30. What we say, and how we feel about it: Focus on Catalan in France

James Hawkey
University of Bristol


All students of sociolinguistics quickly learn that language is highly variable, and that such variability can convey a great deal of social information. Let’s take the United Kingdom as an example. As soon as a native speaker of British English utters the word bath, they are opening themselves up to an onslaught of social judgement… Where is this person from? What kind of school did they go to? What kind of job do they have? What political party do they vote for? The answers to all of these questions are potentially coded within the pronunciation of that one little word: does the speaker use a short vowel, like in hat, or a long vowel, like in car? Is that last consonant pronounced with the tongue between the teeth, or in the same way as the last sound in staff? Just this one tiny word, with its three different sounds, reveals a number of possible pronunciations, each conveying its own social information. So, how do we go about patterning this complex variation? What sort of social categories can predict or govern a speaker’s preference for using one sound over another? And, of particular interest to us in this discussion, do our own opinions about the languages we speak affect the way we speak them?

In order to answer these questions, we need to draw on the most recent scholarly advances from two related sub-disciplines of sociolinguistics: variationism and language attitudes. Variationism seeks to establish correlations between the preferred use of a given variant (so, for example, whether you use a long or a short vowel in bath) and some broader social category (such as your age, social class, etc.) Language attitudes examine the views and belief systems we hold about the language(s) we speak. Social psychologists have long maintained that attitudes are made up of different components, including our emotions about a given object, our beliefs and thoughts about said object, and our enacted behaviours regarding the object. Attitudes and behaviours are therefore closely linked, and much social psychological scholarship seeks to understand how our feelings about something tie into our actions. Bringing all of this together, my work looks at the question of whether our opinions about a language reflect patterns of variation, with a particular focus on situations of bilingualism and language endangerment. My latest book, Language Attitudes and Minority Rights (Palgrave Macmillan 2018), constitutes an in-depth study of the Catalan language in Southern France, and asks whether language variation in Catalan can be correlated with speaker attitudes.

Firstly, I used anonymous questionnaires to paint a clear picture of the language attitudinal landscape in Northern Catalonia (the area of Southern France where Catalan is an autochthonous variety, based around the city of Perpignan). Language attitudes are typically quantitatively measured along two dimensions: status (‘the desire to get ahead in some way’) and solidarity (‘the desire to be accepted by another group’) (Woolard 1989: 90). Of course, it should come as no surprise that, in France, French was seen by almost everyone as the main language of status in the region. For centuries, regional and minority languages in France (like Catalan) have been subjected to processes of marginalisation and its speakers forced or encouraged to move towards French monolingualism. However, one thing that is particularly interesting in Northern Catalonia is that, even for speakers of Catalan (who, according to official statistics, number around 35% of the population, though the real figure is likely to be lower), French is a language of solidarity. That is to say, French is a language that all speakers hold dear as a vehicle of in-group communication and identity, even if these people are also speakers of Catalan. The centuries of French dominance have led to this language leading the way not only as a means to get ahead socially, but also as a strong identity marker. This being said, while French is viewed as the preferred vehicle of in-group solidarity and social advancement across the population of Northern Catalonia, people are not necessarily negative about Catalan: some speakers view Catalan as a language of potential status, social mobility and international advancement. So, what are the consequences for language variation? Do Northern Catalans who view Catalan positively as a language of status speak in a certain way? What about those people who do not view Catalan as a language of status, but instead as a vehicle of in-group solidarity? Do they sound different?

Turning to the study of language variation in the region, I focused on a number of different variables, only one of which we’ll briefly look at here. One of the most salient differences between varieties of Catalan spoken in France and those spoken elsewhere is the ‘r’ sound. In Northern Catalonia, many speakers use the uvular fricative rhotic, more familiar to most people as the guttural ‘French r’. This is basically limited to Catalan as spoken in France, with speakers from other areas using the alveolar tap or trill, recognisable as the rolled ‘r’ sound used in languages like Spanish or Italian. Within Northern Catalonia, this rhotic variable proved highly unusual, since there was very little intra-speaker variation, but large amounts of inter-speaker variation with little apparent accommodation. In short, this means that people were very consistent in their own use of either the ‘French r’ or the ‘Spanish r’ when speaking in Catalan, and that this didn’t depend on their conversation partner. So if people were categorically using one variant or another, and if they weren’t influenced by the person to whom they were speaking, how could this variation be predicted? What was the social meaning carried by the use of the ‘French r’ or ‘Spanish r’? The answer turned out to be language attitudes…

If a speaker rated Catalan highly as a language of in-group solidarity, then this person was more likely to use the uvular (‘French’) rhotic. A sense that the local language is important to interpersonal relations is clearly linked to use of local variants: remember that the uvular rhotic is not found outside this region. Local orientation on an attitudinal level is aligned with usage of local language patterns. In very broad terms, people who feel that their identity is bound up with use of the local language, sound more local. People overtly comment on this kind of variation too, as I discovered during the interviews I conducted in Northern Catalonia:

El meu accent és pas un accent totalment tradicional per la gent del sud. Per exemple, potser és anecdòtic, però les bromes pel fet de fer les erres guturals i no de les fer rulades.

