Cave paintings of Wonjinas
(from A. P. Elkin 1930: 262)

On the Translatability of Australian Aboriginal Myth

© Anthony Pym 1998
 
 

(Text originally published in Catalan as "Qüestionement de la traducció del mite", Mites australians, Ed. Anthony Pym, Calaceite: Caminade, 1990, 37-48.) 
 

If myth, according to Lévi-Strauss, is "the discursive mode in which the value of the saying traduttore, traditore is practically zero" (1958: 232), it might be assumed that nothing could be easier than translating myths. And yet it is not clear whether the argument immediately concerns actual work against difficulty or simply the fact that, in the case of myth, there is no unique author who might be betrayed. The ultimate sources are unavailable; every version is already a transformation of a transformation; and thus no translator can be wrong or unfaithful. More important, the thesis of maximum translatability says nothing about how myths should be translated, nor why contemporary societies should attach any value to such translations. These latter questions require a rather more critical approach, and perhaps far less self-assuredness about the relation between what can be said in theory and what is desirable in practice.

Some well-founded voices of dissent may be pitted against Lévi-Strauss. Malinowski, who considered myth to be the "narration of the primal event which set the precedent for an institution" (1926: 56), thus allows us to posit that a myth can only function in the social place of its corresponding institution. Beyond that place, its relation to action must be annulled and translation can in practice only achieve the movement of dead text. The ethnolinguist Hymes similarly regards each people as the owner of its myths, and does so to the point where the only correct translations or interpretations are those carried out by the people itself:

In a Chinookan myth, for example, any translation, even an abstract, would make clear the presence of a structure "Interdiction: Interdiction violated", and imply that the outcome (a murder) follows from the violation, as so often is found to be the case. Analysis of the myth in terms of its specific development in Clackmas Chinook discloses structures that place an almost opposite significance on the myth. [...] The myth is to be understood in terms of a specifically Chinookan theory of myth such that it is here not the violator, but the one who issues the interdiction, who, in Clackmas terms, is culpable. (1974: 99-100) 

Hymes argues that this particular myth is untranslatable. And yet, in presenting his argument, he has surely translated the supposedly untranslatable part of the myth. Georges Mounin, at least, would seem prepared to classify such explanations as effective translations (1976: 54). Yet Hymes' notes obviously belong to an explanatory discourse that is not in itself immediately mythological: it is all too explanatory. Something has changed in this transition; a particular mode of untranslatability has been encountered.

Exactly what is untranslatable here? Surely, that part of the myth which is explanation and yet remains attached to a particular social institution. The linguistic function concerned is moreover decidedly injunctive (the social distribution of guilt), over and above the other entirely translatable functions able to convey the basic narrative and definitional structure of the myth. The part of the myth that can be owned is thus not the "world-view" as such, but the specific social value of certain descriptions, definitions and categories.

A similar distinction between readily available narrative material and untransferable values is in fact a feature of most myths prior to the moment of translation, as Von Leonhardi found in his dealings with the aboriginal tribes of Central Australia:

There can be no doubt that most of the men sing the chants without understanding them, and that women and children witness sacred rites without having knowledge of their meaning. But the old men, who carry the traditions, know precisely what these acts mean, in all their details, and are able to explain them. (cited by Lévy-Bruhl 1935: 114) 

There can be no doubt as to why the specific values of myths often remain unexplained, even though their textual material is known. Their possession is not only knowledge, but institutional power, which will cease to exist when the mythological significations are exchanged in the public domain or the social institutions themselves disappear. Explanation is thus in itself a mode of profanation.

A further theory places the untranslatability of myth at the opposite end of the scale of linguistic functions, referring not to injunctions but to categories. For Lotman and Uspenski (1973), mythological consciousness is marked by a symbolic logic that is untranslatable onto another plane. The basic argument is not sociological, but depends more on examples that benefit from the relative absence of the verb "to be" in Russian:

1. Mir est' kon'.
2. Mir est' materija.

 The first statement ("world is horse") is considered to be based on the non-hierarchical logic of the proper noun and thus non-equivalent to its metalinguistic description in the second ("world is material"). The difficulty of the transition becomes particularly evident in the English rendering of the utterances, where one feels almost obliged to add the analytical articles "the world is a horse". But even with the articles, one readily senses the radical difference between the kind of category conveyed through symbols and that conveyed through hierarchical logic. As much as the second statement might explain the meaning of the first, it is not mythological.
 

