Definition.

At one level, Gallie’s definition is brief and clear: essentially contested concepts “inevitably involve endless disputes about their proper uses on the part of their users.”Footnote ²⁰ It is this definition per se, rather than any of the seven criteria, which explicitly states that the concepts are, in fact, contested.

Along with this concise formulation, Gallie introduces two additional elements that make the definition more complex, and that must be sorted out. First, he offers a distinction between concepts that are “contested,” as opposed to “contestable.”Footnote ²¹ With the latter, one might expect contestation, yet it is not occurring at the time the concept is analyzed. This outcome potentially reflects what will be discussed later as “decontestation,” and it complicates the task of distinguishing contested from noncontested concepts.

Some commentators on Gallie likewise employ the term “contestable,” and in one instance they intend the same distinction that Gallie introduces. In other instances, it is not clear that this is the case, but their labeling nonetheless points in the same direction and again suggests the difficulty of distinguishing contested from noncontested concepts.Footnote ²² To use one criterion from Gallie’s own framework, we would therefore argue that the category of “contested concepts” is “open.”

A second complicating element is that Gallie at one point refers to Criteria I to V as “establishing the formally defining conditions of essential contestedness.”Footnote ²³ Yet, as will be demonstrated later in the discussion of specific concepts, these criteria may or may not be met even for concepts widely recognized as contested. We find it more productive to think of his criteria not as standing in a well-established definitional relationship to one another, but rather in a relationship that varies somewhat with the distinct constellations of features that characterize each concept.

In conjunction with these contrasting perspectives on definition, it is productive to ask, overall, how many concepts are encompassed in alternative understandings of Gallie’s idea of essentially contested concepts. Jeremy Waldron, on the one hand, expresses concern that “the use of this term has run wild.”Footnote ²⁴ On the other hand, Gallie offers a surprisingly – indeed, unconvincingly – restricted list in one of his summary statements, where he excludes such major concepts as science, law, liberty, and government.Footnote ²⁵ Between these extremes, we would reiterate that the category of essentially contested concepts is open, and that the application of Gallie’s definition and his framework should be governed by whether they yield useful insight into the concept at hand.

Hypothesis versus framework.

Up to this point the term “framework” has been used to characterize Gallie’s overall set of arguments. However, early in his article Gallie refers to his approach as an “explanatory hypothesis,”Footnote ²⁶ and he employs different forms of the term “explain” as he summarizes his approach.Footnote ²⁷ The assertion that Gallie offers a hypothesis is repeated by some authors who have discussed his contribution.Footnote ²⁸

The idea of a hypothesis or explanation is certainly relevant in some respects. One can hypothesize that appraisiveness, internal complexity, and diverse describability partially explain the contestation of these concepts. However, the idea of a hypothesis may in part be misleading because it implies, overall, that Gallie’s approach is either right or wrong. It can also raise the expectation that all of his seven criteria should prove correct, both for the analysis of a particular concept and for analyzing concepts in general.

For this reason, it appears more productive to label Gallie’s set of ideas as an analytic framework – that is, a set of interrelated criteria that serve to illuminate important problems in understanding and analyzing concepts. Gallie elsewhere adopts similar terminology by calling his approach a set of “semi-formal conditions” and a “schematization.”Footnote ²⁹ Like any framework or schematization, it should probably be judged by its overall utility, and not by whether one element or another appears to be correct – as would be suggested by the idea of a hypothesis. Indeed, it might be implausible to expect that all elements are salient or correct all of the time.

Domain. Gallie delimits the domain of his concerns, arguing that his discussion is focused on the role of essentially contested concepts in philosophical and scholarly thinking, and not in real-world politics.Footnote ³⁰

Commentators, however, extend their focus beyond this domain to encompass the world of real politics. Waldron analyzes the “rule of law” as it emerged as an issue in the contested 2000 presidential election in the US.Footnote ³¹ Robert Grafstein views the political character of some concepts as crucial to understanding forms of contestation. Michael Freeden sees the domains of real politics and scholarship as tightly interrelated, but he also makes a key distinction between them: The prospects for resolving conceptual disputes, he believes, are greater in the latter. For present purposes, Gallie’s framework will be treated as relevant to both domains.Footnote ³²

I. Appraisiveness.

An essentially contested concept “signifies or accredits some kind of valued achievement.”Footnote ³³ As Gallie puts it, the term that refers to the concept is an “achievement word.”Footnote ³⁴ And indeed, it is highly plausible that the positive normative valence attached to these concepts is important in spurring debates over their meaning. Two paradigmatic examples of this positive valence offered by Gallie are democracy and social justice. With regard to democracy, Gallie states that:

Most commentaries on Gallie do not dispute that the concepts under discussion are appraisive, yet they introduce three clarifications that usefully extend Gallie’s discussion. First, William Connolly underscores the fact that designating a political system as a democracy is both to “ascribe a value to it,” but also very crucially to describe it. Freeden makes this same point by observing that along with the complex issues of normative appraisal, essentially contested concepts do indeed refer to “brute facts.”Footnote ³⁶ Second, appraisiveness encompasses not only positive valuation but also negative valuation, a point with which Gallie would probably not have disagreed, but that he simply failed to discuss.Footnote ³⁷ Finally, for some major concepts the normative valence may be unclear and may depend on the theoretical framework employed, or on the specific context in which the concept is applied.Footnote ³⁸

Beyond these specific arguments about appraisiveness, we would reiterate the observation made in the introduction. Working with Gallie’s framework does indeed present an opportunity to explore the normative component of political research, at the same time that this exploration is linked to a spectrum of other issues concerning the formation and application of concepts.