My accent isn’t exactly traditional for somebody from the south [i.e. the Autonomous Community of Catalonia]. For example, it might be anecdotal, but the jokes (I get) because I have my guttural r instead of a rolled one (Hawkey 2018: 119).

This variable is merely one example of the complex variation that characterises this (and indeed, every) language variety. The links between variation and attitudes have been, until now, largely overlooked. Variationism has become increasingly focused on locally constructed categories as correlates for language variation: instead of simply saying ‘person X is more likely to use a given variant if they are male/working class/White, etc.’, scholars are now examining the different ideological stances that language variation might imply. My work is taking this a stage further, by using large-scale questionnaires to look at societal-level attitudes and belief systems, and then relating these to language variation. By marrying the different methodological approaches from variationism and language attitudes, it is possible to shed light on how our belief systems influence what we do. In short, perhaps the way we speak is indeed affected by how we feel about language itself, and this conclusion represents an interesting move forward for sociolinguistic scholarship. Moreover, this kind of holistic approach to linguistic communities (particularly those that find themselves in situations of endangerment) allows us to better understand the rich linguistic and cultural diversity that surrounds us, and ultimately, to better connect with other people.



Hawkey, James. 2018. Language Attitudes and Minority Rights: The Case of Catalan in France. Cham: Palgrave Macmillan.

Woolard, Kathryn. 1989. Double Talk: Bilingualism and the Politics of Ethnicity in Catalonia. Stanford, CA: Stanford University Press.


29. Déclin et résilience : les dilemmes de l’arménien occidental cent ans après

Anaïd Donabédian
Inalco, Paris


1.1. Langue, identité et normes

Comme l’ont montré les travaux sur l’idée moderne de nation, la conception de la langue comme support de l’identité s’appuie dès le dix-neuvième siècle sur l’élaboration d’un norme dotée de prestige social et culturel. Aujourd’hui, la langue occupe toujours une place de choix dans la représentation que les Arméniens se font de leur identité. Pourtant, après cent ans de Diaspora, cette représentation de la langue comme un substitut de la patrie perdue est un piège, tant dans les pays òu l’arménien occidental est vernaculaire que là où il ne l’est plus.

Bien après l’entreprise fondatrice, la norme linguistique demeure un point de focalisation chaque fois que l’identité du groupe est menacée. Le Proche-Orient, jusque-là conservatoire de l’arménien occidental, est fragilisé par la réforme des systèmes scolaires (en Syrie depuis les années 1960) ou par les conflits (au Liban dans les années 1970, en Syrie depuis 2011). Le sentiment de déclin a conduit à une sorte de raidissement normatif, sous-tendu par l’idée que la défense de la langue passe par la maîtrise de son évolution. En diaspora, où l’identité arménienne n’est pas aisée à définir, il est tentant, plutôt que de chercher à penser cette identité, d’investir idéologiquement, aux côtés de la religion et de la revendication politique, la langue, parfois au mépris de ses règles internes.

Idéologiser la langue conduit à l’investir d’un sens qui dépasse celui d’un simple moyen de communication: l’enjeu identitaire, dramatisé par le contexte du déclin, en fait un sanctuaire dont l’accès est hautement codifié. La langue littéraire, idéalisée, est toujours plus normative, et le vernaculaire est privé de la légitimité qui devrait être celle du locuteur natif. Ce type de diglossie n’est pas rare, mais ce qui fait sa spécificité pour l’arménien occidental est sa situation de langue dominée et le bi- multilinguisme généralisés en contexte diasporique, qui explique la virulence avec laquelle la normativité se manifeste à divers niveaux de la société. Au Liban, le quotidien Aztag Daily, le plus diffusé en arménien occidental, reçoit régulièrement des courriers de semonce avec citations à l’appui, remettant ainsi en question la légitimité des auteurs des articles incriminés. Des jeunes, lassés de corrections interrompant un échange entre pairs, font sciemment le choix de parler anglais, français ou arabe avec les jeunes qu’ils pensent plus performants qu’eux en arménien littéraire. Les enseignants eux-mêmes sont soumis à des jugements remettant en cause leur légitimité.

