Religion and Reason
General Bditor Jacques Waardenburg, Lausanne
Board of .Advisers
R. N. Bellali, Berkeley - M. Despland, Montreal - \V. Dupré, Nijmegen
S. N. Eisenstadt, |erusalem - C Geertz, Princeton - U. King, Bristol
P. Ricccur, Paris - M. Rodinson, Pans - K. Rudolph, Marburg
L. E. Sullivan, Cambridge (USA)
Volume 42
Walter de Gruyter • Berlin • New York
New Approaches to the Study of Religion
Volume 1: Regional, Critical, and Historical Approaches
Edited by Peter Antes, Armin W. Geertz, Randi R. Warne
Walter de Gruyter • Berlin • New York
m
t
Meaning and Religion
On Semantics in the Study of Religion
by
Jeppe Sinding Jensen
Preamble: Meaning between Essence and Construction
It is common knowledge, among both "lay and clergy," that religions are generally considered to be meaningful phenomena in human affairs. But just what that assertion means is a matter of much contention—it has, in fact, been so for quite a long time. The religious point of view, that of the practitioners of religious traditions, is that the traditions and teachings etc. indicate the ways of things—what they are and have always been. A religious world view furnishes human habitats with forms and modes of knowledge, with ascriptions of meanings, so that we may "know the world." Speaking in very broad and loose terms, a religious worldview will maintain that the knowledge we have is "given" —it consists of true propositions about the cosmos—and, in turn, the cosmos is only such because it consists of true propositions. There are of course many other aspects of religion and religious practice than assertions about the values of propositions, but let that rest for the moment since the main argument here concerns the status and functions of religions as semantic entities, whether these are seen as sanctified "givens" or as human inventions. The entire question concerns religion as a "meaningful" phenomenon and whether the "mean-ingfulness" of the phenomenon rests on conventions or on truths. As such, the question of meaning is intimately related to the case for or against "religious language," a problem which has been prominent in certain parts of the world and in their traditions' struggles with the processes of modernization.1 In that process sanctified "givens" lose their "meaning;" that is, their reference is questioned and the remainder becomes an object of historical-critical commentaries. Certainly, positivists and logical empiricists could point to "religious language"
1 In this context "modern" refers to the loss of credibility of religious/transcendent/ metaphysic discourse and the loss of authority of its referents—so that revelations become "old books." "Modernization" is a common term for this process (in many ways akin to "secularization").
220                                                   Jeppe Sinding Jensen
as being meaningless because it has no true and verifiable referent(s) in the systems of physicalist or naturalist conceptions of meaning, but others point to the possibilities of "religious language" as speaking about other matters which are (perhaps) more existentially true than the "truths" purveyed by the natural sciences. So, there could (perhaps) be more than one kind of truth. If religious language about god was somehow disqualified by lack of reference (i.e. "non-cognitive" in philosophical vernacular), then a possible candidate for reference could be the "human condition," and religious language could then be the expression of existential concerns.'
However, the present exposition of semantics in religion(s) is not primarily concerned with "religious language" and the ontological or epistemological possibilities of "god-talk" or "deep meanings" about human existence. Here, the object of concern is the investigatipn of what it entails to view religions as socio-cultural facts and thus as semantic phenomena. Socio-cultural facts are semantic because of the simple and compelling reason that there would be no human social or cultural "worlds" without communication, without symbolic and linguistic mediations— the products of our "semiosis"— our "sign-making." And all our historical knowledge points to religious traditions being the warrants of symbolic and conceptual stability in social systems—as well as being legitimators of schisms, violence and revolutions, but then that normally means re-instating a conceived "original" conceptual and symbolic state of affairs... This last aspect also discloses the fact that even the most stable and sanctified "givens" are the objects of human manipulation —even a death penalty for changing the traditional materials of myth does not "work" (van Baaren 1972: 200). The histories of religious traditions also present us with seemingly endless possibilities of religious innovations in schisms, sects, heresies and other modes of re-interpretation. The semantics of religious systems are as flexible and multi-facetted as are all other forms of semantics. The question seems to be whether there really is any such specific entity as "religious language" and/or whether the semantics of religious systems are just "plain" semantics of an order similar to other specialized terminological systems, those of, say, politics, sports or economics. When speaking of such more or less ideological systems as politics, sports and economics it is easy to see
2 These debates have been around for quite some time and they shall not be restated here, but see Frankenberry 1999 for interesting remarks on the "two level theories of truth." The contents of a majority of text-books on "philosophy of religion" seem to indicate that one major problem in that field is whether it is "rational to talk about god" —a typical legacy of analytic philosophy. I would say that it can be, but that does not "prove" anything concerning god's separate ontological status, but only that as a feature of our universe it is as rational to talk about god as about any other metaphysical entity, say "art." From a completely different angle it has been pointed out that religious representations have a high degree of "naturalness" (Boyer 2000).
t         Meaning and Religion: On Semantics in the Study of Religion                   221
"human hands at work" and the same applies to religions: that they are local knowledges and technologies concerning the many aspects of human social life, including the "invisible" worlds of traditional cosmologies. But such functional analogies are only possible because religious systems do "say" things, because they are expressive and communicative (they are also other "things" but these will be the topics of other contributions to the debate). Religion is "talk" but it is not idle talk, it is pragmatic and performative, in various degrees and shades, but nevertheless active in the sense that humans can employ the vocabularies and classifications of religious traditions as repertoires of rules and meanings as relevant to their own descriptions of their actions (including thought). As such, religions are ontologically no different from other socio-cultural systems of representations—the crucial difference lies in their purported reference to "otherworldly" agents but as long as we are concerned with religions in a semantic perspective (using texts and statements in whatever medium) that distinction makes no difference because a "semantic definition of religion remains at that same level of meanings represented by the texts and their related symbolic universes, without seeking validation or reference in theories about "something else," be it society, the psyche or the sacred" (Jensen 1999: 409). Why that is so will be a major underlying topic in the following presentation and discussion. It is also noteworthy that most works on theories of religion do not contain the word "meaning" in their lists of contents or subject indexes-and it could hardly be because they simply disregard the "meaning"-aspects of religion.3
1. What Is "Meaning" —the Scope of Semantics
"Meaning" is one of the most tricky words in the English language-a multi-referential one —which would have meant less if it were not such a central term. It is quite apt here (in case the reader not fully recalls or has it at hand) to quote a standard work of authority, The Oxford Thesaurus:
meaning n. 1 sense, import, content, signification, denotation, message, substance, gist: The meaning of the word lucid' is clear. 2 purport, implication, drift, spirit, connotation, significance, intention: You understand my meaning, so I need not explain. 3 interpretation, explanation: What is the meaning of my dream about being attacked by my philodendron? (1997: 279).
Two examples are Critical Terms for Religious Studies (Taylor, ed. 1998) and Guide to the Study of Religion (Braun/McCutcheon, eds. 2000). The latter contains a contribution by Hans H. Penner, which relates to the issues-but it is not made explicit, nor does "meaning" appear in the index.
222
Jeppe Sinding Jensen
Although nouns, these "synonyms" are obviously members of quite different classes of terms, but most are concerned with actions (the list of synonyms of the verb to "mean" is much more extensive, ibid.) of either "intending" something, "pointing" to something, "expressing" something or of something "having importance" or "purpose." And it should come as no surprise then that the controversies over "meanings" are often traceable to a lack of definitional precision.4 Consequently as meaning may have so many different meanings (in 1923 Ogden and Richards listed 22 definitions) I shall refer—henceforth —to "semantic content": that which is commonly meant by word meanings. Word meanings are concepts (most likely) and like other concepts they have two sides: intensions as contents in relations to other concepts and extensions as referents, that which they are about. These distinctions should, however, be treated with caution and only used as a rough guide as to what they "are about."5 This is so because we cannot divide the world into one realm of things and another containing the words about the things, although that is what common-sense (a Cartesian legacy in this case) seems to suggest: Things are in the world and meanings are "in the head"—but that is not how it is anymore. There have been some momentous developments in the study and understanding of meaning, and much of that development was set in motion by the Czech priest Bernhard Bolzano (d. 1848) who may well be said to be the founder of the modem semantic tradition, as he was: "the first to see that the proper prolegomena to any future metaphysics was a study not of transcendental considerations but of what we say and its laws and that consequently the prima philosophia was not metaphysics or ontology but semantics." (Coffa
4       Simon Blackburn's Oxford Dictionary of Philosophy (1994) summarizes the problem like this: "meaning [bold in original] Whatever it is that makes what would otherwise be mere sounds and inscriptions into instruments of communication and understanding. The philosophical problem is to demystify this power, and to relate it to what we know of ourselves and the world. Contributions to this study include the theory of speech acts, and the investigation of communication and the relationship between words and ideas, and words and the world. For particular problems see content, ideas, indeterminacy of translation, inscrutability of reference, language, predication, reference, rule following, semantics, translation, and the topics referred to under headings associated with logic. The loss of confidence in determinate meaning ('every decoding is another encoding') is an element common both to postmodernist uncertainties in the theory of criticism, and to the analytic tradition that follows writers such as Quine" (235-36). Another possibility is to treat meaning as a metaphysical "primitive" and underived notion, as that which involves translatability and which founds human intentionality.
5       For instance: Intensions may serve as referents, because concepts cannot avoid also referring to other concepts, and referents can be the intensions of other intensions, which often happens in metaphors: e.g. a "white dove" is not just a bird but a symbol of peace.
Ä         Meaning and Religion: On Semantics in the Study of Religion                   223
1991: 23). And that suggests that a brief history of meaning might put certain matters in perspective.
