A Model of Cults of Personality

A Model of Cults of Personality Xavier Márquez Senior Lecturer, Political Science and International Relations Programme Victoria University of Wellington Wellington, New Zealand marquez.x@gmail.com or Xavier.marquez@vuw.ac.nz Work in progress. Comments welcome. 1 Abstract This paper presents a general model of the emergence and functioning of cults of personality. Drawing on the work of Collins (2004), I understand a cult of personality as a set of interaction rituals, linked in chains, focused on symbols that refer to a political leader, and saturating a significant part of the public space of a polity. I argue that when interaction rituals focused on leader symbols are embedded in pat o age elatio s, p o esses of flatte i flatio a e likely to emerge. I illustrate this model by looking at three very different cases: the short-lived cult of Caligula in the early Principate in Rome, the cult of Mao in China, and the emerging cult of Hugo Chávez in Venezuela. 1 A version of this paper is to be presented at the 2013 Meeting of the American Political Science Association in Chicago (29 August). I wish to thank Jason Young and Daniel Leese for discussion on the Mao cult, Marcus Frean for general comments, and Andrew Ivory for research assistance. I developed much of the argument in this paper over several years in bits and pieces in my blog, http://abandonedfootnotes.blogspot.com. Some sections still bear the mark of these origins (especially the discussion of Caligula and Mao, which incorporate material that first appeared there). I thus wish to thank the many people (some anonymous) who have commented on these posts, directed me to useful and relevant sources, and encouraged me to research a topic for which at first I had very little previous training. The stude ts i POL“ o Di tato ships a d ‘e olutio s lass o e the ea s also provided much of the impetus for delving deeper into the phenomenon of cults of personality. 1 Cults of personality are rarely taken seriously. They are typically viewed as curiosities of authoritarian rule, with their over-the-top effusions and grandiose imagery merely certifying the narcissism of absolute power. To the extent that they are studied as instruments of control in their own right, with very few exceptions,2 existing work on leader cults has tended to rely uncritically on Weberian atego ies of ha is a a d legiti a to e plai thei e e ge e a d de elop e t,3 even as these categories have had to be stretched considerably in the process (Cohen 2007). In this view, charismatic leaders create cults that boost their legitimacy and thus help secure their hold on power. The phenomenon is thus viewed primarily through the lens of persuasion and the focus is on the techniques and symbols used to persuade a population to accord legitimacy to a leader (Heller 2008, Plamper 2012). Yet even a cursory examination of the historical evidence suggests that such explanations are at the very least hopelessly incomplete. As Wedeen showed in her study of the cult of Hafiz al-Assad (Wedeen 1999, 1998), cults can remain perfectly serviceable instruments of power without inducing anything we could call persuasion, or convincing people of the legitimacy of their leaders. Other researchers have also noted the general unbelievability and absurdity of much cult rhetoric.4 Unbelievable claims may serve an important purpose, as we shall see, but it is clearly not necessary for them to be believed by a large majority of the populatio i o de fo the ult to o k; hate e e ha is s the ult uses to produce submission, they cannot be reduced to the persuasive effectiveness of propaganda techniques.5 2 “ee, e.g., Wedee s pio ee i g studies of the ult of Hafiz al-Assad in Syria (Wedeen 1999, 1998), or the rationalchoice based remarks of Svolik (2012, pp. 80-81). 3 See, e.g., Rees (2005), as well as most of the other essays collected in Apor et al. (2005). 4 Wedeen (1999) reports that Hafiz al-Assad as said ult p opaga dists to e the ou t s p e ie pha a ist ; Hassig and Oh (2009, p. 55) e ou t ho [i]n 2006 Nodong Sinmun published an article titled Milita -Fi st Telepo ti g lai i g that Ki Jo g-il, the e t ao di a aste o a de ho has ee hose the hea e s, appea s i o e pla e a d the sudde l appea s i a othe like a flash of light i g, so ui kl that the American satellites overhead cannot track his movements ; the list of titles gi e to Ni olae Ceauşes u included such stupendous nonsense as the Giant of the Carpathians, the Source of Our Light, The Treasure of Wisdom and Charisma, the Great Architect, the Celestial Body, and the New Morning Star (Sebestyen 2010, 161); i toda s Ve ezuela, du i g the ele tio afte the death of Chá ez, Ni olás Madu o, Chá ez hose successor and eventual winner of the election, claimed that Chávez had appeared to him in the shape of a small i d to less hi [Madu o], a o e t that led the Ve ezuela sati i al site El Chigüi e Bipola the Venezuelan equivalent of The Onion; the site s a e ea s, oughl , the ipola ap a a to eak ha a te a d ite e do ot k o ho to ake a joke of this (Anonymous 2013). The examples could be endlessly multiplied; Svolik (2012, p. 80) provides further examples, and stresses, correctly, that their absurdity does not undermine the cult. 5 Hassig and Oh note, with reference to the hyperbolic rhetoric of the Kim Jong-il cult, that [e] e i No th Ko ea few people have been convinced by this propaganda because since Kim came to power, economic conditions have go e f o ad to o se (2009, p. 57). Even Myers (2010), who credits North Korean propaganda with a high degree of effectiveness, agrees that it is not believed literally; to the extent that it functions to increase commitments to the regime, according to him, it must do so by making national status salient and encouraging people to identify with certain narratives of national grievance, not by persuading people of the charisma of a pa ti ula leade . Hu o a out Ceauşes u s faili gs a d foi les as o o th oughout the o u ist pe iod; jokes about the leader were also common during Hafiz al-Assad s ule. A d the pu li dis ou se of justifi atio used by these cults, in all its hyperbolic silliness, was constantly subverted in private conversation and in myriads of private and individually unconnected rebellions. Even in North Korea, where the cult of Kim family members has long been assumed to have been extremely effective, jokes about the Kim family have been known to circulate in 2 In fact, both stories of preference falsification (cf. Kuran 1991) – people who merely go through the motions of the cult while not feeling any particular attachment to the leader symbols, or even actively work to hide their true feelings6 - and narratives of disillusionment – a child grows up thinking the leader is a god, but something happens after a certain age that disillusions them permanently (Kelly 2005) – seem to be common in cults, though their relative frequency vis à vis stories of genuine devotion is hard to gauge. 7 But even the fact that some people assert the godlike status of the leader after coercive penalties for failure to be insufficiently devoted no longer operate (and hence cannot plausibly be said to be engaging in preference falsification) is not properly explained by appeal to the persuasive effectiveness of propaganda, but must typically make reference to various identity-related motivations.8 It is dou tful, i a ase, that p opaga da p odu es o siste t effe ts o is ead its i te ded 9 audiences in sufficiently similar ways outside of particular ritual contexts. And appeals to charisma as an explanation of cults often seems to confuse cause and effect; beliefs about the charisma of a leader are in many cases clearly an effect, rather than a cause, of the cult, as the most sensitive analyses recognize.10 If cults of personality do not appear to generate committed support in a reliable way (the collapse of the Ceauşes u egi e i just fou da s i ei g the ost spe ta ula e a ple of this failu e , the h do rulers invest large amounts of economic resources in the production of cults, and why do they seem to strategically manage flattery,11 sometimes encouraging it, sometimes dampening it? Mere narcissism is an inadequate explanation: though narcissism may not be an uncommon personality trait among absolute rulers, a rational leader should not encourage the highest possible degree of flattery, since it recent years (New Focus International 2012a, b); and the veneration of Stalin in the Soviet Union appears to have often been quite superficial (Brandenberger 2011). 6 For example, Demick (2009, 97-101) reports the story of a elati el p i ileged stude t, Ju -sa g, who at the time of the death of Kim Il-sung must force himself to cry when he realizes that, despite the fact that he does not feel much for the deceased, his e ti e futu e depe ded o his a ilit to : ot just his a ee a d his e e ship i the Wo ke s Pa t , his e su i al as at stake. It as a atte of life a d death p. . Crying turns out to be easy enough while surrounded by other crying people, for reasons we will discuss later. 7 Plamper (2012), for example, opens his study of the Stalin cult with a story about how a group of Moscow stude ts, ete a of WWII, [f]ea i g the spi itual p ese ce of the leader, turned the Stalin poster on their do ito all a ou d i o de to feel f ee e ough to talk a out thei e pe ie es at the f o t, a d otes that e e so e i ti s of “tali s te o ea ted so st o gl to “tali s death that the suffe ed heart attacks. Even if o e disag ees ith Pla pe s spe ifi i te p etatio of these a d othe sto ies, the e is o de i g that leade s ols i a ults do e o e sa ed to so e sig ifi a t pa t of the populatio ; the sto of Mao s Ma goes (Chau 2010, Dutton 2004) provides another striking example of this phenomenon. We could also note the emergence of impromptu shrines to Chávez in Venezuela as examples of the sacralization of leader symbols even in the absence of large degrees of coercion (Ultimas Noticias 2013). 8 See Jones (2005) for some suggestive evidence regarding the motivations of people who resisted the discourse of de-Stalinization in the Soviet Union. 9 A poi t ell ade i Mittle s (2012) sensitive study of Cultural Revolution propaganda. 10 As Svolik (2012, p. 80) notes, quoting Suny (1998, p. 38): “tali as sho t i statu e, a mediocre speaker, and the ulti ate a of the a hi e a d did ot p oje t a i age of a leade u til o e as eated fo hi . The same could be said of many other leaders who have developed personality cults. 11 On the strategic management of flatter i “tali s ult, see Davies (2005). I provide more examples of such strategic management –Caligula, Mao, and Chávez – below. 3 would seem to prevent him from identifying genuine supporters;12 and leaders who develop cults are typically those who have the longest tenures,13 which would suggest that these are not irrational leaders who are easily deceived by their own propaganda. Indeed, dictators often appear to be perfectly aware of the problem excessive flattery presents,14 and classical advice to beware of flatterers is common, at least in the Western tradition; both Aristotle (Politics 1313b32ff.) and Machiavelli (The Prince, chapter 23) mention the problem flatterers pose to the rational tyrant, for example. Why then do we observe cults? This paper presents a general model of the emergence and functioning of cults of personality that does not rely on Weberian categories of charisma and legitimacy. The starting point of the model is what Randall Collins has called an interaction ritual (Collins 2004). An interaction ritual is a situation of copresence with joint focus: a group of people interact with each other while jointly attending to particular objects, actions, or events s ols . These symbols circulate in chains of interaction rituals (rituals with similar focal symbols) where their meaning and value is established and renewed via processes of emotional amplification. I argue here that cults of personality are best understood as interaction ritual chains in which leader-related objects circulate as focal points for group attention. The second building block of the model is the idea of signalling: politically speaking, cult practices are first and foremost public signals of support for a leader or a regime produced within the context of interaction rituals. But unlike many other forms of support signalling, I argue that cult practices are typically the result of inflationary processes that devalue the credibility of previous support signals and lead to the de elop e t of e t a aga t sig alli g o e tio s that ide the status gap et ee leaders and populations. I call this process flattery inflation (following a suggestion in Winterling's biography of the Roman emperor Caligula, Winterling 2011), and distinguish two varieties: managed and unmanaged. Leaders might strategically manage flattery inflation as a way of monopolizing certain esou es; o su h i flatio ight e e ge f o elo , as the esult of ed uee p o esses, whereby individuals experience heightened social pressures to engage in costly displays of support, leadi g to suppo t as ades. M odel of ults diffe s i this espe t f o “ olik s (2012, pp. 80-81) discussion of personality cults in the context of his more general model of authoritarian consolidation, since Svolik only considers the cult from the point of view of the autocrat (as a signal of invincibility). Finally, I argue that inflationary signaling is most likely to occur when interaction ritual chains focused on leader-related objects are embedded in patronage structures (Martin 2009), that is, social structures in which lower-status individuals are dependent for protection, position, or advancement, on higher-status individuals who monopolize crucial resources. Here I follow previous studies of personality cults (Gill 12 At least one contemporary researcher has formalized it in terms of a di tato s dile a: the ge e al i e ti e those ruled by a dictator have to falsify their preferences makes the dictator more insecure the more powerful he becomes (Wintrobe 2008, 2000, 1990). 13 Data on this in a section of this paper that is under construction. 14 For example, Mao wrote to Ho Chi Minh in June 1966: I advise you, not all of your subjects are loyal to you. Perhaps most of them are loyal but maybe a small number only verbally wish you "long live," while in reality they wish you a premature death. When they shout long live, you should beware and analyze [the situation]. The more they praise you, the less you can trust them. This is a very natural rule (quoted in Leese 2011, p. 168). 