning is a “package deal,” not “a la carte,” in that when we choose a side, we choose all the opinions held by the party representing that side, kit and caboodle.  Thus if workers support a party claiming to be a workers’ party, they are treated as reasoning well; if they do not, it is assumed that, in the absence of other explanation, that they are failing to reason.  Of course, everyone will recognize that a party that claims to be for the workers may not really be for the workers, or even if it is, that the party faces the same problems of incomplete knowledge that individuals face.

However, even bracketing this, such a conception of identity-based politics runs into problems if we have a polity that has what we call “cross-cutting cleavages” (Simmel, 1958 [1923]; Lipset, 1960)—that some workers are Catholics and others Protestant, say, so that it is unclear whether Protestant and Catholic workers should band together and form a Workers’ Party, against Protestant and Catholic capitalists, or whether Catholic workers and Catholic capitalists should form a Catholic party against Protestant workers and capitalists.  Thus the heuristic of choosing sides sometimes brackets what is most important to us—the question of why voters choose the side they do.

Without denying the force of this objection, we may still find that the importance of such “choosing up sides” on opinion formation is not restricted to affiliating with a particular pre-given program.  Sniderman et al. (1991) propose that one way that only somewhat-informed citizens can generate their beliefs is to consider what their enemies are likely to hate, and choose that.  (They call this the “likability” heuristic, but it has more to do with disliking than liking).   There are four things to note about this proposal.  The first is that there is indeed evidence to support it; and the second is that it radically undercuts the classical model.  The third is that we are being forced to take a view of ideation that is compatible with a pragmatist perspective—we need to understand what people are trying to do with their ideas.  And the fourth is that it implicitly returns us to a notion of politics that few American political scientists have found appealing, namely that it is a struggle between camps first and foremost (I will return to this shortly).

But this also leads to an interesting implication—if politics involves the establishment of webs of alliance and opposition, and this in turn is used by political actors to generate opinions, then we may find that Marx’s original conception of the nature of ideology may have a great deal to offer us.  I turn to a brief recapitulation of his argument.

Back to Marx

Here we must remind ourselves of what Marx and Engels’s own position was when, in 1845, they wrote The German Ideology, given that it has been creatively re-read by generations of putative followers with very different goals in mind.  In particular, faced with the general rejection of their program by most European workers, Marxists often formulated various versions of “ideology” that explained why things didn’t go the way they had said (and hoped) they would.  Ideology became (in this later theory) a surprisingly effective way of controlling masses of persons—exactly the opposite of the claims made by Marx and Engels.

For coming from the context of the Young Hegelian movement, where such claims as to the mystifying powers of ideas were rampant, Marx and Engels, in diametrical opposition, denied the importance of such ideas and instead treated them as largely epiphenomenal.  They began their work with a parody of the Young Hegelian way of thinking, which was to assume that our ideas have somehow achieved a position of power over us.  In contrast, Marx and Engels emphasized that if ideas ever seem to be fetters, it is because they are “the mere images of very empirical fetters and limitations, within which move the mode of production of life, and the form of intercourse coupled with it” (1976 [1845-6]: 45).

In any case, what is ideology?  To Marx and Engels, it was organized beliefs at a high level of abstraction; they used the term to indicate include morality, religion, metaphysics, politics, law and judicial theory, and certainly speculative philosophy.  While it is not the case that all beliefs are ideological, these ones are because they are idealized, universalized and detached expressions of actual social relations.  For example, the concept of “freedom” central to German idealist philosophy was, argued Marx and Engels, an ideal expression of the material relations of market orientation that characterized nineteenth century bourgeois society.  Further, this notion was universalized, in that it wasn’t simply freedom-to-buy-and-sell, but freedom tout court that was supposedly being talked about.  Finally, this was detached in that rather than accept that this freedom comes from these material relations, thinkers believed it to occupy a special position in a realm of ideal elements.

The generation of such ideology, while indeed left to specialists, is not the result of a clever conspiracy, but rather a natural expression of the division of labor.  This splits mental from manual labor, leading to ideational production by persons who are themselves detached from production.  The very connection of ideational production and material production explains the detachment of ideas from materiality, as ideational producers, like others, generalize their own experiences (which are now set against those of others, due to the inherent contradictions of the division of labor).[3]

Thus ideology is a generalization of social relations; it is the ideal form of the actual relations, seen from the perspective of one position in this set of relations, but universalized, idealized and abstracted.  Marx and Engels, thinking at the largest scale, were of course concerned specifically with the general relations of production in a social world—those that, seen sociologically, appear as class relations, and that, seen juridically, appear as property relations.  My argument is not that political ideology is some form of these class relations, but rather, that it is to specifically political relations what Marx’s ideology is to the relations of production.  