II. Internal complexity and III. Diverse describability.

These two criteria are tightly interrelated:Footnote ³⁹ the internal complexity of a concept makes it plausible that different users may view, or describe, its meaning in different ways. In fact, this outcome may or may not occur – hence the label “describability,” rather than description. This distinction is parallel to the contrast drawn by Gallie (noted earlier) between contested and contestable.

Gallie summarizes these two criteria as follows. With an essentially contested concept, the “achievement must be of an internally complex character.” It includes a variety of possible components or features – although “its worth is attributed to it as a whole.” Further, the concept is “initially variously describable,” with the consequence that, a priori, “there is nothing absurd or contradictory” in the existence of alternative meanings.Footnote ⁴⁰ Diverse describability may involve an exclusive emphasis on one or another facet of the concept. Alternatively, as Gallie points out, different facets may be emphasized to varying degrees, involving contrasting relative importance.Footnote ⁴¹

These two criteria, as with appraisiveness, are illustrated in reference to the concepts of social justice and democracy. Social justice is variously understood, according to Gallie, in terms of a liberal conception, concerned with individual freedoms; and what might be thought of as a social democratic conception, concerned with providing to the population “the necessities of a worthwhile human life, and the distribution of products to assure such a life.”Footnote ⁴² Like justice, democracy has multiple internal components. Three of those emphasized by Gallie are majority rule, equality in the right of all citizens “to attain positions of political leadership and responsibility,” and “the continuous, active participation of citizens in political life at all levels, i.e. … self-government.”Footnote ⁴³ This internal complexity in turn makes it likely that different users of the concept will characterize – or describe it – in different ways.

Among the commentators, Grafstein explicitly endorses Criteria II and III as highly salient.Footnote ⁴⁴ Some authors recognize their importance, but also emphasize that the salience of these criteria can be mitigated. Thus, as will be discussed in the next section, Christine Swanton maintains that it is possible to judge some meanings as better than others,Footnote ⁴⁵ Norman S. Care offers the idea of “practical closure,”Footnote ⁴⁶ and Freeden develops the idea of “decontestation.”Footnote ⁴⁷ All of these could mitigate diverse describability. Nonetheless, there is little overall disagreement that this criterion (diverse describability), as well as Criterion II (internal complexity), are important.

We would underscore a further point about internal complexity. The claim that a concept is internally complex is meaningful only if the different components are indeed part of the same concept. The idea of a “cluster concept” is sometimes evoked to underscore the claim that these components do, in fact, belong together in a single concept. If, by contrast, the concept is “over aggregated” – that is, it brings together elements that are only loosely related to one another – then it is appropriate to “disaggregate” the concept, which may eliminate, or drastically reduce, internal complexity. The decision to carry out disaggregation is complex, and in the extended example provided later that is focused on the concept of democracy, we do not consider it justified. But in discussions of Gallie’s Criterion II (internal complexity), the possibility that one is working with an over-aggregated concept must be recognized.

IV. Openness.

Essentially contested concepts are viewed as open in their meaning, that is, subject to periodic revision in new situations. The “accredited achievement must be of a kind that admits of considerable modification in the light of changing circumstances; and such modification cannot be prescribed or predicted in advance.”Footnote ⁴⁸ In discussing the openness associated with the concept of democracy, Gallie observes that

Further, Gallie discusses how advocates’ preferred conceptualization interacts with the changing context. Conceptual competitors must consider altered circumstances when defending their own version. New circumstances may be unfavorable to dominant conceptualizations, but advocates may effectively adapt to preserve their preferred conceptualization.Footnote ⁵⁰

Most commentators on Gallie have affirmed the importance of openness. Regarding the overall significance of this criterion, Alasdair MacIntyre views openness or “essential incompleteness” as a key factor that produces essential contestability. Nonetheless, various commentators argue that openness may sometimes be superseded.Footnote ⁵¹ Norman Care points to the possibility of “practical closure,” or “temporary closure,” that may be achieved even in the absence of absolute, objective solutions to debates about concepts that are essentially contested. Even if practical closure does not preclude periodic revisions, it nevertheless suggests that “at least for a time,” openness can be partially overcome. Indeed, without such closure, we cannot “answer for ourselves certain of the basic questions about the character of our institutions and practices.”Footnote ⁵² Relatedly, Gray argues that taxonomies developed within well-defined scholarly frameworks may not be characterized by openness.Footnote ⁵³

With the idea of “decontestation,” Freeden offers a valuable perspective on openness. Just as concepts can be contested, they can also be decontested, in that they achieve a stable meaning within a given framework. Freeden introduces this idea of decontestation in his analysis of “ideologies,” which can be understood both as constellations of ideas that are ideologies as conventionally found in the real world of politics, and also as interrelated systems of meaning that are the conceptual frames of scholarly usage.Footnote ⁵⁴ Freeden carefully presents the idea of decontestation as an element within his approach to analyzing ideologies. Specifically,

Beyond Freeden’s specific formulation, in discussions of contested concepts the idea of decontestation can usefully be understood in a broader sense as well, involving a variety of circumstances under which contestation may be reduced or mitigated. This broader usage will occasionally be followed later.

V. Reciprocal recognition.

This criterion presumes that contending parties acknowledge the concept’s contested character. To some extent, they recognize their adversaries’ use of the different facets of the concept – for example, of social justice and of democracy as discussed earlier – that guide their respective applications. Thus,

In subsequent commentaries, some scholars have accepted the idea of reciprocal recognition. Kenneth Smith finds that disputants may recognize and accept that they are using similar concepts in different ways, involving “mutually contested concepts,” rather than essentially contested concepts. Grafstein argues that the political character of certain concepts leads to an explicit contest among rival users.Footnote ⁵⁷

Others, however, have questioned the relevance of this criterion. Freeden argues that it is irrelevant to essential contestedness, because analysts often employ concepts in a way that may simply differ from another usage, without framing their usage vis-à-vis an alternative meaning.Footnote ⁵⁸ Proponents of a particular conceptualization may not explicitly acknowledge contending variants of the concept, thus violating this criterion. Rather than question a concept’s status as essentially contested, scholars seeking to apply Gallie’s framework should recognize that this criterion is not always pertinent.