4 1980) and flattery in authoritarian contexts (Shih 2008), which have implicitly noted that inflationary signaling seems to be especially likely to occur when personal advancement within a hierarchical organization is dependent on ties to powerful patrons rather than impersonal bureaucratic criteria, and develop a theory of the conditions under which status competition within such patronage structures is likely to result in flattery inflation. The model is intended to explain why cults of personality are found in autocracies across a wide range of politico-economic conditions (in communist regimes committed to ideologies of equality as much as in absolute monarchies not so committed) but are uncommon in established democracies; why leaders sometimes prefer not to encourage flattery inflation, and why sometimes they lose control of the process; why leaders sometimes enforce anti-cult norms while accepting the breaking of such norms as reliable signals of support; why cults both produce genuine commitments from some fraction of the population, and why nevertheless these commitments typically prove fragile and fail to penetrate the entire population. The pape is o ga ized as follo s. The fi st se tio des i es hat a ult of pe so alit is, a d a gues that they should be understood as a specific sort of interaction ritual chain in which the value of the leader-related objects and signals that are the focus of interaction is subject to inflationary pressures. The second section introduces a simple model of signaling and identity in interaction rituals, and explains why we should expect flattery-inflationary processes within patronage structures. The third section illustrates the model with evidence from three specific cases drawn from widely different contexts – the short-lived cult of Caligula in during the early Principate in Rome, the Mao cult in China, and the nascent cult of Hugo Chávez in Venezuela today.15 A final section sketches some implications for the study of power in authoritarian regimes. What is a cult of personality? I defi e a ult of pe so alit he e as a set of interaction rituals, linked in chains, whose focal symbols relate to a particular political leader, and which occupy a considerable fraction of, or even saturate, the public space of a polity. Four points are worth noting about this definition. Fi st, follo i g Colli s se i al o k o i te a tio ituals (Collins 2004), which in turn draws on Goffman (1958), Durkheim, and Mead, I understand a ritual as any practice involving 1) a group of people who 2) interact while jointly focusing on some particular object (including, potentially, the sel es o s ol a d ha i g o o k o ledge of thei atte tio al fo us. ‘itual he e has thus a much broader meaning than formal ceremony; many informal practices qualify as rituals, from everyday conversation to sports games and work meetings, as well as (of course) many more formal ritual practices. The stress of the definition is on the ritual practices rather than on the particular objects that are the focus of such practices and their specific meaning in context. It does not matter, for our purposes, whether the focus of cult practices is on the specific image of the leader, his sayings, or even 15 th A fourth section, using quantitative evidence from a new dataset of cults of personality in the 20 century, is still under construction and not included in this draft. 5 the leader himself; what matters is that ritual practices exist whose focal symbolism refers to a political leader. Second, the definition emphasizes the fact that cults of personality are not constituted by one or a few events, but are composed of chains of rituals where leader-related symbols circulate in a variety of media. In particular, we can usefully distinguish among four different sorts of rituals that can be linked in chains where leader-symbols acquire value and circulate throughout a polity. First, and most obviously, we find spectacular mass gatherings (Hitler at Nuremberg, the mass receptions of the Red Guards for Mao at Tiananmen Square, large-scale election rallies elsewhere) where the leader himself is the focus of attention for masses of people gathered together in large public spaces. Second, there are a variety of relatively regular small-s ale ituals he e the leade s i age o utte a es a e the fo us of atte tio .16 These rituals tend to have a recognized structure and setting; they happen at specified times and places; and are often coercive, insofar as nonparticipation carries negative consequences, sometimes large. Third, there are many transitory rituals which entail the production of leader-related symbols that can be interpreted as signals of support for the leader.17 Finally, there are a number of private or secondary rituals (secondary in the sense that they do not necessarily involve a group, but instead use symbols whose meaning and value is established in more public interaction rituals), focused on images or utterances of the leader, and generally unobservable to others.18 The reason for stressing the need for leader-related symbols to circulate in interaction rituals is that (successful) rituals act as engines of emotional amplification. Through processes described in Collins (2004), interaction with joint focus can produce patterns of bodily synchrony and their attendant emotions, leading to feelings of euphoria (especially among people centrally located within the group) and to the experience of the objects of focus as valuable, even sacred. Following Collins, I shall speak of rituals as charging leader-related objects with value, though as we shall see this value is experienced heterogeneously by ritual participants; not all participants will consider the leader-symbols to be sacred. Indeed, in some circumstances rituals will fail to produce these effects for most participants, leading to feelings of disaffection or boredom, and to the progressive dampening of the value of the leader symbolism, or even its inversion (i.e., where leader-symbols acquire negative value, as with symbols related to Ceauşes u fo ost ‘o a ia s i the s .19 The circulation of symbols in chains of rituals 16 E.g., the itual of aski g fo i st u tio s i the o i g a d epo ti g i the afte oo i f o t of a po t ait of Mao many workplaces during the cultural revolution, o the stud sessio s fo used o Mao s iti gs i the army during the Great Leap forward (Leese 2011). 17 E.g., lo alt da es du i g the ultural revolution (Leese 2011, Jiang 2010, pp. 208-210, photos 232-235), giftgivi g o “tali s i thda (Davies 2005, Plamper 2012), the ado atio of Mao s a goes as eli s (Chau 2010), the construction of shrines to Chavez (Ultimas Noticias 2013), pilg i ages to the leade s to o ausoleu , pla i g a portrait of the leader in a visible place in the workplace, etc. 18 E.g., u i g i e se i f o t of a sh i e ith a pi tu e of Mao i pla e of the t aditio al sto e god (Steinmüller 2010, Landsberger 2002), praying in front of an image of Chávez, turning over a portrait of Stalin in order to have a private conversation (Plamper 2012, p. xiv), etc. 19 See Sebestyen (2010), hapte , fo a e a ples of the egati e ale e of Ceauşes u s i age a o g ordinary Romanians during the 1980s. Tismaneanu (1989, p. 330) epo ts that ‘o a ia s efe ed to the apital it of Bu ha est, disfigu ed i a o da e ith the p eside tial a hite tu al tastes, Ceaush itz, Ceaushi a, o Pa a opolis. 6 is necessary to keep their value high; without circulation in chains of rituals where the same people participate, their emotional charge is likely to decay. The mere existence of leader-focused rituals, however, is not sufficient to speak of a cult of personality; the phenomenon we are interested in is, to a first approximation, the near-saturation of the public space of a polity with such rituals. To be sure, from this perspective the existence of a cult is often a atte of o e o less, he e a su essful ult is si pl a ult that a ages to fill ost of the available ritual space of the polity. Furthermore, cults wax and wane – both in the sense that is symbols can appear more or less sacred to some fraction of the population, and in the sense that its rituals can occupy more or less of the public space of the polity – via inflationary and deflationary processes, as we shall see later, though below a certain threshold we are unlikely to speak of a cult of personality per se and speak simply of discrete rituals associated with leaders (e.g., election meetings in a democracy). It should also be stressed that what we are interested in is not primarily the saturation of public space with images of the leader. Leader-related objects circulate widely throughout the ritual space in successful cults, and (as we shall see) leaders often encourage the wide dissemination of such objects, but a successful cult is not an advertising campaign; leader-related objects must become a focal part of interaction rituals throughout the polity for us to speak of a cult, rather than merely a (widely ignored) background of interaction. We might say that leader symbols command attention in a cult of personality. I make no sharp distinction here between the cult of modern leaders and the cult of religious figures, saints, ancient monarchs, and the like, or any special qualitative distinction between sacred and profane objects. Following Collins, I take it that the production of a distinction between sacred and profane is merely the result of the operation of interaction rituals; successful rituals (i.e., rituals that successfully amplify emotion) produce sacred objects, and in a society saturated with successful rituals focused on particular leaders some fraction of people is bound to experience the leader and leader-associated objects as sacred. Moreover, from this perspective the rituals of cults of personality are also not sharply separated from the rituals of other forms of politics: the mass rallies of a cult differ from the mass election gathering focused on a particular politician only in the degree to which the cult rituals saturate public space, and from the mass rallies of a nationalist movement only in their focal symbolism. (Indeed, national and leader symbolism is often fused together). The advantage of discussing cults as interaction ritual chains is that it allows us to see that worship practices in a variety of contexts are generated by the same mechanisms, even if the techniques used to circulate leader symbols are relatively new and the meaning of the symbols is experienced differently. Moreover, the conceptualization of cults as interaction ritual chains allows us to cut through ideological de ates a out hethe pa ti ula s oli uses of politi s o stitute ults. ‘athe tha e gage i f uitless de ates a out hethe o ot a pa ti ula leade is o as a tuall o side ed sa ed, e simply assume that the sacredness of a symbol is merely the effect of large-scale chains of successful interaction rituals. To use some jargon, sacrality is endogenous to ritual: not a fact about the meaning of the symbols used in some particular practice but an extra-semantic fact about how some symbols e o e e do ed ith alue e otio al e e g i Colli s te i olog as they circulate in ritual chains that amplify their emotional charge. This perspective allows us to shift our vision away from the 7 (often quite entertaining, for a certain temperament) content of the symbols used in cults and towards the actual mechanisms that a ou t fo the su ess o failu e of cults as instruments of power, which we will miss if we focus solely on the cult as a form of persuasive rhetoric or even on the particular propaganda techniques used to make leader-symbols circulate through society. Flattery Inflation The emergence of cults of personality has two closely related aspects. First, an emerging cult may expand over the ritual space of a polity, displacing alternative rituals focused on other symbols; and second, leader-symbols may increase greatly in value for some participants in these rituals: they sa alise. These t o aspe ts of the de elop e t of a ult a so eti es d ift apa t; s ols may increase in value only locally (so that the cult never manages to permeate the ritual space of a polity, but remains relatively localized to some particular group), or rituals might expand without much increase in the value of leader-related symbols (so that the cult remains desultory, or is quickly routinized).20 The second aspect of the process (symbol-value inflation) is the most important one, however, since a regime, especially an authoritarian one, can typically flood the public sphere with leader-symbolism, without necessarily increasing the value of these symbols. (In fact, flooding the public sphere with leader-symbols can sometimes cheapen the value of the symbols, as we shall see). For this reason, we shall focus here on the process through which leader-related symbols increase in value, or flatte i flatio . Let us start by examining a simplified single ritual situation, repeated indefinitely, and focused on leader-symbols. Through processes of emotional amplification described in Collins (2004), some fraction of participants where come to consider the leader- elated s ols sa ed – valuable enough as a token of membership in the group that they are willing to expend significant energies in defending the leader-symbol and preventing others from denigrating it, or otherwise punishing those who display insufficient devotion, while a fraction of participants finds the ritual demotivating and the symbols empty. More formally, we say that a fraction of ritual participants place value s on the leader-symbols (where s might be negative if the ritual charges symbols with negative valence, capable of arousing anger and similar emotions), while a fraction of participants places no value on these symbols. The fraction of e thusiasts and the value they place on focal symbols in a ritual context depends on the structure of the ritual, the previous associations of the leader-symbol (e.g., is the leader-symbol associated with victory in war or economic development? , a d the leade s o the itual s p odu tio tea s a ilit to d a o p e-existing charged symbols of identity to construct narratives that increase 20 In fact, many cults of kings can be understood as routinized rituals of worship of particular offices, rather than of the particular persons who currently occupy them. Here, signals of support and focal symbols are highly conventionalized, and value inflation has essentially stopped; the cult occupies a restricted, though reasonably sig ifi a t, pa t of the itual spa e of the polit , a d e e od k o s the i i u lo alt id e ui ed see below). 