What is political action?

To understand the nature of political relations, we must first answer the question, ‘what is political action?’, for we shall see that these relations are the outcome of specifically political action.  To attempt to answer this question, we can turn to two sources, one historical, and the other contemporary.  That is, we examine where the concept of political action first arose, and we also look for how we use it in contemporary speech; we will prefer the results of this sort of exercise to conclusions arising via deduction from first theoretical principles.

Regarding the first question, I turn to Hannah Arendt’s (1958) analysis of political action in ancient Greece.  Political action—action in the polis—was paradigmatically speech, speech made in the open air.  Second, it was speech that mattered, and it mattered because others could be convinced.  Yet not all needed to be convinced in order to win the day.  Despite the attempt of Plato to turn all of politics into the application of abstract principles of the good, even afterwards, politics required careful attention to the cultivation of a core set of adherents and, in many cases, the acceptance that some others would never be persuaded to join one’s side.  Even in an unorganized, plebiscitary democracy, it is not necessary to sway all, but enough of those who mattered, so that the others could not prevent one’s own proposals from being realized.

And this brings us to a second aspect of politics not emphasized by Arendt, but by another German thinker of decidedly different extraction and sensibilities, namely Carl Schmitt.  Politics, Schmitt (2008 [1932]: 26f) argued, is fundamentally about the division of others into friends and enemies.  Schmitt famously focused somewhat obsessively, as did those of his followers who joined the Nazi movement, on the rejection of the outsider, the enemy, the stranger.  I think we can excise that aspect from the more lasting aspects of his thought.  These are not only the division into friends and enemies, but his emphasis on the fact that no one but the political actor can identify who one’s enemy ought to be.

The brutalist sound of Schmitt’s writing—and his importance for Nazi political thought—may have scared many democratic theorists away from his argument.  But it seems to fit with other, seemingly radically different, understandings of the political, such as that of Arendt.  Because what seems to be distinctive about political action is the assembling of allies into groups to pursue the project of control over whatever degree of organizational apparatus a state has.  

Further, this conception seems to fit the way the term “political action” is used in everyday life.  Of course, where there is a developed political system, we will use the term to refer to anything related to this system, especially insofar as it involves parties.  But more generally, a decision is said to be “political” not only if (as Weber would say) it tends to involve striving for power, but if it specifically does so by making some substantive decision a means for the furtherance of one’s own side at the expense of others.  Indeed, even if the action does not noticeably affect the distribution of power, but only of other good things, we would term it political (or “playing politics”) if it is oriented to the division into friends and enemies.  Formulaically, one might say that when we use politics only to “line our own pocket” (increase our individual material wealth), we engage in “corruption.”  But when we line the pockets of our friends—not just a few close ones, but our specifically political friends—that is politics.

Finally, when we consider the sort of action that characterizes an accomplished political actor, what we find is that, contrary to the implications of Schmitt’s focus on the rejection of the enemy, it often involves increasing one’s stock of friends.  Eliminating the enemy is usually left to generals—it is courting him that is the task of the politician.  That is, if political action involves making alliances between friends, a key way to triumph is to make one of your enemy’s friends (and hence your potential enemy) into a friend.  Thus specifically political relations are the product of political action—they are the webs of alliance and rivalry, friendship and enmity, that constitute political sides.

The result, then, is that political actors, even when they act individually, are (insofar as they are carrying out political action) doing so with an eye to their position in a web of alliances.  In particular, where there is a well developed party system, these alliances take the form of parties.  We go on to explore the nature of such parties, and the implications for ideology.

 

Aggregation and alliance

Parties and alliances

Different theories of political party formation begin from very different premises.  Some of the most elegant of these sets of premises would not be seriously proposed as a historically valid account of party formation.  Yet they can prove useful analytic tools for understanding equilibrium action in a developed party system.  For example, some theories assume that all individuals are distributed on a one or two dimensional space of preferences, and that parties arise to compete for the allegiance of such atomized individuals.  That is, the goal of political action is no different from market purchasing—each individual has a set of preferences, and she makes choices to maximize her utility.  

To derive this approach, consider every political actor to have a “portfolio” of goals that she is pursuing; in our extreme starting point of total individualization, this portfolio is identical to each actor’s preferences (this will change as we pursue the development of parties).  Although this approach does not require any partition between material and non-material interests, material here meaning “narrowly economic,” for purposes of simplicity, we will imagine that this is the case, and that people are able to correctly ascertain their material interests as well.  Further, we will here imagine that actors pursue only their “material” interests, as opposed to abstract and/or transcendent values.  The reason for these assumptions is that, as we shall see, they allow us to begin an analytic pursuit of ideology without assuming its presence, as we would if we were to allow for “ideological interests.”