Wondjinas for Quine
 

Myth may be one of the few areas in which Quine's theory of "indetermination in translation" is something more than an occasionally perverse thought experiment. The imaginary jungle linguist who sets out to translate native expressions into English is confronted by precisely the kind of problem regularly encountered in the ethnological interpretation of myth. Indeed, the problems faced by real ethnologists are sometimes even more radical that Quine's radical example. Not only is there a decided lack of immediate "stimulation", but also, in some circumstances, a lack of srutable response: initially, Wirz (1925) could only guess the symbolic meaning of the term gåri (nevertheless written with a little hat) because to every question posed, the Marind-anim always answered 'yes'. As Hintikka has stressed (1969: 70-71), without at least a working hypothesis distinguishing between assent and dissent, Quine's system of formal categories falls flat on its face, and no utterance may be considered radically translatable. It should be admitted, however, that the association of sentences with their objective references is not entirely of the same order as the interpretation of symbolic meaning: Wirz was eventually able to identify the term gåri with the enormous head-dress worn in certain ceremonies; but this immediate referentiality was not what he sought as the translatable unit of meaning: "Exactly what does the gåri [now object, not term] signify?" (Wirz 1925: 120) In other words, the object itself simply remits to a further language, a system of objects, in which it must be explained. The problem thus quickly becomes not so much of knowing where to begin (Quine's essential problematic), but of knowing where to stop. The term gåri did not lack referentiality, but situational explanation. The appropriate approach was obviously the analysis of contexts: observation of the ceremonies in which gåris were worn led Wirz to suppose that the symbolic meaning he sought was associated with the sexual orgies that took place at the same time. This was of course pure conjecture, based on the formation of context and comparative observation. But its outcome, as we shall see, was not without interest for future extension of the comparative method: Lévy-Bruhl (1935: 138) notes that a native drawing of a man wearing the gåri, reproduced by Wirz, "is very similar to the paintings we find in Grey and Elkin" and shows no mouth.
 George Grey discovered cave paintings of the mouthless figures in 1838, during his expedition to the north-west of Australia, close to what is now the south of Irian Jaya, that is, not far from the Marindi studied by Wirz. Similar specimens from the same region were studied by A.P. Elkin in 1930 and 1933, providing the basis for the above comments offered by Lévy-Bruhl in 1935. The Frobenius expedition of 1938-39 discovered further examples, described by Petri, Schulz and Lommel. Ernest Worms then located more paintings in 1943-44, documented in his 1968 study; and Grey's specimens were re-found by H. Coate in 1946 (studied by Elkin 1946). In all cases, these figures - Wondjinas - have no mouth. Or, as Quine would perhaps suggest, all paintings without a mouth are called Wondjina.

The first interrogations concerned the verbal level: the use of the name. According to Elkin (1930), "wondjina" not only refers to the paintings, but also means "rain, or the power to bring about rain": "Perhaps we can see the wondjina as the reproductive force in nature and in man, associated with rain." The same author later cited the interpretation given by Mr Love (an appropriately named missionary in the region), according whom the Wondjinas were the first men to walk on the earth, creating numerous geological features and entering the ground in the places where their pictures now stand and where their spirits shall remain forever (Elkin 1933). The relation between the paintings and rain was formally, indeed experimentally, attested:

The natives do not try to restore the paintings. However, in certain cases, and especially in the case of the large Wondjina and his children, the head, hair, eyelids and nose are repainted and restored. This should only be done at the beginning of the rainy season, because the addition of paint brings rain. A native who accompanied me repainted a large Wondjina with some charcoal as I was examining the paintings. Several days later, in the middle of the dry season, there was light rain. The natives took great pains to point out the cause of these exceptional rains. (Elkin 1930) 

This was enough evidence for Lévy-Bruhl to associate rain with fertility, to refer to the orgies Wirz had attached to the gåri, and to declare the mythological relation between the gåri of the Marindi and the halos which, in fact, surround the heads of the Wondjinas.

