494 BURKHARD GLADIGOW

then the 'success' of lifelong repetitions may be 'greater' than thesum of the individual, concrete rituals. As in the case of the inter-nal course of a ritual, in the sequence of the repetitions of entirerituals diere lies a gain in experience and 'certainty';34 this sum ofabsolved rituals cannot be denied an individual and a collectivemeaning. Complex rituals are in a certain way also 'learning sys-tems', since the experiences of older performances enter into everynew performance of the ritual and at the same time innovations canbe generated. With an increase in interna! and external complexity,a meaning of rituals 'inevitably' increases, and in large part irre-versibly: a complexity of rituals is a guarantee that their meaningscan be individualized and passed on.

34 Concerning religión as a 'security system' see B. Gladigow, "Sicherheit", MetalaLexikon Religión 3 (Stuttgart, Weiraar, 2000), 305-308.

DEFERENCE

Maurice Bloch

Anthropologists have long been puzzled by the observation that,while it is clear that rituals seem to be (in part at least) meaningfulacts insofar as some kind of non-trivial information is conveyed andis involved for both participants and observers alike, it seems verydifficult to be satisfactorily precise about what this contení might be.Many, including myself, have even suggested that a precise decodingof the message of rituals is necessarily misleading.1 Some have goneso far as to argüe that rituals are simply meaningless,2 though exactlywhat such a claim would amount to is very unclear. One reason forarguing in this way is simply that we are frustratingly and continuallyfaced in the field by informants who say that they do not know whatrituals mean or why they are done in this or that way. Nonetheless,what stops anthropologists from adhering easily to the thesis that rit-uals are meaningless is that these very same informants, who a minutebefore admitted that they did not know what elements of the ritualwere about, add puzzlingly and portentously that thesc elementsmean something very deep and they insist that it is very importan!to perform them in precisely the right way.

In light of this, anthropologists are often satisfied with making therather lame point that rituals convey something or other that is vaguebut somehow powerful. Here I want to follow a tradition in ritualanalysis that, instead of being embarrassed about vagueness, makesit its central concern. This is what I want to be precise about.Furthermore, I want to go far beyond my predecessors, myselfincluded, in arguing that the vagueness of ritual offers us a clue tothe nature of much human social knowledge and learning processes.

1 Bloch 1974; D. Sperber, Le Symbolisme en General (París, 1974); Lewis 1980;Humphrey and Laidlaw 1994.

- Staal 1979.

496 MAURICE BLOCH DEFERENCE 497

Repetition

One feature that has often been noted in discussions of ritual is thepresence of repetition.3 In fact, the term 'repetition' in these discus-sions is used to refer to quite a variety of phenomena, all of whichare commonly preseiit in rituals.

First of all, the same elements or phrases are often repeated inthe same performance. For example, in a type of the Malagasy cir-cumcision ritual, the same phrase can recur several hundred times,perhaps even more. Similarly, in Christian rituals the word 'Amen'is said many times. Secondly, there is the fact that entire rituals areoften repetitions of one another. One weekly mass is in many partsmuch the same as that of the week before. Finally, actors in ritualsguide much of their behavior in terms of what they believe others,or themselves, have done or said on previous occasions. In this sensethey are repeating either themselves or others. Indeed, in Englishany act, whether a speech act or otherwise, that appears to origí-nate fully from the actor cannot properly be called 'ritual'.

It is on repetition of this latter type that I want to concéntratehere. At least some, if not most, of the actions involved in the kindof phenomena mentioned above are understood by actors and observersalike to be repetitions; that is, they are acts, whether speech acts oracts of another kind, that do not completely origínate in the inten-tionality of the producer at the time of their performance. It meansthat what is involved in ritual is conscious 'repetition', either of one-self or, much more often and much more importantly, of others whoone has seen or heard perform the ritual before.

Familiar statements given to anthropologists by participants in rit-uals imply conscious quotation, statements such as: 'we do this becauseit is the custom of the ancestors', 'we do this because it is what onedoes at these events', or 'we do this because we have been orderedto act in this way'.

