New Guinea communities without writing and views of primary orality

Lourens de Vries

published in Anthropos 98.2003:397-405.

Introduction

In his influential book Orality and Literacy, Ong (1982) summarizes in a very lucid fashion the wealth of literature on the nature of primary oral discourse, that is the discourse produced by members of societies in which writing is totally absent. Ong characterizes this discourse as additive-paratactic, formulaic, redundant and context-bound and primary oral modes of thinking are pictured as situational rather than abstract, aggregative rather than analytic, conservative or traditionalist, emphathetic and participatory rather than objectively distanced.

In this article, I first describe Ong's views as representative of still widespread ideas and assumptions about primary oral discourse and then I sketch the picture of primary orality that emerges from corpora of transcribed oral texts of New Guinea. Most of the characteristics given by Ong cannot be found in those corpora, with the important exception of the intentional and action nature of meta-language expressions of primary oral cultures. This intentional and action nature has already been noted by Malinowski (1923).Since views of primary orality are crucially informed by three other academic debates, we need to consider those first[1].

The linguistic debate: orality and literacy in English

The first debate is the debate about oral and written styles in English. For Chafe and Danielewicz (1987: 103; Chafe 1988) processing constraints are the critical factor distinguishing the styles of written and oral texts: 'In other words, there is a strong tendency for casual speakers to produce simple sequences of coordinated clauses, avoiding the more elaborate interclausal relations found in writing. Elaborate syntax evidently requires more processing effort than speakers ordinarily devote to it.' Chafe emphasizes the syntactic fragmentation of spontaneous, casual talk as opposed to the syntactic integration of written texts. In addition to the integration factor, Chafe mentions. just like Tannen (1985), the involvement factor: speech is more involved whereas writing tends to be more detached.

Whereas we still find two parameters in the work of Chafe and Tannen, in the important study of Biber (1988) we find twenty parameters, like situation versus abstract content and interactive versus edited text. Biber investigated around twenty genres of English texts but found no absolute differences between written and oral texts. Whether a given discourse, say an interview, a scientific article, or a personal letter, scored high for a certain parameter could not be predicted on the basis of whether that discourse was oral or written. In other words, Biber's work undermined the whole idea of a style continuum with 'typically oral' and typically written' extremes and that notion was replaced by genre-determined configurations of stylistic characteristics.

Besnier (1988, 1995) did a similar investigation in a non-western context, in the Polynesian society of Nukalaelae. Stylistically, personal letters group with conversations on Nukalaelae whereas written sermons group with political speeches and radio broadcasts. The genre of personal letters is determined by the cultural context of Nukalaelae in which kinship and exchange relations are crucial. In Besnier's work it becomes very clear to what extent genre, the decisive factor of Biber's work, is a culturally and historically determined notion.

The anthropological debate: oral versus literate societies

The second debate relevant to Ong's view of orality is the anthropological debate about universal cognitive dichotomies, like prelogical versus logical cultures (Lévy-Bruhl 1926) or cultures with thought processes tied to the concrete versus those that can abstract from concrete circumstances (Lévi-Strauss 1966). In the work of Goody (1977) and, more recently Olson (1994), such cognitive dichotomies are reformulated in terms of oral versus literate societies. Olson (1994) emphasizes that written language is 'decontextualized' when compared to speech. Speakers rely strongly on the context and situation of their utterances and they expect their listeners to infer information by combining utterance and context. Writers have to be much more explicit and cannot rely on implicatures to a great extent. Olson (1994) links this explicitness and autonomy of writing to cognitive changes in the direction of analytical and critical thinking of a type not possible in the contextualized, more implicit communication of primary oral cultures. The second element in Olson's view is the fact that a written text is available to the reader as a fixed object which can be compared to other texts, interpreted, summarized and so on. Such operations would stimulate the distinction between text and interpretation, and create a different meta-language consciousness, leading to grammars, lexicons and, ultimately, to the development of logic.

