MEANING OR MOANING? AN ETHNOGRAPHIC NOTE ON

A LITTLE-UNDERSTOOD TRIBE

MARK HOBART

Meaning is what essence becomes when it is divorced from the object of re ference and wedded to the word.

Quine - Two dogmas of empiricism.

Was it the Queen of Hearts or Humpty Dumpty who liked changing th e rules to suit their position? An unworthy doubt sometimes creeps into mind that "meaning" is so slippery a word that those who use it may lind th ey are unwittingly wearing Lewis Carroll 's cap. The history of anthropology is littered with th e wreckage of theories, th e ambiguity of th e core conce pts of which was as essential to their initi al appeal as it was to their eventual decline. But anthropologists, understandably, pre fer th eir working co ncepts on the hoof, so to speak, and are suspicious if they are nea tly stuffed for inspection . There is a drawback though to the comfortable stance that what is meant by mea ning should be evident to an idiot. Not only does this let idiosyncratic interpretations of culture pretend to infallibility, but it may make what is being talked about quite obscure. The term itself has a cu rious an cestry. As Harold Bloom rem arks "the word meaning goes back to a root that signifies 'opinion' or 'intentio n' , and is closely related to th e word moaning. A poem's meaning is a poem's complaint ..." (1979 : I, italics in the original; see also Onions I 966). In fact still more lies behind the usage.

A short survey of popular theories of mea ning may help to highlight some o f th e problems, and the unstated presuppositions. Like th e tiger's tail, it is quite possible - if dangerous - to seize upon a convenient notion withou t bothering about what it may entail and commit one to what it may. Ethnography poses a dou ble difficult y.

Mark HomeTypewritten TextThis article appeared in Semantic Anthropology

Mark HomeTypewritten Text

Mark HomeTypewritten Text

Mark HomeTypewritten Texted. D. J. Parkin, London: Academic Press, pp 39-63.

Mark HomeTypewritten TextCopyright Elsevier

Mark HomeTypewritten Texthttp:

Mark HomeTypewritten Text//www.elsevier.com

Mark HomeTypewritten Text

Mark HomeTypewritten Text

http://www.elsevier.com

40 Mark Hobart

Resea rch rcquire s the st udy of indigenous categories and cul tural assumptions, whil e anthropology itself is part of a changing, an d inte rnally diverse, Wes tern academic tradit ion. This makes the problem of transla tion in its broades t sense more se rious than is often recogn ized. It is easy to aSSume that our academic, and cultural categories are self-

42 Mark Hobarl

vernacular li se as a sta rt. Some of th e more obviolls are given in T able I. From this alone " to mean" is roughly synonymous with : intend , sign ify , show, have value (or signifi cance), refer to , sta nd for . Mea ning a lso stre tches to cover ca usa ti on. Clouds are a necessary condi tio n of rain, no t an arbitrary signifier. This issue of the " motiva tion" of signs will crop up in du e course . It should be apparent though that English usage (as those of o th er European languages) may include several senses and distinct kinds of relationship .

Mea ning may also be ap plied to qui te se parate aspec ts of uiscourse. We may need to distin gui sh between th e mean ings of words , sentences and whole tex ts. To. Ri coeur th e whole difference between semiotics and semantics is tha t between simple significa tion (what he dismisses as the "unid imensional approach" ) and the almost infinitely variable rela tionship between subject and predica te by which all propositions are form ed ( 1976: 6ff.) Beyond th a t there is a clear sense in which th e meanin g of sentences cannot be taken out o f con text. Contex t, however, presents some u npleasan t prob lems o f its own. For th e present it is useful to note th at the different levels at which it is possible to speak of " meani ng" are oft en muddled.

If uses of meaning ap pear confused, perhaps an analyt ical approach of the kind favoured in British Or American philosophy may help? The re are at least seven main theories. A short summary may be useful as it separa tes some of the central issues; and if we distance ourselves a little by treating philosophers e thnographica ll y, we find that they unwittingly offer all sorts of clues as to their presuppositions which might o th erwise escape no tice.