1.1. A Very Brief "History of Meaning"
In philosophy most things begin with Plato and Aristotle and so it may here.6 On the early history of "meaning" Dan Stiver says:
The philosophical paradigm rooted in Greek philosophy that has predominated in philosophical thought until modern times has three components. First, meaning lies in individual words. Second, the meaning of words is primarily literal or univocal, which implies that figurative language must be translated into literal language in order to be understood. Last, language is instrumental for thought. The first point implies a kind of verbal atomism, often promoting a neglect of the wider context of words and their use. The second promotes the relegation of figurative language for the most part to secondary status, significant at best not for cognitive purposes bur for persuasive (rhetoric) or aesthetic (poetic) purposes. A corollary of Aristotle's approach thus was that the meaning of figurative language can be grasped only if it can be transposed or reduced to literal language ... The third point recognizes the priority of thought. The idea was that thinking is in some ways a separate process from speaking, as evidenced by the common experience of seeking for the right word to express a thought. The effect also was to make language, philosophically speaking, secondary and less significant. (Stiver 1996: 11)
For centuries, the dominant theories of meaning were "correspondence theories"—based on the idea that meanings are "made" by correspondence between word and thing or concept. In religions, "sacred meanings" are those which corresponded to revelation and to gods' intentions. But such theories also center on the transparency of language, of words and their meaning in search for the corresponding referents, which could be (as in Platonism) the "eternal" forms to be "seen" more than spoken about. Language was an instrument of thought, and the ideals, from Antiquity to Descartes (and on...) were clarity and certainty. These would not, of course, always be the most prominent features of religious statements and thus Thomas Aquinas (d. 1274) held that religious language was analogical and purveyed not literal but a different kind of truth; as in the example of Jerusalem, which is literally a town on earth, allegorically the Christian church, topologicals (i.e. ethically and morally) the soul and anagogically (i.e. eschatologically) the future heavenly city of God.
6 This is but a very brief overview —the "history of meaning" is also contained (and to some extent "hidden") in the general histories of semantics in philosophy and linguistics, of hermeneutics, semiotics etc.
224
Jeppe Sinding Jensen
1.2. Positivisms, Empiricism and Verificationist Theories of Meaning
The predilection for the precise and empirical tracing of the meanings of words was one of the hallmarks of historical philology: the meanings of expressions were to be found in their genealogies and in their origins and such was also commonly the practice of explaining ancient myths. Numerous are the exercises trying to explain what Greek gods "really" were in the beginning.7 This etymological approach culminated with Friedrich Max Müller and the "school" of comparative and nature mythology. It was also largely driven by the discoveries in comparative Indo-European linguistics and thus paved the way for both comparative and historical studies in which meanings are contextually determined. To the philologians, theologians, and historians of religions in the nineteenth century, "origins" were the most cherished objects of reconstructive study, but the historicist paradigm also indicated that things were to be understood in their historical context (in so far as these could be reconstructed). The romanticist perspectives presented, simultaneously, the relations between language, linguistic meaning and culture as larger entities (communities of interpretation) and such a view laid a foundation for the much later ideas of cultures as "systems of symbols."
A very different and strictly empiricist view of linguistic meaning was introduced in the wake of positivism. The general impetus was to view and model natural language (i.e. the form of language that ordinary humans speak) along the lines of formal logic in philosophy and mathematics. This rigorously positivist attitude to "meaning" reached its apogee in the theories of the logical empiricists of the "Vienna Circle" —for instance Rudolf Carnap for whom mathematical logic represented the grammar of the "ideal" language. Semantics analysis consisted in linking symbols with sense impressions and directly observable empirical matter in the world. Judged from such a view, "natural language" inevitably founders —and those parts concerned with moral, political and religious issues were considered "meaningless." Or, as it may also be termed: "non-cognitive" —that is, about "nothing" (scientifically) real. The doyen of empiricism, A.J. Ayer, can thus assert—on "religion as non-sense" — what the difference is between saying that one sees "a yellow patch" or that one sees "God." For, whereas the first concerns:
a genuine synthetic proposition which could be empirically verified, the sentence „There exists a transcendent god" has, as we have seen, no literal significance. We conclude therefore, that the argument from religious experience is altogether fallacious ... The theist, like the moralist, may believe that his experiences are cognitive experiences, but unless he can formulate his „knowledge" in propositions that are
7      See e.g. the now "classic" works: The Greeks and Their Gods by W.K.C. Guthrie; A Handbook of Greek Mythology by H.J. Rose.
Meaning and Religion: On Semantics in the Study of Religion                  225
empirically verifiable, we may be sure that he is deceiving himself. It follows that those philosophers who fill their books with assertions that they intuitively „know" this or that moral or religious „truth" are merely providing material for the psycho-analyst. For no act of intuition can be said to reveal a truth about any matter of fact unless it issues in verifiable propositions. And all such propositions are to be incorporated in the system of empirical propositions which constitutes science. (1952:119-20)
So much for "religious meaning."8 The rather prolix and frustrated efforts of the positivists at formulating theories about possible criteria of meaningfulness for the natural languages in terms of verification of propositions and sentences slowly filtered out in the 1950s. But in the general (mostly Western) public the "scientistic" attitude persists, not least in regard to religion, which is often and in toto considered meaningless because of its lack of "verifiable reference."'
To positivists and empiricists, religious or other symbolic meaning belongs (if anywhere at all) "in the head": it is a psychological phenomenon pertaining to the realm of the subjective, a "something" as individual and private as are other entities of the world of (scientifically ungrounded) opinion and taste. It is true that meanings are not "things found glistening on the beach"10 and this non-'thingyness" seem to prompt uneasiness in some of a more empirical persuasion. Even more so because meanings are but human ascriptions—they are not essences in objects, for meanings are not in things — meanings are produced or constructed. It is a human propensity and a decisive feature of our inten-tionality to endow meanings to whatever "there is" and to what we can imagine, but that does not make the study of meaning any less demanding. But empiricists are still uneasy about this and, for instance, Frits Staal has lamented the lack of direct empirical reference and verification for "meaning":
Although every zoologist knows that an elephant does not have meaning, students of the humanities and social sciences attach meaning to many expressions and mani-
as Ayer does leave some space for the study of these matters, although not in philosophy: "one should avoid saying that philosophy is concerned with the meaning of symbols, because the ambiguity of 'meaning' leads the undisceming critic to judge the result of a philosophical enquiry by a criterion which is not applicable to it, but only to an empirical enquiry concerning the psychological effect which the occurrence of certain symbols has on a certain group of people. Such empirical enquiries are, indeed an important element in sociology and in the scientific study of a language; but they are quite distinct from the logical enquiries which constitute philosophy" (1952: 69).
9       Edmund Leach's (now almost classic) remarks on this problem deserve mention: "Religious statements certainly have meaning, but it is a meaning which refers to a metaphysical reality, whereas ordinary logical statements have a meaning which refers to physical reality. The non-logicality of religious statements is itself 'part of the code,' it is an index of what such statements are about, it tells us that we are concerned with metaphysical rather than physical reality, with belief rather than knowledge" (1976: 70).
10     Clifford Geertz 1995: 62.
226
Jeppe Sinding Jensen
testations of humanity as if they were linguistic expressions ... Underlying such ideas is the assumption that man must make sense. To an unbiased observer, much of that sense seems arbitrarily assigned, evanescent, or due to chance. The available evidence suggests that man can make as much sense as he likes, but does not make intrinsic sense. (1989: 454)
Staal's problem is a very common one when meaning is conceived in the light of a particularistic and referential semantics: for where is meaning if not in the things themselves —as "intrinsic"? Where is the meaning of an elephant if not in the elephant? What is meaning if not a property of the object itself? Logically then, if the "foundation" of meaning is not in the empirically verifiable world of objects, it must be in the minds of actors and observers and so the only decent solution is to talk about meanings as "meanings for someone" inside their skulls. Thus for the frustrated "semantic empiricists" meanings become psychological entities. Furthermore, if and when it turns out that actors in rituals have no interpretative knowledge of the languages, spells and mantras employed —then there seems to be no meaning "at all." However, as it will be argued below, meaning and meaningfulness are not psychological entities or qualities, and actors' or participants' individual and conscious interpretations, exegeses or "points of view" do not furnish the criteria on which to judge meaningfulness as a property of human action. Meaning is not intrinsic in the objective world nor simply a phenomenon in the subjective world—it is "somewhere in between" and it is a third party which mediates between the two. For, as proponents of the "modern turn" would say: for a human to "inhabit" the world amounts to more than merely being physically present, it is also a consequence of our intentional stance—because we are always "making meaning" to ourselves and others and the primary medium for that activity is language. Thus, to inhabit a world also means to inhabit a language and vice versa.
2. The Modern "Turns"
The "modem" turn towards a conception of meaning as a public and inter-subjective phenomenon gained momentum in the first half of the twentieth century when the classical views of meaning as either intrinsic and universal properties or as mental and subjective contents were given up. This development corresponds with the above mentioned shifts of "first philosophy" from metaphysics to epistemology and Bolzano's realization that the next "prima philosophia" must be a philosophy of language and meaning. In retrospect, this movement towards the philosophy of language has been characterized as the "semiotic" rum in philosophy —later also as the "linguistic turn" by Richard Rorty. Responsible for this profound change were a number of key figures: Gottlob Frege (d. 1925) whose work on the distinction between sense and refer-
*
Meaning and Religion: On Semantics in the Study of Religion
227
ence was ground-breaking.11 Charles Sanders Peirce (d. 1914) who introduced the notion of "semiosis" as a designation for the human propensity for sign-making. Also important was the tradition in British analytic philosophy of the study of "ordinary language"—associated with names such as Bertrand Russell, Ludwig Wittgenstein, G.E. Moore and W.V.O. Quine. All of them combined interests in mathematics, science, logic and language—as it actually functions in human life. Their interest in religion, by contrast, was generally quite negative (in the vein of Ayer) so it is somewhat ironical that we now can see how they laid some of the foundations for much more adequate ways of understanding religious semantics and languages.
As important as the semiotic turn on language was the "social turn" towards conventions. When Durkheim and Mauss published their small book on Primitive Classification in 1903 no one could have foreseen the importance of their views: That classifications, and thereby also ideas, "senses," and meanings, were of a social nature and that meaning is thus also a species of "conventions." The main feature of conventions is that they are shared, for they need not be "epistemically" true to be effective in the world of collective representations. They are constructed by humans and in some sense "arbitrary." To which extent the ideas of Durkheim and Mauss were known by the linguist Ferdinand de Saussure is uncertain. However, his thoughts on language accord well with the theories of the two former because he not only sees language as a system but also as pertaining to the genus "social fact." Among his major theoretical achievements is the distinction between the two sides of a linguistic expression: between the "signifier" ("signifiant") and the "signified" ("signi-fié") and the realization that the relation between them is for the most part completely arbitrary. Only the contingencies of linguistic history explains why "arbre" in French is "Baum" in German etc. Saussure was also of the opinion that the theory of the systemic nature of language could be applied to other cultural phenomena as in the oft cited passage on a future "semiology," "which studies the role of signs as part of social life ... It would investigate the nature of signs and the laws governing them" (1993: 15)." Saussure was a linguist with
11     Very briefly: Frege "discovered" that the sense ("Sinn") of an expression is not determined by its reference ("Bedeutung"). Two expressions —in his example "The morning Star" and "The evening Star" —have the same reference (the planet Venus). Kxpressions may (re-)present references in multiple ways and these ways are then the sense(s) of the expressions which determine our thoughts about the matter, whereas the reference determines the truth-issue of the sentence. Important is also Frege's idea that "sense" is public. Most semantic theorizing has been indebted to his ideas, but not all philosophers do (of course) agree on this issue.