VI. Exemplars.

The role of exemplars in Gallie’s framework has generated much confusion. This is due, in part, to his own terminology and to inconsistencies in his presentation. In fact, he employs this idea both narrowly – in the sense of an original exemplar – and broadly. To better encompass these diverse meanings, in this and subsequent sections we employ the label “exemplars,” recognizing that Gallie’s label in his own text is “original exemplar.”

In Gallie’s narrower understanding of exemplars, the contested concept is seen as anchored in an original exemplar whose “authority is acknowledged by all the contestant users.”Footnote ⁵⁹ The link to the original exemplar plays a crucial role in allowing analysts to distinguish between essentially contested concepts and confused concepts. Confused concepts involve disagreements in which the same term refers to two different ideas. The original exemplar anchors the concept, and the issue is therefore not a simple matter of confusion, but rather contestation over the same concept. This idea of a specific exemplar is reinforced by his use of the singular article – that is, “an” exemplar – as if it were indeed one instance;Footnote ⁶⁰ and also by his specific reference to the French Revolution.Footnote ⁶¹

Gallie also presents a broader understanding of exemplars. He notes the “internally complex and variously describable” nature of the exemplar, arguing “it is natural that different features in it should be differently weighted by different appraisers.” And to this he adds that “acceptance of the exemplar’s achievement must have that ‘open’ character which we have ascribed to every essentially contested concept.”Footnote ⁶² Thus, according to Gallie, an exemplar can assume many different forms, including “a number of historically independent but sufficiently similar traditions,” and he makes clear that “the vagueness of this tradition in no way affects its influence as an exemplar.”Footnote ⁶³

Steven Lukes advances an understanding of exemplars that is parallel to Gallie’s broader framing, in that Lukes focuses on multiple, specific instances that constitute the concept’s common core. With regard to power, the concept he analyzes and describes as essentially contested, there are “standard cases of the possession and exercise of power about which all will agree.”Footnote ⁶⁴ Lukes mentions several paradigmatic examples of power, including the rule of master over slaves, the political influence of US steel in the 1960s, and the fictional dystopia of Aldous Huxley’s Brave New World.Footnote ⁶⁵

Critics who have focused on Gallie’s narrower framing of the exemplar have understandably expressed concern about his approach. Ernest Gellner, for instance, claims that “Gallie is, implicitly, betraying his own idea: he talks as if, behind each essentially contested concept, there was, hidden away in some platonic heaven, a noncontested, unambiguously defined and fully determinate concept or exemplar.”Footnote ⁶⁶ Similarly, Freeden criticizes the narrow version of Gallie, arguing that “the postulation of such an exemplar is in effect inimical to the very notion of essential contestability, as it presumes an agreed or correct position from which deviations have occurred.”Footnote ⁶⁷

Other authors in effect acknowledge Gallie’s broader understanding of the exemplar; but this leads them to express concern that if the exemplar is indeed internally complex, variously describable, and “open,” it is unlikely that all parties will acknowledge its authority.Footnote ⁶⁸ It is therefore unclear how it functions as an exemplar. Whereas the challenge that basically focuses on the narrow understanding misses an important part of Gallie’s argument, this second critique is more telling.

A useful response is to return to Lukes’ view: The common core is centered on multiple, paradigmatic examples that do, in fact, anchor the concept. Correspondingly, as noted earlier, “exemplars” is used in the present article to label this component of Gallie’s argument.

VII. Progressive competition.

Gallie argues that a consequence of ongoing conceptual disputes “might be expected to be a marked raising of the level of quality of arguments in the disputes of the contestant parties.”Footnote ⁶⁹ In fact, as with the role of exemplars just discussed, Gallie offers a narrower and a broader understanding of progressive competition. In the narrower version, this criterion specifically involves achieving more complete agreement about the original exemplar.Footnote ⁷⁰ He thus focuses on whether “continuous competition … between contestant users of the concept … enables the original exemplar’s achievement to be sustained and/or developed in optimal fashion.”Footnote ⁷¹ Given the issues just raised about the original exemplar, the meaning of this stipulation is ambiguous.

The broader framing, which omits reference to the original exemplar, appears more promising.Footnote ⁷² Gallie admits that “a general principle may be unobtainable for deciding, in a manner that would or might conceivably win ultimate agreement, which of a number of contested uses of a given concept is its ‘best use.’” Nonetheless, “it may yet be possible to explain or show the rationality of a given individual’s continued use (or in the more dramatic case of conversion, his change of use) of the concept in question.”Footnote ⁷³

Some commentators defend the idea of progressive competition. As noted earlier, N. Garver strongly praises Gallie for his effort to provide order and structure to a particular sort of adversarial discourse.Footnote ⁷⁴ Swanton maintains that despite the internal complexity and diverse describability of contested concepts, it is in fact possible to judge some meanings as better than others, and hence possible to move beyond relativism.Footnote ⁷⁵ Norman Care argues that practical closure benefits social inquiry, and he supports the view that achieving shared understandings of social science concepts does indeed contribute to conceptual coherence in research.Footnote ⁷⁶ For Freeden, of course, decontestation occurs within specific ideologies or conceptual frameworks.Footnote ⁷⁷

Ian Shapiro, while implicitly rejecting Gallie’s overall approach, does point to an avenue for the progressive clarification of concepts. Shapiro is concerned that Gallie’s framework can only produce endless conceptual debates. The focus should instead be on “substantive interdisciplinary knowledge” of the domains to which the concepts apply. For example, the concept of “freedom” must refer concretely to “enabling and restraining conditions” that shape the degree of latitude, or freedom, that individuals experience in their lives. Correspondingly, “many, if not most, of the politically charged questions about freedom” will be substantive, empirical ones that cannot be resolved without this interdisciplinary knowledge.Footnote ⁷⁸ Progressive clarification – which is, to reiterate, the goal of Criterion VII – should be possible, but it requires attention to these empirical issues, and not simply to concept analysis.