Perhaps the most plausible view is th at words are a way of talking about things. In " Deno tation Theory" words have meaning by denoting things in the world , the object being the meaning (Russell 1905 ; c f. Lyons 1977: 177-215 on confusio ns be tween uenoting and referri ng). Matters are not so simple however. For how does o ne speak, for instance, of past events and imaginary objects? It is hard , by this approach, to cope with words like "and" or " if' , which have no physical counterparts, but being logical connectives ought be included in a comprehensiv e theory of mea ning. The stress on physical objects turns ou t not to be ac~iden t al. The same object may be appreciable in differen t ways; and it is common to distingu ish between th e reference and the sense of a term (Frege 1892, translated 1960) which may be variously interpreted but is widely treated as close to th e difference between extension (what a word denotes) and intension (what it connotes in 1.S. Mill's parlance) .' The dichotomy between semiotics (semiology) and henmeneutics can

Meal/ing O f Moaning? 43

be related to these two ways of defining things (cf. Guirard 1975: 4044). Intensional meaning is often expressed in term s of properties which may be described furth er as subjective , objective or co nventional in the ir lin k to an object (Copi 1978: 144). It is possible to trace iJJtension, with its emph asis upon essential properties, back to Greek theories of esse ncc (Quin e in my opening quo tation). So th e link between word s and things is not as straigh t fo rward as mi ght seem ; but the histo ry of th e con nex ion is ancient. If words do no t simply refer to things, what then is mea ning? On one reading:

On ce the theory of meaning is sharply separalcl1 from (he theory of rercrence, it is a short step to recognizing as the prilllary busi.Jl ess of {he \heury of meaning sim ply the synonymy of linguistic for ms anti the analy ticity of statements: meanings themselves, as obscurc inlcrmetliary entities, Illay well he ahcllldoncd. (Quinc 1953: 22 ).

If worus do not simply name things, do they name ideas inste"d ? Th is view, which goes back to Locke (e.g. Staniland 1972: 28-52), was more recently espoused by Sa pir ( 1921) where he tied meaning to the mental images of objects. Images of a thing vary , however, between people; and many words cannot be imagined a t all. One version of "Image Theory" substitutes "concept" fo r "image" and on th is de Saussure based his theory of langu"ge (for good critiques see Black 1968 152-6 ; Kempson 1977: 16-17). For his d istin ction of signiJicanl:signiJie is that of sou nd:concep t (Baldinger 1980: 1-7; Lyons 1977: 96-98). Th e reliance of de Saussure and some of his successors upon a rather stea m-age theory of meaning is rarely made explici t.

The two approaches so far discussed try to fix th e meaning of words. The next set are concerned with sentences, or propositions (what may be wrong with reducin~ the former to th e latter is discussed in Quine 1970: 1-14) . Th ese theories seem to ground themselves in so me fonm of "rea iity" or, as Pu tn"m put it "a world which admits of description by One True Theory" (I98 i: xi). The crud est version , " Ca usa l Theory", tries to derive meaning fro m causation . The mea ning of a sta tement is the response(s) it induces (Stevenson 1944) . One way of whiling away a dull afte rnoon is in inventing expressions to which no sa ne man could possibly respond .

A more serious co nte nder is "Verification Theory" . This se ts ou t to defin e th e me"ning of a proposition by its correspondence with reality. In its classic fo rm "the meaning of a proposition is the me th od of its verifica tion" (Schlick 1936). This view has an obvio us appeal in th e natural sciences; but it is harder to see how it would comfor tab ly fit cu lt ural disco urse. There are man y things which are

44 Mark Hobart

beyond verifica tion even in principle, such as past, or unobservable, events. The origin al version has been refined in various ways (a "weak" versio n of the criterion of verifiability has been proposed by Ayer 1936) perhaps th e best known being Popper's preference for "falsifiable" over "verifiable" . So, for a sentence to have meaning, what it says must in principle be falsifiable by facts. This is potentially a useful way of scrutinizing certain kinds of theory (see the debate be twee n Kuhn a nd Popperians in Lakatos & Musgrave 19 70) but, on at leas t o ne interpreta tion, it wou ld leave every novel, poem or religious belief as mea ningless. It would seem th en that theorie s of meaning may a t best o nly work for a given problem. If so it might be inappropriate to try to apply them genera lly.

The work of the Logical Positivists points to a fascinating problem. Members o f th e schoo l such as Carnap se t out explicitly to produce a system free of metaph ysical assumptions ( the title of one work was "The elimination of metaphysics through logica l analysis of language" 1932, translated 1959), and further held tha t all metaphysical statements were meaningless. It is questionable whether they succeeded in this. If one is empiricist enough it is perfectly possible to regard physical o bjects as me tap hysica l assumptions in their own right. For instance:

Physical objects are co nceptually imported into the situation as conven ienl illtermcdittries - not by definition ill terms o f ex perience, but simply a~ ilTeduciblc posits co mparable, epistemologically, to the gods or Homer .. _in point of e pistemological foo tln g the physical objects and the gods differ only in degree and nut in kind. Bo th sort s or entities ente r our concc ptio n o illy as c ultural posi ts. (Quine 1953: 44).