12     Further theorizing on this issue is presented and assessed in Paul Thibault's (1997) comprehensive account of Saussure's revolutionary contribution to the history of linguistics. See also Harris 2003.
228
Jeppe Sinding Jensen
a (primarily) structuralist interest in signs and sign-systems and he was less concerned with the external or pragmatic aspects of actual language-use and of linguistic change and history. However, it is not difficult to "supply" these aspects to Saussure's thoughts, for they do not in any way preclude such "ex-ternalizations"—they simply were not his primary concern, wherefore it is "unjust" to charge Saussure of not supplying those aspects and therefore, as some have done, consider his theory "static" or "abstract." Furthermore, it is an epistemological premise that we have recourse to such systems in order to perceive, understand and explain the flux of the world —we would be hard pressed to explain a football game if we had no idea of the rules behind it. Thus, Saussure's main accomplishment was to point to the systemic and rule-governed aspects of language and thus indirectly also of other human socio-cultural behavior as well as to the idea that any action (including speech and thought) is always interpreted on the background of some sort of system of which the action can be said to be an articulation. This does not imply that rules and systems are causal, it is rather that they are constitutive and regulative for our conceptions of what counts as a meaningful action or description of something. Saussure's work was truly ground-breaking and later it became the foundation of structuralism.
2.1.1. Semantics and the Social Construction of Reality
The idea that meaning is exhibited by and through " language-in-use" was primarily set forth by two very different scholars: the anthropologist Bronislaw Malinowski and the philosopher Ludwig Wittgenstein. It was an accepted view that when humans wish to convey information there are certain conditions that must be met in order for them to be able to "make sense." Philosophers had for long been debating these matters and one of Wittgenstein's major contributions was to delineate the grounds on which humans can convey meaning. In Alberto Coffa's words on Wittgenstein's project:
A basic assumption of his project was that meaningfulness and meaninglessness are not merely relative to specific language systems; there are, in fact, general conditions that a system of signs must satisfy in order to qualify as a language and therefore also general conditions that determine the failure of meaningfulness. Once meaning is available, we can redistribute it at will, "by convention"; but we cannot originate sense by convention, since no act of semantic convention is possible in the absence of resources to express sense. The point is, in effect, that there is an objective factual difference between a representational system and a mere jumble of symbols, that there are conditions to be discovered-rather than agreed upon-such that fulfilling them is necessary and sufficient for being an information-conveying device. The emergence of sense is not the result of an act of the will but the outcome of acting in conformity with predetermined conditions of meaningfulness. (1991: 316)
Meaning and Religion: On Semantics in the Study of Religion                   229
Wittgenstein is famous for coining the idea of "language-games" to cover the function and use of language as rule-governed and self-contained —like a game. But, if like games they are performed for their own sake then that may lead to strong relativist assumptions, and on religious language-games it could perhaps also be said that if they are worthwhile activities for the participants then that settles the issue of their value and perhaps even "truth." This is the case when truth becomes a function of utility in the pragmatist view that religious people are rational in accepting religious beliefs because it enables them to deal with their existential concerns.13 "Meaning is use" is another slogan associated with Wittgenstein. It implies that a word has the meanings which are warranted by its use —thus a theory about the meaning of words becomes a theory about use; that which the proponents of more formal theories of meaning would conceive of a second pragmative level. But, for the "use theorist" actual language use is the only way to solve the questions of meaning. What you do with a word will tell us how you understand it.14 According to Wittgenstein, a description of the use of a word is a description of the rules for its use: "I can use the word 'yellow'" is like "I know how to move the king in chess," he has stated (1974: 49). Obviously, then, the rules for the production of meaning are also rules for doing things with words and this view counters the idealist (Platonist) ideas that objects have intrinsic meaning "in themselves":
A better account is that to possess a concept is to acknowledge certain cognitive moves as justified. Grasping concepts is acknowledging norms. By analogy to the slogan that meaning is use, one may say that concepts are cognitive roles ... to possess a concept is to acknowledge a pattern of epistemic norms. (Skorupski 1997: 48)
A recognition that language produces more than meanings and that words have a pragmatic force also occurred to Bronislaw Malinowski during his field-work among the Trobrianders and especially in connection with his detailed
3     As one source — best left unmentioned —triumphantly declares in its final page: "The debate of whether religious beliefs are rationally acceptable is over." Nancy Franken-berry, in a critical assessment of "neo-pragmatism," very pointedly remarks: "The wavs in which a particular religion may be shown to function in the benign and salutary-manner approved by its adherents helps to obscure the fact that it may also function to express and reinforce superstition, irrationality, fanaticism, sexism, infantilism, and eschatological abstentions from real moral and political tasks" (1999: 519). Also, on rationality in and "on religion" see Jensen/Martin, eds. 1997.
4     On this aspect of the management of the Wittgensteinian heritage Avramides notes: "In contrast to the formal theorists, the use theorists put central emphasis on speakers and what they do in their account of meaning. They are not content to let mention of speakers and their intentions be relegated to the level of pragmatics. The debate is over the core; use theorists see themselves as offering an account of semantics ... According to these philosophers one cannot abstract away from the imprecision of natural language, but must study language in its natural habitat, so to speak" (1999: 62).
230
Jeppe Sinding Jensen
analyses of magical rituals. In his ambition to "grasp the native's point of view" in "his world" he emphasized the importance of context for the understanding of utterances and because context is more often than not a real, practical communicative situation, it was logical to conceive of language use as a form of social practice.15 The primary function of language is not the expression of thought but its role as an active force in life and a basic precondition for concerted human action. This also entails that humans produce meaning while performing socially and that meaning is imbued in social action. For, as he says about myth: it is "a reality lived," "a narrative resurrection of a primeval reality," "a hard-worked active force," and "a pragmatic charter of Primitive faith and moral wisdom." (1948: 101). On the basis of Malinowski's view we may conclude that social action is—somewhat tautologically —only social action because it involves meaning and because it does so, we are in a position to understand the situation, context, other actors' intentions etc.
The view that language is a "force" was launched systematically by J.L. Austin in his now famous work on "How to Do Things with Words," posthumously published in 1962. He was tired of the prolix philosophical debates concerning the problems of truth and reference of words, meanings and utterances. So, Austin investigated what it is that we do when we speak, in "speech acts" —in "performative utterances" — which have "illocutionary force" because we do something "in the saying." For instance, when we "promise," we do not simply say that we promise, but we also make a promise. Such utterances contrast both with common "locutionary" utterances concerned with ordinary meaning and reference as well as with "perlocutionary utterances" which are intended to produce certain effects or functions. The theory (later to be developed e.g. by John Searle) is quite interesting for the study of religion because it makes us realize how much (to which extent) religious languages are not simply locutionary (descriptive) statements about the state of the world, they are often far more involved in expressing, creating and acting. Propositions and utterances in religious languages are more often either "illocutionary" in that they state "I/we herewith do" such and such, e.g. praying is not just talking, it is also an act or they are "perlocutionary" in the
15 These considerations also led Malinowski to reconsider the issue of translation and in his ethnographic theory of language, the requirements of translation include "the native's" contextual and practical considerations. See e.g. Duranti 1997, Ch. 7 on "Speaking as social action" (214-18 on Malinowski). Cf. also his view of the social context as the locus of the "real meaning" of myth: "It is clear, then, that the myth conveys much more to the native than is contained in the mere story; that the story gives only the really relevant concrete local differences: that the real meaning, in fact the full account, is contained in the traditional foundations of social organization; and that this the native learns, not by listening to the fragmentary mythical stories, but by living within the social texture of his tribe" (Malinowski 1948: 115).
Meaning and Religion: On Semantics in the Study of Religion                   231
sense that something could be achieved by the saying of such and such, e.g. in sacrificial discourse. A very interesting aspect in relation to the study of religions is that Austin saw how the illocutionary purport of an utterance should not be evaluated in terms of truth, but in what he termed the "felicity" conditions concerning how well an utterance fills in with other sentences in a certain set of actions and conditions. This is easily recognized as highly important in the use of religious and ritual language: the right things have to be said in the right circumstances, they are "framed" in a particular way because participants have a metapragmatic awareness about the well-formedness of the situation —even when they may not be able to provide any explicit rationale.'6 In a certain sense Austin's work can be seen as a pre-cursor for the later development of discourse analysis in its insistence on the importance of contexts and on what he terms the "condition of convention." Thus, the meaning of an illocutionary utterance depends on it being said or performed in the appropriate setting, and this in turns requires the recognition and mastery of socio-cultural conventions—not only by the speaker as author but also by the audience and its tacit expectations and knowledge.
Given the view of the importance of convention, the realization of how linguistic meaning is a co-determinate factor in human constructions of their worlds is the next step. Two names are prominent in that respect, Edward Sapir and Benjamin Whorf, who became famous (or notorious) for their "linguistic relativity thesis" concerning the socializing force of language on culture and thought, i.e. on the causality of language on "worldviews" (including concomitant metaphysics). The theory states briefly that the forms and ranges of "meaning" in words and word-patterns in a language determine the ways in which its inhabitants think; thus it is also a theory of "linguistic determinism."'7 Language is the primary instrument which allows us to make sense of the world(s), that is, to ascribe meaning to it (them). We should, however, not do so in a "humpty-dumpty" fashion because no "concerted action" would then be possible. On the contrary, meaning ascription must be conventional and religious discourse is the prime example of that. The "fragility" of religious discourse is also a fact—one need only contemplate the innumerous schisms, sects, and other fractionings of religions and religious groups. For, characteristically, a specific group rallies around a specific discourse —that goes for Trobrianders as well as for Arminians...
16     Impressive ethnographic examples of such complex semantic situations have been provided by Victor Turner in a range of publications on ritual actions and processes, e.g. 1969.
17     The relativist "language at work" hypothesis entails, as Sapir stated, that the "worlds in which different societies live are distinct worlds, not merely the same world with different labels attached" (in Duranti 1997: 60).