This anthropological debate on universal cognitive dichotomies of cultures is a thing of the past now, partly because of relativistic theoretical developments that leave no room for universal , a-historical dichotomies, and partly because of the results of empirical research on the cognitive effects of the introduction of writing. Scribner and Cole (1981) for example studied the effects of writing among the Vai of Liberia. Three writing systems are used by that community, the Arabic script for religious instruction, the English script for school, government and the workplace and the syllabic Vai script for informal use at home, such as personal letters and shopping lists. The Vai script is informally transmitted in the home. There are Vai people without formal training who write just in the syllabic Vai. Those persons, although literate, turn out to score low for tests involving abstract categorizations but high for tests in which words have to be formed from syllables. In other words, there are cognitive effects of writing but those effects are not universal, but local, determined by the local cultural and institutional contexts in which writing systems function.

The philological debate: formulaism

The third debate is the philological debate on the oral nature and origin of texts from the Antiquity as we find in the work of Parry (1928) on Homer and Kelber (1983) on the Gospel of Mark. Parry emphasized the high density of epithets, formulas, standard themes and other formulaic elements in the metrical epics of Homer. That formulaic nature of the Homeric texts was attributed to the oral performance of such texts in which performing artists produced their text by drawing from a rich store of fixed formulas, standard themes and epithets, stitching such elements together in various ways depending on the occasion and circumstances of the performance. Kelber (1983) also observes formulaic elements and other oral narrative style features in Mark such as triads (three times Jesus wakes up the sleeping disciples, three times the prediction of Jesus's death), the use of the narrative present tense and the use of the third person plural instead of the passive. Kelber (1983) sees those oral features as concessions of a writer to an orally oriented audience.

Link between the three debates and views of primary orality

Many elements from the three debates mentioned play a role in the book Orality and Literacy by Ong (1982). Ong (1982: 49) describes the oral mode of thinking as situational rather than abstract, aggregative rather than analytic, conservative or traditionalist, emphathetic and participatory rather than objectively distanced. These cognitive characteristics stem from the anthropological debate on universal dichotomies as represented in the writings of Lévy-Brühl (1926), Lévy-Strauss (1966) en Goody (1977).

The textual and stylistic features mentioned by Ong (1982) include additive rather than subordinative, redundant, formulaic and absence of sequential parallellism. Ong's explication of the feature additive style makes clear that the debate on speech and writing in English is crucially linked to Ong's portrayal of primary oral discourse. Referring to Chafe (1982) Ong writes that he misses in primary oral discourse 'the analytic, reasoned subordination that characterizes writing' and he continues with the observation that 'written discourse develops more elaborate and fixed grammar than oral discourse does because to provide meaning it is more dependent simply upon linguistic structure, since it lacks the normal full existential contexts which surround oral discourse and help determine meaning in oral discourse somewhat independently of grammar.'

The characteristics formulaic and no sequential parallellism in Ong's portrayal of primary orality have their origin in the philological debate on the so-called Homeric question as Ong (1982) himself points out. That debate, in which Milman Parry was the central figure, is regarded by Ong as fundamental to the modern discovery of primary orality. Below I will return to the presumed formulaicity and absence of sequential parallellism in primary oral discourse.

Summarizing we can say that in Ong's picture of primary orality the additive parataxis from the linguistic debate, the context-bound concreteness from the anthropological debate and the formulaicity from the philological debate come together.

Primary orality in transcribed texts from New Guinea

Let us see to what extent we can find back the characteristics ascribed by Ong (1982) to primary orality in transcribed oral texts from primary oral communities in New Guinea such as the collection of Korowai texts published in Van Enk & de Vries 1997.

As far as the cognitive characteristics from the anthropological debate go, like 'situational rather than abstract', there is little left to verify since cognitive anthropologists themselves have rejected universal dichotomies in which literate societies were contrasted with illiterate societies in oppositions like rational versus traditional and abstract versus concrete thinking. Such univeral hypotheses have been rejected on both theoretical and empirical grounds and replaced by more limited hypotheses about local, culture-specific cognitive effects of writing systems like Scriber and Cole's research on the cognitive effects of Vai writing systems referred to above.