8 the emotional energy of ritual participants while focusing on the leader-related symbols.21 Indeed, we could even understand the quantity as a easu e of hat itual pa ti ipa ts ight all the leade s ha is a; ut it is i po ta t to st ess that ha is a he e easu es a skill which may or may not belong to the leader as such, rather than a personal quality of the leader that induces certain reactions in ritual participants independently of pre-existing identities, ritual structure, and the circulation of symbols though ritual space. E e a highl ha is ati leade f o the poi t of ie of so e itual participants may appear a oo to othe s. I stead of speaki g of the leade s ha is a, the , I shall speak elo of a itual s po e as a su a easu e of oth the f a tio of pa ti ipa ts ho as a result of the ritual) end up placing a high value on its focal symbols, and of the value they place on these symbols.22 The more diverse the pre-existing identities of ritual participants, the more difficult it is for a given ritual to affect all participants equally (i.e., the smaller the fraction ). I assume that the appeal of a particular symbol (including leader-symbols) for a particular person depends on all the other symbols that make up the pe so s ide tit , hi h is i tu uilt up i hai s of i te a tio ituals of g eate o lesse po e . This implies that under conditions of free exit and entry into rituals, where participation in particular rituals does not carry other compensation than heightened emotional energy, and ignoring the material pre-requisites of participation in rituals, the equilibrium outcome would be segregation by identity, as people migrate towards the rituals that exert the strongest emotional pull for them (or, put otherwise, the rituals whose symbols have the highest value or emotional charge for them).23 Under these circumstances (which approximate those prevailing in most modern electoral campaigns), we should normally see no large-scale cult formation.24 Where, then, should we see flattery inflation? It is worth stressing that it is not obviously rational for a leader to encourage flattery inflation. Excessive flatte dis upts the leade s a ilit to disti guish et ee ge ui el lo al follo e s a d pu e opportunists, and leaders themselves are often aware of this.25 The problem is especially acute whenever the leader wishes to distribute a scarce good with value outside of the ritual situation (protection, economic resources, or status) to his followers, and wishes to condition the distribution of the good on genuine loyalty,26 that is, to supporters who genuinely identify with him as leader (rather 21 For example, cultural symbols often drew on pre-existing identities; for careful reconstruction of the ways in which they did so, see Mittler (2012). 22 A more precise statement of these ideas might associate particular rituals with the production of a specific distribution of values for the focal symbols of the ritual among participants, ranging from negative values to high levels of emotional attachment. We could even represent this as a function mapping ritual participants to symbol values for a specific ritual . Indeed, below I use just such a representation. 23 As Collins (2004) puts it, [h]u a eha iou a e ha a te ized as e otio al e e g t opis . “o ial sou es of EE di e tl e e gize eha iou ; the st o gest e e gizi g situatio e e ts the st o gest pull and i di iduals do not experience such situations as controlling them; because they are being filled with energy, the feel that they [a e i ] o t ol … Whe EE is st o g, the see i ediatel hat the a t to do. 24 We can represent this process as follows. Assign each person a natural number greater than or equal to 2. Assign each focal symbol a prime number value greater than or equal to 2. 25 See footnote 14 for one striking example from Mao. 26 There is a prior question here concerning why a patron might wish to reward mere loyalty (identification) rather than other characteristics of the clients, like competence at particular tasks. In actual fact we should expect patrons to reward some mixture of characteristics including loyalty, competence, and other factors, but 9 than, for example, with the group as a whole or some potential rival). This situation arises generally in patronage relations, where a powerful patron distributes some scarce resource in exchange for loyalty, and expects support from his clients that is forward-looking and diffuse rather than specific or contractual, i.e., expects identification with his goals.27 But genuine identification is unobservable; the leader can at best condition the distribution of the valued economic resources or status on the value group members place on the leader s p efe ed symbols.28 The idea here is that the valuation a group e e pla es o the g oup s fo al s ols ithi thei solida it ituals is evidence of identification with the objects referenced by the symbols. A client in this situation must credibly signal commitment to the leader by implicitly placing a price, as it were, on symbols associated with the patron, if they hope to gain access to the goods that the patron can allocate (or to avoid punishment, if the good in question is protection). The higher the price, the greater the commitment signaled. The te h i al te fo this should e sho i g that ou a e – e.g., sending congratulatory messages, placing the portrait of the leader in a position of honour, reverently a d isi l stud i g his o ds, uoti g his sa i gs, o se i g pa ti ula ta oos o e i g the leade s i age, et . . The ost of a sig al i this o te t a e easu ed oth i te s of the e o o i resources or time expended and the status loss incurred, since in a ritual context one can signal the value one places on the leader by dramatizing the distance between oneself and the leader. Indeed, from the perspective of group members, self-abasement is often an imperfect substitute for time or other resources used to signal loyalty (see, e.g., Shih 2008, on abject humiliation as a signalling cost), though in some (perhaps most) cases the patron will prefer to condition the allocation of the good not on mere self-a ase e t ut o so e sig that the lie t ide tifies ith the pat o s ideologi al goals e.g., adi alis i a i g out the pat o s p oje ts, de u iatio of the pat o s e e ies, et . .29 The problem, from the point of view of the patron, is that conditioning the allocation of the good on the value a group member places on the focal symbols of the group, and in particular on the leader symbols, is likely to attract not just genuine enthusiasts (people who genuinely identify with the leader, and already place a very high value on the leader symbols), but also opportunists (people who value the good the pat o allo ates o e tha the leade s ols, a d do t ge ui el ide tif ith the leade . If they value the patron-allocated good sufficiently, opportunists can mimic the signals of loyalty of enthusiasts; but since the support promised to the patron is diffuse and forward-looking, there is a commitment problem implicit in the relation: opportunists may renege on their commitments when the going gets though. In the process, moreover, they devalue the signals of the enthusiasts, which cease to identification with the pat o a d he e ith the pat o s goals a e espe iall salie t i a situatio s – e.g, when the goals of the patron are quite radical or easily subverted. For a formal treatment of the loyaltycompetence tradeoff, see Egorov and Sonin (2011), who show that dictators will often prefer to reward mere loyalty over competence. Their model assumes, however, that loyalty is generally observable. 27 For a full treatment of patronage relations, see Martin (2009, chapter 6). 28 The patron can condition the allocation of the good on the value a group member places on any of the g oup s focal symbols within solidarity rituals – Marxist-Leninism, Arab nationalism, the Bolivarian revolution – not just on the leade s ols. We ill etu to the uestio of the pat o s choice of symbols later. 29 Kung and Chen (2011) discuss the case of rewards to radicalism du i g the g eat fa i e i Mao s ti e. Fo reasons we shall explore in a moment, rewarding radicalism seem to have attracted opportunists rather than true believers. 10 be informative regarding their identity; opportunistic signaling by people who are not genuine true believers both cheapens the identity of enthusiasts, and makes it difficult for the patron to condition the allocation of the good on the basis of genuine loyalty. We can odel the situatio as a au tio , ith a t ist. People ake a lo alt id d a atizi g i visible ways the value they place on the leader, and the patron allocates the good to all those who meet so e th eshold lo alt p i e, which in turn will depend on the nature of the good allocated and the pat o s goals. The twist is that the more a person identifies with the patron, the more they are willing to e gage i alt uisti pu ish e t of those i di iduals ho id too lo , si e the s ols i uestio are tied to their identity; too low a loyalty bid by other participants in the ritual context is interpreted as disrespect. The more successful the interaction ritual is at imbuing the pat o s symbols with value for these enthusiasts (i.e., the higher s), the more likely it is that enthusiasts will punish those they perceive as opportunists; and the higher the number of enthusiasts (the higher ), the lower their costs for punishing disrespect. Moreover, depending on the specific ritual context, they are likely to have better local information on who is an opportunist than the patron. Under what circumstances should a patron distribute the good according to the price followers place on his symbols, and under what circumstances should he find an alternative allocation mechanism to allocate the good? There are at least three circumstances in which a patron might choose to encourage a high minimum loyalty bid. First, there are cases in which the patron is a le to e gage i p i e dis i i atio , distributing goods that are of limited alue to oppo tu ists. These ill t pi all e heap status s ols – medals, mere proximity to the leader, prizes and honours. If the patron distributes these goods, we should see a separating equilibrium: only those people whose identity is sufficiently tied to leader symbols through the rituals in which they participate will credibly bid high enough for them to the leader, since cheap status symbols are ex hypothesi of limited value to the opportunists (whose identity is not bound up with these symbols) whereas they are in fact valuable to the enthusiasts (we are excluding here the possibility of oppo tu ists ho a u ulate heap ho ou s e ause the a e fo a d-looking, and believe these cheap honours will help them later on). But this is obviously not the case in the vast majority of patronage relationships; many of the goods that the patron must normally distribute – the perks of political office with all its opportunities for material enrichment, for example – appeal to both the opportunist and the enthusiast, and have both material and symbolic appeal. The second circumstance in which a patron might choose to encourage flattery inflation requires the assumption of an initial situation of status equality between patron and clients. Suppose patron and client are of similar status but differ in their ability to control of specific material resources. Assume further that status equality precludes the existence of direct command relationships between patron and client. The wider the status gap, the more likely that direct command relationships can be established. But flattery dramatizes a status gap between leader and follower; through self-abasement, the follower indicates a recognition that the patron is of higher status than himself, and in fact expresses a readiness to take orders from the patron. What the patron wants, however, is the promotion of a social norm that recognizes a qualitatively higher difference in status between himself and his clients, which is only possible if all clients engage in self-abasement, not just the unreasonably loyal or the occasional sycophant. In this case, the patron might encourage flattery inflation (distributing the good 11 only to those who make high loyalty bids) as a way of encouraging the constitution of this sort of norm even if he suspects most people signaling support are actually craven toadies, who must weigh the status cost of signaling loyalty against the value of the good they receive. The patron thus trades off knowledge of the type of follower (whether genuinely loyal or not) against increased status (which allows him to substitute relationships of direct command for diffuse support). Finally, consider a situation in which the patron in chief – the leader - fears competition from clients that have their own clients, and controls resources that are proportional to the number of people directly signaling loyalty to himself. Here encouraging flattery inflation – the signaling of direct loyalty to himself - makes it harder for elite competitors to mobilize the loyalty of their clients against himself or even against his preferred policy goals. The indirect, intransitive patronage structure is modified by adding direct links between the patron and the lower levels. We can formalize these intuitions a bit more. Consider the first case. The patron is interested in the | probability ( , where is the loyalty bid by a particular individual. B Ba es theo e , this is equivalent to: ( Or, since ( | ( , ( | ( ( | ( | ( If the patron does not offer a good sufficiently valuable to the opportunist to be worth the loss of status involved in flattery, and does not punish lack of identification, it straightforwardly follows that ( | | , and hence that ( = 1; this is the situation where the patron only distributes cheap status goods of no interest to the opportunist, leading to a separating equilibrium where only true believers signal loyalty to the leader. Consider now the second case, which we will call the pu e s opha case. Here, most ritual participants are opportunists; the good distributed by the patron is highly valuable to opportunists, or the patron is able to punish disrespect within the ritual context; and existing rituals do not produce large degrees of commitment even among people who identify with the leader (i.e., ritual power is low). In this case, there is no minimum loyalty bid that will discriminate sufficiently between opportunists and true believers; an opportunist can always at h a t ue elie e s lo alt id. I deed, it is easo a le to suspect that only unprincipled opportunists would be willing to incur the status loss involved in making a very high loyalty bid in these circumstances, given the low power of the ritual to produce genuine identification with the leader. But since a low can indicate either an enthusiast or a disloyal member of the group, i.e., someone who identifies with a different set of symbols, enthusiasts cannot compete fo the pat o s goods deli e atel aki g lower loyalty bids than others. And the patron cannot count on enthusiasts to identify and punish opportunists, since ex hypothesi there are few genuine enthusiasts, which greatly increases their costs of punishing opportunists, and at any rate opportunists 12 can also denounce enthusiasts as opportunists when making loyalty bids. If the patron does condition the distribution of the good in this context on signaling loyalty to the leader, then, the stage is set for full-blown flattery inflation; and the eventual outcome is a general lowering of the status of ritual participants vis à vis the leader (who may or may not be the patron), and the allocation of the good to the most craven toadies. Signals of commitment to the patron are likely to increase in value (and thus status is likely to decrease for those aki g the sig als u til the at h the alue the lie ts put o the resources the patron can allocate, unless the patron takes measures to discourage flattery, or conditions the distribution of the good on other criteria (e.g., on meeting impersonal, rule-based criteria). This situation, at any rate, is likely to arise where there is a large disparity between status and esou es: if the pat o a d the lie t a e i itiall lose i so ial status, ut the pat o s esou es a e much larger than those of the client, the client can only offer status in return for these