232                                                   Jeppe Sinding Jensen
2.1.2. Symbolic Studies
When religious "worlds" are conceived as symbolic constructs, the view of meaning becomes one of symbolic production and, extending contemporary metaphor, of symbolic capital management. The idea of "symbolic worlds" was, if not originally then influentially set forth by Ernst Cassirer (d. 1945) in his general theory of culture drawing on a variety of inspirations from the German philosophical tradition, not least from Immanuel Kant but also from a tradition of Lebensphilosophie which "taught the superiority of immediate experience over reflection, emotion over reason, synthesis over analysis, past over present, and so on" (Strenski 1987: 31). This ideal is found and expressed as an emotional and an experiential unity in the wholeness and coherence of myth and mythical worldviews, which, for Cassirer, are not so much rational but rather emotional expressions of a unity of feeling.18 The importance of this view, not considering so much the problems it introduces, is displayed in the stress it puts on the relevance of experience in the construction and maintenance of socio-cultural systems of meaning—where semantic theories often seem to downplay the importance of semantic "practice" (because they separate semantics and pragmatics). It should be quite obvious that meaning may be obliquely displayed in many symbolic forms connected to, say, ritual activity and that the question of semantics involves much more than the piecemal deciphering of meanings and truth-values of explicit propositions. Thus, according to Cassirer's Philosophy of Symbolic Forms it is a basic characteristic of humans that they build worlds in and through symbols and in different registers such as science, philosophy, art, religion etc. These different modes of knowledge are distinct, they have their own "range" and their own intrinsic function which all together contribute to a polydimensional understanding of the world and a "unity of cultural consciousness." Sympathetic and comprehensive as this view may appear it would also seem to suggest the existence of a set of mutually exclusive discourses: the view that there are separate languages for the various domains and the registers in which they are expressed, and that, in the end, they would become incommensurable and untranslatable. But, Cassirer avoids a normative judgment as to their epistemological status
18 As Strenski notes on Lebensphilosophie: "This feeling of the unity of life connoted an especially sentimental attitude to mindlessness perhaps unknown outside the tradition of German romanticism" (1987: 31). Nevertheless, this attitude had remarkably important consequences for the ways in which many scholars of religion conceive of their subject and its place in the academy as well in human life in general —for this attitude accords well with the view that religion(s) may, in the modem world, complement science as being more existentially "true" or "relevant." No further references should be required on this ... the number of scholars having expressed such views (which trace back to these "philosophical" assumptions) are legio.
Meaning and Religion: On Semantics in the Study of Religion                   233
and the "unity of cultural consciousness" assures the construction of a meaningful (and thus intrinsically semantic) world in and as a totality." Similar reflections were advanced by Suzanne K. Langer, who combined aspects of Wittgenstein's philosophy of language with Cassirer's theories of symbols in a synthesis she termed the "new key." This was a linguistic-semiotic turn in the interpretation of culture(s) where she introduces an important distinction between discursive and presentational forms, rather similar to Peirce's distinction between the iconic and the symbolic. This distinction is quite interesting in relation to the importance of more imagistic modes of religious representations in spectacular rituals or experiential ordeals in initiations —in situations where the discursive and semantic elements play a minor role (Whitehouse 2000). However, even such actions and experiences which are not linguistic, or non-semantic, are (or may become) objects of descriptions, explanations, and interpretation.20
The theories concerning religions as symbolic "worlds" are well-known and widely accepted. Although they come in various guises they all depend upon the semantic aspects, that is, on the realization that the "worlds" must inevitably be construed as worlds of meaning—whether the emphasis be placed on construction, re-production, or functions (etc.). In fact, the idea of human "life-worlds" seems to be one area in the study of religion where there is some cumulative growth of knowledge, i.e. the results, ideas, and perspectives of earlier theorists are not overturned as much as they are accepted, employed, and elaborated upon. More recently, the theories of Peter L. Berger and Thomas Luckman have demonstrated tenacity as a productive perspective. It is also well-known that a related view was introduced by Clifford Geertz in his program for the analysis of religion as a "cultural system." In short, "the lifeworld perspective" of religion as a semantic phenomenon is forwarded by a range of theorists from Pierre Bourdieu to Jürgen Habermas. This perspective may thus be characterized as a kind of "normal paradigm" (e.g. Paden 1988). It should be noted, however, that most of these theorists are more interested in functional analysis rather than the systemic properties of the "worlds"—but that and similar aspects are the subject matter of structuralist investigations.
19     The Philosophy of Symbolic Forms is no light reading —the reception of Cassirer has been scant and skewed in the study of religion and it could deserve a correction. See Capps 1995: 210-15 for a very brief overview.
20     This line of thought was later taken up also by Nelson Goodman in his Languages of Art: An Approach to a Theory of Symbols from 1976 and in Ways of Worldmaking (1978) where he states: "We can have words without a world but no world without words or other symbols." Goodman also notes that ail "making" realty is "re-making" —"from worlds already on hand" and "the search for a universal or necessary beginning is best left to theology" (1978: 7). He (rightly) considers "the search for a first world thus to be as misguided as the search for a first moment of time" (ibid., n. 8).
234
Jeppe Sinding Jensen
2.2. Structuralism
Structuralism was and still is a very strong "paradigm" in the human sciences. In some fields it has become more or less relegated to the history of theory— but for the study of religion it must be said that the productivity of structuralist thought is so impressive that this approach to religious materials is not likely to be given up or discarded. In fact, as religious "worlds" so aptly lend themselves to structural analyses it seems likely that a continued refinement of structuralist theorizing may in fact take place within the study of religion. The history and characteristics of structuralism in the study of religion, and in the human sciences in general, need not be rehearsed here, as they are amply accounted for elsewhere (e.g. Lechte 1994; Jensen 2000 for further references). In this connection the most noteworthy feature of structuralism is its insistence on the importance of analyzing semantic matter as a feature of relations, and it is thus a decidedly holistic approach. It is also important to note the systematizing and formalist ambitions of much structuralist theory and practice. And although these ambitions have sometimes been met with a critique (sometimes justified) of leading to sterile, static and intellectualist constructions, it must also be said that probably the most rewarding results of the structuralist endeavor are demonstrations of the systemic character of products of the human mind, be they language, mythology, or religion in general. In this sense, structuralists have contributed much to the "scientific" turn in the study of religion. The more formalist structuralist approaches are found in semiotics, where the heritage from linguists such as Roman Jakobson and Louis Hjelmslev has been expanded, transformed and turned into very productive analytical tools by e.g. Umberto Eco, Algirdas J. Greimas, Yuri Lotman and others. Being a formalist trend in the study of meaning, they have attempted to institute a "science" of meaning —with schemata almost resembling "periodic tables" of semiotic objects, and they have attempted to demonstrate how the construction and organization of meaning is constrained by features of these very same systems.21 One of the problematic aspects of structuralist and semiotic theory was the relation to hermeneutics. At the most general level, a major issue of contention is concerned with the ontological and epistemic status of systems of meaning and their use and involvement in human interpretive practice. To most structuralists and semioticians, the status of such systems of meaning would be that they are objective features of the world, that they are imminent, self-contained, and exist as such irrespective of individual subjects' use of them. This is understandably hard to accept for those holding a subjectivist and individualist view of meaning, interpretation, and understanding. In an almost
21    For a more comprehensive introduction to these scholars Lechte (1994) is an invaluable guide. See also Peregrin 2001 and Caws 1997.
Meaning and Religion: On Semantics in the Study of Religion                   235
caricatured sense, we can say that for hermeneuticists it is we who, as human individuals, make meaning and for structuralists and semioticians it is meaning, as an intersubjective phenomenon, which makes us human individuals. In order to create meaning, to understand others as well as ourselves, we need these symbolic systems —pure introspection, if not dead, is at least seriously challenged.22
That humans are always "situated interpretants" and that interpretations are always pragmatically motivated has been stressed by later "post-structuralist" critics. The consequences of that criticism have probably been somewhat overrated. For it comes as no surprise that meanings are not stable, that everything that makes sense is only so because of human intentional practice and that this is again motivated and informed by power and all other kinds of interests. Some of the more interesting aspects of "posť'-theories is their insistence on the reception-perspectives and on the ways in which meaning is pragmatically constituted and employed. To the study of religion, this move from the study of "origins" and later "developments" has been turned towards perspectives on use and meaning as products of effective history. It is a move from the comforts of the view that origins and essences have causal power as traditions which create history to the much more unstable view that humans exploit their symbolic resources (including "tradition") in order to make meaning and understand themselves, not only existentially but also in order to sustain power, dominate and generally further their own interests (which are, however, also constrained by the resources...). An interesting aspect of many "use"-, "reading"- and "reception"-theories is their demonstration of how meaning is always created in relation to preexisting patterns —all cultures favor certain constructions of meaning and downplay others. Cultures are in that sense "semantic processors" which set some limits for what can be said and understood. Cultures and religions speak in different modalities and they furnish the "grids" along which items and events may become intelligible.23 These views are (evidently) functionalist-but that does not make them either false or trivial. Furthermore these perspectives make it possible to analyze the "politics" and "economics" of meaning(s) — what people do with words and what words do to them. That kind of analytic activity has now become familiar under the label of "discourse analysis" —which appears
22     One of main proponents of the combination of structuralism and hermeneutics is Paul Ricoeur —see e.g. the instructive volume edited by John B. Thompson (Ricoeur 1981).
23     As cultures and religions are models of and for the world, they are discursive systems, regulative of a number of aspects of social production and organization of meaning: on the epistemic (what can be known, assumed, doubted etc.), on the alethic (what is true, necessary, possible), the axiological (on values, what is good/bad) and the deontic (what is mandatory, prohibited, permitted). One does not need much familiarity with any religion to see that this seems to be the case.