There is more. Worms, who appears not to have read Elkin's early papers, reports two myths associated with the Wondjina pantings. According to the first, the images represent the spirit of the great snake Walangor, which is concentrated in caves. According to the second, the paintings were born from the eggs of the snake Oongod (whom the ethnologist accorded a fitting transcription). But there must be some doubt as whether these are two distinct myths. The second name would appear to englobe the first since, as Elkin notes, "in some tribes 'wondjina' is an equivalent of 'ungud', meaning 'belonging to the dreamtime or simply definitive': sometimes, to the question 'What is that?', the reply is simply 'It is ungud' (it belongs to the dreamtime)." It would seem that Worms, like Quine's linguist and perhaps like Lotman and Uspenski, took the whole for the part, or mistook the non-reply "ungud" for a proper name.

Whatever the fantasies of ethnologists, there can be no doubt that the Wondjinas are associated with snakes. Worms reports that, in one cave, several mouthless figures were surrounded by no less than forty-two paintings of snakes. The association with fertility would seem fairly clear. Yet still there is more: the snakes concerned - be they Wondjina or Ungud - are specifically water snakes, and their spirit is that of the rainbow (cf. Testart 1978: 109-116). Looking closely at the published reproductions of these figures, it is possible to imagine that the halos around the Wondjinas' heads are in fact snakes, arched in the form of rainbows, and not particularly the gåris about which Lévy-Bruhl was so enthusiastic.

But even this apparent solution does little to exhaust the explanatory possibilities of comparison. For example, an Aranda rain-prayer published by Strehlow (in Nevermann et al. 1968: 235) tells of how Kantjia, the spirit of the rains, who is always surrounded by a curtain of rain, keeps the water of clouds in his hair (his hair is rain), and stops it from falling by means of a tight headband that he only releases at the beginning of the rainy season. Or again, Hernandez (1961: 115-116) relates a Kimberlies myth that tells of how a man who wept so bitterly that Galoru sent thunder and lightning, creating a flood (tears are like rain; the Wondjina's eyes are drawn like its hair). The ethnologist adds that Galoru is a rain-spirit, and that drawings of his head show a surrounding rainbow. And so on. Given this comparative context, Lévy-Bruhl's association of Wondjinas with gåris must seem far-fetched, but remains unfalsifiable.

The point is that explanation through comparison (which is also the basis of structuralist approaches) does not reduce mythological material, but instead tends to accumulate it. In this particular case, comparison may certainly reinforce associations with rain and fertility rites, but such partial explanations merely reveal further absences requiring further explanation. One would have to find out, for example, why the Wondjinas have no sexual organs, and why some of them have a conspicuous mark on their chest. And so on. The simple fact is that ethnological explanation is in itself essentially narrative prose; it is designed to tell us stories about other cultures. Indifferent both to the apparently definitive nature of myth and to the initial difficulties of Quine's jungle linguist, the real problem of ethnological explanation is that it sets up a process of semiosis that it is then unable to stop.

Why no mouth? For Lévy-Bruhl, the question was banal and uninteresting, since "mouthless figures are to be found in numerous prehistoric drawings and paintings, and even on many ancient monuments in the Mediterranean region" (1935: 136). This may be so, but could not the geographical extension of the phenomenon serve to multiply doubt, making its explanation all the more vital? Elkin of course asked why there was no mouth, but received a definitive non-reply: "They simply replied that it was impossible to draw a mouth. Apparently, the effectiveness of the painting depended on this absence" (1930).