resources (since, ex hypothesi, he has no genuine loyalty to the patron), and flattery inflation may occur even in the absence of encouragement by the patron (i.e., be unmanaged flattery inflation). The final case requires some additional assumptions. For simplicity, let represent the cost of signaling support for the leader or chief patron, and let it range over the real numbers. This cost is visible to other people in the ritual context. Negative numbers indicate signals of support for other potential leaders, hile positi e u e s i di ate sig als of suppo t fo the hief pat o . A i di idual s al ulatio of ho u h to id fo lo alt ho high to ake his o he sig al ) depends essentially on three factors: first, the difference between the private value they place on the leader symbols (denoted by , which we might call the degree of identification and which we can stipulate is negative for supporters of other leaders) and the cost of the signal they produce, a quantity which they wish to minimize (the la ge the diffe e e, the o e disho est the sig al, and the larger the identity costs); second, the expected value of the goods they might receive from the patron for a given bid (which is increasing in the value of their signal relative to the average signal, under the assumption that the good is relatively scarce); and third the expected costs of punishment, which also depend on the average loyalty signal: the larger the difference between o e s lo alt sig al a d the a e age lo alt sig al, the highe the chance that people bidding higher will either i te p et o e s lo alt sig al as punishable disrespect (and hence as an opportunity to signal loyalty themselves) and the lower the average costs of punishment for people signaling high (since more people will be available to act as punishers).30 Schematically, we might say that the utility of making a loyalty bid is: ( ( | ̅ (| | ( |̅ Where ̅ represents the average loyalty bid, represents the private valuation the agent puts on the goods being distributed if he or she actually receives them, and represents the agent s private valuation of the costs of punishment if he or she is actually punished. More informally, the first term in this equation represents how much the agent values the goods the leader can provide; the second how important ho est sig ali g is to the age t; the thi d ho u h he fea s the pu ish e ts othe s ight inflict on him for not making a high enough bid for support for the leader. This sets up a pure game of 30 This is a simplification. In practice, the probability of punishment depends on the network structure of interaction, not just on the average signal of loyalty. 13 expectations, where how much to bid for loyalty depends on how much others are bidding. As in other threshold models of collective behavior, the final equilibrium depends on the network structure of interaction (which we are assuming away here) as well as on the underlying distribution of private valuations of the leader symbols ( ) and the underlying distribution of private valuations of the good being distributed ( ).31 But the higher the expected average signal, and thus the expected benefits and e pe ted osts of pu ish e t elati e to the osts to ide tit of sig ali g disho estl , the g eate the incenti es to id high fo lo alt ; conversely, the lower the expected average signal (e.g., the more people expect others to be uncommitted followers, or opinion to be polarized between different leaders), the greater the incentive to minimize costs to identity and signal honestly. The question we are interested in: how is this sort of expectation about the average signal – set? There are two cases worth noting. The first is a case of managed flattery inflation, where the leader can affect the distribution of private valuations of leader-symbols. Here the leader suspects large numbers of people actually have a high degree of identification with him, or he has some control over the rituals affecting the actual degree of identification with him that other potential leaders do not have. For example, he might be able to promote additional rituals through which his symbols circulate and are further charged with value, increasing the private valuations for some substantial number of people. By promoting such rituals, he can increase the identity costs of bidding low, and thus increase the expected average signal of loyalty. If, moreover, the value of the resources the leader can distribute is increasing in the average signal (that is, in the number of people directly signaling support for him, and the apparent intensity of that support, as when the resource in question is political office; in other words, is increasing in ̅ ), we have a positive feedback loop: the more intensely people signal loyalty to the leader, and the more people do so, the more the expected value of the benefits the leader distributes increase, the highe the e pe ted osts of pu ish e t fo iddi g lo , and the greater the average signal of support. As this process unfolds, the value of signaling direct loyalty to any other leader decreases, and hence the possibility of successful challenges to his position from other potential leader-candidates, since clients with low private values of s find it very hard to coordinate in the absence of any signals of support for alternatives. The process continues until the value of the benefits the leader can distribute stabilizes, or the power of the rituals that charge his symbols with value is exhausted. A leade ho suspe ts he has a ge ui e ese oi of suppo t a o g his lie ts lie ts a thus encourage flattery inflation to generate a source of personal support independent of his immediate clients. But note that if he does not actually have this support, or cannot change the underlying distribution of valuations of leader-symbols via ritual manipulation, the process will quickly stall; he will not be able to generate a critical mass of i te se suppo t that tu s e e o e else. As in other threshold models, if the distribution of s and y has gaps – if some subgroup greatly disvalues the leader and cares little for the resources he is able to distribute – the cult may not expand over the entire population, or the equilibrium outcome may be polarization. The second is a case of unmanaged flattery inflation. Assume there is some exogenous shock either to the reputation of the leader (e.g., the leader appears responsible for the independence of the nation) – 31 For a fuller discussion of threshold models of this kind, see Granovetter (1978). 14 increasing s for individuals – or to the value of the benefits he can distribute (e.g., he suddenly comes to control new economic resources) – or even to the network structure of interaction, increasing the probability of punishment by people who value the leader highly (e.g., free train travel enables enthusiasts to spread from major cities). If the shock to the underlying distribution of private valuations s or y is high enough, this may induce some people to bid higher for loyalty, which in turn increases e e o e s e pe ted pu ish e t osts fo iddi g lo e , a d so o , u til the e flatte e uili iu reflects the underlying distribution of resources or the private values of the leader symbols. It is important to recall that the value of the leader symbols, even for genuine enthusiasts, is not a fixed quantity, but depends on both the ritual context and any external associations the symbols may acquire through their circulation in ritual space. If the power of a particular ritual decays – if, for example, the leader-symbols become associated with some negative economic outcome and lose value even for true believers, or the ritual becomes too routinized or even ceases to operate – the a ea lie pu e fe o e uili iu ight tu i to a pu e s opha e uili iu . Similarly, if the conditions for the allocation of the good change, we might see relative deflation in the value of leader symbols. These are negative exogenous shocks to the private valuations s. The problem is that once the expectations are set, they may be hard to reverse: if everyone expects a high average loyalty bid, even major decreases in private valuations of the leader symbols may not much affect the actual loyalty bid each individual is willing to make. The simple model may be formalized further, and extended in various ways. Though I will not pursue various possible formalizations here, it is worth mentioning one important extension. Recall that the ritual situation described above is embedded in a chain of ritual situations with greater or lesser power to produce identification with the leader-symbols via emotional amplification, and with greater or lesser fo us o the leade s i age. No , e e if ost of these ituals ha e lo po e a leade ight still a t to flood the ritual space with his symbols, not because he trusts that these will induce commitment (increase the average value of ), but because by increasing the minimum loyalty bid he discourages the emergence of alternative symbolic foci of mobilization. Though there is no commitment on behalf of the leader, there is no commitment on behalf of anyone else either. This induces inflation (the devaluation of signals of loyalty), and encourages opportunists, but the opportunists can be useful insofar as they can signal loyalty by identifying principled opponents of the leader (those who will not meet the minimum loyalty bid).32 At this point, however, we shall gain greater understanding of the dynamics described by the model by looking at a number of specific cases. Three examples: Caligula, Mao, Chávez We can illustrate the abstract model of the cult developed above by looking at a few specific cases in more detail. I have chosen to look at three cases that are quite different in their social, economic, 32 Wedeen (1999, 1998); documents precisely this strategy on the part of Hafiz al-Assad. The ritual space of Syria was flooded with leader-related symbols, but it was common knowledge that none of these symbols helped to produce any great degree of commitment and identification with the leader; what they did make impossible was the emergence of alternative symbolic foci, and they facilitated the identification of principled opponents of the regime. 15 cultural, and political contexts to show how the processes of cult formation are independent of many of these variables, though of course the focal symbols of each cult vary. The short-lived cult of Caligula emerged during the early Principate in Rome (Caligula was emperor from 37AD to 41AD), an agrarian, highly class-di ided so iet he e the idea of a he o ult as idel k o ; the ult of Mao e e ged in a society undergoing a communist and agrarian revolution explicitly committed to collectivism and egalitarianis , a d though it offi iall e ded ith Mao s death, it has had a lo g afte life, o ti ui g to provide symbols for private rituals; the cult of Chávez emerged in a broadly capitalist society with relatively democratic institutions, and it shows no signs of deflating yet, despite his death. In all three cases we can observe the sudden take-off of flattery inflation as ritual contexts interact with patronage structures, despite wide variations in culturally valued symbols, economic structures, and political institutions within these three societies. At the same time, they illustrate different sorts of cult d a i s: the Caligula ult is a e a ple of pu e s opha ; the Mao ult illust ates the d a i s of runaway flattery inflation through the conjunction of very successful interaction rituals and a large degree of coercion; and the Chávez cult displays the dynamics of a separating equilibrium (cult and anticult) in a situation where exit from leader-focused interaction rituals is possible and even common. Caligula I d a he e o Alo s Wi te li g s supe , ut e isio ist, iog aph of Caligula (Winterling 2011).33 From Augustus Caligula s g eat g a dfathe , the fi st e pe o o a d, the imperator was the most powerful person in Rome, partly due to his control of the Praetorian Guard, and partly due to the economic resources the imperial household had come to control. At the same time, the emperor depended (at least early on) on the senatorial aristocracy to rule the empire. The 600 member senate constituted the group from which the emperor needed to draw the people who could command the legions, coordinate the taxation of the provinces, and in general govern the empire and keep him in power. The emperor could differentially favour members of the senatorial aristocracy (by promoting them to various highstatus positions), but segments of the aristocracy could also conspire against him and potentially overthrow him, selecting a different emperor, especially since principles of hereditary succession were not clearly institutionalized. Nevertheless, though senators as a group might dislike a particular emperor, they did not necessarily agree on any given alternative (much less on any alternative acceptable to the Praetorian Guard, hi h ould ulti atel eto the se ato s hoi e), and at any rate individual senators could always benefit from convincing the emperor that some other senators were conspiring to unseat him (via maiestas or treason trials, in which the convicted were executed and their property confiscated). Senators thus faced some coordination costs in acting against even a hated emperor. These obstacles were not insurmountable (conspiracies did take place, and sometimes succeeded), but they were not insignificant either. Yet despite the disparity in military and economic resources between the emperor and the members of the aristocracy, emperors and senators did not at first have widely different social statuses, and the senate remained the central locus of the distribution of honours in Roman aristocratic society. Senators jockeyed over relative status in the many rituals of Roman life (marked by such things as the seating order in the circus or the theatre, the order of voting in the senate, the lavishness of their hospitality in 33 An earlier version of this section appeared in my blog. 16 their private parties, elections to political office, the number of their clients, etc.) while recognizing the primacy of the emperor, but they remained notional social equals. Augustus was known as the princeps, literally the fi st itize he e the ea l ‘o a E pi e is o all alled the p i ipate ; the standard republican offices were filled more or less normally and retained their meaning as markers of status (though elections were often rigged, when they were held at all, to produce the results decided in advance by the emperor); the senate voted triumphs and special festivals in honour of particular people a d e e ts, a d te h i all o fi ed the e pe o s o positio ; e e the title imperator originally meant nothing more than military commander (though it came to be applied exclusively to the princeps or certain members of his family). Most importantly for our purposes, the first two emperors (and many later ones as well) did not (and could not, for reasons that should become clear shortly) compel the sorts of marks of obeisance typical of earlier Helle isti o a hies, he e the status dista e between the rulers and the members of the