If the philoso phical elders ,Ire not unanim ous, it seems at least th at most have strong, and partly assumed, beliefs of a distinctive kind.

All this might seem far from anthropological te rra firma (if that it be). Not only is knowledge of Our own ideas beginning to seem increasingl y relevant to a study of mea ning, le t alone in oth er cultures ; but it seems that our ideas are collective represen tat ions which impose stark limits on what we think. This co mes ou t dearly in the most elegant of the reality-based of th e "Correspond ence Theori es". Rather simply put, a true proposition is in correspondence with reality, a fabe one not. The argument developed by Tarski (1944) and Davidson (1967) is too complex to discuss here (for good accounts see Lyons 1977: 10-13,154-173; Kempson 1977 : 23-46). Several points are relevant though. First translation is held to be possible by virtue of it being possible to specify co nditions of truth valid for all possible worlds (presumably this ought to

Meaning or Moaning? 45

include the e thn ographer's cu ltu re of s tud y; one trusts this is not an impossible world). Seco nd the theory applies to sentences, not propositions, so it is necessary to remove the ambiguity of the form er. To cope with this uemand , it is necessa ry to focus on the truth o r falsity of se ntences under a given interpretation . Othe r sentences may have indetenninate refere nce. So, to fix the mea ning of a sentence , we have to posit, however temporarily, a se parate interpretation , or specify a refe rence. If ambiguit y still remains, Ihis is held to be the fault of the component expressions, or of grammatical structure (Lyons 1977: 169-70) . Language it seems must be made transparent whatever the cost. Procrustes and his bed-technique seem kind by comparison.

The dirficulties of correspondence theory have been nea tly put by Gellner:

Language runctiolls in a variety or ways o ther than "rererring to objects" , Many objects are simply not there, in fl ny obvious physica l sense, to be located : how could Olle, by this method , es tablish the equiva lences, if they exist, between abstra ct or negative or hypothetical or religious ex pressions? Again , lIlan y "objects" are in a sense createu by th e language , by Ihe manner in which its terms carve up the world or experience. Thus the mediating third party is simply not to be found: either it turns ou t to be an elusive ghost ("reality"), or it is just one further language, with idiosyncracies of its own which are as liab le to distort. in translation as did the original language of the investigator. (1 970: 25).

The difficulties include then how truth is to be und erstood and the problems in moving from sentences in actual (natural) languages to notionally context-free true propositions. Th e loss is that all religious, moral and aesthetic statements become beyo nd the pale , which leaves us poor anthropologists driven back to eco logy, with even such trusty standbys as power looking distinctly gJeen at the gills.

The last approach we need to consider puts meaning finnly within culture and habits of language use; for which reaso n perhaps it has a degree of popularity among anthropologists. After proposing, in his complex "Picture Theory", that meaning was achieved by a homology between reality and the structure of language, Wittgenste in emerged with his second, or "Use Theory" (1958, 2nd edn 1969; 1953, 2nd edn 1958; in each case the latter difrers slightly). It has kinship links with verification theory in the stress upon method, but improves on it by locating meaning in the use of words in a language. So meaning is not a kind of object in th e natural world, but a part of cu ltural convention. Language is used in a rather special sense though. For, in any society , there are man y different

46 Mark Hoban

systems of verbal signs, e~ch with rules of proper use. Meaning depends th en not o n a pa n-c ultural convention, but upon employment in a particular context (Wittgenstein 1969: 17). Wittge nstein re fers to each se t , an d "also the whole, consisting of a language and the acti ons into whieh it is woven (as) the 'language game'" (1 958: I, 7 my parentheses). These ga mes include: giving orders, describing objects, reporting events , forming hypo theses, making up stories, translating, praying etc. (1958: 1,23 ). "Th e speaking of language is part o f a n activity, or of a form of life" (1958: 1,23) and "what has to be accepted, th e give n, is - so one could say - form s of life" (1958: II, 226e). Differe nt sets of terms ca nnot be directly compa red; for lang\iage usedepends on a context.

Oncc again Gellner is convenie ntly on hand to note the drawbacks.