Therefore, the inevitable implication of such statements is that,both for participants and onlookers, it is not just the specific pre-sent spatio-temporal context that frames the intentionality of the actsof the ritual actor and that is relevant to understanding them fully,but also the past spatio-temporal context of specified, or unspecified,previous occurrences of the repeated or quoted acts. As Garoline

Leach 1966; Bloch 1974; Rappaport 1974; Lewis 1980.

Humphrey and James Laidlaw have put it, in a way that echoes apoint I made in an earlier article,4 "ritualization transforms the rela-tion between intention and the meaning of action."5

When, during a circumcision ceremony, a Malagasy sprays waterby way of blessing on those present, everyone knows that he or she isdoing this kind of action (the spraying) in this way, because this is'what one does', that is, it is the tradition. This means that whateverthe eider feels at the time and however he perceives the situation willbe insufficient to explain, and is well known to be insufficient to explain,why he uses water at that moment he does as he does. Comparethis with a situation where he merely reaches for water from astream. In this case most observers would find, though not neces-sarily rightly, that, given their background knowledge, the twin factsthat the person was thirsty and that he saw the water in front ofhim, that is, his beliefs and desires (in the psychological-philosophicalsense of the terms), was all there was to it.

Deference, Understanding, and Truth

Rituals therefore are acts of repetition or quotation. Such a remarkplaces ritual within what externalist philosophers have identified asa central aspect of human thought and communication6 and whichhas been called by some 'deference',7 that is, reliance on the author-ity of others to guarantee the valué of what is said or done. Whatmakes such an observation particularly interesting for anthropologistsis diat deference fundamentally alters the relation between under-standing something and holding it to be true. It seems common sensethat one must understand something in order to hold it to be true.This is not the case, however, when deference ís involved, especiallywhen deference is linked to quotation.

We can say roughly that, in pragmatic theories along Griceanlines,8 understanding meaning is seen necessarily to require not only

4 Bloch 1974.5 Humphrey and Laidlaw 1994, 90.6 H. Putnam, "The Meaning of 'Meaning'", H. Putnam, Plúlosophical Papers 2:

Mmd, Language, and Reality (Cambridge, 1975), 215-271.7 T. Burge, "Contení Preservation", Phtlosopkical Reoieuc- 102 (1993), 457-488.8 H.P. Grice, "Utterer's Meaning, Sentence-Meaning, and Word-Meaning", J.R.

Searle (ed.), The Philosopky of Laxguage (Oxford, 1971), 54-70; D. Sperber and D.Wilson, Relévame. Cammuincation and Cognition (Oxford, 1986).

498 MAURICE BLOCH

knowledge of the lexicón and of the syntax employed but also theunconscious reading of the mind of the speaker and of what (s)heintends as (s)he utters the sounds. Without such 'mind reading', thewords are, at the very ieast, so open to a wide range of ambiguitiesthat it is impossible for the hearer to process them successfully. Sucha theory is all the more interesting in that it makes the understandingof language directiy dependent on what many would now argüe isthe key distinguishing feature of Homo Sapiens Sapiens, the so-called'theory of mind' that enables a person to 'read' the mind of othersand that sepárales humankind fairly sharply from all other animalspecies.9

Quotation implies an obvious modification oí the simple Griceanprincipies just considerad. It forces the hearer to try to read not onlythe mind of the speaker but also the mind of the speaker beingquoted. Given the meta-representational ability of human beings, thisis easily done, even if we are dealing with further degrees of meta-representation.10 In this case, once the quoted sentence is understood,its truth can be considered.

Quotation offers another possibility, however. This is a kind ofabandonment of the examinatíon of the truth of the quoted statement,because one is only concemed with the fact that the statement hasbeen made and that the speaker has been identified. If this speakeris worthy of trust, one can assurne that what has been said is trucwithout making the effort of understanding. In such a case, defer-ence is combined with quotation, and it accounts for the rather oddpossibility that one may hold something to be true without fullyunderstanding it. If one trusts the source suffkiently, understandingis not necessary for the truth to be accepted, as is illustrated by thefollowing example from Gloria Origgi." She tells us of a followerwho is convinced of the truth of a statement made by a leader whoasserted that there are too many neo-Trotskyites in their party, eventhough she knows that she has no idea what a neo-Trotskyite mightbe. She will then be happy to transmit the information to another

:1 D. Premack, "Does the Chimpanzee Have a Theory of Mind? Revisited",R. Byrne and A. Whitten (eds), Machiavellian Inlelligence. Social Expertise and the Evolutwnof Intellect m Monktys, Apes and Humans (Oxford, 1988), 160-179.