Others have asked the same question, and their replies have not been without a kind of mythological reason:

"I am noman, my name is noman" 
but Wanjina is, shall we say, Ouan Jin 
or the man with an education 
and whose mouth was removed by his father 
because he made too many things 
whereby cluttering the bushman's baggage 
vide the expedition of Frobenius' pupils about 1938 
to Auss'ralia 

Ouan Jin spoke and thereby created the names 
thereby making clutter 
the bane of men moving 
and so his mouth was removed 
as you will find it removed in his pictures

        -Pound, Canto LXXIV, first of the Pisian Cantos [1948] (1975)


It was perhaps through Ezra Pound's extensive comparisons (here conflating Homer, the Frobenius expedition and a Chinese figure of silent wisdom) that a certain explanation of the mouthless figure found its way into the poetry of John Berryman (Of 1826):

I am so wise I had my mouth sewn shut

A similarly definitive answer was at least intimated by Elkin: "If a thing is ungud, or brought about by ungud, there is nothing more to be said about it, or at least nothing more is going to be said". And again in the constant "yes" given in reply to Wirz's questions about the gåri. The absence of the mouth is non-reply, negation of the proper name, the escape of Odysseus.

I do not know why the Wondjinas have no mouth. But I suspect it is to prevent them from telling everything to prying ethnologists. As Ortega commented, "every society hides something in order to say something else" (1937: 336). Despite all attempts at translation, sacred knowledge may thus remain sacred, and an essential absence may continue to stimulate both the curiosity of ethnologists and the instituional power of tribal elders. It may even bring rain. Let us leave this absence where it lies - at the base of a process of non-tautological semiosis overlooked by Lotman and Uspenki -, and remark that only poetic metaphors have offered anything approaching a definitive explanation.

Homogeneous translation is cultural appropriation

It is a peculiar fact that, in a fairly mythological science like contemporary theoretical physics, a significant text may come from virtually any university in the world and effectively participate in a discussion with wholly international frames of reference. Since the categories and definitions concerned are explicit, they can belong to an international culture, independently of whatever they might be construed as meaning within individual societies. The cultural distance involved is socially vertical (the initiated as opposed to the uninitiated), and not horizontal (that is, not based on material distance in time and space).

Something must surely be wrong when this same kind of international participation is imposed on a mode of discourse that, as we have seen, is marked by a high degree of implicit and even secret meaning. When ethnological comparison jumps from tribe to tribe, society to society, and even continent to continent, the kind of explanation attained seems more appropriate to its own international extension as European science than to the social specificity of the myths supposedly explained. This is not to deny that myths have long migrated from culture to culture, enabling striking parallels to be discovered in the narratives of societies that are otherwise far apart (even in the case of Australia's geographical separation, there can be no doubt that there were numerous cultural contacts with the islands to the north). Concrete transfer between societies is not in itself a sufficient basis for the assumption of automatic homogeneity. The hypothetical sameness necessarily projected prior to comparative analysis can be equated neither with a universal entity like the "primitive mind" or "mythological consciousness", nor with a constant mode of explanation. It is therefore potentially dangerous to suppose that there should be a homogeneous or constant strategy for translating myths.

In the case of Australian aboriginal culture, it is too late to undo the effects of European homogenisation. Alain Testart (1978) admits it is now impossible to base historical readings of Australian myths on traditional sources: the only record of the myths of pre-conquest Australia is the corpus of texts and observations compiled by European ethnologists over the past century or so. That is, the myths that might most legitimately be associated with an ideal like prelogicity are in fact those recorded in European languages, mostly English and German. Strictly aboriginal narrative is now at least as rare as full-blood Aborigines. Many of the traditional concepts have been mixed with Christian theology, and even basic terms of reference have been adapted to circumstances. Worms notes that, although there are many aboriginal words for the mythological time of creation, "many of the natives themselves now use the English translation 'to dream, dreaming, the dreamtime'" (in Nevermann et al. 1968: 169). European colonisation has already imposed a first level of homogeneity on the narratives previously belonging to numerous different tribes and numerous different languages. On the social level, it is now impossible, and not particularly desirable, to restore the previous heterogeneity. Does this mean one should therefore translate in a homogeneous way?