If "meaning::::; use". then "use ::::; meaning" .. . if the meaning of expressio lls is their em ploy mcnt . lhcn.in tum. it is of the esse nce o r the employment of ex pressions (an o h y

48 Mark Hobart

the metaphors , requires an intuitive leap to grasp whole and part together (cf. Ricocur 1981: 57, o n the subjec t "entering into" th e knowledge of th e objec t). Similar problems apply between sentence and text; and presumably be tween tex t and culture . To approach meaning requires "pre-understanding" by the interpreter (Bultmann 195 7: 113 ; cr. Betti 1962: 20-1 who objects to this whole idea) or the context of any utte rance. So, once again, context descends as th e deus ex machil1a to reso lve the seemingly intractable problems or meaning. Text being philoprogenitive , it has spaw ned co n-text and , ror general ed irication, pre-text and inter-t ext (Culler 1981 : 100-118 ; ir pre-text = pre-understanding, does text = und ers tanding?). Gelln er, among'o th ers, has made the point that th e distinguishing reature o f most soc ial anthropology - typified by runctionalism - is its stress on contex t in analysis ( 19 70 ; 1973a ; 1973b). It is not unamusing, there rare, to see hermeneuts and philosophers find the answer to their problems in a concept which anthropologists have been enthusiastical ly dissec ting in numberle ss specialized ways ror decades. The re are, it would seem, Frankenstcins aroot hoping to breathe lire back into th e dismembered corpse or context.

In view of the difficulties in getting th e semantic band wagon onto th e road , it will hardly come as a surprise that much of the successful work has been phrased in tenn s, not of meaning, but of significat ion . (Wilden has argued that sign irication is simply the digital coun teIJlart or meaning in analogic cod ing, 1972.) There will be, I rear, a sense or deja vu wh en it tums out that th ose wh o agree that language, and indeed culture, should be approached semiotically disagree as to how signification is to be understood (see Lyons 1977: 95-119; also Baldinger 1980). The problems may be exemplified by a short look at th e work or de Saussure , because or his great impact in an thropology. Ju st how closely apparently unrelated schools a re actua ll y providing alternative rormu lations or similar problems comes out in the rollowing ci tation.

. .. semiotic syste ms are "closed", i.e., without relations to external, nonsemiotic reality . Th e defi nition of the sign given by Saussure already implied this postulate : insteiJd of being defined by the external relation between a sign

50 Mark Hobart

(cf. Boon 1979), philosophical roots. As Benoist has made plain, the problem was aired as long ago as Plato's dialogue, the Cratylus, as to whether the relation of names and things is natural or conventional; whether thcy are based in physis or nomos (techne).

Hermogenes versus Cratylus, Saussure versus Pierce: western knowledge since the Greeks has always put, and tried to solve, the question of the relalionship between culture and nature. Is culture rooted in nature, imitating it or emanating direct from it? Or, on the contrary, is culture at variance with nature, absolutely cut ofT from it since the origin and involved iII the process of always transforming, changing nature? The matrix of this opposition between culture and nature is at the very matrix of Western metaphysics. Metaphysics constitutes it, or, in virtue of a circular argument, whose name is history, is constitued by it. (1978: 5960).

At every turn the close link between meaning, or signification, and notions of essence, truth and so forth have lurked near the surface of discussion, Benoist brings out clearly just how much cunent debates depend on conveniently forgotten, or worse unrealized, philosophical conundrums. Our intellectual ostrich seems to bury his head ever deeper.

No account of signification would be complete without reference to the work of Levi-Strauss, the more so as he has often been held to dismiss meaning as unimportant to his style of analysis. Sperber has, rightly, questioned how seriously the parallel between linguistics and structuralism should be taken. For

. despite a terminology borrowed from linguistics, symbols are not treated as signs. The symbolic signifler, freed from the signiflcd, is no longer a real signifier except by a dubious metaphor whose only merit is 10 avoid the problem of the nature of sYlllbolislll , not to resolve it. (1975: 52 )'

Further

the fundamental question is no longer "What do symbols mcan?" but 'How do they mean?' . (but) the questillil 'how' presuppos~s the knowledgc of 'what'. Saussurian semiology therefore docs not in principle constitute

52 Mark Ho bart

ques tion that tha t impenetrable writer, Derrida , to my mind makes one of his most use ful suggestions. It is a pervasive "metaphysics of presence" which crea tes these seeming paradoxes , or contradictions.