'" D. Sperber (ed.), Metarepresentations. A Multidisdplinmy Perspeclive (Vancouver Stucliesin Cognitive Science 10; Oxford, 2000).

" G. Origgi, "Croire sans Comprendre", Cahiers de Philosophie de L'Unioersité deCaen 34 (2000), 191-201.

DEFERENCE 499

without understanding it. This might seem an unusual scenario, buta moment's reflection will confirm that we are all, to varying degrees,in much the same sort of situation most of the time.

Deference and Social Life

What is particularly interesting for anthropologists about an examplesuch as the one just given is that not only do such occurrences cropup continually but that their occurrence is not random in the courseof social life. Situations in which the truth of certain propositionsare to be accepted through deference, and therefore not necessarilyunderstood, are socially and culturally organized and regulated. Livingin a partially institutionalized form of life—which is what it meansto live in society—means that there are moments, concepts, contexts,in which one may examine the whys and wherefores and moments,concepts, and contexts in which this is inappropriate. For the reasonwe have seen above, this means that the latter need not be understood.

Thus, social life 'manages' the occurrence and the nature of def-erence through different institutional devices and thus establishes atthe same time an economy of the necessity of understanding. It isclear that living in a socially organized system, even the apparentlymost ad hoc system, nonetheless involves moments of compulsory def-erence in the sense used above. There are moments when there arelimits not only to understanding but also to the appropriateness ofattempting to understand. This means that all normal human com-munication involves a mixture of searching for meaning (our ownand diat of others) and also not searching, moments of understand-ing and not understanding. When young children exhaust their par-ents by endlessly asking why-questions, they may well be trainingtheir judgment of when to search and when not to search.

We have seen why deference makes it possible to hold somethingtrue without understanding it, but there is also a reason why sociallife makes this abandonment of the search for meaning common.Namely, the experience of living in a historically constructed systemmeans that deference continually occurs without it being possible toidentify easily to whom one is deferring. As a result, intentionalitycannot be 'locked' onto an intending mind, and therefore understandingcannot be 'clinched'. People around us and we ourselves are clearlydeferring to others. But if we were so unwise as to want to examine

500 MAURICE BLOCH

these others more closely, they would turn out to be deferring toyet others, and so on, without the process liaving any clear termi-nus. This is because humans live and act within a set of conven-tions that are no doubt the product of a long historical process ofcommunication and quotation and that are experienced as 'given',that is, without specific minds intending them. These are the con-ventions that have been so internalized that they have become com-pletely unconscious. Anthropologists sometimes cali them 'culture' or'habitus' and sometimes by other ñames, such as 'structure'. In otherwords, we are continuaüy deferring to others but we do not catchsight of the minds to which we are deferring. For what we are read-ing are not simply human minds but historically constructed humanminds. We do not simply understand others and ourselves; we always,to varying degrees but semi-consciously, understand that peoplearound us are deferring to invisible and indeterminable others andthat therefore we should limit our attempt to understand them.

Of course, such indeterminate deference is unconscious much ofthe time, though not always. The example of the follower who acceptsthe belief about the neo-Trotskyites ¡s a case where it is quite pos-sible that the act of deference becomes conscious, although here theperson deferred to is clearly identified. What difference this con-sciousness of deference means, has not been, to my knowledge, muchexplored in pragmatics or in philosophy, but since it is so prominentin ritual and religión and so closely linked with the question of exe-gesis, we shall have to consider the question.

Thus there are three elements in human communication whichcan be combined: 1) quotation and deference, 2) consciousness ofdeference, and 3) lack of clarity on the person to whom one is defer-ring. When all three are present, we have the phenomena that inanthropological English are commonly referred to as ritual. Becausethe combination of these three elements is likely to lead to limitedunderstanding, it is not surprising that this state of affairs is frequentin ritual.

Deference and

Now we have the tools to examine what all this might mean for rit-ual and religión. At first, I shall examine two apparently simple def-erence scenarios, both of which correspond closely to Origgi's example.