I emphasize that I do not believe this pure model of atomized decision making to have any descriptive utility, but find it remarkably useful as a thought experiment.  First, if individuals were able to choose to maximize their material interests, there is no reason that they would need to appeal to ideology at all.  Their justifications of their action, should these be required, could be made honestly on the basis of what is sometimes called “pocketbook” interests.  

Now let us continue this analytic account by allowing for party “aggregation,” basically by following the logic of Chhibber and Kollman (1998, 2004), who examine nationalization in terms of the strength of party attachments across regions.  For them, nationalization refers to an agglomerative process whereby local candidates throw in their lot with one another and, crucially, are recognized by voters as doing such.  This suggests a useful, if historically inaccurate, analytic reconstruction of the process of party formations, which we can term “topic clustering.”  That is, all individuals are originally located in a topos, a spatial position, and some of these positions, originally distinct, become aggregated to make a larger area.  We imagine that all persons are distributed in a space, call it “social space,” such that those who are closer to one another are more likely to share both their actual interests, and their perceived interests.  From this simple set-up, we can model the development of a party system.

Intersection and union

Each actor may initially be considered to pursue his own individual interests, but also, as a means to this end, to want to form alliances with others.  We will imagine that there are two ways in which this alliance can be cemented, which we can call “logrolling” and “suppression.”

“Logrolling” is a term from American politics to refer to when two actors or two parties make an exchange relation over their support for certain issues (Buchanan and Tullock, 1999 [1958]).[4]  If there is one person or party (A) which cares a great deal about issue X, and prefers outcome x1 over x2, but is largely indifferent to issue Y, and another party (B) which cares a great deal about issue Y, and prefers outcome y2 over y1, but is largely indifferent to issue X, then it makes sense for the two to join forces on a program (x1, y2).

“Suppression” is a term used by Mische (2009) for the political practice required to cement an alliance by A and B who share some, but not all, interests.  Using the approach to the relation between persons and ideas associated with Breiger’s (1974) conception of duality, Mische proposed considering set-theoretic intersection a possible tactic to facilitate alliance.  That is, if A‘s goals are the set {a, b, c, d, e} and B‘s goals are the set {c, d, e, f, g}, it would make sense for A and B to join forces on a program of {c, d, e}; to do so, however, A would need to suppress interest in a and b, while B would need to suppress attention to f and g.  Why?  Because we assuyme that some members of A don’t approve of f (or g), which is why this is not part of A‘s program; ditto B and a and b.  Note that while logrolling adds some (relatively trivial) “interests” to an actor’s “portfolio,” suppression removes some.  Suppression therefore tends to make an actor’s portfolio more abstract, while logrolling makes it more complex.

Now there is certainly evidence that political elites perform logrolling and suppression with alacrity when necessary.  But things may be quite different for their supporters, if these are necessarily brought along to defend the resulting platform.  The supporters are not always privy to the historical sequences, the back-room deals, or simply the worldly wisdom that has led to an alliance’s position, and yet they may need to be in a position of defending these to others, or to themselves.  It is my claim here that ideology is citizens’ way of comprehending the nature of the alliances in which they find themselves.

Parties as contours

Imagine that we let this process continue—at any time, two groupings fuse, to make a single one.  We began with a very simple dyadic alliance between two actors.  Let us now imagine that, faced with other dyadic alliances, one dyad wishes to join with another.  And then one of these alliances will fuse with another alliance, and so on and so forth.  With each iteration, the joint dynamics of suppression and logrolling should lead the ideology to become both more abstract and more complex, respectively.   

We also imagine that the fusing groups are “adjacent” in the social space (that is, there is no third party that is “between” and separating the two).  In many cases, the fusion process will stop far short of two parties, though in single member, “first past the post,” district elections, as Duverger (1963 [1954]) has shown, there is a strong tendency towards a two party solution.  Note that there is no reason to imagine that the resulting groups are simple shapes, such as spheres or cubes.  The precise distribution of persons in this space (whether it is more or less even), like the precise “path dependence” of the historical process that has occurred (which alliances happen first), can lead the emerging alliances to take strange shapes.  Each party, in other words, can be thought of as a contour that snakes through the space in some way.  A party system, similarly, can be understood as the set of contours that divides persons up into a set of mutually exclusive and exhaustive classes.  We have assumed, by the adjacency criterion, that these contours are all single curves, and hence that each party is one continuous shape.

For example, imagine that persons are distributed in some two-dimensional space, though we make no assumptions about the nature of the dimensions (thus there don’t necessarily need to be “two” actual principles organizing people—all that matters is that their pattern of likenesses and differences is one that can be represented in a two dimensional space), and we choose two dimensions only for convenience.  Persons who are near each other in space tend to agree on what they want, and people who are far apart tend to disagree.  Figure 1 presents an example of a party system consisting of two parties.