The problem may be seen to lie in the western tendency to construe being (what exists) in ternlS of what must be experienced as present. The notion of meaning, Derrid a argues, stems from this metap hysics. For we tend to think of meaning as something present to the awareness of a speaker (one migh t add the idea of awareness itself is compounded of presence) as what he " has in mind" (Culler 1979: 162) without recognizing how metaphorical our observation is. The difficulty is th at th e image of container (mind) and contents (meanings, thoughts, ideas e tc.) is dangerously misleading as there are no grounds seriously to hold this position except as metaphor. Yet the two notions co nvenie ntly imply one another (Derrida 1979: 88ff.). If the rela tion of signifier and signified is not simple substi tuti on , but rat her involves mutual supplementation as well, then it no longer becomes self-ev id ent tha t the proper sense of words, ra ther than the figura tive, is original (Benois t 1978: 29) . Put another way , how much is the priori ty we give to Iite r~1 meaning over, say, me taphorical due to our sense that the former is somehow more "real" or present? The supposed paradoxes of semiotics become expressed in terms of figurative speech.' It seems tha t we must pack our bags yet agai n.

So far two themes seem to run through approaches to meaning. Each theory tends to be grounded in another domain , so displacing the focus of inquiry. Saussurean schools of thought escape this in part and make clear the dichotomy between internally and ex ternally defined models. More ge nera ll y, whatever the approach, a t each turn we are faced with problematic distinctio ns which have their roots in the history of Western phil osophy : the reality of the physical or the mental ; the relation of focus and context; na tural law agai nst cultural preferences: the essential o r the nominal.

Figures of speech would seem to by-pass the hybrid problems of fo rm versus content by being cen tred abou t content , Or mea ning. They offer a classification of possible fo rms of resemblance, and associa ti on, and so a potentially unambiguous language of critica l evaluation. Th is promise obviously depends all exactly what figures of speech, or tropes, are or do; and the assumptions on which they rest. With the tropological phase a t its peak, figures of speech are being hailed as the new philosopher's stone - gall-stone to some and the problems tend to be shoved aside. Tropes may be brought to bear on almost anything not only within the study of discourse, but

Meaning or Moaning? S3

they ~re used to threaten the fo undations 01" our knowledge. They are seen as the key to epis temological shi fts (Foucau lt 1970) ; they ma y be constitutive of aU our thought (Ortony 1979 ; Lakoff & Johnson 1980): to the delight of many they offcr to turn LeviStrauss's gay new structural dog into a mangy mongrel with a prom iscuous pedigree reaching back to Quintilian and Ari stotle (Cu ll er 198 1; Derrid a 1976, 1979 ; Sapir 1977, cf. Crocker 1977).

Th e problems start when we try to find out quite what tropes are. Rh etori cians commonly ho ld the vast range to be red ucible to four main forms: me taphor, metonymy, sy necdoche and irony (t he order is important as a sequence to Foucault) . In Sapir's scheme, metaph or has two varieties: inte rn al based on shared properti es; and external (or analogy) whcre properties a re secondary to the fo rm al congru ence of relationships. Th is latte r, he argues. is central to LeviSt rauss's analyses (\ 977). Metonymy is often trea ted as contras ted to metaphor: contiguity not shared property (Cu lle r 198 \: 189ff.). Synecdoche is the possible permutations of wholes (genus) and parts (species), and underpins classifications (Sa pir 1977 : 12-19) . Irony is often held to sta nd apa rt. There are two obvious questions. What kinds of relationship fa ll to each trope? And how are the tro pes related? For Sapir cause and effect, fo r insta nce, are metonymic (\ 977: 19-20) ; for Burke they are clea r examples of synecdoche (\ 969: 508). The difficul ty stems from how the majo r tropes themselves are to be defined. J akobson reduces synecdoche to metonymy (1956). The Belgian rhetoricians in Liege , Group 11 , after detailed review of th e fi eld , concluded that all metaphor ca n be redu ced to synecdoche (1970 French edn ; 1981 Eng lish). In the same year however, Genette traced synecdoche, metonym y and all other trop es back to metaph or (1970). Sin ce then Eco has completed the confusion by derivi ng all metaphor from spurned metonymy (1979). One might be fo rgiven for thinking th at wh om God wishes to destroy, He first makes mad.

Why should such dis tinguished scholars disagree so s trikingly? One reason is that the classical sources themselves s ta rted from different positions (Aristotle 1941 ; Quintilian 192 1). What kind of entit y (sic) are tropes in fa ct? Often they are trea ted as a simple classification of types of association: "buttern y-

r"

54 Mark Hobart

means by

. wlDeh being properly speaking is ex pressed in so far as it is expressed through several twists. several tropes. The system or cat.egories is the system of the ways in whic h being is construed . ( 1979: 91)

Figures of speecll seem to bring us back to the old problems of what are properties? what is essential? what indeed is accident? Derrida would push it further and see trop es as underlying those "basic" kinds of category - subs tance , action , relation , space, time, accide nt - through which westcrn philosophers try to capture being.