By contrast, other scholars strongly question Gallie’s claim that conceptual disputes can lead to an improved quality of argumentation. They express concern that opposing positions may be poorly argued and outside the bounds of reasonable discussion. Eugene Garver writes that “there are some – perhaps too many – cases in which dignifying one’s opponent by treating opposition as competition over an essentially contested concept would be foolish. Charity in interpretation is not an unconditional duty.”Footnote ⁷⁹ Garver goes on to quote Aristotle’s argument that addressing such opponents “can only result in a debased kind of discussion.”Footnote ⁸⁰ Similarly, Freeden maintains that conceptual debates may be of poor quality and they may impoverish rather than enrich the contested concepts. Hence, these debates may be regressive and part of the concept’s meaning may be lost or abandoned.Footnote ⁸¹

Gray takes the discussion in a related but somewhat different direction, finding that a belief in progressive competition conflicts with the normative source of disputes, that is, the criterion of appraisiveness. For Gray, the strong normative resonance of essentially contested concepts results in divergent, intractable positions and values, precisely those not conducive to standardization or harmonization through continued debate.Footnote ⁸²

We would add that whether progressive competition is possible and valuable must be assessed in part on the basis of concrete examples. In the extended example presented on the rule of law, there is some division among commentators, but Waldron makes a strong case that progressive competition has occurred. The extended discussion of democracy considers a related pattern of progressive cooperation among scholars. These examples illuminate Gallie’s Criterion VII, which on another level can and should also be evaluated in terms of more generic criteria.

Fitting the definition: An essentially contested concept? Gallie’s own discussion of democracy certainly suggests that it fits the category of an essentially contested concept, and there would appear to be little disagreement on this point. However, this opinion is not unanimous. One interesting dissent is presented in David Beetham’s outstanding book on Defining and Measuring Democracy. Focusing on alternative definitions, Beetham writes of the

Although this chapter of Beetham’s book is carefully argued, and though it is certainly possible for the idea of contested concepts to be overused,Footnote ⁸⁵ we would question this assessment. A substantial degree of decontestation (in the broader sense) has occurred in current research on democracy, as will be emphasized later. Yet there has been great change in meaning and contestation over time, and there is no reason to assume that such contestation will not arise again in the future. Thus, as originally emphasized by Gallie, democracy appears to stand as a contested concept.

I. Appraisiveness.

Gallie was very much on target in calling democracy “the appraisive concept par excellence.”Footnote ⁸⁶ Major works on democracy by Robert Dahl and Giovanni Sartori, as well as Guillermo O’Donnell and Philippe Schmitter,Footnote ⁸⁷ begin with strong statements that a central motivation in studying democracy derives from its positive normative valence.Footnote ⁸⁸ But the appraisal is not always positive. Huntington and Rosanvallon have convincingly traced the varying normative assessment of democracy, pointing to a negative valuation in the West during important parts of the nineteenth century and at key points in the twentieth century.Footnote ⁸⁹

II. Internal complexity and III. Diverse describability.

As employed in the comparative politics literature, democracy meets Criteria II and III. The complexity of the concept comes out clearly, for example, in O’Donnell and Schmitter’s four dimensions of liberalization, encompassing (1) civil liberties; (2) democratization, in the specific sense of meaningful electoral competition with universal suffrage, and fair and open elections; (3) democratization of social institutions and economic processes; and (4) democratization in the sense of the extension to citizens of substantive benefits and entitlements.Footnote ⁹⁰

In order to label different combinations of these four postulated dimensions, scholars often create new terms. Many have qualified the noun democracy, yielding a proliferation of “democracy with adjectives.”Footnote ⁹¹ In O’Donnell and Schmitter’s analysis, examples of democracy with adjectives that reflect these four dimensions include political democracy, limited political democracy, popular democracy, social democracy, and socialist democracy.Footnote ⁹²

Analysis of these dimensions raises again the issue of disaggregation. How do we know that they are components of the same concept? Perhaps these are really four different phenomena that should be conceptualized separately; that is, the liberalization of national political regimes, the democratization of these regimes, the democratization of social institutions and economic processes at a more microlevel, and the extension to citizens of substantive benefits and entitlements. In the literature on national political regimes, it has certainly been argued that concepts may become more analytically useful when they are disaggregated.Footnote ⁹³ Perhaps it is most productive to conceptualize these components separately, rather than as aspects of democracy.

We would argue that disaggregation is often useful, but this does not undermine Gallie’s overall argument as applied to the body of research under discussion here. This literature is indeed about democracy, and the scholars and political actors of concern here do indeed think of the struggles in which they are involved as being about democracy. Of course, this concept does have multiple components. Some components are emphasized more than others by certain authors, and some components may appear to be more or less relevant in particular contexts. But these are routinely conceptualized as facets of democracy.

IV. Openness.

Here again, as noted earlier, Gallie’s language is vivid: “democratic targets will be raised or lowered as circumstances alter, and democratic achievements are always judged in the light of such alterations.” In the field of comparative politics, carefully analyzing diverse conceptualizations of democracy across contexts, and establishing analytic equivalence, is a challenge of wide concern.