Quine (1960: 58) notes that Malinowski "spared his islanders the imputation of prelogicity by so varying his translations of terms, from occurrence to occurrence, as to sidestep contradiction." That is, the strategic variation of terms overtly works in the interests of homogeneous translation, ultimately making natives speak like ethnologists. Quine goes on to support the ethics of this strategy: "Wanton translation can make natives sound as queer as one pleases. Better translation imposes our logic upon them". Thus, presumably, "the world is a horse" is a better translation than "world is horse". But could it then be considered ethically inferior to truly impositional translation as "the world is material"? To what extent is the projection of "our logic" a strategy for the extension of international science, for the cultural appropriation of all myth in the guise of universal human dignity?

What exactly is "our logic" in such cases? It would be possible, for example, to extract only basic narrative structures and transform all aborginal myths into a series of cute children's stories, as has been done in the coffee-table versions marketed by Mountford and Roberts (1969). Alternatively, a translator could overtly choose the path of ethnological explanation and get lost, as we have seen, in the midst of skyscraper footnotes. Or again, the sacred nature of many myths, especially as related through chants, lends itself to essentially poetic translation which, like the Wondjinas, may even find some kind of explanation in verse. In this line of argument, Meschonnic (1973) believes that fidelity to rhythm should act as a defence against annexation.

Yet myths translated according to any one of these strategies will inevitably finish up sounding like children, ethnologists or "queer natives" who communicate by beating drums. None of these logics can hope to be entirely adequate, since none can transfer the diverse forms that myth has within the familiarity of the source society. It should be borne in mind that myths are circulated not only in their recititative versions, but also through images, song, dance and drama. They are at once children's stories, the points of departure for endless secret elaborations of knowledge, and the stuff of poetic utterance. The translator's problem is not that of choosing an incorrect discourse, but the fact that contemporary logics of genre and modes of artistic production makes such choice obligatory. Whatever is done, the translated text will convey only certain aspects of the source material. A certain degree of cultural appropriation is unavoidable.

Perhaps there can be no theoretical solution to the problem of how to translate myths. All radical translational alternatives prove more or less unsatisfactory. Non-transfer, the simple refusal to have anything to do with mythological material, would lead to the total disappearance of traditional myths along with traditional societies. Non-translation, which amounts to simply repeating proper nouns, would eventually deprive myths of all possible understanding. And the ideal of total translation, which would use all conceivable strategies to convey numerous versions corresponding to the diverse meanings a myth may have had in the society of origin, is destined to lose the essential mythological quality of significant absence.

Relatively little can be done about the noble or ignoble reasons why industrialised societies seek to accumulate the cultural products of non-industrialised societies. The training and dispatching of ethnographers and translators is perhaps necessarily a very one-sided affair. However, the strategic relations between source and target texts do not exhaust the mediational possibilities of translation. There are also strategies that concern the relation between the target text and the intended target-culture receiver.

Close inspection of Lévi-Strauss's maxim of translatability reveals that it is in fact a theory of reading: "a myth is perceived as myth by all readers" (1958: 232).

When compiling a small selection of Australian aboriginal myths for translation into Catalan (1989), I found myself unable to apply any coherent theoretical principle that could justify the ethics of the undertaking. All the texts sounded infantile, so I added a modern revolutionary myth (dating from 1963) in which Jesus becomes Jinimim and Noah's ark becomes the survival of aboriginal society after the drowning of the whites. But the selection remained completely prosaic, so I included a rain-prayer and an epic in verse. The genres appeared to be falsely separate, so the epic was also presented in a partly contradictory prose version. The myths seemed too distinct, so I made the same character reappear in different contexts. There was no explanation, so I threw in a Catalan version of the above essay, to explain the absence of footnotes. The result is perhaps no more than a bizarre pedagogical exercise. But the practical strategy that slowly evolved is perhaps not quite so naïve. It is enough to imagine the presuppositions with which a reader would approach such a text, and then to use the order of presentation and conflicting modes of translating to set up and frustrate an expanding horizon of expectation. Hopefully, this bastard strategy will allow something of value to be left unsaid.


 

Last update 9 February 1999