 

 

Now in this case, we see that the parties seem to be oriented to one dimension (even if there is no clearly namable “thingness” to this dimension, like a degree of some quality), and it would seem very plausible that actors, attempting to understand the logic of the political system, would rely on unidimensionality.  That is, they would speak of others being (for example) “to the right” or “to the left” of them.  In other words, the ideology of a dimension (such as liberal-conservative) would arise as actors’ theory of the principles of their own action.  What would best express their set of alliances is a single dimension (even though, as we have seen, they are actually in a two-dimensional space).

In other cases, however, contours are not drawn such that a “dimensional” understanding seems plausible.  This then leads to a challenge for political actors who need to theorize the logic of their party.  This sort of complexity often arises when parties develop as an agglomeration of smaller clumps, most notably, local parties.  

For example, the contours shown in figure 2 are not compatible with a unidimensional subjective representation of the party system; it is not even possible for them to use something like “moderation” versus “extremism” as would be the case if they were arranged as concentric circles.  How can they come up with any understanding of what joins the members of a party?  In many cases like this, it seems that party members will simply fall back on the issue of whether or not they are currently in power.  Those in power may believe that they are united by their “competence” (which basically means that they are in a position to make decisions, some of which turn out to work reasonably well), while those who are out of power may believe that they are united in terms of their resistance to “tyranny.”[5]  Such allies, should they come into power, may be honestly confused at how they suddenly turn out to always have had antithetical views.

 

 

In sum, this analytic account—one that starts from simplistic and unrealistic preferences—suggests that parties may develop as contours that connect persons who to a great extent vary from one another in their interests and goals.  Although this derivation is fanciful, the resulting image of parties, I contend, is not.  Now we can compare this analytic derivation to a more historically plausible one, regarding the origin of political parties.

Parties from scratch

Now historically, it seems that in those cases in which parties arise “from scratch” (before the development of a democratic institutional infrastructure deliberately designed to channel party formation in particular directions), we still find the development of local structures of opposition, generally based on pre-existing vertical structures, whether kin, land-ownership, or patronage (Barth, 1965; Martin, 2009).  There are then alliances of such local parties across these regions, as elites begin to make arrangements so that they can coordinate against common enemies.  Many party structures then develop as odd assemblages of different groups across regions.  As masses get more involved, and parties begin to appeal to category-based interests of actors (such as class, religion) as opposed to particularistic ones (such as being a dependent of such-and-such an elite family), parties then develop as patchworks of different types of category across different regions.

This patchwork nature is clearest in large countries with two-party systems.  Thus in the United States, the major parties have always been alliances of very different sets of interests—for example, the Democratic party from the late 19^th century through the 20^th involved a coalition between anti-black whites in the South and blacks in the North, farmers in the South and industrial union members in the North.

Thus each party may be understood as an amalgam, a clumping of different groups, an accretion of ties of alliance.  The logic of this amalgam is only partially consistent, as it is equivalent to sailing on a ship that is constantly being changed and rebuilt—some parts are old and no longer of any use, but have not yet been changed, while others are brand new and, though they fit poorly with much of the old, are expected to be guiding the development of the future structure.

This raises a serious practical problem for adherents, namely how to conceptualize the nature of their party and therefore the principles of their political action.  For political action, we recall, is paradigmatically about the favoring of friends.   But the citizen does not possess a roster of other party members, complete with each’s occupation, religion, education, and so on, let alone knowledge of what these others want.  Hence she is faced with the question, who is my neighbor, my ally, anyway?

A structural anecdote

Let me give a (true) anecdote to explain my meaning.  I once saw a pickup truck in my home town that had two bumperstickers on the rear.  One had a representation of the American flag, and words next to it:  “One nation, one flag, one language.”  The other side had the Confederate flag.  This is the flag used by the short-lived Southern confederation of states during the Civil War, when they tried to break away from the Union in order to preserve their “peculiar institution,” that is, slavery of Africans and their descendants.  They wanted there to be two countries, and two flags.  Indeed, the truck itself had two flags on it!  Yet the other sticker emphasized the importance of only having one flag and one country.  This seems, in a way, to be the acme of political inconsistency, and might be understood as demonstrating the complete incapacity of the owner to participate in any sort of meaningful politics.

But quite the contrary, it demonstrated a mastery of the political landscape.  Displaying the Confederate flag in the United States does not imply anti-black racism.  However, it does imply a lack of concern with being “called out” as a racist—it implies fearlessly embracing aspects of American political culture without apology, even though these are associated with racism.  In other words, this flag does not prove racist animus (though racist animus might well be sufficient to produce the desire to display the flag), it does demonstrate anti-anti-racism.  Whether or not it is anti-black, it is certainly anti-Northern-liberal.