DEFERENCE 501

The first concerns learning the Quran in Muslim schools and thesecond concerns spirit possession.

Reading or reciting the Quran, which is the central purpose ofMuslim education, apparently involves a simple type of quotation onthe part of the student since the speaker is merely quoting one sin-gle other intentional mind: that of God to which he defers totally.'2

Ideally, the student should learn the Quran perfectly by heart andso become a totally transparent médium, just like Mohamed him-self. He should become a sort of tape recorder, so that his inten-tionality, and thus his understanding, disappear or become irrelevantto the text. As a result, the speaker or the hearer can focus entirelyon the presence of God in the words. The student should effacehimself as much as possible.

Another example of such ideally 'transparent' quotation is spiritpossession. Theoretically, the utterers of sound have totally surren-dered his or her body, and especially his or her vocal organs, to thebeing who temporally possesses them. In this case, too, the sourceof the emission of the sound should ideally disappear. Asking thestudent of the Quran, or the médium, to explain his or her choiceof words or contení, that is, to provide an exegesis, would clearlybe to deny their complete deferral.

These two examples may seem simple, but in fact they involvetwo quite different elements. Both the pupil learning the Quran andothers around him or her believe that what is proposed there is trucand that they must assert it, whether they understand it or not. Themédium has so effaced him- or herself that the assertions that comefrom his or her mouth must be true, precisely due to their spiritualsource irrespective of his or her understanding. This is straight-forward. However, one might expect that such practices would simplyplace at one remove the effort to understand. Having gotten pastthe pupil or the médium, it should be possible to concéntrate onunderstanding God or the spirit. This, however, does not seem tobe the case ethnographically. In such practices the act of deferraltakes center stage, and everybody joins with the pupil or the médiumin abandoning their intentionality and in making themselves trans-parent to whomever's words they are quoting, which strangely fadeout of focus.

12 D. Eickelman, "The Art of Memory. Islamic Educación and Social Reproducción",Comparaíim Studies m Socicty and Histon 20 (1978), 485-516.

502 MAURIGE BLOCH DEFERENCE 503

In such cases we have two of the elements isolated above: defer-ence and the consciousness of deference, even though this may ulti-mately disappear. What is not present, however, is the third elementdiscussed above: the indeterrnination of the originating mind. It isclear that it is God who is the source of the Quran or it is GreatGrandmother who is the spirit. But what happens when such definitiondisappears? It is to this that I now turn.

Ritual

It is the presence of the third element which characterizes much rit-ual ancl, more especially, those ritual elements that are most stronglyresistant to exegesis. In such instances quotation, and therefore def-erence, is obviously taking place, but it is not clear who is beingquoted or deferred to.

As noted above, a very common experience among anthropolo-gists who ask why someone is doing something in a particular wayin a ritual is that these questions are answered with such phrasesas: 'It's the tradition', 'It is the custom of the ancestors', or 'It goesback to early history'. Now, these apparently frustrating answers arenonetheless interesting in many ways, for they combine explicitnessconcerning deference and awareness of the imprecisión about whoexactly is the originating mind behind the practice.

If the participants, or the observers, engaged in such rituals as themass, the Malagasy initiation ritual, or making the sign of the cross,try to work out who intended what they are doing to be exactly so,they are going to be in a difficult situation.

The search for original intentionality is in itself perfectly reason-able, and although frustrating, almost inevitable. After all, we aredealing with people with human minds, that is, with an animal whosemind is characterized by an intentionality-seeking device that is nor-nially exercised ceaselessly (one might almost say obsessively), some-times consciously but often unconsciously, and that enables them toread the minds of others and thus coordínate their behavior withthem. But in a ritual these poor animáis, including the poor anthro-pologists, appear to be faced with an impossible situation becausethe search for intentionality leads them ever further back, to evermore remote authorities, but without ever settling anywhere withany finality. This is the predicament both oí participants who might

unwisely ask themselves why on earth they are doing this or thatand of mere onlookers who ask the same question.

This Kafkaesque nightmare of being endlessly referred back toother authorities can be renclered bearable only in one of three ways.