L’insécurité linguistique est la perception négative qu’un locuteur a de sa propre pratique linguistique, du fait d’une stigmatisation explicite ou imaginée de la part d’une majorité linguistique porteuse d’une variante plus prestigieuse. D’abord introduit par Labov pour évoquer les situations de diglossie en anglais, le phénomène est décrit plus récemment dans la dynamique de la mort des langues. L’inconfort ressenti face au jugement réel ou imaginaire de l’interlocuteur affecte sa pratique. Quand le locuteur n’a pas les moyens d’échapper à cette angoisse, il peut s’agir de stratégies comme l’hypercorrection (production d’une forme erronée visant à corriger une faute imaginaire). Mais en situation plurilingue, lorsque le locuteur dispose d’un autre code pour communiquer avec son interlocuteur, la stratégie d’évitement est souvent à la fois économique au plan linguistique, et gratifiante en termes de prestige.

L’évitement de l’arménien et le passage à la langue dominante est plus souvent lié à une compétence insuffisante pour s’exprimer de manière complète. Les mécanismes à l’oeuvre sont dans ce cas le code-switching, passage d’une langue à l’autre dans la même interaction, ou parfois de code-mixing. Quant à la partie de ces communautés qui n’en a pas eu la transmission, la langue, tout en étant un « impératif moral », est hors de portée, et l’identité est investie par d’autres voies. L’attachement au folklore, observé chez les générations précédentes, est supplanté par le militantisme politique lié à la reconnaissance du Génocide, c’est-à-dire un domaine présentant un enjeu idéologique à la hauteur du défi indentitaire.


1.2. Transmettre la langue : le grand défi

En contexte éducatif, le discours sur la langue varie de pays à pays selon que l’arménien occidental y est plus ou moins vernaculaire, on relève des constantes, comme la notion de ‘hayabahbanum’ (préservation de l’arménité), rarement justifiée (que préserver, comment, pourquoi, dans quel but?). Difficile à traduire, la notion peut même être idéologiquement douteuse pour la société environnante, et les écoles sont souvent tiraillées entre cet objectif interne à la communauté, et un discours officiel qui vise à former des citoyens bilingués équilibrés.

La dévernacularisation de l’arménien occidental, dont le danger semble préoccuper jusqu’aux élites éducatives et culturelles du Liban, suppose de modifier les modèles d’enseignement : il ne s’agit plus d’enseigner la littératie à des locuteurs d’arménien (langue maternelle), il faut désormais avant tout transmettre les competénces de base en langue parlée (langue seconde). Les moyens des écoles ne permettant pas de créer des classes différenciées,  elles doivent choisir un modèle pédagogique sans tenir en compte de la diversité des profils de leurs élèves. Ainsi, au Liban, les enfants dont la compétence vernaculaire en arménien est jugée insuffisante ne sont pas pris en charge dans les écoles arméniennes, tandis qu’en France, l’évaluation du niveau des élèves s’appuie sur des compétences scolaires surestimées par rapport à leurs compétences réelles en langue.

Cette situation conduit à des impasses en termes de résultats (fétichisation de l’alphabet au détriment des compétences communicatives, compartementalisation linguistique, perte de prestige) et en termes de cohésion sociale (l’absence de réflexion collective générant frustrations et incompréhensions entre enseignants, administrateurs, familles et élèves, l’école devient un lieu de tension associée à l’arménien).

L’ambigüité entre hayabahbanum et transmission de la langue dans l’objectif fondateur des écoles se reflète aussi dans les motivations des familles choisissant l’école arménienne : les primo-arrivants pour les facilités de communication ; d’autres familles y trouvent un refuge contre la mixité sociale et ethnique qu’ils perçoivent comme une menace; d’autres enfin souhaitent une sociabilisation arménienne pour leurs enfants, reproduisant souvent leur propre expérience scolaire. La demande vis-à-vis de la langue est variable et d’autant plus floue que les familles ne sont pas arménophones (et ne peuvent évaluer les compétences acquises par l’enfant). Les familles ne choisissent généralement pas l’école arménienne sur la base de son niveau général d’enseignement, mais en contrepartie, la compatibilité avec le système local, qui rend leur choix réversible, les sécurise. De fait, aucune école ne choisit un modèle totalement innovant et sans contrat avec l’Etat (comme les écoles immersives bretonnes Diwan en France). L’école arménienne semble donc intrinsèquement traditionnelle, parents comme administratifs (et souvent financeurs) étant souvent mus avant tout par le désir de reproduire un modèle du passé, et peu soucieux de repenser les modèles pédagogiques.