The other bumpersticker, however, comes in response to certain political initiatives to ease the barriers to American citizens, residents, and possibly others who read (or speak) Spanish but not English.  Whether it is printing all government documents in Spanish as well as English, offering bilingual instruction in the schools, or printing street and highway signs in Spanish, this movement has been pushed largely by political liberals.  It is opposed both on practical grounds in some cases (for example, the increased cost of equipping schools for multilingual instruction), but also for reasons having to do with the implicit position of different groups in a status hierarchy—whether English speakers lose their implicit priority and their capacity to feel “at home” everywhere.  

What is key is that the Democratic party has tended to get the lion’s share of support from both Blacks and Spanish speakers (with the exception of Cuban refugees), and has pursued the policies that are generally considered advantageous to both.  In placing these two, seemingly contradictory, bumperstickers on his truck, our unknown actor was successfully indicating his opposition to the liberal coalition. (Further, given that his truck’s perfect paintjob demonstrated that it was a stranger to serious work, its pointlessly excessive size also embodied opposition to environmental conservation.)

In sum, the argument here is that what ideology is, is actors’ theorization of their politics, that is, it is their attempt to come up with an abstract representation of the political alliance system in which they are in, and the nature of their opponents.  They may be logically inconsistent, but are politically consistent (and teleologically consistent), when they develop a set of vaguely interrelated themes that help them always orient to their friends in a positive manner, and to their enemies in a negative manner.

Political reasoning in practice

We are now ready to return to the puzzle we started with.  We saw that political reasoning doesn’t follow the logic which was first hoped would characterize an informed, but ideologically driven, populace, namely that (ideological) values + beliefs = opinions.  Instead, we saw that ideology seems to provide people not with values, but with beliefs.  But how does this happen?  It seems that the “knowledge” that ideology gives us is that which would justify our side and strip our enemies of their justification.  

To return to our running example (citizens trying to decide whether to support a policy that would provide assistance to the out of work poor and/or blacks), the classic conception imagines a person beginning with the value of equality, adding the facts about discrimination (say) and producing support for the policy.  But those who oppose the policy do not claim to be any less enthusiastic about the value of equality, and, unless we simply dismiss their protests on the basis of the fact that they reject the policy (a pathological form of science, in which we prove our claim by throwing out any information that doesn’t fit it), we have a puzzle.  That puzzle, of course, is solved by the fact that the conservatives disagree about the world of fact, not that of values—they “know” that the beneficiaries of the program are undeserving.

When we think about it, how do either of our hypothetical ideologues have any information about the worthiness of the poor?  They both get it from the nature of the alliance system in which they are embedded.  The rule is, simply put, “me and my friends are good” and “those others are bad.”  Thus it seems like the actual calculus of opinion formation is “sides + self-concept = opinion.”  

It would be reasonable to object that our allies are not assigned to us at birth; we are free to choose them, and so rather than alliance being the cause of our conceptions, our conceptions may be the cause of our alliance, as we choose our side based on how we evaluate the members of the coalition.  There is no need to deny that this can happen… but there is not much evidence that it is a major contributor to the variance we are examining.  First, party identification is basically assigned to us at birth, in the sense that partisanship correlates highly between parents and their children.

Now to some degree, this comes because other aspects of individuals that are associated with party (region, ethnicity, religion, occupation) are associated across generations.  Yet there is something further about partisan attachments that resists change.  And when people do change their partisanship, they are often aligning the party choice with the rest of their life.  You don’t get a membership card in a left party with your sociology PhD, but you nearly might as well.

And when we choose a side, we find that it does come with alliances already built in.  Not all political actors will necessarily accept that package deal.  But to the extent that they do not, they are hampered as political actors.  That is, the American Democrat who concedes that regulation of handgun sales is unconstitutional, the Republican who admits that opposing abortion rights is unconstitutional, are going to be less impressive fighters for their side than those who have no such doubts.  And that means that the urban Southern democrat, the rich Republican woman (in these cases), may need to figure out how to encompass the programs of their unchosen, and unwished-for, allies.

But more important, if they are to have a true ideology, they must have a coherent theorization of what unites their sides—and it is my argument that this is nothing other than the sides themselves, idealized, universalized, and detached.  This might be plausible to most when they consider the nineteenth century European systems—the “socialism” of the “socialist” party is the worker-intellectual alliance; the “liberalism” of the “liberal” party is the capitalist-tradesman alliance.  My argument is, however, that this is generally true, and that this is how it is possible for ideology to provide people with an understanding of how they should decide more specific issues.