The first is the most straightforward. One can attempt simply toswitch off the intentionality-seeking device, an attitude that could bedescribed as retreat or 'putting on hold' or 'letting things be'. Thisswitching off requires some effort since, given the way our mindworks, it is unnatural, but it can be done nonetheless. Saying thatyou do what you do, or say what you say, because of 'tradition'may in some cases be nothing more than an expression of this atti-tude. The refusal to look for intentionality, however, presents theparticipants with a disappointing propositional thinness. It is as if,when one is very tired and kept awake by a hubbub of voices, oneapparently makes out somebody or other saying 'raindrops are Jesus'srnacaroons'. In such a situation, the person might make no efTort todiscover the intentionality of the speaker and hope to go to sleep asquickly as possible. The only thing that the person has from the sit-uation is the realization that it involves the use of proper language,therefore probably that it has potential for speaker meaning. Glearlyhere there is no understanding, and it is far from clear whether any-body in such a situation even holds the proposition to be true.

The second possibility is much more common, but will also appearin a number of somewhat exceptional situations, one of which isbeing faced by an over inquisitive anthropologist, though not theonly one. Then, for some reason, it will seem necessary to make aneffort to understand what is going on. At first, one is tempted tosearch in the dark recesses behind the producer of the ritual acts,whom we know, after all, is only quoting someone, somewhere, whomight have meant to mean something. (Doing this without payingattention to informants in anthropology is called 'functionalism'.) Butit's dark back there: as soon as someone seems to come into focus,he or she becomes transparent since the person seen reveáis anotherperson behind hirn or her. He or she is only deferring to someoneelse, further back, who, when focused on, becomes similarly trans-parent, and so on. Finally, we give up searching for meaning, thoughnot in the same total way as the giving up discussed in the first case,and for the foliowing reason.

All this frustration occurs only due to the difficulties encounteredwhile searching for the intentionality of the initiator of the message.

504 MAURICE BLOCH DEFERENCE 505

By contrast, the intentionality of the speaker, the singer, or the actorin the ritual is not more problematic than those of the spirit médiumor the pupil learning the Quran discussed above. The intentionalityof all these people can be read simply as deferral, and this act isgreatly valued. The search for intentionality is therefore switched tothe unproblematic examination of the intentionality of the transmit-ter, the situation that Humphrey and Laidlaw describe for the JainPuja.13 And vvhen people tell us that they do not know what suchand such a phrase means, or why such and such an act is per-formed, but that it is being said or performed in this way becauseone is following the customs of the ancestors, they are surely tellingus that what they are doing, saying, singing is, above all, deferríng.

This brings us to the third possibility. Namely, that even this solu-tion to the problem may be unsatísfactoiy. In rare, but important,moments people are going to ask themselves, or others, why thingsare done or said in this or that way, and they will not give up inspite of the apparent difHculties encountered in their search. Theirmind reading instinct will just not leave them alone. Thus, one wantsto attribute speaker meaning to what is going on, but to do thatone must inevitably créate some sort of speaker. A normal speakeris not available since such a speaker would become transparent assoon as he or she is considered and will therefore perform the dis-appearing act discussed above.

The solution to the problem of wanting to lócate meaning with-out having normal originators of that meaning is to merge all theshadowy transparent figures into one phantasmagoric quasi-personwho may be called something like 'tradition', 'the ancestors as agroup', 'our way of doing things', our 'spirit', our 'religión', perhapseven 'God'. These are entities to which 'mincls' may be attributedwith some degree of plausibility, thus apparently restoring intentionalmeaning to the goings on of ritual. The apparent specificity of suchentities thus appears at first to solve the problem of the indeterminacyof the intentional source. After all, we are familiar with the attribu-tion of human-like intentional minds to things like mountains or deadpeople,14 so why not attribute such a mind to an essentialized tra-

13 Humphrey and Laidlaw 1994." P. Boyer, "Whaí Makes Anthropomorphism Natural. Intuitive Ontology and

Cultural Representations", The Journal of the Roval Anthropological Institute n.s. 2 (1996),83-97.