Tropes seem to have more immedia te uses in creating words and images, where non e were before. "Metaphor plugs the gaps in the literal vocabulary" and so is a form of catachresis - the putting of new senses into old word s (Black 1962: 32). Such extensions may be almost totally constitutive when conceiving the relation between events in temlS of ideas like time. This raises the possibility that language itself may be construed metaphori ca lly. In English , Reddy has argued that our impressions of language are largely structured by the image of language as a container, the contents being ideas, thoughts, feelings, or indeed meanings (1979 : 289). By pointing out that alternatives are available to what he calls the "conduit metaphor" , Reddy makes a strong case for the catachretic nature of many of our core concepts (those noted above). Foucault has sough t to generalize this kind of argument by applying it to how we structure relations between classes - such as the sane and mad (1967) - or even how our basic ideas of what constitutes an explanation are made up (1970). In this view it is the image of irony which is now dominant. This produces a contrast between a surface and an interior, such that the superficial is to be explained by a deeper structure, as in th e Freud ian model of mind or in structuralism. The doubling will allow alternative styles of analysis: the search for formalism (perhaps Needham 1978; 1980) as against some (hidden) meauing (Geertz 1973). Metaphor seems then to make up how we see th e world and how we set about studying it - eve n if we are not sure what metaphor is.

This is not quite the end of the tale. With their Nietschean heritage , the French post-structuralists - in a mood of fin d'epiSlhnesee no escape from the web of words, or tropes, what Jameson called "the prison-house of language" (1972). This gallic gloom may be a little premature. To reverie Davidson (1980) , let us wonder whether, to use a principle of lack of charity, the turgid and convolute style of these writer; does not serve to obscure their own Achilles' heels.

Mean;'lg or Moaning? 55

If re lationships and abstract issues are conceived catachretically , why should this not hold good as well for the relationship between image and referent in Foucault's and Derrida's own models? We seem near to the self-refe rential paradox. What exactly is the relationship between an epistemc and what it stru c tures; or language and to what it relates? There is no reason that this must be confined to th e image of a prison by which thougllt is kept in place. I t seems that their discourse carries within it its own unexplored metaphor. Lovelace's prisoner wondered whether iron bars made a cage. Why should they not make a jemmy for a burglar, or a jail-breaker?

There is another way in which tropes may be understood. They may be treated as ways of perceiving relationships and situations from different per;pectives. As such they may cove r far broader areas than formal categories and may represe nt general processes of thinking. Burke, for instance, sees the four major tropes as exa mples of the more encompassing operations of: perspective, reduction , representation and dialectic (J 969). This would go SOme way to explaining how, if they are treated as classes to be defined extensively , they run in to problems.

Th is is not, however, how tropes are understood by most writers. Around th e terms for the main tropes seem to cluster all those metaphysical problems in Weste rn thought which have dogged meaning throughout. Even the classification is fluid. For the same distinctions may be linked to different figurative terms according to one 's point of view. For instance, the accidence often associated with metonymy may link the latter to me taphor (through essential as against contingent properties), or to sy necdoche (contingent opposed to necessary connexion). At one level tropes come close, it would seem, to simple modes of discrimination and associatio n. As th ey are defined in so many ways it is hard to find neat fit, but the four master tropes involve recognition re spectively of re se mblance, relationship, classification and contrast' If the familiar problems are posed (what is being talked about : essences, properties, names etc.?), we seem to be back to the rondo of confused classes of the rhetoricians.

Studying Western thought with th e aid of tropes may be highly informative. For both are home-grown within the same culture. On what grounds, one might ask, is it legitimate to export them to the tropics? A horrible possibility occur; as to why structu ralism should have the appearance of being so widely applicable. Is it that the main tropes are truly cross-{;uJtural? This has yet to be shown; and there are endless disputes as to how they are to be defined anyway. Or is it just that most (perhaps all?) cu ltures have certain cogni tive