236                                                   Jeppe Sinding Jensen
as a promising direction for the study of meaning in relation to religion (Albinus 1997).24
3. Semantics Currently: Meanings "Without Reference"/ "Anti-verificationism"
The idea of meanings beings meanings without having clear and explicit references to things "in the world" may seem counter-intuitive—and, indeed, it is. Our "intuitive" semantics and linguistics repeatedly "convince" us that we when we are offered coffee, we should not expect tea. A note of whatever currency refers to a certain amount and not to others. The label on a product refers to what is inside the package. "Mother" means mother and "Ladies" means that "Gents" should look for facilities elsewhere. Our "intuitive" ontologies and semantics perceive the world in that way—words are referential and "deictic": the objects that the words point to are their meanings. So, "sugar" is sugar, and it does not matter one bit if you translate it into "Sucre", "Zucker" or whatever it is called in Arabic or Hopi... and in this view, the meaning of a statement consists in its verification. This "picture" theory of meaning goes well with the intuitive assumptions of an "experiential realism" —the view that there is a "phenomenological bedrock driven by perception and physiology that provides an interpretative anchor for the words we use, and enables us to understand each other" (Edwards 1997: 256).25 Thus, it is problematic (to say the least) to make the move from an analysis of language to an analysis of "reality"—not least because our analysis of "reality" must be made through language. There is no way in which we can understand the world apart from any mode of describing it. We may also say that "theoretical concepts" in the sciences do not directly refer to things that scientists have observed, rather that these concepts enable scientists to analyze, discuss and produce coherent accounts which we as humans are able to understand. That requires, more often than not, that the world (all things included) must be narrativized, and it
24     Discourse analysis is a truly cross-disciplinary field —but like so many others it has hardly been communicated in the study of religion. A perceptive introduction by a major scholar in the field is Fairclough (1992).
25     As Hans Penner says —on the question of "meaning" in religion-we are mostly told: "to look for a reference. Words refer to things —you name it: sensations, certain stimuli, psycho-neurological states, needs, the numinous, the given and so on. Thus, if I can demonstrate what religion refers to I can tell you what it means. This is the famous correspondence theory of meaning, now also labeled as the 'realist' theory. It is the implicit theory in most studies of religion from Emile Durkheim to Victor Turner. This theory is critically wounded but has not yet been laid to eternal rest" (Penner 1999: 474).
Meaning and Religion: On Semantics in the Study of Religion                   237
is thus fair to hold that scientific languages are purpose-specific extensions of natural language. Similarly, we could say that religious languages are extensions of ordinary language—and thus not of a wholly different nature. Some think that religious languages may be given or inspired by supernatural agents, gods (etc.) and that they are therefore of a "godly" nature; but it is probably more likely (we might imagine) that it is the gods who speak the languages of humans. At least, they must if "they" want to communicate with us humans.
3.1. Coherence Theory and Holist Semantics
The problem with human communication and language is that they consist in much more than just pointing to things and naming entities in physical space-time. Language may convey meanings about metaphorical assertions, imaginary situations, the future, a dream-time to which we have no access and —strangely enough—still be "meaningful." The possibilities of such sense-making seem puzzling. When the idea that meaning consists in verification of correspondence to matters of fact in the world or private states of mind has been given up, something else must be available for us to account for how meaning is created, used, discarded etc., and there is some agreement on coherence as being the necessary condition for sense-making: For any statement to make sense, it must be part of a pattern, network or system in which it makes sense—and so far this is a legacy from several sources, e.g. Saussure and Wittgenstein. The view that meaning is thus inherently dependent upon other meanings in a total pattern, network or system is habitually termed "holism" or "holistic." Hans Penner characterizes the theory in this manner: "Roughly and briefly, this theory rejects the principle of reference and emphasizes consistency and coherence of ideas, symbols, archetypes, and the like" (1999: 474). It is quite easy to imagine skeptic attacks on a holist semantics which emphasizes coherence above correspondence — in the sciences, it could lead to adherence to circularity and non-falsifiable hypothesis and (this has not so much been contemplated by philosophers) in the field of religion, it could lead to the defense of the "rightness" of religious beliefs. On the problem of the "coherence theory" Hans Penner states: "The flight into idealism is not the only problem here. The theory entails the shuddering thought that your beliefs could be completely consistent/coherent and totally wrong. The skeptics are now beginning to laugh" (1999: 474). The relativist menace is evident and imminent.26 However, there is more to the theory than coherence among a set of
26 These issues have been the topic of heated debates among such well-known philosophers as Hilary Putnam, Donald Davidson and Richard Rorty — not least because the issues are very complex and difficult to access. Penner's contribution (1999)
238
Jeppe Sinding Jensen
beliefs. It involves the totality of beliefs, and it is, as Donald Davidson has emphasized, unlikely that we are all massively mistaken in our beliefs about the world. Neither are we so differently oriented in our conceptual schemes that they become radically alternative and untranslatable. Our general picture of the world cannot be totally mistaken, according to Davidson, for that general picture "informs" all of our beliefs, even those that may turn out to be false. The truth or falsity of beliefs is, to the semantic holist, not something that is given in relation to things in the world, as that is "mistaken." For as Penner explains (on Davidson): "The truth (meaning) of a sentence has nothing to do with reference, intermediary entities, or ideas, or bits of the world that make a sentence true" (1999: 479). Very briefly, the idea is that sentences are not true or false in virtue of extra (or non)-linguistic objects or "facts" —they have what meaning they have in relation to the totality of a language.27 In relation to the study of religion this means that "religious beliefs and actions are to be studied as a system, a semantic structure. It makes no sense, for example, to speak of the meaning of a religious belief or action in isolation from the whole systems of beliefs, rituals, etc., of a religious tradition. Beliefs make sense only in relation with other propositional attitudes. Beliefs are holistically structured." (Penner 1999: 498).28 Another version of semantic "anti-realism" is found in Michael Dummett's theory. He defines a speaker's knowledge of the meaning of a sentence—the speaker's understanding of the sentence—as knowledge of the conditions under which it can be asserted. This has become known as the theory of "assertability conditions."" As already noted, realist semantics would consider meaning to reside in the truth conditions of the sentence, that is, in verification of the facts which "make the sentence true." Dummett explains the differences between realists and "anti"-realists on their view of the truth of assertions:
The anti-realist accuses the realist of interpreting those statements in the light of a conception of mythical states of affairs, not directly observable by us, rendering them true or false. According to the anti-realist, what makes them true or false are the
is a strong effort in trying to explain the consequences of holist semantics for the study of religion. See also my own contribution (Jensen 1999),
27    It is impossible to do justice here to the complexity of the argument. Besides Donald Davidson's own production on the issue, of which 1990 is instructive (and relatively accessible), Evnine (1991) should be mentioned.
28    Indeed, Penner also notes that this is not really so new: "Structuralist studies of myth and religion have stressed this warning for decades. I am afraid that it has yet to be taken seriously by most scholars who are full-time students of religion" (1999: 498).
29    This rendition of Dummett's semantics (mostly based on Frege and Wittgenstein) is almost unrecognizably condensed, which is all the more problematic because there is, among philosophers, general agreement on only one thing about his philosophy, and that is that it is difficult. Dummett's ideas are well presented in his anthology Seas of [.(inguage (1996).
*
Meaning and Religion: On Semantics in the Study of Religion
239
observable states of affairs on the basis of which we judge of their truth-value. On the realist's interpretation, these merely provide evidence for the truth or falsity of the statements, or constitute an indirect means of judging them true or false; the anti-realist retorts that they are the most direct means there could be. (1996: 469)
Thus, "anti-realists" do not assert that there is no real world—or any such nonsense. Anti-realists are sensible people—who are "taking the sciences at full value" (as Nelson Goodman says), also when studying things that are not normally considered so realist. Being a form of "meta-language" the study of religion must treat religions as semantic systems in which it is culturally meaningful to speak of things that make little sense epistemologically. In that light, semantic anti-realism and holism appear attractive whether in Dummett's version or in Davidson's and Putnam's. These semantic theories may well be employed in the study of religion as they may assist us in ensuring that the course chosen is philosophically justifiable. However, the most uninteresting issue in philosophical debates about semantics, as viewed from the perspective of a scholar of religion, is the truth question, and that is, ironically, that which interests philosophers the most. Most scholars of religions—be they anthropologists, philologians or others—do not really care about truth, or, when they work on any number of mutually incompatible cosmologies, they tend to consider it a waste of time to discuss which local ontology may be more epistemologically defensible.30
As already noted in relation to Wittgenstein's notion of language-games, it is easy to slip from an idea of the autonomy and internal coherence of particular meaning-systems ("cultures") to an idea of incomparabüity or incommensurability. "Meanings" and interpretations not only differ from person to person but from culture to culture with no means of judging their importance—it would, in the end, be the "positivist's nightmare" revisited and the use of the term "meaning" might as well become meaningless and we shall all be dragged down the "hermeneutical vortex" (as some, e.g. Donald Wiebe, iiave already seen it becoming). One reaction to the apparent threat of subjectivity in the realms of meaning and interpretations is to limit talk to that which concerns the "factual." Apart from the fact that this (empiricist ambition) has been amply proved not to work in general or in scientific language, it certainly would have impaired any understanding of what religious and other normative "meanings" are about and how we may know them. It seems that "we can acknowledge that the normative is a domain of understanding, something we can judge of—but yet that norms are still like rules in this respect: we do not
30 However, in a world where religious traditions increasingly speak "for themselves" across traditional boundaries and in all media, it is not at all irrelevant that scholars of religion engage the truth-issue in relation to the validity of scientific versus religious discourse—see, e.g. Murphy 2000.
240                                                   Jeppe Sinding Jensen
find them in the world. They are presupposed in cognition of a world" (Skorupski 1997: 54). That there are limits both to interpretation and of pragma-tist license is quite obvious in many familiar settings —such as in visiting a restaurant where (this is part of the "frame") it is conventionally expected (even by the most post-modernist) that there be a high degree of correspondence between menu, orders, the food served (and tasted...) and the amount indicated on the bill.
3.4. Cognitive Constraints
Experience tells us, and scholars of religion know this more than many others, that the range of imagination of human minds seems quite limitless. And yet, for all its variability the same human mind appears (in what concerns religion) to revolve around or return to certain themes, and therefore there are limits as well as there are certain recurrent features in the sense that religions really do seem alike in many ways. These similarities may of course be the product of the scholar's own imagination so that we focus solely on things which "appear" according to a prefigured pattern: Religions look alike because we have determined that just such configurations as exhibit certain traits are labeled religion... and thus what we ensure is really just the circularity of our argument. That charge is, however, unavoidable, simply because (and this is somewhat overlooked) when we recognize something, it is a model which we recognize and not an object "in itself." Given this precondition for recognition, however, we are still able to say that religions exhibit common and recurrent semantic (as well as cognitive) features and properties amidst all the observable variations at the surface level.3' There are indications that we may legitimately theorize about universals in religions and that these universals are not bound to religious ontologies, as former generations of religiously motivated univer-salists would have it. Universals "in a new key" are, when it comes to religion, both of a semantic and of a cognitive nature (Jensen 2001).