Concerning the characteristics from the linguistic debate, the central feature here is the additive parataxis, a characteristic from the area of clause combining, the way speakers and writers combine clauses into larger syntactic and textual units. Ong (1982) uses the term additive parataxis to refer to simple juxtapositions of clauses, with coordinating conjunctions providing the links. As an example of additive parataxis, Ong (1982) quotes Genesis 1 verses 1-5 in the Douay translation of 1610. That translation reflects the additive parataxis of the Hebrew original according to Ong:

'In the beginning God created heaven and earth. And the earth was void and empty, and darkness was upon the face of the deep; and the spirit of God moved over the waters...'

According to Ong, the meaning relations between the clauses in such additive paratactic style are not coded in the utterances themselves but have to be inferred by the receptor on the basis of the context, a typical oral style which contrasts with the complex, interclausal relations found in written style, explicitly coded in the utterances by forms of subordination. Notice that Ong (1982) takes a written text, from a literary tradition in ancient Israel, to illustrate primary orality. Ong (1982) sees a massive oral residue in ancient Israel and that enables him to take such texts as examples.

However, to study primary orality there is a much directer way: we can study the discursive practices and the languages of really primary oral cultures such as still to be found in some more isolated parts of New Guinea. There are well-documented oral languages that have subordination as the most frequent, unmarked way of clause combining, for example Yimas of Papua New Guinea. Foley (1991: 497) writes that Yimas 'contrasts with many other Papuan languages in making less extensive use of clause chaining.. The most common type of clause linkage in Yimas involves nominalization, both finite and non-finite'.

It is indeed true that clause chaining is by far the most widespread type of clause combining in the oral languages of New Guinea. In clause chaining languages, single verb clauses of usually simple internal structure are chained together to form long chains of clauses. The last clause of the chain has independent or final verb forms whereas the preceding clauses have medial or dependent verb forms. The medial verb forms express interclausal relations of switch-reference and temporality. Switch-reference distinctions pertain to the question whether the next clause in the chain has a different or the same agent as the preceding clause whereas temporality relations pertain to relations of sequence and simultaneity of the events denoted by the verbs in the chain. These medial morphological oppositions often play a major role in indicating textual coherence relations.

The functions of syntactic processes of subordination and coordination in Indo-European languages, to express interclausal and textual relations, is carried out to a significant extent in language groups such as Semitic and Papuan languages by morphological processes. For Hebrew I refer to recent work by Baayen (1997) about the textual role of morphological oppositions in Biblical Hebrew verbs. If one looks at Semitic or Papuan languages from the perspective of Indo-European languages, one easily overlooks the morphological expression of interclausal coherence, since the morphological categories involved (like switch-reference in Papuan languages and the opposition between prefix- and suffix verb conjugations in Semitic languages) are unknown in Indo-European grammars. 'Literal' translations of Papuan or Semitic clause chains often give an impression of a loose, additive 'and..and..and' style, falsely suggesting a low level of interclausal integration in the Semitic or Papuan source texts because the translations do not capture the morphological processes of interclausal integration. The obligatory indication of switch-reference of Papuan utterances for example is hard to translate explicitly and naturally in Indo-European languages.

The philological debate on Homer is the source for the following three characteristics ascribed by Ong to oral cultures: a formulaic style, a highly redundant style and the absence of sequential parallellism. With the last feature Ong (1982: 147) means that oral narrators would have the tendency to start a story in medias res: 'Oral narrative is not greatly concerned with exact sequential parallelism between the sequence in the narrative and the sequence in extra-narrative referents. Such a parallelism becomes a major objective only when the mind interiorizes literacy.' The Homeric bards may have had the tendency to use a in medias res narrative style, there is overwhelming evidence for many oral languages for the opposite tendency, namely that discourses adhere to a very strict sequential parallellism. Practically all translators in New Guinea and also in oral Amerindian languages (Callow 1974: 43) struggle with the opposite problem, namely that the target cultures tell stories in strict chronological order and that translations retaining in medias res forms of their sources are difficult to understand.