traditional elite had been much larger than in Rome: proskynesis (prostration), kissing the feet or the robe, worship as a god, elaborate forms of address, clear hereditary succession, etc.34 I fa t, Augustus i pa ti ula e t out of his a ot to sig al a so t of i te tio to e o e a ki g, that is, a ruler like the Hellenistic monarchs, despite the fact that the Roman polity had obviously e o e a o a h i all ut a e, so ethi g that as o o k o ledge a o g all e e s of the elite. He lived in a relatively small house on the Palatine hill; stood for office in the normal way, and sometimes resigned it; and let the senate conduct the business of the republic in appearance, cleverly signaling his i te tio s so that se ato s ould ea h the ight esult i.e., the esult Augustus a ted . The reason for this cautious behavior is that signaling an intention to become a king (that is, to widen the status distance between himself and the senatorial aristocracy) reduced the senators coordination costs. This as, afte all, hat happe ed to Julius Caesa Augustus adopti e fathe . By behaving in ways that signaled an intention to become a king in the Hellenistic sense (a rex), he threatened to destroy the foundations of senatorial status in the Republic, i.e., to drastically humiliate them vis à vis the emperor. Rex was a symbol that could still mobilize passions against those who tried to appropriate it; any credible signals of an intention to re-establish kingship seem to have greatly lowered the coordination costs of dissatisfied senators for conspiring against the emperor. Yet the strong norm against the appropriation of kingship symbols by powerful individuals was not e ough to p e e t the e e ge e of the ult of Caligula a fe de ades afte Augustus death. The problem was that the social status of the emperor was not fully commensurate with the resources he controlled. Senators as a group preferred this situation of status equality. But individual senators could benefit (both materially and in status terms) from credibly signaling special loyalty to the emperor, who as of ou se pat o i hief, a d o t olled status a d material resources that only he could allocate to the senators. Such signaling could take two forms. First, senators could inform on each other. Yet excessive denunciations also increased the risk of actual conspiracies (as senators anticipating that they might be denounced could attempt to take power) and devastated the elite on which the emperor relied, so 34 To get a sense of the cult of Hellenistic monarchs, see Eckstein (2009). 17 emperors often attempted to curb excessive conspiracy-mongering. But senators could also directly flatter the emperor, attempting to show how much they valued his person by costly acts that inflated his status relative to theirs. The problem here (for the emperor) was that any particular form of flattery quickly became devalued, and the emperor lost the ability to distinguish genuine supporters from nonsupporters. Moreover, flattery inflation tended to diminish the collective social status of the senatorial aristocracy: the more the emperor was praised, the more the senators were abased. For example, in Roman elite society the morning salutatio was an important indicator of status: friends and clients visited their friends and patrons in the mornings, and the more visitors a senator had, the higher his status. But nobody could afford not to visit the emperor every morning, or to signal that they were not eall f ie ds ith the e pe o ; every member of the senatorial aristocracy was in a sense the emperor s client. So the morning salutatio at the e pe o s eside e tu ed i to a ush of hu d eds of se ato s, all of the jostli g to get a little it of the e pe o s atte tio , a d all of the p ete di g to e the e pe o s f ie ds, ega dless of thei p i ate feeli gs. Similarly, in principle, the senate retained some discretion in allocating honours to the emperor (triumphs, titles, etc.), but individual senators could always sponsor extraordinarily sycophantic resolutions in the hopes of gaining something from the emperor (offices, marriages, etc.), and other senators could not afford not to vote for such resolutions due to the risk of being made the target of denunciations. In sum, flattery inflation was, from the point of view of the senators, a tragedy of the commons: as each senator tried to further his relative social status within the aristocracy, they tended to devalue their collective status. And it was not necessarily a good thing from the point of view of the emperors either, who could not easily distinguish sycophantic liars and schemers from genuine supporters, and who often disliked the flattery. So the emperors tried to dampen it or manage it to their advantage. Winterling distinguishes three different responses. First, as noted earlier, Augustus managed flattery inflation through ostentatious humility. Everybody could then pretend that things remained the same even though they all knew that Augustus was ultimately in charge. But this required indirectly signalling his intentions so that senators had enough guidance to know what to vote for and who to denounce without ordering them to do anything (which would have resulted in a catastrophic loss of status for the senators, potentially risking a conspiracy – a direct atta k o the se ato s ide tit ). But less a le politi al ope ato s like Ti e ius, Augustus immediate successor) could not use indirection as effectively. Tiberius apparently detested flattery, but he was at the same time unable to clearly communicate his intentions to the senate. His inability to master the complex signalling language that Augustus used led to inflationary pressures on flattery, i.e., on the value of symbols associated with his person relative to the value of the symbols associated with senatorial status. As the level of flattery increased, Tiberius was led to use increasingly blunt instruments to tame it, like moving to Capri permanently and banning the senate from declaring certain honours. We ight thi k of Ti e ius st ateg as the e ui ale t of i posi g p i e o t ols o e tai kinds of loyalty bids. Yet by imposing such controls and removing himself from Rome, senatorial energies were instead turned towards denunciations, which were made especially easy by the fact that senators e e o sta tl aki g istakes a out hat Ti e ius eall a ted. The o e denunciations, moreover, the less actual conspirators had to lose, leading to a poisoned and dangerous 18 at osphe e, espe iall as fa tio s of Ti e ius fa il s he ed o e the su cession. Most potential heirs did not live long; Caligula was the last man standing. Caligula s as e t to the th o e did ot e d the flatte -inflationary pressures induced by the fundamental power disparity between the imperator and particular senators. To manage these pressures (which, as discussed above, make it difficult for the patron to discriminate between genuine and non-genuine supporters, especially when, as in this case, the patron can be reasonably sure that most group members are opportunists rather than genuine supporters), Caligula first tried the Augustan policy of indirection and ostentatious humility, and was reasonably good at it. But for reasons that are not entirely clear (though Winterling suggests they were related to attempts on his life), he seems to have changed tack in the third year of his reign to deliberately encourage flattery hyperinflation. He did this, in part, by taking the senators literally: when they said that he was like a god, he demanded proof of this, thus forcing them to worship him as a god and humiliating the particular senator. Or when he was invited to dinner, he forced senators to ruin themselves to please him. And he demonstrated contempt for their status by the way he behaved in the circus and elsewhere. (The famous story of how he named his horse a consul can be understood as one such insult). Yet the senators could not retaliate by revealing their true feelings; their coordination costs had increased and their individual incentives were always to flatter Caligula. Strategically speaking, the point of this seems to have been to lessen his dependence on the senatorial aristocracy and to move the regime towards a Hellenistic model. (Winterling discusses some suggestive evidence that Caligula might have been planning to move to Alexandria, an obviously symbolic move to the historic capital of Hellenistic dynasts). Encouraging runaway flattery inflation made conspiracies harder to pull out (since even the most innocuous comment could be used to betray the other conspirators) but also succeeded in completely humiliating the flatterers (in this case the senatorial aristocracy) and lowering their collective social status vis à vis the ruler. At the endpoint of this process (represented here by the status hierarchy of the Hellenistic dynasts), ambiguous language is no longer necessary to manage the relationship between the patron and his clients; their competitive selfabasement has widened their status distance so much that direct orders are no longer out of the question. It is also possible that by encouraging runaway flattery inflation, Caligula might have hoped to institutionalize the principle of hereditary succession, which would also have contributed to a shift in power from the aristocracy to the imperial household. (It does not seem to be coincidental that cults of personality even in the modern world appear to be associated with forms of hereditary succession even in regimes that are not in principle hereditary, like North Korea or Syria). This was a high-risk strategy: deliberate humiliation, by striking at the core of the symbols of senatorial identity, also made some senators more willing to run the risk of conspiracy; even as these risks increased, the costs to identity of allowing Caligula to humiliate them so directly also increased. In fact, according to Winterling, the humiliation of the aristocracy not only led to the downfall of Caligula, but also contributed to his characterization by later a isto ati ite s as the ad e pe o : Caligual s ols had become objects of hatred, since they never had a ritual context that charged them with positive emotional energy. 19 Mao Fo dis ussio of Mao, I d a p i a il o Da iel Leese s Mao cult (Leese 2007, 2011).35 ag ifi e tl esea hed studies of the Leese argues that the cult first emerged during the later years of the Chinese civil war as a mobilizing device. It was consciously promoted by the top leadership of the CCP (not just Mao) in reaction to the growing cult of Chiang Kai-shek on the Guomindang side, and seen even by people who had doubts about overly personalizing Marxism as a way to unify the party against their enemies. From this point of ie , the ult appea ed as a fo of hat Leese alls a di g ; a d it as spe ifi all u tu ed ithi the pa t th ough the p a ti e of g oup stud of pa t histo , hi h p ese ted a thi al a ati e of the Lo g Ma h u de Mao s o e t leade ship. Group study among committed cadres focused on Mao-related symbols is an archetypical interaction ritual; and since the cadres already identified with communism, it is likely that these rituals were relatively successful in increasing the value of Maorelated symbols among them. As Leese notes, loyalty signalling in these rituals could be used both to marginalize certain factions (e.g., the group of Soviet-t ai ed ad es a ou d Wa g Mi g, ho had “tali s fa ou and to motivate party and army members in the continuing struggle with KMT forces: those insufficiently devoted to the Mao faction could be identified relatively simply, since most participants would not have been opportunists at this point, and hence would not ha e t ied to fake thei sig als. But to the extent that the cult also mobilized non-party members (non-ritual participants), it would have done so mainly through general propaganda campaigns, an arena where it had to compete with similar publicity by the KMT, at least in o tested hite a eas. With the i to of the CCP these o ilizi g a d u if i g fu tio s of the ult e a e less i po ta t, though the pa t of ou se o ti ued to o t ol the pu li displa of Mao s image, and the cult could still be used as one of the instruments of centralization employed by the CCP (e.g., against Gao Gang in 1953- , ho de eloped his o egio al ult i Chi a s o th-east and was eventually purged). This is ot to sa that the e as o de a d f o elo for cult practices. Since the CCP was in part a huge hierarchical patronage machine with few formal mechanisms for promotion, signalling loyalty through praise – sending congratulatory telegrams to Mao, for example, even when these were discouraged by the CCP leadership – was a useful means of career maintenance and even advancement (for similar dynamics in contemporary China, see Shih 2008). But praise of the top leaders was tempered both by the fact that it was embedded in a larger discourse where Stalin, not Mao, was the pre-eminent leader of the communist world, and by the fact that the top leadership of the party seems to have consciously discouraged extreme praise, perhaps because it feared (not unreasonably, as it turns out) concent ati g po e i Mao s ha ds. The ult thus appea s he e ot o l as a o ilizatio de i e pushed from the top, but as the unintended consequence of loyalty signalling by lower levels of the party, which tended to keep the overall level of flattery relatively high, and inflationary pressures steady; and it was clearly fuelled, though not fully explained, by the undoubtedly high popularity of the party and the prestige of Mao as its leader during the early 1950s. 35 An earlier version of this section also appeared on my blog. 20 The death of “tali , Kh ush he s secret speech, and other political developments disrupted this initial equilibrium, in which the expression of loyalty to Mao had not yet crowded out all other signals of loyalty to the party and the revolution, and had not yet colonized public space to the extent to which it did during the Cultural Revolution. For one thing, the death of Stalin eventually had the effect of displacing foreign leaders from their pre-eminent position in public displays, leaving Mao to monopolize an ever larger and more central share of public space. But at first Kh ush he s spee h fed i to a p o ess of liberalization of the public sphere which had begun somewhat earlier. Criticism of the cult and other fo s of dog atis as ai ed i high pla es, a d suppo t fo olle ti e leade ship expressed. At any rate, the party was (with good reason) confident in its popularity at this time, and prepared to relax its control over the public sphere. In particular, Leese takes the Hu d ed Flo e s a paig of to e a (botched) attempt at genuine liberalization, though Mao himself later described it (rationalized it?) as a t ap, a a to lu e s akes out of thei holes. Whatever the original intention might have been, both Mao and groups within the party came to think that liberalization had gone too far: cadres became demoralized and confused, critics started attacking the party and Mao directly (worryingly for Mao, even senior party figures like Peng Zhen and Liu Shaoqi joined in), a d Mao s p estige suffe ed: The failure of the