Now, when it is agreed upon that religions "look alike," the next question will address the possible causes of the likeness. A likely answer to the question could concern some kind of general human psychological or mental mechanism, function or module which would then be responsible for the constructions and working of religious systems of meaning. In that respect the problem somehow resembles that more well-known problem concerning the origins and  "causes" of language. The theories about modularity and the
31 In what concerns the cognitive aspects of this problem see Boyer 2000 which contains a very convincing analysis of how and what goes into the cognitive construction ot religious representations.
f         Meaning and Religion: On Semantics in the Study of Religion                   241
"innateness" of language competence (primarily conceived by Noam Chomsky) have been met with mixed reactions. There are, however, other more interesting aspects of Chomsky's theory than the idea of innate modules, and that is his view of generative grammar. As first formulated (in 1957) it was a grammar of rules, of syntax only, but later criticism and developments led to the inclusion of semantics and an idea of a generative semantics has evolved for the purpose of explaining how "deep structures" may account for the construction of meanings at "surface" level. That is, the purpose is to elicit the mechanisms responsible ("transformational rules") for the ways in which semantic meanings are set forth. So, when it is also agreed that religions are semantic phenomena it is quite plausible that similar mechanisms can be elicited for religious behavior, including the production of meaning.32
3.5. Meaning and "the Mental"
As already mentioned, many of the contributors to the classical quandaries over "meaning" in the study of religion viewed meaning as a mental fact or property. That reaction was perfectly understandable as most of them developed their views on the basis and perspectives of subjectivist philosophies and psychologizing hermeneurics. Many of us (humans) tend to think that meaning is in our minds—it is an intuition, and seemingly a plausible one, for the meanings I have are mine—are they not? So where else could they be? It was argued above that this view is flawed for many reasons and the idea that meaning (in the semantic sense) is a mental property cannot be upheld. On the other hand, it is also obvious that things semantic are related to things mental — if there were no brains there probably would not be any languages or symbols (etc.) either. Languages also give indications of "how we think" —as witnessed in the work of e.g. George Lakoff (1987) on complex systems of linguistic and cognitive classifications. Now, although asking and, perchance solving, semantic linguistic questions, the "cognitive way" does amount to a theoretical reduction (that is: formulating the problems of one "domain" in the theoretical idiom of another), it is not quite so simple to answer the question whether and to which degree this kind of operation also implies an ontological or epistemological reduction. In one sense we may say that (semantic) meanings really are, simultaneously, inside minds as well as in language outside minds. It depends on the kind of description we use. For, we may say that they are
- 2 Iliese advances in semantics have not fully been appreciated and applied in the study of religious semantics, but it seems that there is ample potential. The story about generative grammar and semantics is told and explained by Leech (1990: 343-59). A noteworthy example of the study of rules and syntax in ritual practice on an inspiration from generative grammar is Lawson/McCauley 1990.
242
Jeppe Sinding Jensen
ontologically independent, but also that they have some epistemological equivalence because we only know the mental through the domain of the linguistic and the semantic. That is, it is only as presented in "intersubjective semantic stuff" that we may gain and formulate knowledge about the cognitive and the mental. Furthermore, the state of the problem and the kind of discussion depends upon which kinds and levels of "meaning" are addressed. Some are more basic than others, for it may well be that "meaning at the most basic levels is supported and driven by general, not specifically linguistic, cognitive operations" (Faucormier 1997: 190). But, then again, as Faucormier also points out: "the simplest meanings are in fact not simple at all" (188).33
3.6. What and Where Is "Meaning" —on the (Emergentist's) Division of Labor
As witnessed by some of the previous notes it would be —by many —a welcome addition to our knowledge of ourselves and the world if we could reduce meaning and semantics to something more "basic" by means of which we could "prove" things and produce irrefutable evidence (the "scientific method" etc.). The only question is whether the results of such operations, favored by "semantic eliminativists" (as they are called), are still concerned with "meaning" or should they more likely be regarded as concerned with something else? Should the ambition of "reducing meaning" to nano-electric or micro-chemical (or some such) functions of/in the brain of an individual pronouncing the sentence "Rhubarbs are delicious" actually succeed some day (which it may — given the speed of scientific discovery), then the results of such analyses are more likely than not not to be products of "meaning" in the semantic sense. On the other hand, as long as the results of these investigations would be communicated between humans and understood by them, the non-semantic would have to be translated back into semantic realms. Scientists involved could then re-examine by reduction their ideas about their reductions, and then they would never run short of matter to investigate.34 It is or should be possible to
33     As Faucormier further explains concerning simple meanings: "They rely on remarkable cognitive mapping capacities, immense arrays of intricately prestructured knowledge, and exceptional on-line creativity. They also rely on the impressive, and poorly understood, human ability to resolve massive underspecification at lightning speeds" (1997: 188). There is much more to be said on that point oř view than may be referred to here, but Fauconnier's basic idea is that language primarily "serves to prompt ... cognitive constructions by means of very partial, but contextually very efficient, clues and cues" (ibid.).
34     A similar operation can be performed by the reduction of gastronomy into organic chemistry. The analyses of the constituents of nutrition would (probably) be correct under that new description, but "it" would cease to be gastronomy which still retains
i        Meaning and Religion: On Semantics in the Study of Religion                  243
retain the "level of meaning" as one which may both be reduced and not be reduced without the latter position being responsible for upholding some "mysterious" ontology for matters semantic. Facts about meaning and intention need not be reducible to other facts of a naturalistic (or "gravitational") kind — perhaps it is not even possible, for as Putnam states: "The problem, of course, is that what the semantic physicalist is trying to do is to reduce intentional notions to physicalist ones, and this program requires that he not employ any intentional notions in the reduction. But explanation is a flagrantly intentional notion."35
3.6.1. Linguistic Ontologies and Epistemologies
The ontologies of meaning are (so far) quite mysterious. It is obvious that languages exist and that they are different. Also, it turns out that they are translatable and anything described in one natural language can be described in another natural language with some degree of precision —as well as with some loss in precision. But, and this is the really mysterious part, the physical sounds "emitted" in linguistic practice are not meaning producing in themselves. They are only prompts and cues which make brains work in certain cognitively and culturally preconfigured ways. We cannot (so far) make an audio-spectral analysis of the meaning of a sentence, let alone of sentences embedded in social practice. But the question then apparently still remains whether it would be possible to reduce, and thus perhaps explain in idioms of current scientific practice, semantic meaning to something non-semantic? That idea depends on and proceeds from a dualism concerning things meaningful, or as formulated by Jane Heal:
to the idea that the semantic arises from, or is constituted by, some kind of appropriate complexity in the non-semantic. For want of a better word, f shall say that he or she is committed to the reducibility of the semantic to the non-semantic. But it is to be remembered that what is involved is reducibility in some extremely broad sense. The difficulties of dualism have given a bad name to the whole idea of non-reductive
its own level of description (at least among, say, the French and Italians and other sensible nations). 35 As quoted in Hale/Wright, eds. 1997: 442. Putnam launches similar attacks on "meaning reductionists" in other places: e.g. against the idea of there being "innate semantic representations" he says: "A Chomskyan theory of the semantic level will say that there are 'semantic representations' in the mind/brain; that these are innate and universal; and that all our concepts are decomposable into such semantic representations. This is the theory I hope to destroy" (1988: 5). And: "Mentalism is just the latest form taken by a more general tendency in the history of thought, the tendency to think of concepts as scientifically describable ('psychologically real') entities in the mind or the brain ... this entire tendency ... is misguided" (1988: 7).
244
Jeppe Sinding Jensen
accounts of meaning (in the very broad sense of 'reduction' just gestured at). The bulk of philosophical writing on meaning (in the analytical tradition) has thus been concerned to pursue the radical interpretation strategy. But are dualism (in which a hidden and separate meaning is inferred behind the non-semantic surface) or a reductive materialist view (in which it is discerned in the patterns of the non-semantic) the only options? What if we abandon the assumption common to the materialists accounts and dualism, namely that meaning is not observable, while retaining dualism's commitment to non-reductionism? This gives us a view on which meaning is a public and observable property of certain sounds, marks or movements, but a non-physical one. But it is not part of the predominantly quantitative and value-free conceptual scheme we have built up for describing, predicting and explaining the behavior of inanimate objects; rather, it belongs to a different but equally fundamental area of thinking namely the one we use in our relations with other persons. This line of thought is favored by those with Wittgensteinian sympathies. If we accept this view it is likely that the idea of the imagined starting-point for radical interpretation, a starting-point in which a person knows plenty of non-semantic facts but no semantic facts at all, will come to seem incoherent. The starting point for any thinking is one in which we are observationally aware of the world as containing both semantic and non-semantic facts. (1997: 178-79)
These arguments presented by Heal indicate the difficulties in linking the mental as a stratum (or several such?) of neurological facts with the "level" of semantic meaning. And yet, there can be no doubt that such a level (or several?) are somehow connected to our physical capabilities as humans. The questions of how we install "culture in mind" and how culture, as a complex of semantic properties and function, works and what it "really" consists of, remain (so far) unsolved problems —although suggestions are not lacking—and some of them seem quite promising.36 For instance, as pointed out by Terrence Deacon, different linguistic tasks are processed more or less in various regions in the human brain:
Producing a metaphoric association requires selecting words with common semantic features, whereas producing a metonymic association requires shifting attention to specifically alternative features. This is why there may be a posterior cortical bias to metaphoric operations and a pre-frontal cortical bias to metonymic operations. (1997: 306)
Thus we may assume that: "The symbolic functions, the grammatical and representational relationships, are not processed in any one place in the brain, but arise as a collective result of processes distributed widely in the brain, as well
36 One very interesting suggestion comes from Bradd Shore in his Culture in Mind-Cognition, Culture, and the Problem of Meaning (1996). "Interesting" because Shore, as an anthropologist, is not unfamiliar with the kinds of problems facing the study of religion on "cultural models," "mental models," "instituted models" etc. It is beyond the scope of this presentation to elaborate further on Shore's work—but it deserves a more thorough application and testing in the study of religion(s).
Meaning and Religion: On Semantics in the Study of Religion                   245
as with the wider social community itself" (Deacon 1997: 309). But—whatever PET scannings reveal of brain activity, they will probably not (but who knows?) be capable of decoding such meanings as pass through the reader's mind when reading just these pages... Obviously, when speaking, listening, reading, writing and otherwise processing semantic materials, humans do employ their cognitive abilities; thus meaning-"metabolism" is a mental activity, and yet, that description does not exhaust the topic. Meaning is not only in minds, it is also, and perhaps more significantly, between minds, which then put together (or synthesize) the semantic materials in such ways that they engender meaningful events in our minds. Religious activities, imageries and utterances are prime examples of humankind's propensities and proclivities for such activities for it seems that these activities produce some kinds of "well-being," kinds of cognitive "flow," kinds of "blissful" effervescence on the more benign side but also, since religion is not just a "nice thing," on the side of the "tremendum": kinds of experience which satisfy thirsts for power, dominance, horror, violence, etc.