A central aspect of openness is change over time in the political, economic, and social systems being compared. The fundamental shifts in the meanings and normative valence of democracy in the West, traced by Rosanvallon and Huntington through the nineteenth and twentieth centuries, are substantially linked to the emergence over many decades of a middle class, the rise of industrial society, and the appearance of the working class as a major political actor. In the second half of the nineteenth century, for example, the term democracy was commonly used in characterizing regimes that maintained property requirements for voting, that did not encompass the working class in the electoral arena, and that – given exclusion of women from voting – certainly lacked universal suffrage.

Another example is the challenge of comparing democracy in the late nineteenth century and the latter part of the twentieth century. The specific arguments examined in Ruth Berins Collier’s study of the working-class role in democratic transitions and in Bruce Russett’s study of the democratic peace hypothesis undertake comparisons of democracies and democratic transitions in the two eras.Footnote ⁹⁴ By the standard of the late twentieth century, there were no democracies in Europe in the late nineteenth century – in part because of the limited suffrage in the earlier period. Yet these scholars identify cases in this earlier period that are appropriate to include in the comparison, given the frameworks employed by these analysts. Rather than apply the same standard for democracy in both periods, these authors focus on whether a country is democratic by the norms of that historical period. Recognizing that empirically democracy has a different meaning in the two eras, they adapt their comparison accordingly.

Openness is also an issue in comparisons among countries. One example arises in the analysis of some Latin American regimes that might initially have been presumed democratic. However, owing to the continuing power of the military, the traditional oligarchy, or external economic and political actors, it appears that the elected government did not to a reasonable degree have authority to rule. In light of this issue, some scholars have revised their definition of democracy, establishing a new cut-point according to which these countries are not treated as such.Footnote ⁹⁵

In all of these examples, “democratic targets” have indeed been adjusted to changed circumstances, to state this in Gallie’s own language.

V. Reciprocal recognition.

The relevance of this criterion is illustrated in both the political and scholarly domains. In the political sphere, a suggestive example is the debate between the government and the opposition in Pinochet’s Chile. Pinochet called Chile a “protected democracy,”Footnote ⁹⁶ a label intended to characterize the Chilean regime as far less harsh than was widely believed. This label was obviously adopted in recognition of the larger debate on democracy at that time, in Latin America and beyond. By contrast, both domestic and international opponents of Pinochet (including the pope) referred to the regime as “dictatorship.”Footnote ⁹⁷ Pinochet, in turn, strongly protested such descriptions. He “railed at critics who accused him of operating a ‘fascist dictatorship.’”Footnote ⁹⁸

Later, with the end of the Pinochet period, some commentators were concerned that the military retained excessive influence, to a degree that impugned the democratic character of the new regime. Some employed Pinochet’s label, calling Chile a “protected democracy.”Footnote ⁹⁹ Of course, Pinochet’s use of this term conveyed his view that between 1973 and 1990 Chile was closer to being a democracy than critics recognized. By contrast, in the post-1990 period, analysts used this same term to underscore their view that Chile was further from being a democracy than some commentators maintained. Notwithstanding the extreme conflict that characterized the Pinochet period, opponents were thus to some extent able to recognize their adversaries’ use of the concept.

In the sphere of scholarship, reciprocal recognition may be found in a context as obvious as book reviews. Reviews of O’Donnell and Schmitter’s volume debated their use of a procedural definition of democracy.Footnote ¹⁰⁰ Arthur MacEwan raises standard concerns about alternative meanings of democracy, arguing that it is essential to develop a more complete conceptualization and analysis of the interconnections between political democracy, on the one hand, and economic and social democracy, on the other.Footnote ¹⁰¹ Daniel Levine points to similar themes, and further argues that the procedural conceptualization of democracy is too much influenced by a concern with promoting transitions to democracy, and insufficiently concerned with recognizing that the meaning of democracy depends centrally on the ideals and struggles of those who promote it.Footnote ¹⁰² Revisiting the O’Donnell–Schmitter text, it is evident that they had considered these alternative conceptualizations, but chose not to adopt them.Footnote ¹⁰³ Thus, rather than talking past one another, these analysts are to some degree making choices among a mutually recognized set of alternatives.

VI. Exemplars.

The idea of exemplars, in its narrower version of the original exemplar discussed earlier, on occasion is relevant to the concept of democracy. Athenian democracy is evoked in discussions of the New England town meeting,Footnote ¹⁰⁴ and in the current debate in the US on deliberative democracy.Footnote ¹⁰⁵ At certain points in the evolution of French political thought, Athens has been a relevant exemplar, although much of the time it has not.Footnote ¹⁰⁶

The widely discussed international diffusion of democracy certainly reflects the role of exemplars in the broader sense – although the impact of these exemplars must be demonstrated through careful empirical assessment.Footnote ¹⁰⁷ The idea of “waves” of democracy illustrates a global version of this perspective.Footnote ¹⁰⁸ On the other hand, many relevant exemplars have been proximate to the cases under consideration – temporally or spatially, or in terms of established patterns of international communication. Thus, presidential democracy in the US has certainly been an important exemplar in Latin America, as has British parliamentary democracy for many former British colonies.

Not only established democracies, but also transitions to democracy, can be exemplars. Schmitter, building on Laurence Whitehead’s analysis, observes that

For instance, in Latin America in the 1980s, the prior democratization of Portugal, Spain, and Greece was certainly an inspiration. And one can find negative as well as positive exemplars. The difficult and prolonged democratic transition in Brazil, extending from the 1970s well into the 1980s, may have discouraged those promoting democratization in Argentina, and possibly also in Chile. Further, while many political actors in South America – as well as scholars concerned with the region – found much to admire in the Nicaraguan Revolution launched in 1979, they specifically did not see it as the appropriate model for South America. They were convinced that following the Nicaraguan path had a high likelihood of triggering a regression to the military authoritarianism from which they were trying to escape. For this reason, Nicaragua was a negative exemplar.