With a formulaic style Ong means the frequent use of epithets, proverbs, fixed sayings, standard themes and so on. Indeed, anyone reading Homer will be struck by the formulaic style. This style has been linked to the demands of the meter: to get metrically smooth lines the oral performers of texts from the Homeric tradition could choose from a large stock of epithets so that they could pick the epithet which best suited the metric scheme of the line. But Ong goes further than this metric motivation, for him there is a mnemonic motivation: oral cultures cannot fix their insights and knowlegde in writing and in order to be able to fix knowlegde in the memory, knowlegde is stored in easily storable rhythmic, formulaic form. Oral cultures think in rhythmic, formulaic fashion, according to Ong (1982).

However, in the primary oral cultures formulaic elements do not appear to be significantly more prominent than in societies with writing. Rather, at least in some oral societies, like the Awyu societies in south-east New Guinea, one has to look hard to find evidence for formulaic elements. It is quite possible that the idea that primary oral societies think in formulaic fashion has its roots in the fact that literate observers cannot imagine other ways of preserving and transmitting complex knowlegde than through language, in the form of easy to memorize verbal forms. But it is easy to find other forms of preserving and transmitting knowlegde in most oral societies. The complex cosmology of the Marind and Asmat of Papua is represented in very elaborate ritual actions, masks and woodcarvings. The totemic ancestors of Korowai clans are remembered by foodtaboos concerning the totem animal. Of course, there are myths which relate the same notions in verbal form but these myths have no fixed linguistic form: it is the content, the intention of the mythology which counts. The symbolic and dramatic representation of those contents in rituals and carvings, together with the telling of the myths, ensures the transmission and continuation of knowlegde. It is not that the myths form the basis for the rituals, rather rituals and myths express the same basic cosmological themes, in a mutually reinforcing way. In the Awyu cultures which I studied even the magic 'formula's' had no fixed linguistic form. In such magic expressions, the only fixed elements are proper names of ancestors and spirits. These names are the fixed heart of the magic expressions since the names evoke the powerful entities they are associated with.

Apart from additive parataxis, formulaism and the absence of sequential parallellism, redundancy is stressed by Ong (1982) as a characteristic of primary oral discourse: recapitulation, repetition and a low information rate. When reading the reader can go back to what he has just read, the listener gets only one chance. When writing, the writer can endlessly rephrase what he wants to write, the speaker has to process his new utterance while performing the present utterance. Repeating information and other strategies of reducing the amount of information per utterance all help the speaker to win time to process his next utterance and the listener to process the present utterance.

The verification of the claim that primary oral discourse has a high degree of redundancy, by studying corpora of transcribed texts of oral societies, is not so easy since the notion of redundancy is hard to operationalize. What is redundant at the propositional level, may be informative at the interactional or rhetorical levels. Ong (1982) seems to refer to redundancy at the propositional level. If we study textcorpora from New Guinea like the Korowai texts edited by Van Enk and De Vries (1997) from the perspective of propositional redundancy, two, at first sight conflicting, things strike the observer. In the first place, tail-head linkage and generic verb linkage of clause chains, two conventional recapitulative devices occurring very widespread in New Guinea. Especially in narrative texts these repetitive linking devices occur very frequently. The other striking thing is the complaint by almost all scientific editors of such oral texts that so much information is left implicit and is supposed to be known to the addressees that it is often impossible to know who is doing what and where the events of the story are taking place.