rectification a paig [the Hu d ed Flo e s a paig ] led to a selfge e ated isis of faith i ... the CCP s go e a e, a d the espo si ilit as lea l to e pla ed o Mao. He thus fa ed t o edi ilit gaps : The a paig had ta ished his i age as omniscient hel s a of the Chi ese ‘e olutio a o g pa t e e s, a d the a paig s indecisive enactment led non-party members to question his authority over the CCP (Leese 2011, p. 63). Mao s loss of p estige can be interpreted as an exogenous devaluation of Mao-symbols. The promise, i itiall elie ed, that iti is ould ot e pu ished lo e ed the lo alt ids of less-than-truebelievers (i.e., reduced their signalling costs), and the general relaxation of controls made it possible to circulate other symbols within the space of interaction rituals. This not only threatened the identity of true believers, who suddenly had to face criticism of their Mao-symbols, but presented an opportunity for others to mobilize through interaction rituals focused on different symbols. In response, Mao attempted to reassert control over the circulation of symbols within the public sphere. He not only placed loyal supporters in control of the propaganda apparatus, but also formulated a distinction between a o e t cult of personality (indicated by the term geren chongbai 个人 崇拜) and an i o e t cult (indicated later by the term geren mixin 个人 迷信) that made it possible for him to arbitrarily punish insufficiently loyal supporters. The distinction sidestepped the theoretical problem aised Kh ush he s iti is of ults edefi i g good ults as a o ship of t uth, ut it as t a spa e tl d i e Mao s u de sta di g of the ult as a e t a u eau ati sou e of power that did ot el o its e og itio ithi the pa t elite (Leese 2011, p. 69). In other words, if there had to be solidarity rituals focused on some revolutionary symbols, Mao indicated that it better be his symbols as the ep ese tati e of t uth, ega dless of pa t ie s. As Mao said, uoti g Le i , it is ette fo 21 e to e a di tato tha it is fo ou. 36 Mao clearly saw that, in a context where most cadres were susceptible to the appeal of the symbols of the revolution (including symbols related to himself), encouraging the cult would provide him with a source of authority that was independent of the party, and that the intensification of leader-focused rituals would be a good inst u e t fo p o oti g a li el , e otio al li ate that ould oti ate people to take a g eat leap fo a d to a d o u is , just as the cult had served to motivate party members and soldiers during their struggles against the KMT. To use our terminology, Mao understood that when group members are already predisposed to incorporate ritual symbols into their identity construction, interaction rituals serve as powerful emotional amplifiers. But as e ight e pe t, as the pat o Mao aised the previously held in check, skyrocketed. As Leese notes: i i u lo alt id, flatte i flatio , ... ith the alidatio of a o e t ult it as ot e essa a o e to p aise the ki g the hole ti e, ut, so to sa , ithout e pli it p aises , as Paul Pellisso , court historian of Louis XIV, once wrote. During the early years of the PRC, praise of Mao Zedong in public discourse had a d la ge ee u ed ith Mao s o se t. But afte Ma h , efe e es to the Pa t Chairman and his thought witnessed a huge upsurge in the media, although in comparative perspective the excesses were dwarfed by the Cultural Revolutionary rhetoric (Leese 2007). Cadres wishing to prove their loyalty could now stop worrying too much about the question raised by Khrushchev of whether cults of personality were compatible with Marxism-Leninism, and hyperbolic praise of Mao a d his latest li e soo e a e a e essa i st u e t of a ee ai te a e a d advancement within the CCP, though at the beginning such praise was still carefully defined as praise of the t uth hi h just happe ed to e e odied i the person of Mao and his works). The praise soon came into conflict with reality, however. The burst of flattery encouraged by Mao led to a flood of o pletel fi ti e u e s of oth ag i ultu al statisti s a d ultu al a tifa ts i o de to signal adherence of the p o i ial ad es to the Pa t Ce t e (Leese 2011, p. 73, see also Kung and Chen 2011 for more on the consequences of Mao's implicit encouragement of radicalism). But the great famine of 1958-59 could not be hidden by mere propaganda; for those affected by the catastrophe, the evidence of the senses was of course in direct contradiction with the claims of Mao and his flatterers, which again halle ged Mao s p estige a d edi ilit a d offe ed oppo tu ities to disaffe ted people within the party. Here we have another exogenous shock to the value of the Mao-symbols. This challenge was the most serious yet to Mao s positio , i pa t e ause the fa i e fo e ted dissatisfa tio ithi the People s Li e atio A , hose soldie s ould ot e full isolated f o reports coming in from family members about the situation in the countryside. (Not even the Central Bureau of Guards, the unit in charge of guarding the leaders of the party, was immune to unrest). Even more seriously, Marshal Peng Dehuai, who had enormous prestige within the PLA, became severely iti al of Mao s poli ies. This as a i tole a le halle ge to Mao s positio , ho fea ed a oup; a d 36 Mu h late , Mao told Edga “ o that Kh ush he s failu e to de elop a ult had led to his e e tual pu ge Politburo members (Leese 2011). 22 though Peng was eventually purged, the need to regain control over the army was pressing. Lin Biao (the youngest PLA Marshal) proved the man for the job. For one thing, Lin was not shy about praising Mao, and knew how to wield the charge of insufficient adherence to Mao Zedong thought against his enemies within the party and the military. In fact, he was a le to shift the o s p e aili g at the top of the CCP so that adhe e e to Mao )edo g thought became the sole criterion of loyalty; the minimum loyalty bid could now only refer to a highly restricted set of symbols. In practice, this meant that any statements critical of Mao – uttered at any time in the past – could be used as incriminating evidence of disloyalty, and used in factional disputes which nearly destroyed the party, and served to purge many people at the top. Moreover, since most of the top party leade ship i Beiji g e e t ue elie e s – loyal followers of Mao, who held him in high esteem – rather than mere opportunists, they were also ill-positioned to defend themselves if they became the targets of attack; they could hardly count on help from other true believers, who would be apt to consider offering help to people accused of disloyalty as a form of disloyalty itself, a kind of countersignalling evidencing insufficient identification with the leader.37 Li used the ult ot o l to p ote t hi self f o the i ious ou t politi s of the CCP, ut also to discipline the army and tamp down dissatisfaction among the soldiers. The main tool he used to a o plish this o je ti e as si ila to the o igi al fo s of g oup stud that had ee used at the e egi i gs of the ult, e ept o e a o l fo used o Mao s iti gs a d o e formalized. The li el stud a d appli atio of Mao )edo g thought as i p a ti e edu ed to lea i g to e ite a d use uotatio s f o Mao s o ks as pe suasi e tools. But these interaction rituals nevertheless seem to have been very effective in stabilizing the value of the Mao-symbols among the troops and increasing the minimum loyalty bid. It s o th taki g a lose look at ho these ituals o ked. To egi ith, ontacts between the troops and their families were monitored, but they were not necessarily directly censored. Instead, reports of dist ess i the ou t side e e tu ed i to tea hi g o e ts – focal points of ritual - that extolled the necessity of staying the course and blamed unfavourable weather or the deviations of local officials from the correct line. Elaborate performances making use of all kinds of media – big character posters, theatre, films, poetry, etc. – e alled the itte ess of the past efo e the o u ist t iu ph a d e tolled the s eete ess of the p ese t though, as o e offi ial oted, ost o pa iso s of the present sweetness referred back to the period of the land reform, whereas remarks about the Great 37 Interestingly, though Lin knew how to signal his unconditional loyalty (in costly, even humiliating ways sometimes) he seems to have been no true believer himself. On the contrary, he seems not to have liked Mao much, and to have promoted the cult in part as a way of protecting himself from the treacherous shoals of politics at the apex of the CCP; he had see i Pe g Dehuai s ase ho e e the e est hi t of iti is ould e tu ed by Mao (and others) against the critic, with severe repercussions, and was determined to avoid a similar fate. Leese uotes a p i ate ote of Li s o Mao s politi al ta ti s: Fi st he ill fa i ate ou opi io fo ou; then he will change your opinion, negate it, and re-fabricate it – Old Mao s fa ou ite t i k. F o o o I should e a of it p. .B Li as adept at a ti ipati g Mao s position and changing his opinion as soon as he sensed that the old opinion was no longer operative. 23 Leap Fo a d e e i li ed to e a st a t a d ithout su sta e , p. , hile p ese ti g e a ples of communist mart s fo e ulatio . The fo us as o ge e ati g e otio e e e i g ha dships and then channelling that emotion against the enemies of the communist project to achieve bonding. The e e e also a paig s to e ulate soldie s of Mao )edo g thought, which encouraged status competition among soldiers who were already disposed to value revolutionary symbols (the heroes of Mao Zedong thought, like Stakhanovite workers in the Soviet Union, received media attention and other rewards – heap status goods, i the terminology of the model above), and hence provided a positive i e ti e to adopt the o e t so t of ide tit a d eha iou , o ple e ti g the egati e i e ti es provided by peer pressure in group study sessions or other collective interaction rituals. The combination of peer pressure, genuine emotional experiences, and threats of discipline for e al it a e as lea l po e ful, et the pa t as a a e of the da ge s of people e el a ti g as if the elie ed.38 Indeed, advice from higher ups indi ated that ad es e e ot to i sist o fo alities su h as the eepi g of pa ti ipa ts as de o st atio of thei si e it (Leese 2011, p. 100). But the very fact that such advice had to be given at all probably shows that lower-level cadres did insist on such performances (increasing the minimum loyalty bid) just to be safe (not, perhaps, primarily to eed out pote tiall oppo tu isti soldie s, ut to sig al thei own loyalty to their immediate superiors). From the army, the more intense forms of the cult spread to the broader population over time, accelerating as the Cultural Revolution started. But while ritual participants in the army could be assu ed to lie ostl o the t ue elie e side of the spectrum, the assumption did not necessarily hold fo the ge e al populatio . Co side the sto of ho the Little ‘ed Book sp ead. The Little Red Book (the Quotations of Chairman Mao) was at first confined to the army, but demand for it outside its confines was soon enormous. For one thing, political study campaigns in the countryside (which increased in the 1960s) required a focal text to mobilize people properly, and the Quotations provided one. But, as Leese astutely observes, the main thing that the Quotations offe ed as the possi ilit of empowerment for non-pa t e e s p. . This is not to say that most people in the general population were not genuinely devoted to Mao, or were complete opportunists; but, on the (plausible) assumption that the proportion of true believers was lower in the general population than in the army, the book could be used by people whose identities were not so tightly bound to the symbols of the revolution or to Mao to crack the ode that e a led one to act more or less safely within the highly unpredictable environment of the early cultural revolution. Indeed, as Leese documents later, during the early Cultural Revolution Red Guards would sometimes set up te po a i spe tio offi es o the streets and ha ass pedest ia s a out thei k o ledge of Mao s o ks. The book, in other words, provided guidance about how to make a credible minimum loyalty bid, which contributed to high demand for it; and the party enabled this demand by basically diverting the resou es of the e ti e pu lishi g se to to p i ti g Mao s iti gs, at the e pe se of e e othe p i t ite , i ludi g s hool ooks p. . 38 Yet this did not lead Mao to try to dampen flattery inflation; despite the danger of sycophancy, the emotional amplification ritual achieved proved useful. 24 Other rituals were of course important to the spread of the more intense forms of the cult outside the army. The eight ass e eptio s of the ‘ed Gua ds i e e the ost spe ta ula of these, though in some ways the least interesting. Though the Red Guards became the vanguard in the spread of the cult throughout Chinese society during the cultural revolution, the actual number of people who pa ti ipated i these e eptio s ould ha e ee uite s all elati e to Chi a s total populatio , ost of them impressionable young students who took the advantage of free train travel to get involved in something bigger than themselves. Under the circumstances, it is unsurprising that many of them reported ecstatic experiences on seeing Mao, which in turn cemented their identities as Red Guards; this sort of interaction ritual seems likely to produce this outcome fairly reliably, independently of any ha a te isti s of the supposedl ha is ati figu e. As cult rituals spread and the chaos of the Cultural Revolution deepened, however, the party lost control o e its s ols. Leese efe s to this as the pe iod of ult a a h ; in our terminology here this is the point where the patron loses control over the inflationary process. Mao was no longer able or willing to take a i i u id as suffi ie t e ide e of lo alt , leading to runaway flattery hyperinflation. Diffe e t fa tio s of ‘ed Gua ds sta ted usi g Mao s i age a d o ds i i o pati le a s, a d e cult rituals emerged from the grass roots, sometimes from the enthusiasm of the genuinely committed, sometimes seemingly as protective talismans against the uncertainty and strife of the period. Everybody appealed to Mao to signal their revolutionary credentials, but there was no longer anyone capable of settling disputes over the credibility of these signals. Mao himself was not very helpful; whenever he spoke at all, his messages were often cryptic and did not really settle any important disputes. The cult as o a ‘ed Quee a e of asteful sig alli g, athe tha a a efull ali ated tool of mobilization or discipline, driven by a complex combination of genuine desires to signal loyalty and ide tit a d fea s fo o e s se u it . As Leese notes, failure to conform to the arbitrary protocols of the ult put people at isk of ei g se te ed as a a ti e ou te e olutio a a d do u e ts a cases in which minimal symbolic transgressions resulted in incarceration or even death; in these i u sta es, the e is o safe i i u lo alt id, si e the pat o is u illi g o u a le to defi iti el take a id. By 1967, for example, statues of Mao first started to be