VII. Progressive cooperation as a variant of progressive competition.

Progressive competition entails the idea that reciprocal recognition among contending users of a concept may possibly lead to improving the quality of arguments. Among Gallie’s critics, this criterion has been met with substantial skepticism, with a number of commentators convinced it does not occur.

This section discusses a particular – and possibly more plausible – form that progressive competition may take, which we will call progressive cooperation. Within the democracy literature this pattern encompasses some sharp shifts in focus, but maintains continuity in analytic concerns on a broader level. In other words, there is sufficient continuity that the concept may be seen as decontested (in the broader sense), yet with important shifts in meaning and application. This is in effect a case of cooperation among successive scholars – of a kind commonly interpreted as reflecting progress in scholarship.

This trajectory of progressive cooperation may be seen as inaugurated with the idea of a procedural definition of democracy, introduced by Schumpeter. He criticizes what he calls the “classical doctrine of democracy” – based on the idea of the common good and the will of the people. He asserts that “there is no such thing as a uniquely determined common good,” and the idea of “the will of the people … vanishes into thin air.”Footnote ¹¹⁰ At a key point in the argument, he discusses major negative political consequences of these vague conceptions – consequences that were evident in the rise of totalitarian regimes in prior decades. As an alternative, he proposes a definition based on “competition for political leadership.” “The democratic method is that institutional arrangement for arriving at political decisions in which individuals acquire the power to decide by means of a competitive struggle for the people’s vote.”Footnote ¹¹¹

In a subsequent step in this trajectory of progressive cooperation, Dahl makes a similar argument, although its components fit together in a different way. He considers the ideal situation of a regime that is “perfectly or almost completely responsive for all its citizens,”Footnote ¹¹² which is in part parallel to Schumpeter’s idea of “the will of the people” as a criterion for democracy according to the classical theory. However, rather than simply delinking the term democracy from that meaning, Dahl proposes that this hypothetical, ideal situation be attached to the term democracy. He uses the term polyarchy in referring to his procedural definition – a definition that is far more elaborate and detailed than the one proposed by Schumpeter. Thus, Dahl provides a label that retains within his framework this ideal, if unattainable, form of democracy, while at the same time developing a more complete procedural definition. Indeed, this latter definition is sufficiently concrete and detailed that scholars have subsequently operationalized it as a cross-national measure of democracy.Footnote ¹¹³

In a further refinement within the procedural tradition, O’Donnell and Schmitter seek to map out the political alternatives open to potentially democratizing regimes in Latin America and Southern Europe in the 1970s and 1980s. They explore how different types of democracy can be understood in terms of their four dimensions noted earlier – two involving procedural issues and two involving substantive outcomes. Within this framework, they propose a particularly clear and usable typology based on these dimensions, advancing their analytic goal of privileging the procedural side of democracy by using the label “political democracy.”Footnote ¹¹⁴

A further step in this sequence has emerged following the period of initial transitions to democracy and corresponds to a widely held set of concerns with the quality and deepening of democracy. Within this body of work, O’Donnell acknowledges the importance of Dahl’s earlier procedural definition as a standard against which regimes should be judged. Yet he opens the discussion to a wider understanding of democracy, involving equality before the law and the protection of democratic citizenship under the rule of law. These can certainly be understood as democratic “procedures,” and in that sense they are squarely within the procedural tradition. At the same time, however, they are outcomes that the governments embedded within democratic regimes can produce through public policy.

The final step considered here, in this trajectory of progressive cooperation, concerns the distinction between access to power and the exercise of power. This distinction is intended to provide new leverage in analyzing issues of the democratic state and the rule of law just discussed. Sebastian Mazzuca suggests that the basic line of argument among these several authors focuses on access to power,Footnote ¹¹⁵ yet O’Donnell’s research on the rule of law under the democratic state is about the exercise of power. The idea of access to power fits extremely well within standard conceptions of procedural democracy. By contrast, Mazzuca proposes a very different conceptualization for the exercise of power, suggesting that it is better understood in terms of Weberian types of administration: that is, patrimonialism versus bureaucracy. He evokes this different analytic frame, which provides new leverage and brings into focus a distinct set of similarities and contrasts among countries. Thus, as part of this sequence of progressive cooperation, Mazzuca succeeds in capturing O’Donnell’s distinctive focus on the exercise of power – that indeed departs from standard procedural definitions – and successfully and productively places it in an alternative framework.

What has happened with this sequence of progressive cooperation? Within the framework of a basically decontested (in the broader sense) understanding of democracy, a sequence of scholars has produced a series of linked conceptualizations and insights. With each step, one can see varying combinations of a sharpening of analytic framing, a response to changing circumstances, and a shift in focus that to some degree takes the discussion in a new direction, while building on prior steps.