These two at first sight contradictory aspects of New Guinea oral narratives, the redundance of the frequent recapitulative linkage devices and on the other hand the avoidance of redundance by leaving out known of inferrable information both have to do with the distribution of information in clause chains. The medial clauses in the chains have a high information load: whereever possible, it is avoided to identify participants with nominal phrases or pronouns. The addressees have to combine switch-reference and agent cross-referencing morphology on the verb with context and situation to identify participants of medial clauses. Not only participants but also the time and place framework is left implicit whereever possible in medial clauses. The clause chains are connected by tail-head linkage and generic verb linkage (De Vries 1989; Dik 1997: 438-440). In tail-head linkage the last clause of the chain (the tail clause) is repeated as the first clause (head) of the next chain. During tail-head linkage the speaker talks much slower and repeats information. Thus the tail-head linkages provide conventional resting points between two chains for the speaker to process his new chain and for the addressee to process the information from the previous one. The highly redundant recapitulative linkages are badly needed to compensate for the high rate of information in the medial parts of the chain.

The next fragment (1)-(3) from an oral Kombai text on the journey of the dead to Romalü (De Vries 1993a: 118-119) illustrates some of the points made above:

(1) Gu o mena mi-n-e-neno

you water this drink-TR-IMP.SG-QUOTE.PL

luwa-n-o-n-a khana

say-TR-3PL.NF-TR-CONN(DS) mouth

rakheba-n-a b-adiya-n-o-n-a

open.3SG.NF-TR-CONN(DS) DUR-give-TR-3PL.NF-TR-CONN(DS)

ya khokhufa mofene-n-e khali buma-rabo

yes just that-TR-CONN parcel big-very

ne-y-a ba-nege

eat.3SG.NF-TR-CONN(DS) stay.3SG.NF-until

fa-rofode mena

put.aside.3SG.NF-SUB.reason TOP

ami-f-e-ne-n-a

drink-1SG.INT-CONN-QUOTE.SG-TR-CONN(DS)

marulara akhukhofa

next vomit.3SG.NF

'"You must drink this water" they say and he opens his mouth and they give him the

water and, yes, because he just ate the very big parcels until he put it aside, he wants

to drink but he vomits.'

(2) Akhukhofa-n-a me makho bürü khowaru

vomit.3SG.NF-TR-CONN(DS) but there place corpse

rofoba-n-o mogo marora luwo

watch.3SG.NF-TR-CONN people next corpsefluid

khwoima-ma-n-a makho

come.out-SUPP.3SG.NF-TR-CONN(DS) there

Romalü-bürü kha-rof-ode dunorof-a

Romalü-place go.3SG.NF-SUB-reason food-and

okh-a b-adiya-n-o-n-a

water-and DUR-give-TR-3PL.NF-TR-CONN(DS)

kha-nege fa-n-a

go.3SG.NF-until put.aside.3SG.NF-TR-CONN(DS)

gone-n-a

overfull.3SG.NF-TR-CONN(DS)

bo-khugi-nene

DUR-become.black.3SG.NF-QUOTE.PL

akhukhofa-n-a khura luwo

vomit.3SG.NF-TR-CONN(DS) therefore(?) corpsefluid

bo-khwoi-ma-ma-n-e-nera

DUR-come.out-SUPP-SUPP.3SG.NF-TR-CONN-QUOTE

mene umo-ma-n-o-n-e

this tell-SUPP-TR-3PL.NF-TR-CONN(DS)

'He vomits and the people who watch the corpse there, say when the corpsefluids come

out repeatedly: "Since he went to Romalü's place, they gave him food and water until

he put it aside because he was overfull and (his skin) turns black and his vomits and

that is why the corpsefluids come out repeatedly", this they usually say and...'

(3) Mena-mo-ra jadi mena-khu Romalü mena

this-do.SS-and therefore this-also Romalü TOP

dunoro adiya-ra ba-nege

food give.SS-and stay.3SG.NF-until

ba-nege ba-nege marulara

stay.3SG.NF-until stay.3SG.NF-until next

luwo bo-khwoima-ma-nene

corpsefluid DUR-come.out-SUPP.3SG.NF-QUOTE.PL

mene umo-ma-n-o-n-e

this say-SUPP-TR-3PL.NF-TR-CONN(DS)

'"Thus this constant giving of food by Romalü makes the corpsefluids come out", they

usually tell this and....'