dition? And here we are faced with a situation that is somewhatsimilar and somewhat dissimilar to that of the second case discussedabove. It is similar in that the message is held to be true regardlessof whether it is understood. Again, the act of deference is consciouslypresent and valued in and of itself. However, the act of deferencedoes no hold center stage as much as in the second case, becausespeaker meaning becomes an alternative poiiit of interest. Nonetheless,this is no ordinary speaker meaning to the extent Üiat the 'speaker'is no ordinary mind but is instead an essentialized conflation. Infact, I would propose that the precisión of our understanding varíeswith the degree to which the phantasmagoric initiator is cióse ordistant in nature to ordinary minds. Thus the mind of an entitycalled 'the tradition' would be more difficult to interpret than thatof an entity called 'the ancestors', but the latter are themselves prob-ably more difficult to interpret than a singular spirit, simply becauseplural minds are not what we are equipped to understand readily.

The three variants just discussed are, of course, not distinct in timeor place. Individuáis may slide from one to another during a par-ticular ritual, and not everybody present will do so completely. How-ever, the form of the ritual and the entities invoked will ensure thegeneral organization in most people's minds of relative degrees ofunderstanding. This is because the problems of attributing clearmeanings to what is done all result from the central fact that ritualinvolves high degrees of deference.

However, as we have seen, deference is a common aspect of humanlífe. It occurs whenever we do something, or believe something tobe true, and rely thereby on the authority of others—something wedo constantly. If people are always partly, but very significandy, liv-ing in a sea of deference, this is largely an unconscious fact. But itis nevertheless a fact that hovers not very far from the level of con-sciousness, and which can, and often does, cross' into the level ofconsciousness. As Hilary Putnam stressed,15 people are almost con-scious of the fact that they are constantly relying on the under-standing of others and that they normally act in terms of beliefs thatthey do not fully understand but that they hold valid because oftheir trust in the understanding of others. People therefore allowthemselves to depend on others. By and large, this is a good feeling,

Putnam, "The Meaning of 'Meaning' ".

506 MAURIGE BLOCH

while at other times it is oppressive. But when one is in trouble anddoes not know what to do, one allows oneself to be taken over bythe knowledge and the authority of others. It is only sensible to doso, and there is not much else that one can do.

Now I am arguing that ritual is just that, in a rather extremeform. Rituals are orgies of conscious deference. But if this is so, thesearch for exegesis is always misleading. This is not because it isimpossible. Glearly, exegeses exist, whether prívate or shared, whetherthe secret of experts or available to all, whether conscious, semi-con-scious, or unconscious. But the exegeses are beside the point of thecentral character of ritual: deference.

DYNAMICS

Bruce Kapferer

This discussion of the dynamics of ritual concéntrales both on the innerprocesses of ritual and the dynamics of the relation between ritualand the realities that are part of its larger context. I do not opposedynamics to statics, for what appears to be static, repetitive, orunchanging in ritual is nonetheless a product of the particular dynam-ics of its action, which has the capacity to effect changes in the expe-rience of participants, as well as within the wider social and politicalcontexts in which rituals are enacted. Overall, the concern with thedynamics of ritual is with the organizing or structurating practicesor techniques through which it intervenes pragmatically (often con-stitutively) in lived experience.

Any exploration of the dynamics of ritual turns on the issue ofwhether ritual is a distinct phenomenon in its own right. Does theconcept of ritual in fact occur as an actual discrete phenomenon, oris the concept merely an anthropological construct with its roots inthe European Enlightenment, colonialism, and the emergence of ascientific and technological age?1 Undoubtedly, it is ritual as a concept,as an analytical tool for understanding psychological and social phe-nomena, that is vital. The invocation of the term 'ritual1 to describean increasingly vast array of practices engages a particular analyti-cal attitude. Ritual exists in the anthropological imagination. Howit is imagined has implications for the analysis that follows the nam-ing of an event as being ritual. The definitions of ritual outline howrituals are conceptually imagined and how ritual as a relatively dis-tinct practice is recognized. Such definitions are contingent on theo-retical commitments in a diversity of disciplines from anthropology(where the concept has had a central status) vía theology and per-formance studies to ethology and socio-biology. In other words, ana-lytical or theoretical orientations independent of whatever thephenomenon may be in itself (as an actual empirical practice somehow

See Asad 1993; Bell 1997; Tambiah 1981.