I. Appraisiveness.

Waldron asserts that the rule of law “is clearly an appraisive concept: it is deployed by almost all of its users to enter a favorable evaluation of the regimes or situations to which it applies.”Footnote ¹²³ In the Florida debate, participants with very different views justified their position by appealing to the rule of law, drawing on the highly positive valence of the concept. For example, when Florida Secretary of State Katherine Harris (a Republican) certified the controversial result in favor of George Bush (another Republican), she stated “the true victor in the Florida election is the Rule of Law.”Footnote ¹²⁴ Reflecting a very different sentiment, Supreme Court Justice Stephen Breyer’s dissenting opinion argued that “although we may never know with complete certainty the identity of the winner of this year’s Presidential election, the identity of the loser is perfectly clear. It is the Nation’s confidence in the judge as an impartial guardian of the Rule of Law.”Footnote ¹²⁵

While Waldron views the rule of law’s normative valence as positive, an alternative view may be formulated. Martin Krygier discusses the sharply contrasting viewpoint that the identity of power-holders is the key issue, with the rule of law functioning as an “ideological mystification” behind which the exercise of power may be hidden.Footnote ¹²⁶ Within that framework, the rule of law could be negatively appraised. Whether one agrees with this perspective, it is clearly another instance in which the normative valence of the concept is dependent on alternative analytic framings.

II. Internal complexity and III. Diverse describability.

Both the basic juridical requirements for the rule of law, and the institutional arrangements that help ensure its achievement, are complex – as is made clear in the discussion of openness. Hence, Criterion II is met for both. However, as just noted, Waldron argues that these juridical requirements, while complex, are a matter of substantial consensus, such that the criterion of diverse describability is less relevant. By contrast, different actors have distinct views over the wider institutional arrangements that help to protect and implement these juridical procedures. For these institutional arrangements, complexity does yield diverse describability. For the rule of law, then, Criteria II and III do not coincide.

IV. Openness.

Waldron emphasizes that, within the context of the Anglo-American legal system, one component of the law is basically settled – that is, not open – whereas another component is subject to uncertainty and contestation in the context of the Florida election. He argues that important participants in the debate envisioned the rule of law as encompassing, on one level, a standard menu of juridical attributes that in principle characterize a well-functioning Anglo-American legal system.Footnote ¹²⁷ According to one formulation of this standard menu, the rule of law should be (1) general, (2) publicly promulgated, (3) prospective, (4) intelligible, (5) consistent, (6) practicable, (7) not too frequently changeable, and (8) congruent with the behavior of the officials of a regime.Footnote ¹²⁸ Although jurists differ on the exact content and wording of these attributes, Waldron views this aspect of law as basically settled and uncontested. Following the usage of Norman Care, it is subject to practical closure.Footnote ¹²⁹

Waldron identifies a second and far less settled component of the law – the institutional task of “designing a political system in which the laws rule rather than men,”Footnote ¹³⁰ that is, a system that will best realize and protect the basic juridical attributes just noted.Footnote ¹³¹ Correspondingly, contending advocates vociferously argue about basic, interconnected issues of constitutional design such as the separation of powers, the primacy of legislation as opposed to judicial supremacy, discretionary rule by judges, and the role of popular sovereignty.Footnote ¹³² According to Waldron, at this level, the rule of law is the “solution to a problem we’re not sure how to solve,” that is, the problem of crafting institutions so as to protect the eight juridical features just enumerated.Footnote ¹³³ This aspect of law is far more open.

Events in 2000, according to Waldron, reflected four general areas of contestation about these institutional practices: (1) the exercise of official discretion by political actors, (2) the regulation of discretion by vague standards, (3) the parties’ reliance on the legal process, and (4) the role of courts in resolving political disputes.Footnote ¹³⁴ In each area, opposing sides claimed that the rule of law would best be upheld by a set of institutional practices that they favored.

When debating the role of courts, for instance, some argued that the rule of law required that the controversy be resolved through the hierarchy of courts.Footnote ¹³⁵ One example of this perspective is commentator Michael Kramer’s defense of Gore’s appeal to the court system. Kramer argues that “the way out of this chaos is through the Rule of Law,” and specifically that the Democrats should “use the courts as the courts are meant to be used – to resolve civilly [what are] otherwise intractable issues.”Footnote ¹³⁶

The opposing side, in contrast, viewed the delay that could result from prolonged litigation as undermining the rule of law. This perspective is reflected in Republican advocate James Baker’s view that “at some point, the Rule of Law must prevail and the lawyers must go home.”Footnote ¹³⁷ Baker’s comment points to a basic tension between two potentially opposing objectives in the legal system: formulating prompt decisions, which achieves finality under the law; as opposed to devoting sustained time and effort in an attempt to achieve full fidelity to the law.Footnote ¹³⁸

Because the rule of law can encompass a host of attributes and institutional practices as discussed earlier, there exists ample basis for contestation. Each contender can legitimately advance contrasting understandings of the rule of law, given its internally complex character. In other words, both sides of the debate can claim allegiance to some important aspect of American legal practice and tradition, and they can defend their conceptualization on that basis. Thus, openness and its connection with contestation are clearly manifested in the events of 2000 in Florida.

V. Reciprocal recognition.

In legal contestation, a central concern of each side is to anticipate, understand, preempt, and respond to the position of the other side. Reciprocal recognition is a foundation of the legal process and is inherent to the kinds of contestation that occurred in Florida. Waldron himself never explicitly discusses reciprocal recognition, although he implicitly acknowledges its significance in legal debates. For example, he argues that “we imagine people advancing and defending (and criticizing and modifying) rival conceptions” of the rule of law.Footnote ¹³⁹ The reciprocal recognition among rival conceptions is especially evident in the role of precedents in case law, to which we now turn.