The examples (4)-(5) from a Korowai totem text (Van Enk & de Vries (1997)) illustrates another aspect of many New Guinea oral narratives, the frequent use of topic markers (TOP):

(4) nokhuf‑efè mül‑khuf‑efè buom

we‑TOP former‑time‑TOP kind.of.lizard

buom buom‑da‑é apa

kind.of.lizard kind.of.lizard‑NEG‑EXCLM only

ye‑nggé bembüo‑lo lai‑dakhu‑fekho

his‑friend lizard‑FOC break.3SG.REAL‑SS‑and

'As for us, in former times, the white house‑lizard, the white house‑lizard..., not the house‑lizard, only its friend, it was the little brown wood‑lizard('s eggs) which hatched out and..'

(5) lai‑bando‑khe‑nè lefu‑fefè wola ‑khip

come‑bring‑go‑SS some‑TOP earth‑over

gil‑bolüp fu‑dakhu

ceremonial.festival.place‑place put.3SG.REAL‑SS

lefu‑fefè bando‑khe‑nè nokhu dif‑e‑kha‑la

some‑TOP bring‑go‑SS our this‑TR‑CONN‑LOC

fu‑dakhu wa‑mo‑bo‑kha‑fefè

put.3SG.REAL‑SS thus‑SUPP‑3SG.PERF.REAL‑CONN‑TOP

wof‑è nokhup laibo‑kha‑fefè

that‑CONN us be.born.3SG.REAL‑CONN‑TOP

'..some he brought away and put them on a festival place, others he transported and placed them in our area, and given that he has done so..., that is (the story of) our origin...'

An important characteristic of primary orality mentioned by Ong (1982) which does show up also in New Guinea is the intention and action nature of primary oral metalanguage, of expressions in primary oral languages about language. This specific nature of primary oral metalanguage has already been observed by Malinowski (1923) who remarks that language for oral peoples is a mode of action.

When studying the metalanguage of primary oral peoples of New Guinea like the Kombai and Korowai, the first thing striking the observer is the absence of metalanguage expressions denoting segmental notions like speech sound or word, word in the sense of a semantic and syntactic unit which can be isolated, be pointed at, as in the English metalanguage expressions 'you chose the wrong word' or 'repeat that last word, please' or 'I could not that word in the dictionary'. That segmental notion of 'word' has not been lexicalized in Kombai or Korowai. Ong (1982) was justified in pointing out the tremendous impact of writing on metalanguage and on thinking about language. Segmental and reified notions of letter, word and sentence obviously reflect the metalanguage distinctions of people that write.

Positively, primary oral metalanguage has two characteristics, intentionality and an action nature. The action nature of oral metalanguage can be illustrated by the Kombai noun lu. Lu is used to refer to messages, quarrels, promises, talk, languages, calls of birds. When used to refer to elements of language, it is always to actions that are performed through language, never segmental units of language. The intentionality is connected to the action nature: speech acts have intentions. Since the segmental notion of word is absent, there is no distinction between the wording of an intention and the intention itself: there is just intention, a promise or a threat or a statement or a wish.

One of the consequences of the absence of the distinction between the words produced and the intention behind those words, is that the distinction between direct speech, as a representation of someone's 'own words' and indirect speech, as the representation of someone's intention without the pretention that someone is quoted verbatim, is absent in primary oral communities. Inasfar the distinction direct versus indirect speech is made in the oral languages of New Guinea, it is made in terms of deictic orientation.

An interesting consequence of not distinguishing between wording and intention is that in representing the speech of other people, primary oral speakers ascribe words to people which they have may never have said and which are not pretended to have been spoken by those people. An example would be

(6) Luwano kho mofena yafe.rambo-neno

they.said. man that good.very-QUOTE.PL

'They praised that man.'