3.7. Levels of Semantics
To clarify what I think must be an inevitable stratification of matters semantic, I shall briefly refer to the concept of "downward causation." In a theory of "bottom-up" causation, things happen automatically at the higher levels-much as in chemical experiments. That being so, the other way of processing ("top-down") is much more interesting when we talk about meaning and semantics, for it is only in the presence of such functions that we can see more general information and intelligence at work. Translated to our discussion concerning semantics in religion, it could answer the question of how meanings effect social and cognitive "power," that meanings are more than "mere" epiphenomena-unable to influence the world (Jensen 2001: 256-59). It is in fact quite plausible that language, and thus semantics, is responsible for rather thorough "re-shaping" of not only cognitive contents, properties, and mechanisms but also of brains (as neuro-physiological and chemical etc. entities). As Mark Turner wrote:
If we use the old metaphoric conception of the brain as an agent who "deals with" language or as a container that for a moment "holds" language while examining it for storage or discard, then it is natural to think of the biology of the brain as unchanged by its dealings with language. But if we use instead the conception of the brain as an active and plastic biological system, we are led to consider a rather different range of hypotheses: The brain is changed importantly by experience with language; language is an instrument used by separate brains to exert biological influence on each other, creating through biological action at a distance a virtual brain distributed in the individual brains of all the participants in the culture; early experience with language affects cognitive operations that go beyond language. (1996: 159-60)
246
Jeppe Sinding Jensen
What this means in terms of religion is that it lends theoretical credibility to the more old-fashioned idealist (but intuitively plausible) view that religions somehow condition the ways in which we think: that they as "semantic engines" are co-responsible for the ways in which we process information and construct meaning.37 In most traditional societies, culture, religion and language have been learned ("installed") simultaneously so that meanings are mulri-"meshed" in the classificatory architecture. As already pointed out above—languages (and semantic meaning) are concerned with and involved in much more than simple description. This is a field in which much remains to be done —thus there is ample space for "new approaches" along these lines. Although nothing retrospectively appears more antiquated than prophesying, it does seem that the fields of semantics and cognition appear as challenges to the study of religion, not the least in combination 0ensen 2002).
4. Religion as a Socio-"Cultural System"
What difference does all this make? It has been common-place for a long time to talk (with Clifford Geertz) of religions as "cultural systems."38 In a more conventional view of these, they are pre-eminently semiological and of a thoroughly linguistic nature. Along the view of an objectivist semantics (such as Geertz'), these signs in life, which constitute the systems, are "as public as marriage and as observable as agriculture" (1973: 91). That these are public is a result of their being done, by their being "staged," by participants in whose conceptions of things they have to be made in a way that "counts as" something intentional
37    Gilles Fauconnier reaches a similar conclusion (although not about religion): "When meaning construction is taken into account, the fundamental cognitive issues of learning and evolution appear in a different light. Clearly, what children learn is not language structure in the abstract. They acquire entire systems of mappings, blends, and framing, along with their concomitant language manifestations" (1997: 189). An example from my own fieldwork: When Muslim mothers shout "Kafir!" (meaning: "unbeliever," "heathen") to children misbehaving they involve and invoke much more than a "no!"
38     It should be noted and remembered that Geertz' launching of the "cultural systems" program was a way in which to translate "the linguistic turn" in philosophy into anthropological theory and methodology. It has proved a very successful move although criticism was inevitable —against his idea that anthropology should become a primarily interpretive endeavor and thus, in the eyes of some, a less respectable scientific undertaking. However, after some decades of debate, it seems that the battle over "interpretation versus explanation" is (largely) over. Even the most empirically minded concede that interpretation is involved in all scientific activity and the most hermeneutically committed also acknowledge an element of explanation in interpretation.
Meaning and Religion: On Semantics in the Study of Religion                  247
(Searle 1995). In that sense the cultural and public displays of meaning are a way of objectivizing and projecting widely distributed cognitive models — including schemata, frames and "scripts." That is one of the ways in which symbols "have meaning" in religions, i.e. they "trigger" or evoke response by linking to other concepts "in the mind" among those who participate in the joint projects of having those concepts, values, and thoughts. In other words, those who share the "meanings" and the knowledge they amount to. Thus, when speaking about where we should "look for meaning" we may in fact look in more places—in the doings, writings, sayings that are publically available and in the heads of participants and interpreters as well. But since interior states are not always so easily accessible, the meanings and references of discourse are much more amenable and tractable at the public levels—such as in texts. Although not all, nor perhaps even many, participants of "cultural systems" have textual meanings "in their heads" they will always carry meanings that may be "text-ihed," i.e. made into narrative and text. This a very simple corollary of the fact that we make sense of the world by talking about it and that knowledge which can not be transposed into some kind of narrative is (probably) not knowledge at all. So the "cultural systems" approach seems to be able to hold true and tractable for much more than religion (cf. also Geertz 1973, 1983).39
Throughout the history of philosophy one important topic of debate has been the relation between words and concepts. The question of meaning is intimately related to this debate and the same three (or four) positions are (logically) available. It is possible to transpose the stances on concepts to those on meaning and see how they may or may not make sense: 1) A "realism" concerning meanings according to which they would exist "in themselves" seems to be a non-sensical option; 2) an "empiricism" which derives meanings from experience is likewise questionable; 3) a "rationalism" positing meanings as psychologically or mentally "innate" fares a little better in light of advances in cognitive linguistics, at least as a necessary "sub-stratum, but 4) the "nominalist" view that they are the results of human intentionality, ascriptions which are defined as properties of their "relational" positions in comprehensive semantic systems and networks and thus "nothing but" conventional becomes the most probable solution to the problem. This holistic view in which mean-
39 In this I must confess to side with those who think it perfectly possible to study "externalized" linguistic and other "meaning-laden" facts in a theoretically informed way. Does this sound strange? Remember how Noam Chomsky "pushed" the study of language (back) into speakers' minds and made it a subdivision of psychology by saying that only internalized or "I-language" as "a structure in the mind" is the proper object of linguistic analysis. On the relations of such views to the study of, e.g. religious rituals see the discussion in Lawson/McCauley 1990, Ch. 4: "A cognitive approach to symbolic-cultural systems."
248
Jeppe Sinding Jensen
ings are eminently available as "externalized" objects in a collectivist methodology does not impair the view that meanings also exist and function as cognitive entities with "evocational" potentials. On the contrary, the first thesis presupposes the latter. But, when the emphasis is strictly on the externalized semantics that are available to us directly as articulations of discursive formations, the question of the individual appropriation of these systems of meaning become somewhat less relevant. There is nothing wrong (inherently) with a theory in favor of the study of externalized meanings and an "objectivized semantics," for these are the meanings which "go into peoples' heads" (further, the "internalist's" problem is that s/he cannot communicate the internalist point of view except in externalist terms).40 Obviously to many, a two-tiered model of meaning(s) requires the problematic acceptance of a dualist ontology in which there exist both meanings as semantic entities and cognitive entities as properties of the mind and its functions.41 But is it really so problematic? For, when all is said and done, it also appears that a consequence could be that we stop talking about meanings as something between and relating "words and objects" but rather as a "something" which exists between words and historically situated humans —who use words to make sense of the world.
Concluding Remarks
Concerning, and in spite of, the importance of semantic questions in the study of religion, it is noteworthy just how little attention the problem has attracted. Perhaps this indicates the necessity of a "semantics of religion" as an addition
40     An interesting, but not too common feature in general in religions is the Muslim notion of "Hafeez", meaning someone who has learned the Quťan by heart, i.e. "internalized" the whole body of semantic material in it as a text. The "argument from individual appropriation" is not altogether fallacious, since what we are being told and what we tell ourselves are really consequential for the way in which we react to new information. This should really not come across as anything new—but perhaps we are now closer to presenting a credible account of how it is so. In order to elicit internal systems, the analyst needs to construct external versions of, say, the internal competence and an "idealized speaker's" competence in French —is most likely —to be found in French grammar and syntax. The constructions of idealized speakers as well as of grammars and syntaxes are all normative endeavors.
41     To what extent there really exist two (or more?) levels in an ontological sense is an object of debate. In this relation I think there is good reason to support Donald Davidson's idea of what he terms "anomalous monism": That there is but one reality and that the various "layers" are related but in such ways that strict causal laws and explanations are not applicable. See e.g. Evnine 1991 on this issue. A more modest proposal is to view the two "worlds" of the mental and the semantic as epistemically diverse or even that whichever parallels we posit are made so for heuristic reasons.
Meaning and Religion: On Semantics in the Study of Religion                   249
to the range of "new approaches." I, for one, think so.42 For, the irony of the situation as described here—both historically and currently —is that we seem to be in a position where we are unable to account for that one matter, which appears to be responsible for the most crucial difference between humans and other creatures. It is so eminently intuitive for us to make sense, to simply "know" that there are things that are more or less meaningful in this world. Speaking about religion we may also conclude that religions are means and ways of making sense—more or less. This is a so much taken-for-granted intuition or so commonplace a conviction that it hardly qualifies as a theoretical position. But it should at least be acknowledged as a starting point for further theorizing. If we wish to go beyond the acceptance of meaning as what we could call a "trivial mystery" then we may have to revise our ideas of what science can be about and include within its range such things as "meaning." I — for one—think that this is the only way to make the study of meanings meaningful. Or, a scientific community which is unable to see the interesting challenge of providing us with an account of what it means to mean will simply remain meaning-less itself (just to toy a bit with the term). On the other hand, the study of religion which has always—as seen by most of its practitioners — been concerned with things meaningful could very well be in a position not only to learn from other fields or sciences but also to contribute directly towards a solution of the "mystery" because religion as well as "meaning" both seem to have been with the human species ever since the "symbolic revolution" eons ago. It is quite probable that religion is a product of the human propensity to "make meaning" and to make it in such a way that meanings appear as natural, intuitively available and —not least—stable, as if "given" so that religion becomes a warrant of semantic stability—or strife.