built, something that CCP leaders, and Mao himself, had discouraged in the past, and still officially frowned upon. The statues were typically built by local factions without approval from the central party, and they were all 7.1 meters high and placed on a pedestal that as . ete s high, fo a total height of . ete s. De e e = Mao s i thda , Jul = the Pa t s fou di g date, Ma = the egi i g of the ultu al e olutio . People a i ed at this precise convention for the statues without any centralized direction, merely through a signalling p o ess . Late Lo g Li e the Vi to of Mao )edo g Though Halls e e uilt o a g a d s ale, agai without approval from the central party. Billions of Chairman Mao badges were produced by individual o k u its o peti g ith ea h othe , hi h e e the sel es su je t to size i flatio [a]s the la ge size of the adges a e to e asso iated ith g eate lo alt to the CCP Chai a , … adges ith a diameter of 30 centimetres and greate a e to e p odu ed, p. ; )hou E lai ould g u le i 1969 about the enormous waste of resources this represented. Costly signalling demands kept 25 escalating; some people took to pinning the badges directly on their skin, for example, and farmers sent lo alt pigs to Mao as gifts pigs ith a sha ed lo alt ha a te . New rituals and performances emerged too. Leese dis usses the uotatio g asti s, a se ies of g asti s e e ises ith a sto li e ased o Mao s thought a d i ol i g p aise of the eddest ed su i ou hea ts, a d lo alt da es, hi h, like the uotatio g asti s, as a g ass oots i e tio desig ed to ph si all sig al lo alt , a d hi h sp ead e e to egio s he e pu li da i g was not part of the common culture a d thus led to o side a le pu li e a ass e t p. . Pe haps o e of the ost i te esti g a ifestatio s of this i flatio a p o ess as the sto of Mao s mangoes, which seems so bizarre that it has received significant scholarly attention (Chau 2010, Dutton 2004). These a goes, gi e as a gift fi st to Mao a d the f o Mao to the Capital Wo ke a d Peasant Mao Zedo g Thought P opaga da Tea e a e the fo us of i te se i te a tio ituals, including processions were replicas of the mangoes were displayed, and ceremonies where the mangoes were boiled and the resulting mango water was drunk ceremonially by the members of various factories in Beijing. But this should not surprise us: among people who were already committed to Mao symbols like the e e s of these Mao )edo g Thought P opaga da Tea s othe o je ts elated to Mao would already have had a significant charge, while among people who were not true believers, the mangoes (rare fruit in China at the time, so deserving of some curiosity) could well serve as occasions to participate in interaction rituals that signalled loyalty at a relatively low cost. The general point to note about most of these rituals, for our purposes, is that they were not authorized by the CCP Centre, and that many of the supposed leaders of the cultural revolution (e.g., Kang Sheng, Jiang Qing, and occasionally even Mao himself) tried to curb their practice, or at best only grudgingly autho ized the afte the fa t. F o thei pe spe ti e, these g ass oots p a ti es a d ituals e e objectionable because they could not be controlled directly by them. But it would be a mistake to think that because these practices were not directed from the top, that they were therefore genuine expressions of love for the Chairman. Motivations were of course various, and there were certainly some people who were true believers – those who adopted the ide tit of ‘ed Gua ds, fo e a ple – ut hate e people s oti atio s a ha e ee the e e lea l do i ated the eed to sig al loyalty against a background of others who were also furiously trying to signal loyalty for their own manifold reasons. The clearest evidence of signalling behaviour is in fact the uniformity of the language used to flatte Mao do to the le el of si gle ph ases o e thousa ds of te ts, p. 184: boundless hot love, [o ou dless ado atio fo a less loaded t a slatio ], the reddest red sun in our hearts, etc.); the language of flattery was a code to be mastered, not a way of expressing deeply held emotions. And this escalation in the minimum loyalty bids was reinforced by the presence of a small core activist group – the Red Guards at first - that was quite capable of inflicting punishment, directly or indirectly, on those who did not conform. The party did eventually regain control over the symbols of the cult by increasing coercive penalties for diverging from approved signals. In essence, the party determined that only some arbitrary signals would be accepted as minimum loyalty bids, and all other signals would be rejected. As Leese notes, [d]e iatio s f o the p es i ed outi es e e ega ded as dislo al ehaviour and thus potentially e ge de ed d asti o se ue es p. ; but by re-establishing clear criteria about which signals 26 would be accepted as loyalty bids and which would not, the party gradually eliminated the uncertainty d i i g the ‘ed Quee aces of the earlier period. By 1971, the party had regained some control over cult symbols, Lin Biao had fallen from grace, and the party even engaged in some flattery deflation, helped somewhat by the death of Mao in 1976. (Interestingly, there was not a great deal of spontaneous public grief at the time; as Leese notes, most people were probably rather cynically disenchanted with Mao by then. The old rituals of the cult had lost their emotional power, presumably aided so e hat the sta da dizi g effo ts of the party, which routinized and formalized them). Yet the symbols of the Mao cult still circulate today in various sorts of grassroot rituals, though without any clear dominant meaning; the meaning of the symbols has fragmented, and without official encouragement or patronage resources, we find that Mao-related rituals are only one of many potential symbols to express a variety of identities and motivations – nationalism, mild rebellion against the 1989 status-quo, nostalgia, religiosity, etc. Chávez In contrast with the Caligula and Mao examples, no study of the cult of Chávez exists yet, in part because the cult continues to develop. The discussion in this section is therefore much more preliminary in character, and relies more on primary sources. For much of the 14-year period during which Hugo Chávez was president of Venezuela (1999-2013), there seems to have been no attempt to systematically develop a cult of personality around him. Indeed, Chávez himself favoured a cult of Simón Bolívar, the 19th e tu leade of Ve ezuela s independence movement. Bolivarian symbols were put into renewed circulation through such actions as ha gi g the a e of the ou t to Boli a ia ‘epu li of Ve ezuela, o issio i g a e like ess of Bolívar, and disinte i g Bolí a s e ai s to o du t a tele ised autops to dete i e hethe o not Bolívar (who is thought to have died of Tuberculosis) had instead been murdered. Nevertheless, though these symbols were thus made to circulate in the public sphere, the ritual context provided by o sta t a paig i g the e e e ajo ele tio s i Chá ez ea s i po e , i ludi g ele tio s to a constitutional assembly, a recall referendum, and a referendum on several important constitutional a e d e ts , a d Chá ez constant presence in the Venezuelan media (through his ability to o a dee ai ti e ia ade as; sho s su h as Aló P eside te – a call-in show lasting sometimes as much as six hours, where Chávez took calls from ordinary people, gave orders to his ministers, discussed ideas, and sometimes sang or recited poetry; his theatrical denunciations of other world leaders in international forums; and even his presence on Twitter) inevitably put Chávez-related symbols into circulation, and amplified their emotional charge. The mass meetings typical of electoral campaigns are well-suited to producing charged symbols of identity, especially in the hands of a skilled practitioner like Chá ez, ho as outi el des i ed as ha is ati (Merolla and Zechmeister 2011). Chá ez charisma, from our perspective here, was simply his ability to reliably producing intense interaction rituals by drawing on pre-existing symbols of identity; and as the focus of such rituals, he inevitably became an object of intense affection, even adoration, among a subset of people who participated in such rituals. Moreover, throughout its time in power, the Chávez government encouraged the formation of a number of popular organizations that, intentionally or not, provided a fertile context for the emergence 27 of Chávez-related interaction rituals. For example, early during his time in office, Chávez called for the fo atio of Cí ulos Boli a ia os Boli a ia Ci les , ea h o sisti g of a ou d -20 people (though some were much larger) sworn to defend the revolution in light of the ideals of Simón Bolívar. These Ci les e gaged i o u it o k e.g., oke i g a ess to go e e t se i es as ell as politi al work – demonstrations, participation in mass meetings, etc. – for Chávez, and most members strongly identified with him; they joined in response to his call, and were unified primarily by loyalty to him (Hawkins and Hansen 2006); indeed, the circles played an important role in the counter-demonstrations against the abortive 2002 coup that briefly removed Chávez from power. Later on, similar groups known as the U idades de Batalla Ele to al a d the attalio s of the Mi a da Ca paig pla ed a i po ta t role in the 2004 recall referendum and the 2006 presidential election, respectively (López Maya and Lander 2011). Members of these groups did not typically control large patronage resources, and hence did not at first contribute to the expansion of the Chávez cult throughout the population, but observers did point very high levels of attachment to Chávez-related symbols among group members (López Maya and Lander 2011, p. 74), including the spontaneous use of religious language to describe their relationship to Chávez (Hawkins and Hansen 2006, p. 120). Some other organizations created by the Chávez government did control substantial economic and symbolic resources, such as the well-k o Misio es so ial p og a es p o idi g lite a lasses Misió ‘ó i so health se i es Misió Ba io Ade t o a d su sidized goods Misió Me al , and there is evidence that these goods were preferentially allocated to groups and areas that signalled loyalty to Chávez. Thus, Hawkins, Rosas, and Johnson (2011) report that the location of Misiones was highly correlated with areas of electoral support for Chávez and the PSUV (and not correlated with the socioeconomic status of potential recipients), and that non-supporters of Chávez were at the very least made to feel uncomfortable enough by committed participants to exclude themselves from participation in many of these programs. For example, most of the teachers in the Misión Róbinson ee o itted suppo te s of Chá ez a d the Boli a ia e olutio , a d a used lass ti e to promote the movement, turning the programme into another set of interaction rituals where Chávezrelated symbols circulated and were positively charged with value. Similarly, community radio stations (whose formation was encouraged by the Chávez government after the 2002 coup) that aligned themselves with the Chávez government found it easier to obtain resources from the state, while stations that maintained a non-partisan attitude or welcomed non-Chavistas have been subject to social p essu e, atta ked as es uálidos o si pl ie ed ith suspi io othe Cha ista o ga izatio s (Fernandes 2011). More generally, there is also some evidence that people who signed the recall referendum petition against Chávez in 2004 have been punished in various ways by the loss of jobs or the denial of state services (Hsieh et al. 2011), and some anecdotal evidence that in communities that have benefitted greatly from government programmes, expressions of gratitude to Chávez are expected, and people who are suspected of being opposition supporters find themselves socially pressured to show evidence of loyalty. In theory, the distribution of material resources on the basis of loyalty signals, and in particular of signals of personal identification with Chávez- elated s ols, should ha e i eased the i i u lo alty id a d led to flatte i flatio . But si e the so ial p og a es i uestio al ead te ded to ta get 28 likely Chávez supporters, and levels of coercion in the Venezuelan public sphere have never been high pu li iti is of Chá ez a d the p o eso is o o ; e e a o g Cha istas iti is of Chá ez government policies was possible and even common among some sectors of the movement,39 though criticism of Chávez himself by self-identified Chavistas was much harder to find), what we see during most of the Chávez era had the characteristics of the separating equilibrium described above: with free entry and exit into rituals, devoted supporters of Chávez tended to gravitate to precisely those organizations where Chávez-related symbols circulated with a positive valence, while people who did not identify with Chávez and his revolution tended to gravitate to organizations and ritual contexts where such symbols circulated with negative valence. The result was polarization over the divisive Chávez symbols, not the monopoly over public space characteristic of a genuine cult. Nevertheless, after Chávez disclosed that he was sick with cancer in 2011, and especially after he left for Havana on 10 December 2012 to seek treatment for the last time (after winning the 2012 presidential election but before taking the oath of office), many observers (including some people sympathetic to the broader goals of the Chavista movement, like the Venezuelan historian Margarita López Maya) ega to ote that a p o ess of deifi atio of Chá ez see ed to e u de a (Díaz 2013). There were videos in heavy rotation on state TV whe e Cha ez e lai s that he de a ds a solute lo alt e ause he is ot a i di idual, he is a e ti e people, o he e people p o ide testi o ials of thei g atitude for Chavez and identify themsel es ith hi o so Chá ez ; PSUV militants (including senior party members, like Maduro, Disodado Cabello [president of the National Assembly], Tareck El Aissami [Aragua state governor], and Elías Jaua [foreign minister]), issued statements declaring that they are the sons and daughters of Chávez, that they owed everything to him, and that they are eternally loyal to Chávez. Large numbers of ordinary Chavistas tweeted thei lo alt a d o e fo Chá ez health, efe i g to hi as i o a da te o a de a d thus e phasizi g thei su o di atio a d absolute loyalty.40 The government even staged a e ti e i augu atio ee o he e thousa ds of Cha istas took the oath fo the a se t Chávez, symbolically embodying him; and, once Chávez died, the government staged a spectacular funeral were hundreds of thousands of people mourned him. The mausoleum where Chávez is interred immediately became a site of pilgrimage, with government encouragement; several public Chávez shrines appeared in Caracas, one of them in the headquarters of the state oil company, and statues of Chávez, previously unknown, started appearing in many places.41 What we see here is a classic case of flattery inflation within the Chavista movement: an increase in the circulation of Chávez-related symbols throughout the public sphere – made possible, in part, by the u h g eate hege o of the go e e t o e the edia elati e to the ea lie pa t of Chá ez ti e i offi e (the days when most of the media was opposition aligned or controlled were long gone; see MacRory 2013, Kornblith 2013) – plus an intensification of loyalty bids on the part of people who were already associated with the Chavista movement. This was not, however, contrary to the 39 See Ellner (2013) for examples of such criticism from within self-identified Chavista sectors. Fo Madu o s statements, see Maduro (2013); fo El Aissa i s state e ts, see Anonymous (2012). 