The reporting Kombai speaker of example (6) does not claim that the words kho mofena yaferambo 'that man is very good' have been uttered by the reported speakers. He just claims that they have praised that man. Reported speech in Kombai always has the form of a direct quote but the quoted words always function as the portrayal by the reporting speaker of the intention of the reported speakers. Non-verbalised intentions are represented in exact the same way as verbalised intentions. This means that an expression as (7) has two readings out of context:

(7) ami-f-e-ne-n-a

drink-1SG.INT-CONN-QUOTE.SG-TR-CONN(DS)

'I want to drink' -quote of one speaker

= 'he wants to drink' or:

'he says/said that he wants/wanted to drink'

To disambiguate and to indicate that the intention was verbalised, a anaphoric or cataphoric demonstrative may be inserted ('luwa mene, ami-f-e-ne-n-a = he said this, I want to drink'). Forms like (6) and (7) are called quotative forms.

Another consequence of the intentionality character is that in the transmission of oral stories, magic formulas and so on, the intention is transmitted and not the precise wording. Instead of fixed, formulaic wording we find highly variable wording: intentions are transmitted and not the wording of those intentions. In magic 'formulas' it is the names of ancestors and spirits that count, secret names which are very much taboo; these names form the invariable core of otherwise variable magic spells and chants.

Foley (1997:185-186) following Rumsey (1990) suggest that the distinction between wording and intention, between saying and meaning and between direct speech (reporting what was said) and indirect speech (reporting what was meant, intented) is the basis for the CONDUIT metaphor, a folk theory of language that has these elements (Johnson 1987):

1.ideas or thoughts are objects

2.words and sentences are containers for these objects

3.communication consists in finding the right word-container for your idea-object, sending this filled container along a conduit (e.g. writing) or through space (e.g. speech) to the hearer, who must then take the idea-object out of the word-container.

The absence of the distinctions between wording/intention and saying/meaning consequently means that the CONDUIT metaphor of language is absent in the metalanguage of oral societies. In the Australian aboriginal groups studied by Rumsey (1990) the meaning of a word is not situated in its reference to something in a pre-given world but includes the world of socio-cultural understandings of its contexts of usage. ‘Speaking a word brings forth a world in which meaning is enacted and validated.’(Foley 1997:186)

Summarizing this section, we can say that very few universal claims about primary orality can stand the New Guinea test. The characteristics of primary orality as additive-paratactic, redundant, formulaic, bound to concrete contexts and so on did not emerge from empirical study of specific primary oral societies in their historical unicity but seem to result from universalistic projections on these societies of pictures from various academic debates such as the Homeric debate, the debate of written versus oral style in English and the anthropological debate on cognitive dichotomies in terms of literate versus illiterate societies. The only feature laying some claim to universality is a specific type of metalanguage in primary oral contexts in which the distinction between wording and intention is not made. This has important consequences for the way language is conceptualized. Words nor meanings are things that can be looked up, segmented and related as word-containers containing idea-objects. Rather, utterances are inseparable from the intentional actions, like questions, commands, promises and praises that they perform.

Literature

Baayen, Harald R.

1997 The pragmatics of the 'Tenses' in Biblical Hebrew. Studies in Language 21:2.245-285.

Besnier, N.

1988 The linguistic relationships of spoken and written Nukulaelae registers. Language 64: 707-36.

Besnier, N.

1995 Literacy, Emotion and Authority. Cambridge: Cambridge University Press.

Biber, D.

1988 Variation Across Speech and Writing. Cambridge: Cambridge: University Press.

Callow, K.

1974 Discourse Considerations in Translating the Word of God. Grand Rapids: Zondervan.

Chafe, Wallace L.

1982 Integration and Involvement in Speaking, Writing and Oral Literature'. In: D. Tannen (ed.), Spoken and written language: Exploring orality and literacy (Advances in Discourse Processes, Vol.IX, 35-54). Norwood, NJ: Ablex.

Chafe, Wallace L.

1988 Linking intonation units in spoken English. In: J. Haiman and S. Thompson (eds), Clause Combining in Grammar and Discourse, 1-27, Amsterdam: Benjamins.

Chafe, Wallace L. and J. Danielewicz

1987 Properties of spoken and written language. In: R. Horowitz and S.J. Samuels (eds), _Comprehending Oral and Written Language,_83-113, New York: Academic Press.