56 Mark Hobart

operatio ns in some form ? 1 t would be hard to imagine a socie ty with no notion of resembla nce (and so presumably the rudim ents for making connexio ns which look like metaphor). Might it not be the ostensible congru ence of these kinds of operation which all ows apparent translatability? Is the stru cturalis t cla im to be able to decode myth accu rately from Indo nesia to South America then false, because th e constru cts it uses regis ter only gross parall els? 1 suspect so. One would be fooUlardy indeed to assume, fo r instance of rese mblance, that exactly what it is about a thing, or event, tha t enables it to be compared is necessarily the same in all cu ltures. In short, can we presume that o th er cu ltu res have precisely the same fo rmulations of rese mblance, relationship, class, contrast and so on? Or are their views of wh at is essen tial, accidental, necessa ry and more sufficiently identical to our own that transla tion is unproblematic? There is sufficient prima facie evidence that ideas vary quite enough - in classical Indian metaphysics as one example (lnden 1976; Potter 1977) - for it to be folly to assum e one's ow n cultu ral constru cts apply across cultures instead of argu ing the case. There is, after all , no reason why translation shou ld be an all-or-nothing busi ness. Why can the re not be degrees of understandi ng and misunderstanding? Part of the t rouble comes, it seems to me, from treating the no tion of "com munication" as simple fact, not sometimes as ideal only partly achieved. Because we dimly perceive so mething through the crud e hom ology of formal opera tions in different cultures, we should not dupe ourselves that we understand very much. 1 think this is why tropes seem to offe r a panacea, and make the formidabl e problems of translation look spuriously easy. If this is so, the sooner we move into a post-tropological age the better. Weste rn philosophers may be excused ethnocentrism . Can anthropologists?

What , if anything, comes ou t of this look at th e philosophy of meaning? The most surprising feature is how much is assumed, and how much of this disclaimed. The elimination of metaphysics perhaps, like marital fidelity, devou tly to be desi red - seems less an ac tuali ty than wishful thinking. Does what we know fare any better then than what we ho ld to exist? There is, 1 think , a case to answer that the lenses in our academic spectacles are forged more figuratively th an we often chose to adm it . On these grounds the slightly facile image of an ethnography of philosophen; will have served its purpose if it has helped to change a tired perspective.

There are other bugbea rs afoot. Wittgenstein's idea of the language game may have its drawbacks , but it does describe rather well what

Meaning or Moaning? 57

academics sometimes do. Can we really talk of theories in general, fo r instance of meaning, when some of the more successfu l work at best in limited situations? (This may be a sim ple aspec t of Quine's point ( 1951) that th e entities which any theory assum es to exist are those which co nstitute the range of the theory's variables.) Theory may have to be very narrowly defined where successors on a single subject inte rpret it such that it has different ranges of application. For example Burke's processua l view of figurative language saw it as framing most thought; whereas Sapir read Burke, or figures of speech, as a formal classifIcation of symbO lic associations. A more disturbing problem is what exactly is implied in the apparent universality of application of our theore tical constructs. Is it, in fact, evidence of the superior power of ou r analytica l frameworks? The scale of Western aca demic resources are so grea t (Gellner 1973a), compared to th e societ ies most anthropologists study, that it is possible to oblitera te the nua nces of a cu lture while seeming to explain it . What criteria are we to use to decide between riva l theories , or transla tions (Hesse 1978)? It is easy to import our own principles of elegance, metaphysics or whatever to resolve th e matter. In th e end, how sensitive will Levi-Strauss's analysis of South Ame rican my th tum out to be , and how mu ch /ese-majeste?

The tension between alternative positions may be renected in differences be tw een philosophy and anthropology. Hollis ha s sta ted one aspect of the problem clearly. He has argued that we are obliged to assume identica l criteria of rationality in other cultures, as we would , in fact, be unable to know it , should altemative logics ex ist (1970). An ana logous argument cou ld presumably be made for meaning, but its implications are frightening. Wh at would be the point of anthropology if, a priori , we could never know if other cultures had different ideas of reason or meaning, even if they did? Part of the impasse stems from different concerns. As J understand Ho llis the philosophe r's br ief is to argue for parsimony, to preven t the world becoming unnccessarily , even hopelessly, complica ted. The more empirical anthropological brief is to keep as much as possible of the subtle ty, and lack of clarity eve n, of cultural discourse as she is spoke. Might it be that we try fa lsely to generalizc issues beyond the en terprises in which thcy were postulated? At any rate the cost to anthropology if we accept Ho llis's argument is so high that it might well be preferable to sacrifice universal notions of rationalit y, meaning or whatever instead.