41 It is worth noting that a Chávez- e a de ee a ed the use of Chá ez fa e ithout autho izatio ; though the regulation had never been properly enforced (as evidenced by the Chávez paraphernalia it is possible to buy in many parts of Caracas), statues and busts of Chávez seem to have been off-limits while he lived. 40 29 conventional analysis offered in the Venezuelan press and elsewhere, an attempt to secure the legiti a of Madu o, hose legal lai to the offi e of P eside t had ee uestio ed y the opposition; Chávez-related symbols were too divisive in Venezuelan society to secure any sort of a epta e o the pa t of people ho e e ot al ead o itted to the , espe iall gi e Chá ez absence. What seemed to be driving the process of flattery inflation at the top of the PSUV hierarchy was the fact that the absence of Chávez made it difficult for committed militants to evaluate the credibility of loyalty signals. Many observers have noted a division – the extent and nature of which is a matter of some controversy – between different factions associated with Chavismo, one of them conventionally associated with National Assembly president Cabello, who had roots in the military wing of Chavismo and has often been seen as more pragmatic, and another conventionally associated with now president Maduro, ho had o e tio s to the la o o e e t a d the i ilia i gs of the Cha ista 42 movement and has often been seen as more ideological. With Chávez incapacitated (and soon dead), a struggle was underway to define the future of the movement and the aims of the revolution. But under the circumstances, there was no one who could credibly arbitrate between potentially disparate goals and visions of socialism or revolution – no one who could set a credible i i u lo alt id. The o l symbols to which top leaders could appeal to maintain the support of the grassroots and prevent public divisions on the eve of a new election were Chávez-related symbols. The signals from the leaders of the movement in turn stimulated flattery inflation at the grassroots, while Chávez sickness acted as an exogenous shock to the value of the symbols, making them more salient to the identity of true believers. But unlike in the previous cases, no goods external to the ritual were available to promote this process; the intensification of loyalty signalling and the increase in value of Chávez symbols were driven by the e oge ous th eat to the g oup s ide tit ep ese ted Chá ez i pe di g death a d the i te sified circulation of these symbols. Ritual and Belief in Cults of Personality: Concluding Thoughts In this paper, I have attempted to show that cults of personality are best conceived as interaction ritual chains focused on leader symbols that induce varying degrees of loyalty signalling from ritual participants. From this perspective, classic questions like whether people believe in cult propaganda no longer make a great deal of sense. The most we can say is that, depending on the power of the rituals involved, some greater or lesser fraction of the population will endow leader symbols with a very high emotional charge, enough so that we might be able to say that the symbols are sacred to them, and hence that they are likely to commit resources to the defence of the leader and to the achievement of his objectives. To ask further whether they believe in the cult propaganda is to miss the point; it is not belief that drives attachment to the leader, but participation in sufficiently powerful rituals. As the rituals decay or are replaced by rituals with different focal symbols, we expect the charge of ritual symbols to diminish as well. By the same token, some fraction of the population will find the rituals empty and fail to endow leader symbols with much value, but may nevertheless feel forced to signal that they place a high value on these symbols in order to gain access to desired resources allocated on 42 For a more nuanced picture of the divisions within the Chavista movement, see Ellner (2013). 30 the basis of loyalty signals or to avoid punishment by committed individuals. Their behaviour is also not driven by belief or lack of belief in the cult. This perspective also suggests that not only are Weberian categories of legitimacy and charisma unnecessary to explain cults of personality; they are also likely to be misleading. Cults are not large-scale exercises in persuadi g a populatio of a ule s ight to ule, ut p o esses i hi h leade s ols circulate widely enough, and acquire sufficient value for some sector of a population, that certain signals of support become universally likely to be produced. The phenomenon of interest is less the content of the particular symbols that circulate throughout a ritual space than the moments of symbol-value i flatio o deflatio . We ould speak he e of a o eta ie of ults of pe so alit , i hi h ule s attempt to manage the value of their own symbols so as to achieve effects of power. References Anonymous. 2012. "Tareck El Aissami tomó posesión de la Gobernación de Aragua." El Periodiquito, 21 December. Anonymous. Pajarito endi e a Maduro y… no sé… de verdad no se puede ha er un histe on esto, cagada de país, 2 April 2013. Available from http://www.elchiguirebipolar.net/02-042013/pajarito-bendice-a-maduro-y-no-se-de-verdad-no-se-puede-hacer-un-chiste-con-estocagada-de-pais/. Apor, Balázs, Jan Behrends, Polly Jones, and E. A. Rees, eds. 2005. The Leader Cult in Communist Dictatorship: Stalin and the Eastern Bloc. London: Palgrave Macmillan. Brandenberger, David. 2011. Propaganda State in Crisis: Soviet ideology, indoctrination, and terror under Stalin, 1927-1941. Stanford and New Haven: Hoover Institution and Yale University Press. Chau, Ada Yuet. . "Mao s T a elli g Ma goes: Food as ‘elic in Revolutionary China." Past & Present no. 206 (suppl 5):256-275. doi: 10.1093/pastj/gtq020. Cohen, Yves. 2007. "The Cult of Number One in an Age of Leaders." Kritika: Explorations in Russian and Eurasian History no. 8 (3):597-634. doi: 10.1353/kri.2007.0032. Collins, Randall. 2004. Interaction Ritual Chains. Princeton, N.J.: Princeton University Press. Davies, Sarah. 2005. "Stalin and the Making of the Leader Cult in the 1930s." In The Leader Cult in Communist Dictatorship: Stalin and the Eastern Bloc, edited by Balázs Apor, Jan Behrends, Polly Jones and E. A. Rees, 29-46. London: Palgrave Macmillan. Demick, Barbara. 2009. Nothing to Envy: Ordinary Lives in North Korea. New York: Spiegel & Grau. Díaz, Sara Carolina. 2013. "López Maya: La diosificación de Chávez busca legitimar a Maduro." El Universal, 13 January. Dutton, Michael. 2004. "Mango Mao: Infections of the Sacred." Public Culture no. 16 (2):161-187. Eckstein, Arthur M. 2009. "Hellenistic Monarchy in Theory and Practice." In A Companion to Greek and Roman Political Thought, edited by Ryan Balot, 247-261. Malden and Oxford: Blackwell. Egorov, Georgy, and Konstantin Sonin. 2011. "Dictators and their Viziers: Endogeneizing the LoyaltyCompetence Trade-off." Journal of the European Economic Association no. 9 (5):903-930. doi: citeulike-article-id:12588190 doi: 10.1111/j.1542-4774.2011.01033.x. 31 Ellner, Steve. 2013. "Social and Political Diversity and the Democratic Road to Change in Venezuela." Latin American Perspectives no. 40 (3):63-82. doi: 10.1177/0094582x13476002. Fernandes, Sujatha. 2011. "Radio Bemba in an Age of Electronic Media: The Dynamics of Popular Communication in Chávez' Venezuela." In Venezuela's Bolivarian Democracy: Participation, Politics, and Culture under Chávez, edited by David Smilde and Daniel Hellinger, 131-156. Durham: Duke University Press. Gill, Graeme. 1980. "The Soviet Leader Cult: Reflections on the Structure of Leadership in the Soviet Union." British Journal of Political Science no. 10 (02):167-186. doi: 10.1017/S0007123400002088. Goffman, Erving. 1958. The presentation of Self in Everyday Life. Edinburgh: University of Edinburgh, Social Sciences Research Centre. Granovetter, Mark. 1978. "Threshold Models of Collective Behavior." American Journal of Sociology no. 83 (6):1420-1443. doi: 10.2307/2778111. Hassig, Ralph, and Kongdan Oh. 2009. The Hidden People of North Korea: Everyday Life in the Hermit Kingdom. Lanham: Rowman & Littlefield Publishers, Inc. Hawkins, Kirk A., Guillermo Rosas, and Michael E. Johnson. 2011. "The Misiones of the Chávez Government." In Venezuela's Bolivarian Democracy: Participation, Politics, and Culture under Chávez, edited by David Smilde and Daniel Hellinger, 186-218. Durham: Duke University Press. Hawkins, Kirk Andrew, and David R. Hansen. 2006. "Dependent Civil Society: The Circulos Bolivarianos in Venezuela." Latin American Research Review no. 41 (1):102-132. doi: 10.1353/lar.2006.0008. Heller, Steven. 2008. Iron Fists: Branding the 20th-Century Totalitarian State. New York: Phaidon Press Inc. Hsieh, Chang-Tai, Edward Miguel, Daniel Ortega, and Francisco Rodriguez. 2011. "The Price of Political Opposition: Evidence from Venezuela's Maisanta." American Economic Journal: Applied Economics no. 3 (2):196-214. doi: 10.1257/app.3.2.196. Jiang, Jiehong. 2010. Red: China's Cultural Revolution. London: Jonathan Cape. Jo es, Poll . . " I e Held, a d I “till Hold, “tali the Highest Estee : Dis ou ses a d “t ategies of Resistance to De-Stalinisation in the USSR, 1953–62." In The Leader Cult in Communist Dictatorship: Stalin and the Eastern Bloc, edited by Balázs Apor, Jan Behrends, Polly Jones and E. A. Rees, 227-245. London: Palgrave Macmillan. Kelly, Catriona. 2005. "Grandpa Lenin and Uncle Stalin: Soviet Leader Cults for Little Children." In The Leader Cult in Communist Dictatorship: Stalin and the Eastern Bloc, edited by Balázs Apor, Jan Behrends, Polly Jones and E. A. Rees, 102-122. London: Palgrave Macmillan. Kornblith, Miriam. 2013. "Chavismo After Chávez?" Journal of Democracy no. 24 (3):47-61. doi: 10.1353/jod.2013.0050. Kung, James Kai Sing, and Shuo Chen. 2011. "The Tragedy of the Nomenklatura: Career Incentives and Political Radicalism during China's Great Leap Famine." American Political Science Review no. 105 (01):27-45. doi: 10.1017/s0003055410000626. Kuran, Timur. 1991. "Now Out of Never: The Element of Surprise in the East European Revolution of 1989." World Politics no. 44 (1):7-48. doi: 10.2307/2010422. Landsberger, Stefan R. 2002. "The Deification of Mao." In China's Great Proletarian Cultural Revolution: Master Narratives and post-Mao Counternarratives, edited by Woei Lien Chong, 139-184. Lanham: Rowman & Littlefield Publishers. Leese, Daniel. 2007. "The Mao Cult as Communicative Space." Totalitarian Movements and Political Religions no. 8 (3-4):623-639. doi: 10.1080/14690760701571247. Leese, Daniel. 2011. Mao Cult: Rhetoric and Ritual in China's Cultural Revolution. Cambridge: Cambridge University Press. 32 López Maya, Margarita, and Luis E. Lander. 2011. "Participatory Democracy in Venezuela: Origins, Ideas, and Implementation." In Venezuela's Bolivarian Democracy: Participation, Politics, and Culture under Chávez, edited by David Smilde and Daniel Hellinger, 58-79. Durham: Duke University Press. MacRory, Robbie. 2013. "Dilemmas of Democratisation: Media Regulation and Reform in Argentina." Bulletin of Latin American Research no. 32 (2):178-193. doi: 10.1111/j.1470-9856.2012.00770.x. Maduro, Nicolás. 2013. Lealtad con Chávez: La Patria está segura. edited by Ministerio del Poder Popular para la Comunicación y la Información. Martin, John Levi. 2009. Social structures. Princeton, N.J.: Princeton University Press. Merolla, Jennifer, and Elizabeth Zechmeister. 2011. "The Nature, Determinants, and Consequences of Chá ez s Cha is a: E ide e F o a “tud of Ve ezuela Pu li Opi io ." Comparative Political Studies no. 44 (1):28-54. doi: 10.1177/0010414010381076. Mittler, Barbara. 2012. A Continuous Revolution: Making Sense of Cultural Revolution Culture. Cambridge, Mass.: Harvard University Asia Center. Myers, B. R. 2010. The Cleanest Race: How North Koreans See Themselves and Why It Matters: Melville House. New Focus International. Kim Jong-un jokes: what do North Koreans think? 2012a [cited 2 August 2013. Available from http://newfocusintl.com/kim-jong-un-jokes/ ]. New Focus International. New Focus Exclusive: 'Kid Warrior' Kim Jong-un, 9 August 2012 2012b [cited 4 August 2013. Available from http://newfocusintl.com/new-focus-exclusive-kid-warrior-kim-jongun/]. Plamper, Jan. 2012. The Stalin Cult: a Study in the Alchemy of Power. Stanford and New Haven: Hoover Institution and Yale University Press. Rees, E. A. 2005. "Leader Cults: Varieties, Preconditions and Functions." In The Leader Cult in Communist Dictatorship: Stalin and the Eastern Bloc, edited by Balázs Apor, Jan Behrends, Polly Jones and E. A. Rees, 3-26. London: Palgrave Macmillan. Sebestyen, Victor. 2010. Revolution 1989: The Fall of the Soviet Empire. New York: Knopf Doubleday. Shih, Victor Chung Hon. 2008. "'Nauseating' Displays of Loyalty: Monitoring the Factional Bargain through Ideological Campaigns in China." The Journal of Politics no. 70 (04):1177-1192. doi: citeulike-article-id:9294371. Steinmüller, Hans. 2010. "How popular Confucianism became embarrassing: On the spatial and moral center of the house in rural China." Focaal no. 2010 (58):81-96. doi: 10.3167/fcl.2010.580106. Suny, Ronald Grigor. 1998. The Soviet experiment : Russia, the USSR, and the successor states. New York: Oxford University Press. Svolik, Milan. 2012. The Politics of Authoritarian Rule. Cambridge and New York: Cambridge University Press. Tismaneanu, Vladimir. 1989. "The Tragicomedy of Romanian Communism." East European Politics & Societies no. 3 (2):329-376. doi: 10.1177/0888325489003002007. Ultimas Noticias. 2013. "FOTOS: Inauguran capilla "Santo Hugo Chávez" en el 23 de enero." Ultimas Noticias, 29 March. Wedeen, Lisa. 1998. "Acting : Symbolic Politics and Social Control in Syria." Comparative Studies in Society and History no. 40 (03):503-523. doi: 10.1017/s0010417598001388. Wedeen, Lisa. 1999. Ambiguities of Domination: Politics, Rhetoric, and Symbols in Contemporary Syria: University Of Chicago Press. Winterling, Aloys. 2011. Caligula : a biography. Berkeley: University of California Press. Wintrobe, Ronald. 1990. "The Tinpot and the Totalitarian: An Economic Theory of Dictatorship." The American Political Science Review no. 84 (3). doi: 10.2307/1962769. 33 Wintrobe, Ronald. 2000. The Political Economy of Dictatorship. Cambridge and New York: Cambridge University Press. Wintrobe, Ronald. 2008. "Dictatorship." In Readings in Public Choice and Constitutional Political Economy, edited by Charles Rowley and Friedrich Schneider, 345-367. Springer US.