Dik, Simon C.

1997 The Theory of Functional Grammar. Part 2. Complex and derived constructions. Edited by Kees Hengeveld. Berlin: Mouton de Gruyter.

Feldman, Carol Fleisher

1991 Oral metalanguage. In: Olson and Torrance (eds), 47- 65.

Foley, William A.

1991 The Yimas language of New Guinea. Stanford, CA: University Press.

Foley, William A.

1997 Anthropological Linguistics. An Introduction. Oxford: Blackwell.

Goldman, Laurence

1993 The culture of coincidence: accident and absolute liability in Huli. Oxford/New York: Oxford University Press.

Goody, J.

1977 The domestication of the savage mind. Cambridge: Cambridge University Press.

Johnson, Marc

1987 The Body in the Mind: The Bodily Basis of Meaning, Imagination and

Reason. Chicago: University of Chicago Press.

Kelber, Werner

1983. The Oral and the Written Gospel: The Hermeneutics of Speaking and Writing in the Synoptic Tradition, Mark, Paul and Q. Philadelphia: Fortress Press.

Lévi-Strauss, Claude

1966 The Savage Mind. Chicago: University of Chicago Press.

Lévy-Brühl, Lucien

1926 How Natives Think. New York: Knopf.

Malinowski, Bronislaw

1923 The problem of meaning in primitive languages. In:C.K. Ogden and I.A. Richards (eds), The Meaning of Meaning: A Study of the Influence of Language upon Thought and of the Science of Symbolism, 451-461. New York: Harcourt, Brace.

Olson, D., N. Torrance & A. Hildyard (eds).

1985 Literacy, language and learning: The nature and consequences of reading and writing. Cambridge: Cambridge University Press.

Olson, D. & N. Torrance (eds)

1991 Literacy and orality. Cambridge: Cambridge University Press.

Olson, D.

1994 The world on paper: the conceptual and cognitive implications of writing and reading. Cambridge: Cambridge University Press.

Ong, Walter

1982 Orality and literacy: the technologizing of the Word. London: Methuen.

Parry, Milman

1928 L'Epithète traditionelle dans Homère. Paris: Société Éditrice Les Belles Lettres.

Reesink, Ger P.

1987 Structures and their Functions in Usan, a Papuan Language of Papua New Guinea. Amsterdam: John Benjamins.

Rumsey, A.

1990 Wording, meaning and linguistic ideology. American Anthropologist

92:346-361

Scribner, S. and M. Cole

1981 The Psychology of Literacy. Cambridge, MA: Harvard University Press.

Tannen, D.

1985 Relative focus on involvement in oral and written discourse In: Olson, D., Torrance, N. & Hildyard, A. (eds), 124-47.

Vries, Lourens de

1989 Studies in Wambon and Kombai. Ph.D. dissertation University of Amsterdam.

Vries, Lourens de

1990 Some remarks on direct quotation in Kombai. In: Harm Pink­ster & Inge

Genee (eds), Unity in diversity. Dordrecht: Foris.

Vries, Lourens de

1993a. Forms and Functions in Kombai, an Awyu language of Irian Jaya. Pacific Linguistics Series B, no. 108. Canberra: Australian National University Press.

Vries, Lourens de

1993b Clause combining in oral Trans-New Guinea languages. Text 13:4.481-502.

Vries, Lourens de & Gerrit J. van Enk

1997 The Korowai of Irian Jaya. Their language in its

cultural context. Oxford/New York: Oxford University Press.

Abbreviations

CONN connective

DL dual

DS different subject (switch-reference)

DUR durative

EXCLM exclamative

FOC focus

IMP imperative

INT intentional

LOC locative

NEG negative

NF non-future

PL plural

REAL realis

SG singular

SS same subject (switch-reference)

SUB subordinator

SUPP support-verb

TOP topic

TR transitional sound

[1]In representing those debates I follow Foley (1997: 417-434).