VI. Exemplars.

Like many of the commentators, Waldron suggests that what is in effect the narrower understanding of an (original) exemplar is too limited and that the functioning of case law is founded on a wide range of precedents and exemplars. In his words, “the reference back to the achievement of an exemplar may be too narrow an account of what gives unity to a contested concept.”Footnote ¹⁴⁰

Extending Waldron’s discussion of exemplars, we may observe that a central idea underlying the role of multiple exemplars in the American legal system is the idea of reasoning by analogy.Footnote ¹⁴¹ This refers to the tradition of reaching legal judgments by comparing a disputed instance with a prior instance in past cases or settled law, identifying similarities and differences, and deciding whether to treat the two instances the same.Footnote ¹⁴² As with the concerns raised about exemplars in Gallie’s framework, the issue arises as to whether this practice of employing case law (i.e., exemplars) in fact gives structure to legal reasoning. Of course, judges can potentially choose among conflicting precedents on self-interested or arbitrary grounds, suggesting that prior exemplars are not at all constraining. Critics of legal reasoning-by-analogy are concerned with precisely this problem, which is a major issue in legal scholarship.

VII. Progressive competition.

Waldron, unlike some of those who criticize the rule of law debate, argues that what appears to be conceptual disarray is actually evidence of a process that can lead to more analytically rigorous concepts.Footnote ¹⁴³ He suggests that the debate over the concept in the aftermath of Florida forced rivals to carefully consider their understandings of the concept and to specify its meaning more adequately.

In particular, because the contenders had to reconcile their own understandings of the rule of law with complex legal debates over vote counting and the propriety of Supreme Court intervention, participants were forced to think systematically about the foundations of this concept.Footnote ¹⁴⁴ Waldron describes this dynamic as follows:

In Waldron’s view, as a result of these controversies, legal theorists and practitioners have seriously reflected on the novel implications of the Florida debacle for the relationship between the rule of law and judicial power. The participant’s “understanding of the rule of law – or the difficulties associated with the rule of law – was in fact enriched and deepened in the context of partisan, perhaps even cynical invocations of the ideal in Florida.”Footnote ¹⁴⁶

One set of commentaries on the Supreme Court decision agrees with Waldron, in that it points to subsequent reflection on the legal issues involved in the Florida crisis. Yet Gary Rosen, conservative writer and managing editor of the periodical Commentary, impugns the seriousness of this reflection, accusing conservative commentators of “hypocrisy”:Footnote ¹⁴⁷ Although these commentators approved of the immediate decision in Bush v. Gore, they expressed grave misgivings about the court’s reasoning:

This acceptance of seemingly contradictory propositions may be understood in two ways, one more optimistic regarding progressive competition, one more cynical. Regarding the former: given the idea expressed earlier by Baker – that the rule of law in part involves arriving at expeditious judicial decisions – the prompt Supreme Court decision reflects one important aspect of legality, that is, finality. At the same time, insights derived from the ongoing discussion among legal commentators of underlying legal principles shows a concern with the other facet of the rule of law – that is, fidelity. In this optimistic account, both of these priorities have been addressed.

A pessimistic interpretation vis-à-vis the idea of progressive competition would point to the fact that these conservatives “won,” given that in the midst of a severe electoral crisis, their preferred candidate became president, even though the decisive vote count in Florida could be (and was) readily challenged. This brings the discussion back to the idea that the partisan exercise of political power – through control over key public institutions in Florida and the majority position on the Supreme Court – rather than the rule of law, decided this case. It may simply reflect the triumph of partisanship over judicial coherence.

This pessimistic view is reflected in the commentary of prominent legal scholar Cass Sunstein, who reacted to Rosen’s article with the following observation:

What inferences should be derived from this discussion? One possibility is that two major aspects of the rule of law – expeditious decisions vis-à-vis carefully reasoned decisions – are genuinely in conflict. This would call for deeper reflection on, and communication about, the rule of law – which could certainly be seen as progressive competition. Alternatively, corresponding to Sunstein’s argument, the outcome simply reflects the balance of political forces. In other words, all that mattered was who had more political power. This certainly does not involve progressive competition.

Optimistic conclusion?

On balance, Waldron finds that Gallie offers a valuable perspective for analyzing these legal issues, and in particular Waldron takes a positive view of the possibilities for progressive competition over the rule of law. The level of reflection that contestation provokes, he argues, is deepened when participants understand the implications of essential contestedness.Footnote ¹⁵⁰ Theorists could benefit in several ways from understanding the source of contestedness, as well as its consequences. First, participants who in effect recognize essential contestedness in a conceptual debate might not needlessly worry about the proliferation of conflicting definitions because they would understand that these debates cannot be fully resolved. Second, they can closely scrutinize how these evolving debates reveal inconsistencies and ambiguities in existing conceptualizations. For Waldron, awareness of essential contestedness improves the efficiency of conceptual debates, in the sense that unrealistic attempts to settle intractable arguments will be less likely and the benefits produced by these disagreements will be more readily recognized.

By engaging with Gallie’s framework and applying it to the rule of law, Waldron seeks to show that pessimism regarding the analytical utility of this concept is unfounded.Footnote ¹⁵¹ The rule of law, he argues, is not an empty phrase. It has meaningful content, some of which is basically settled and some of which is still fiercely contested. The debate over its meaning has structure along the lines that Gallie suggested, and furthermore, this debate can be productive. There may be strong temptation to decry analytically “messy” formulations of the concept. But Waldron optimistically suggests that debates over the rule of law have probably improved users’ understanding of this concept.Footnote ¹⁵²

Finally, Waldron argues that scholarly analysis focused on a concept like the rule of law can and should be connected with real-world usage. It must in part consider the practice of law and the functioning of legal institutions. He argues that “philosophical analysis of an important term cannot afford to distance itself too much from the use of the term on the streets: otherwise the analysis is not the analysis of anything interesting, from a political point of view.” Certainly, as in the Florida election, the “common usage of a loaded term like the ‘Rule of Law’ does not necessarily reflect careful philosophical analysis.” Yet good scholarly analysis can inform ordinary usage, and it certainly must take ordinary usage into account.Footnote ¹⁵³