Many things point to religion as being a field of human activity which could lend itself to the most rewarding forms of inquiry in relation to these matters. So, where the modern empiricist would see religion as meaning-less and a result of human folly and superstition it may in fact be the opposite: that religion is such a powerful means of "making meaning" that it deserves serious attention. That is: serious scholarly attention. And then —the study of religion might also attract more and more serious attention from other fields—what it can hardly be accused of in its current situation. The "problem of meaning" is far from being solved but that only makes the study of it all the more meaningful —also within the provinces of the study of religion.
42    Fortunately, I find myself in limited but good company: Hans H. Penner (e.g. 1999) and Terry F. Godlove Jr. (1997,1999) work along these lines.
250
Jeppe Sinding Jensen
Bibliography
Albinus, Lars (1997), "Discourse Analysis within the Study of Religion: Processes of
Change in Ancient Greece," in: Method & Theory in the Study of Religion 9 (3): 203-
32. Avramides, Anita (1997), "Intention and Convention," in: Hale/Wrigth, eds.: 60-86. Ayer, Alfred Jules (1952), Language, Truth and Logic. New York: Dover Publications. Baaren, Th. P. van (1972), "The Flexibility of Myth," in: Ex Orbe Religionům. Studia Geo
Widengren. Leiden: E.J.Brill: 199-206. Boyer,  Pascal  (2000),  'Functional  Origins of Religious Concepts:  Ontological and
Strategic Selection  in Evolved  Minds,"  in: Journal of the Royal Anthropological
Institute 6 (2): 195-214. Braun, Willi/McCutcheon, Russell T., eds. (2000), Guide to the Study of Religion. London:
Cassell. Capps, Walter H. (1995), Religious Studies. The Making of a Discipline. Minneapolis,
Minn.: Fortress Press. Caws,   Peter   (1997),   Structuralism.   A   Philosophy for   the   Human   Sciences.   Atlantic
Highlands, N.J.: Humanities Press. Coffa, Alberto (1991), The Semantic Tradition from Kant to Carnap. Cambridge: CUP. Davidson, Donald (1990), "The Structure and Content of Truth," in: journal of Philosophy
LXXXVII (6): 279-328. Deacon, Terrence (1997), The Symbolic Species. The Co-evolution of Language and the
Human Brain. Harmondsworth: Penguin. Dummett, Michael (1996), Seas of Language. Oxford: Clarendon Press. Duranti, Alessandro (1997), Linguistic Anthropology. Cambridge: CUP. Durkheim, Emile/Mauss, Marcel (1963 [1903]), Primitive Classification. Chicago: University of Chicago Press. Edwards, Derek (1997), Discourse and Cognition. London: Sage Publications. Evnine, Simon (1991), Donald Davidson. Cambridge: Polity Press. Fairclough, Norman (1992), Discourse and Social Change. Cambridge: Polity Press. Fauconnier, Gilles (1997), Mappings in Thought and Language. Cambridge: CUP. Frankenberry, Nancy (1999), "Pragmatism, Truth, and the Disenchantment of Subjectivity,"  in: idem/Hans H. Penner, eds., Language, Truth, and Religious Belief.
Atlanta: Scholars Press: 507-32. Geertz, Clifford (1973), The Interpretation of Cultures. New York: Basic Books. -----(1983), Local Knowledge. Further Essays in Interpretive Anthropology. New York: Basic
Books. -----(1995), After the Fact: Two Countries, Four Decades, One Anthropologist. Cambridge,
Mass.: Harvard University Press. Godlove, Terry F. Jr. (1997), Religion, Interpretation, and Diversity of Belief. The Framework
Model from Kant to Durkheim to Davidson. Macon, Ga.: Mercer University Press. -----(1999), "In What Sense Are Religions Conceptual Frameworks?," in: Frankenberry/
Penner, eds.: 450-72. Goodman, Nelson (1978), Ways ofWorldmaking. Hassocks, Sussex: The Harvester Press. Hale, Bob/Wright, Crispin, eds. (1997), A Companion to the Philosophy of Language. Oxford: Blackwell.
Meaning and Religion: On Semantics in the Study of Religion                   251
Harris, Roy (2003), Saussure and His Interpreters. Edinburgh: Edinburgh University Press.
Heal, Jane (1997), "Radical Interpretation," in: Hale/Wright, eds.: 175-96. Jensen, Jeppe Sinding (1999), "On a Semantic Definition of Religion," in: Platvoet et al., eds., The Pragmatics of Defining Religion. Leiden: Brill: 409-31.
—  (2000), "Structure," in: Willi Braun/Russell T. McCutcheon, eds., Guide to the Study of Religion. London: Cassell: 314-33.
—  (2001), "Universals, General Terms and the Comparative Study of Religion," in: Numen 48: 238-66.
—   (2002) "The Complex Worlds of Religion: Connecting Cultural and Cognitive Analysis," in: Ilkka Pyysiäinen/Veikko Anttonen, eds., Current Approaches in the Cognitive Science of Religion. London: Continuum: 203-28.
lensen, Jeppe Sinding/Luther H. Martin, eds. (1997), Rationality and the Study of Religion. Aarhus: Aarhus University Press.
Lakoff, George (1987), Women, Fire, and Dangerous Things. Chicago: University of Chicago Press.
Langer, Suzanne K. (1942), Philosophy in a New Key. A Study in the Symbolism of Reason,
Rite, and Art. Cambridge, Mass.: Harvard University Press. Lawson,   E.   Thomas/McCauley,   Robert  N.   (1990),   Rethinking  Religion.   Connecting
Cognition and Culture. Cambridge: CUP. Leach, Edmund (1976), Culture and Communication. The Logic by Which Symbols Are
Connected. Cambridge: CUP.
Lechte, John (1994), Fifty Key Contemporary Thinkers. From Structuralism to Postmodemity.
London: Routledge. Leech, Geoffrey (1990), Semantics. The Study of Meaning. Harmondsworth: Penguin. Lotman, Yuri M. (1990), Universe of the Mind. A Semiotic Theory of Culture. London: LB.
Tauris & Co.
Malinowski, Bronislaw (1948), "Myth in Primitive Psychology," in: idem, Magic, Science and Religion and Other Essays by Bronislaw Malinowski. New York: Doubleday: 93-148.
Murphy, Tim (2000), "Speaking Different Languages: Religion and the Study of Religion," in: Tim Jensen/M. Rothstein, eds., Secular Theories of Religion. Current Perspectives. Copenhagen: Museum Tusculanum Press: 183-92.
Ogden, C.K./Richards, I.A. (1923), The Meaning of Meaning. London: Routledge & Kegan Paul.
Paden, William E. (1988), Religious Worlds. The Comparative Study of Religion. Boston, Mass.: Beacon Press.
Penner, Hans H. (1999), "Why Does Semantics Matter," in: Frankenberry/Penner, eds.: 473-506.
Peregrin,   Jaroslav   (2001),   Meaning   and   Structure.   Structuralism   of  (Post)Analytic
Philosophers. Aldershot: Ashgate. Putnam, Hilary (1988), Representation and Reality. Cambridge, Mass.: The MIT Press. Ricoeur, Paul (1981), Hermeneutics and the Human Sciences. Essays on Language, Action
and Interpretation. Ed., trans., and introduction by John B. Thompson. Cambridge/
Paris: CUP/Editions de la Maison des Sciences de l'Homme. Saussure, Ferdinand de (1993 [1915]), Course in General Linguistics. London: Duckworth. Searle, John R. (1995), The Construction of Social Reality. New York etc.: The Free Press.
252
Jeppe Sinding Jensen
Shore, Bradd (1996), Culture in Mind. Cognition, Culture and the Problem of Meaning. New
York: Oxford University Press. Skorupski, John (1997), "Meaning, Use, Verification," in: Hale/Wright, eds.: 29-59. Staal, Frits (1989), Rules without Meaning. Ritual, Mantras and the Human Sciences. New
York etc.: Peter Lang. Stiver, Dan R. (1996), The Philosophy of Religious Language. Sign, Symbol and Story.
Oxford: Blackwell. Strenski, Ivan (1987), Four Theories of Myth in Twentieth-Century History. London: The
Macmillan Press. Taylor, Mark, ed. (1998), Critical Terms for Religious Studies. Chicago: University of
Chicago Press. Thibault, Paul J. (1997), Re-reading Saussure. The Dynamics of Signs in Social Life. London:
Routledge. Turner, Mark (1996), The Literary Mind. The Origins of Thought and Language. New York:
Oxford University Press.                                                                           '
Turner, Victor W. (1969), The Ritual Process: Structure and Anti-Structure. Ithaca, N.Y.:
Cornell University Press. Whitehouse, Harvey (2000), Arguments and Icons. Divergent Modes of Religiosity. Oxford:
OUP. Wittgenstein, Ludwig (1974), Philosophical Grammar. Oxford: Blackwell.
Religion in Context
A Discussion of Ontological Dumping by
Kirsten Hastrup
In anthropology we have been confronted with a number of more or (mainly) less elegant definitions of religion along with other social and cultural phenomena. Among the more elegant ones is Edward B. Tyloťs definition of religion as "belief in spiritual beings" (1871). Less elegant is Clifford Geertz' attempt at locating religion in the social domain rather than individual belief; for him, religion is (or was, in 1966), "a system of symbols which acts to establish powerful, pervasive, and long-lasting moods and motivations in men by formulating conceptions of a general order of existence and clothing these conceptions with such an aura of factuality that the moods and motivations seem uniquely realistic" (Geertz 1966: 4). A century separates the two, and the difference in emphasis is a symptom of a particular development of anthropology, becoming increasingly sensitive to context and consequently ever more uncertain about its own terms.
My aim is neither to rehearse a series of definitions nor to trace a particular development, however, but to discuss why exhaustive definitions are epistemo-logically impossible, and why elegance may in fact prove a legitimate yardstick of individual concepts. Although I will use examples from the study of religion in both ancient and modern times, the discussion is theoretical rather than empirical, and should be seen in the light of general trends in anthropology, set within the larger horizon of the human and social sciences (K. Hastrup 1995, 1999). The principal object is scholarly understanding itself, that is the process by which we seek to comprehend and represent whatever part of human life we are currently studying.
My suggestion is that scholarship always works and advances by way of persuasive fictions or naturalized illusions one of which is "religion." To call it an illusion is not to claim it to be objectively false, but to point to the power of conceptual categories, including "religion." The idea is not to rehearse the old debate on nominalist versus realist definitions, nor to land us in yet another constructivist camp. The point of the exercise is to highlight the nature of scholarly understanding itself, by way of the study of religion. After the demise of