On e aspect of the me taphysics of prese nce is that argument is sometimes read as claim to truth . This is strange. Usually o nly works of such monumental dullness that no one ca n be bothered to

58 Mark Hobart

question (or read?) them remain unchallenged. The better the argument, often the more it provokes debate and eventually its own reflllement or rejection. In this spiri t let me phrase a conclusion in an extreme form, not beca use it is correct bu t in the hope that it will stimulate others to produce better.

In taking mea ning as the theme , I chose one of many loose threads which threaten to unravel the sweater of contemporary an thropological equivocation, cynically called theory. My arguments are hardly new (Evans-Pritcllard , among others, has put the case far more subtly). If they have any value though, then failure to consider the possibility that other cultures have other philosophies is, at the least, a ghastly epistemological blunder. Western philosophy seems hopelessly caught in its own toils and anthropology is - as I am sure its wiser proponents realized - our one chance of escaping the sheer tedium of our own thought . Anthropology stands little chance though so long as it is bent upon castrating itself on every ru sting knife of in tellectual fashion . As every anthropologist knows, the life of the subject hangs on ethnography, as this is our outlet from onanistic ethnocentrism . Ethnography is not much use if it is no t cri tical; and this critisicm has two aspects. Its obvious face is the reflective considera tion of what to select from the richness of human action and discourse . Its hidden, almost shunned, face is the possibility of reflecting on our own categories, our self-evident assumptions _ call them philosophy or metaphysics if one will - by which we can question the dreary shuffle which passes for the " rational" grow th of knowledge. This is not a fo rlorn search for Shangri-La. For as most ethnographers know, we glimpse through the dark glassily real , and seemingly different , world s. Otherwise there is the grim prospec t of peddling tilched fashions which become, like the Cheshire Cat, long on face , short on body. Then talk of meaning

turns to moaning.

Notes 1 . My special thanks must go to Proressor Martin Hollis who suggested this

ex pression and offered many use ful COIl1ment s;

60 Mark Hobart

Ayer, AJ. 1936. Language, Truth and Logic. London: Gollancz; 2nd edn 1946 Harrnondsworth: Pelican.

Baldinger , K. 1980. Semantic 711eory. Oxrord : Black wells. Benoist, 1.-M . 1978. Th eStrucrural Revolu tion. London: Weidenreld. Betti, E. 1962. Die Hermeneutik als allgemeine Methodik der Ceisteswissen

schaften. Tub ingen: Mohr. Black, M. 1962. Models and Metaphors: Studies in Language and Philosophy.

Ithaca and London: Co rnell University Press .

--1968. The Labyrinth of Language. Lond on : Pall Mall

Bloom , H. 1979 The brea king or ro rm . In Deconstruction and Criticism (Bloom

et al.). London : Routledge & Kegan Paul.

Boon, J.A . 1979 Saussure/Pierce a propos language, society and culture. In

Semiotics of Culture (eds) I.P . Winner & J. Umik er-Sebeok . The Hague:

Mouton. Bultmann, R. 1957. History and scato[ogy . Ed inburgh : Edinburgh Universil y

Pre ss. Burke, K. 1969. A Cram mar of Motives. Berkeley and London: University or

California Press. (arnap, R. 1959. The elimination or mctaphysics through logical ana lysis o r

language (trans.) A. Pap. In Logical Posirivism (ed.) A.J . Ayer. London :

Allen & Un win. Collingwood , R.G. 1939. An Autobiography . Oxford: O xford University Press.

__ 1946. 711e Idea of History . Oxrord: Oxrord University Press.

Copi, 1. 1978 . Introduction to Logic (5th edn). London : Collie r-Mac Millan .

Crocker, 1.C. 1977. The social runetions of rhe to ri,1 forms. In Th e Social

Use of Metaph or: f:ssays on the Anthropology ofRhetoric (eds) J.D. Sapir & J.c. Crocker . University of Pennsylvunia Press.

( uller, J. 1979. Jacques Derrida . In Strucruralism and Since (e d.) J. Sturrock. Oxford: Oxford University Press.

- - 1981 . The PurSliit ofSigns. London: Ro utledge & Kegan Paul.

Davidson , D. 1967 . Truth and Illea nin g. Syn these 17, 304 2 3.

__ 19 80. Essays on Actions all d Even ts. New York: Oxrord University

Press. De rri

62 Mark Hobart

Keg(ln Paul. Oni ons, C.T. (ed.) 1966. The Oxford Dictionory of English Etymology . Oxford :

Clarendon Press. Orlony , A. (Ed.) 1979. Metaphor and 71wughl C