Thus to return one last time to the running example, Martin and Desmond (2010) found that ideologues of high political information were more likely than others to be wrong about the proportion of the US poor who are black, seriously overestimating this.  That is, their knowledge was that-which-helps-us-know- -what-we-want-to-fight-about.  But even more, when they were presented with a vignette that presented them with the ontology which their opponents would have believed to be the case, they paused significantly before answering.  That is, they recognized that some situations are different from others—good for their enemies.  

We saw above that it was difficult to imagine how any person could truly come up with an answer to many of the questions that respondents are presented with in surveys.  How can anyone answer with confidence a question asking us to generalize about the subjective nature of the members of a vaguely defined class?  Yet ideologues do, and they come up with different answers.  But where do they get these from?  If we consider the positions on a continuum from running liberal to conservative as “politicized” in the sense of being oriented toward political conflict, then we may propose that the “knowledge” that comes with an ideological position is that which best facilitates this politicization.  It is not simply that people believe that which furthers their “interests,” although there are undoubtedly tendencies in this direction.  It is that ideology leads people to “put into the world” ontologies that facilitate opinion formation such that they favor allies, and oppose enemies.

 

Conclusion

The classic sociology of knowledge attempted to link the ideas associated with groups, especially large scale strata of society, such as classes, to their position in the overall social structure.  This effort famously suffered from two grave problems. The first is known as the problem of imputation (see Child, 1941), and is largely a technical problem—it is very difficult to know what “the group” actually thinks.  Imagine that we are attempting to determine the ideology of workers.  Do we look at what the leaders of workers’ movements say?  They may be different from other workers (indeed, they may not be workers themselves).  And they may not say what they actually think, but what will accomplish their goals.  And if we use books to determine what they thought, we may be finding things that have to do with the characteristics of texts first and foremost, and less about what was key to the workers’ ideology.

These problems can to some extent be mitigated by the use of survey data, though this has its own serious limitations and interpretive difficulties.  But there has been little enthusiasm for the project of a classical sociology of knowledge using survey data, presumably because of the second problem.  This is the one that Mannheim pointed to, one which we might now call that of the “mutually assured destruction” that came from reducing others’ claims to their position in social structure.  To the extent that the sociology of knowledge became swept up in the quest to “unmask” others—to show that their pious ideals were “really” self-serving, driven by material interests—the analytic tools turned out to be too good.  Even those armed with critique are vulnerable to it.  The all-around critique ends up destroying “man’s confidence in human thought in general” (Mannheim, 1936 [1929]: 45).

This totalizing approach undermined itself and hence was abandoned, even though it had not been demonstrated to be incorrect.  But it may be that the problem was not so much in the logic as the application—the assumption that knowledge in any sphere of activity was rooted in a global position may have been overly convenient, and it could well be that (as Bourdieu [e.g., 1969 / 1966] assumed), the relation between any ideational production and social structure is specific to positon in a particular field.  If so, then political ideology is unlikely to be related to general “class position,” except insofar as this is mediated by alignment with a particular political side, especially a political party.

My argument is that this restricted version of the classical approach is in fact correct, and that this explains characteristics of political ideology that are otherwise obscure:  the fact that its planks are mutually supporting despite the presence of logical contradictions; the importance of prescription despite the valuations expressed being uninterpretable in any literal fashion; the generative nature of ideology despite the fact that what it seems to provide is an ontology.  And, quite elegantly, focusing on the pragmatic difficulties that confront actors as they struggle to make sense of their position in a largely unwished-for web of alliances, we find that the core of Marx’s understanding of ideology is the most reasonable explanation for the resources that actors have to guide their political action.

 

References

Abelson, Robert P., et al.  (1968),  Theories of Cognitive Consistency, Chicago, Rand McNally.         [ Links ]

Adorno, Theodore. W., Else Frenkel-Brunswik, Daniel J. Levinson, and R. Nevitt Sanford  (1950),  The Authoritarian Personality, New York, Harper and Row.         [ Links ]

Arendt, Hannah (1958),  The Human Condition,  Chicago,  The University of Chicago Press.         [ Links ]

Baldassarri, Delia ( 2012), The Simple Art of Voting. The Cognitive Shortcuts of Italian Voters, New York, Oxford University Press.         [ Links ]

Barth, Fredrik (1965),  Political Leadership among Swat Pathans,  London, Berg Publishers.         [ Links ]

Billig, M. (1984), “Political ideology: social psychological aspects”, in H. Tajfel (ed.), The Social Dimension, volume 2, Cambridge, Cambridge University Press, pp. 446-470.         [ Links ]

Bourdieu, Pierre (1969 [1966]), “Intellectual field and creative project”, Social Science Information, 8, pp.189-119.         [ Links ]

Bourdieu, Pierre (1984 [1979]), Distinction.  A Social Critique of the Judgment of Taste, translated by Richard Nice,  Cambridge, MA,  Harvard University Press.         [ Links ]

Breiger, Ronald L. (1974), “The duality of persons and groups”, Social Forces, 53, pp. 181-190.         [ Links ]

Buchanan, James M., and Gordon Tullock (1999 [1958]), The Calculus of Consent. Logical Foundations of Constitutional Democracy, Indianapolis, Liberty Fund, Inc.         [ Links ]

Campbell, Angus, Philip. E. Converse, Warren E. Miller, and Donald E. Stokes (1960), The American Voter, New York, John Wiley and Sons.         [ Links ]

Chhibber, Pradeep, and Ken Kollman (1998), “Party aggregation and the number of parties in India and the United States”, American Political Science Review, 92, pp. 329-342.         [ Links ]

Chhibber, Pradeep, and Ken Kollman (2004), The Formation of National Party Systems, Princeton, Princeton University Press.         [ Links ]

Child, Arthur (1941), “The problem of imputation in the sociology of knowledge”, Ethics, 51, pp. 200-219.         [ Links ]

Converse, Philip E. (1964), “The nature of belief systems in mass publics”, in David E. Apter (ed.), Ideology and Discontent,  New York, Free Press, pp. 206-61.         [ Links ]

Crawford, Jarret, and  Eneda Xhambazi (2013), “Predicting political biases against the Occupy Wall Street and Tea Party movements”, Political Psychology doi: 10.1111/pops.12054.         [ Links ]

Dawson, Michael (2001), Black Visions. The Roots of Contemporary African-American Political Ideology, Chicago, The University of Chicago Press.         [ Links ]

DiMaggio, Paul, John Evans and Bethany Bryson (1996), “Have Americans’ social attitudes become more polarized?”, American Journal of Sociology, 102, pp. 690-755.         [ Links ]

Downs, Anthony (1957), An Economic Theory of Democracy, New York, Harper.         [ Links ]

Duverger, Maurice (1963 [1954]), Political Parties. Their Organization and Activity in the Modern State, New York, John Wiley.         [ Links ]

Eagleton, Terry (1991), Ideology. An Introduction, New York, Verso.         [ Links ]

Federico, Christopher M., and Jim Sidanius, (2002), “Sophistication and the antecedents of whites’ racial policy attitudes”, Public Opinion Quarterly, 66, pp. 145-176.         [ Links ]

Feldman, Shel (1966), Cognitive Consistency, New York, Academic Press.         [ Links ]

Festinger, Leon (1957), A Theory of Cognitive Dissonance, Stanford, Stanford University Press.         [ Links ]

Goren, Paul (2004), “Political sophistication and policy reasoning: a reconsideration”, American Journal of Political Science, 48, pp. 462-478.         [ Links ]

Goren, Paul (2005), “Party identification and core political values”, American Journal of Political Science, 49, pp. 882-897.         [ Links ]

Higgs, Robert (1987), Crisis and Leviathan. Critical Episodes in the Growth of American Government, New York, Oxford University Press.         [ Links ]

Hinich, Melvin, and Michael Munger (1996), Ideology and the Theory of Political Choice, Ann Arbor, University of Michigan Press.         [ Links ]

Holt, Michael F. (1999), The Rise and Fall of the American Whig Party, New York, Oxford University Press.         [ Links ]

Jacoby, William (2006), “Value choices and American public opinion,” American Journal of Political Science, 50, pp. 706-723.         [ Links ]

Jacoby, William, and Paul Sniderman (2006), “The structure of value choices in the American public”, paper presented at the Annual Meeting of the Southern Political Science Association, Atlanta, GA, January 5-7.         [ Links ]

Jost, John (2006), “The end of the end of ideology”, American Psychologist, 61, pp. 651-670.         [ Links ]

Jost, John, Brian Nosek, and Samuel Gosling (2008), “Ideology: its resurgence in social, personality, and political psychology”, Perspectives on Psychological Science 3, pp. 126-36.         [ Links ]

Kerlinger, Fred N. (1984), Liberalism and Conservatism. The Nature and Structure of Social Attitudes, Hillsdale, Erlbaum.         [ Links ]

Klueger, James, and Elliot Smith (1986), Beliefs about Inequality. Americans’ Views of What Is and What Ought to Be, New York, Aldine de Gruyter.         [ Links ]

Kurtz, Kenneth, Dedre Gentner, and Virginia Gunn (1999), “Reasoning”, in Banjamin Martin Bly, and David Rumelhart, Cognitive Science, San Diego, Academic Press, pp. 145-200.         [ Links ]

Lane, Robert E. (1962), Political Ideology,  New York, Free Press.         [ Links ]

Lane, Robert E. (1973), “Patterns of political belief”, in Jeanne Knutson (ed.), Handbook of Political Psychology,  San Francisco, Jossey Bass, pp. 83-116.         [ Links ]

Lau, Richard, Richard Smith, and Susan Fiske (1991), “Political beliefs, policy interpretations, and political persuasion”, The Journal of Politics, 53, pp. 644-675.         [ Links ]

Lipset, Seymour Martin (1960), Political Man, Garden City, New York, Doubleday.         [ Links ]

Mannheim, Karl, (1936 [1929]), Ideology and Utopia, translated by Louis Wirth and Edward Shils, New York, Harcourt Brace Jovanovich.         [ Links ]

Martin, John Levi (2009), Social Structures, Princeton, New Jersey, Princeton University Press.         [ Links ]

Martin, John Levi (2011), The Explanation of Social Action, New York, Oxford University Press.         [ Links ]

Martin, John Levi (2014), Thinking Through Theory,  New York,  Norton.         [ Links ]

Martin, John Levi, and Matthew Desmond (2010), “Political position and social knowledge”, Sociological Forum, 25, pp. 1-26.         [ Links ]

Marx, Karl, and Frederick Engels (1976 [1845-6]), The German Ideology, in Collected Works, volume 5, New York, International Publishers.         [ Links ]

Minsky, Marvin (2006), The Emotion Machine. Commonsense Thinking, Artificial Intelligence, and the Future of the Human Mind, New York, Simon and Schuster.         [ Links ]

Mische, Ann (2009), Partisan Publics. Communication and Contention across Brazilian Youth Activist Networks,  Princeton,  Princeton University Press.         [ Links ]

Peffley, Mark A., and Jon Hurwitz (1987), “A hierarchical model of attitude constraint”, American Journal of Political Science, 29, pp. 871-890.         [ Links ]

Riker, William H. (1982), Liberalism Against Populism, San Francisco, W. H. Freeman.         [ Links ]

Rokeach, Milton (1968), Beliefs, Attitudes, and Values. A Theory of Organization and Change, San Francisco, Jossey-Bass.         [ Links ]   

Rumelhart, David (1989), “Toward a macrostructural account of human reasoning”, in Stella Vosniadou, and Andrew Ortony, Similarity and Analogical Reasoning, New York, Cambridge University Press, pp. 289-312.         [ Links ]

Schmitt, Carl (2008 [1932]),  The Concept of the Political. Expanded Edition,  Chicago,  The University of Chicago Press.         [ Links ]

Simmel, Georg (1958 [1923]),  Soziologie, in Collected Works, vol. II, Berlin, Duncker and Humblot (4th edition).         [ Links ]

Sniderman, Paul M., Richard Brody, and Philip E. Tetlock (1991),  Reasoning and Choice. Explorations in Political Psychology, Cambridge, Cambridge University Press.         [ Links ]

Tedin, Kent L. (1987),  “Political ideology and the vote”, Research in Micro-Politics, 2, pp. 63-94.         [ Links ]

Wacquant, Loïc (1999),  “The double-edged sword of reason”, European Journal of Social Theory, 2, pp. 275-281.         [ Links ]

Wacquant, Loïc (2002), “Scrutinizing the street: poverty, morality, and the pitfalls of urban ethnography”,  American Journal of Sociology, 107, pp. 1468-1532.         [ Links ]

Zaller, John R. (1992),  The Nature and Origins of Mass Opinion,  Cambridge, Cambridge University Press.         [ Links ]

 

Received: August, 30, 2014 Approved: October, 31, 2014

 

Notas

[1] Some of the ideas here were first worked out in collaboration with Matt Desmond, and I am grateful to him for continuing dialogue, friendship and inspiration.

[2] This often involves the idea that certain issues are “owned” by certain parties, in contrast to issues for which multiple parties can each make different, but still plausible, claims to have a valid solution. The party that “owned” the issue associated with the rejection of the party-in-power is expected to profit from this rejection more than other parties.

[3] “Everyone believes his craft to be the true one. Illusions regarding the connection between their craft and reality are more likely to be cherished by them because of the very nature of the craft....The judge, for example, applies the code, he therefore regards legislations as the real, active driving force” (notes made by Marx in the manuscript at the end of the Feuerbach section, in 1976 [1845-6]: 92).

[4] The term comes from woodsmen’s practice of helping each other roll their felled logs from one place to another, a difficult act for one alone.

[5] A very good example of this is the American Whig party of the 1840s-50s. Composed of “out” factions across all regions of the United States, and, due to recent changes in patterns of international economics, suddenly bereft of any coherent economic ideology that might unite them, they became honestly convinced that they had a moral objection to “tyranny,” which for them simply meant continual Democratic dominance of the national government. See Holt (1999).