2. La croisée des chemins

Alors que la transmission familiale régresse de manière critique dans la diaspora occidentale, l’école est plus que jamais essentielle pour l’avenir de la langue. Pourtant, aux prises avec des impératifs idéologiques et conservateurs, les écoles peinent à opérer un diagnostic réaliste et un plan d’action adapté. Certains incriminent l’absence de structures de gestion de la diaspora arménienne, mais sans fournir des alternatives convaincantes.

Je propose de voir là plutôt la nécessité d’un tournant que je qualifierais de post-ideologique, dans le contexte d’une postmodernité touchant déjà l’ensemble des pays occidentaux, et dans une certaine mesure le Liban, mais n’a généralement pas droit de cité dans les institutions communautaires. De ce fait, les représentations collectives liées à l’identité créent donc potentiellement chez les Arméniens de la diaspora une fracture entre deux grilles de lectures irréconciliables : un criticisme postmoderniste appliqué l’identité locale de citoyen, et une sanctuarisation de l’identité arménienne. Il en résulte une polarisation entre les porteurs de l’idéologie, souvent les plus actifs dans les institutions de la diaspora, et la masse des membres de la diaspora, les premiers oubliant souvent, surtout en occident, que les seonds ne sont pas captifs des structures communautaires et peuvent s’en détourner.

Cette situation reflète ce que je schématise ici comme le « paradoxe du ghetto ». Ce dernier, désignant de manière figurée le modèle proche-oriental dans sa dimension dirigiste, est capable d’assurer la transmission de la langue, mais pas de la penser : le raidissement et la normativité peuvent y accentuer le déclin au lieu d’y remédier. Ce sont les limites de l’idéologie même qui a été jusque-là associée au maintien « miraculeux » de la langue pendant un siècle. A l’opposé du « ghetto », le point de vue post-moderniste sait innover sur la base d’un diagnostic réaliste, mais sa masse critique des locuteurs est fragile : il peut penser la transmission mais est moins bien armé pour l’assurer. Ces deux approches sont donc  condamnées à l’échec si elles ne se nourrissent pas l’une de l’autre : la première pour mieux répondre aux défis réels et éviter les mouvements centrifuges à un moment critique de son historie, la seconde pour et éviter la rupture de transmission. La rencontre entre elles porte en elle est la possibilité de réussir le tournant post-vernaculaire sans perdre la transmission du vernaculaire.

Des initiatives récentes ont émergé dans ce sens[1] (innovation pédagogique, traduction d’ouvrages majeurs de la culture mondiale, développement de Wikipedia, camp d’été créatif). Il s’agit de renverser le « devoir de langue » pour en faire un « désir de langue », notamment à l’âge critique où les jeunes éduqués en arménien s’éloignent pour construire leur propre univers dans la société environnante. En contexte post-vernaculaire, cet objectif est en effet aussi essentiel que la norme est inutile. Ce cercle vertueux propose également une issue au paradoxe qui fait de l’enseignant d’arménien un mal aimé de l’école arménienne, alors même que sa discipline est aux fondements du projet de l’école. Reconnaître que l’arménien occidental, rarement monnayable par des examens nationaux, n’a pas de valeur marchande, permet également de repenser la mission de l’enseignant. Libéré des contingences des autres classes, il doit ancrer dans l’enfant le désir d’arménien, celui qui lui permettra de saisir les opportunités futures pour l’améliorer, et qui lui donnera le désir de transmettre la langue à la génération suivante.

Passer du paradigme du devoir au paradigme du désir : c’est sans doute là, bien au-delà des enjeux linguistiques aussi, que siège la résilience qui nous occupe ici. Une résilience qui pourrait bien expliquer pourquoi ce tournant n’était pas possible plus tôt. La résilience, celle de la langue et des communautés, serait-elle une piste post-moderniste pour repenser l’articulation entre arménien occidental et identité arménienne en diaspora, cent ans après ?

[1] Largement portées ou encouragées par le plan d’action de la Fondation Calouste Gulbenkian en faveur de l’arménien occidental.


28. Amerindian Silence

Montserrat Ventura i Oller
Universitat Autònoma de Barcelona


The silence of languages

Silence is cultural, just as words are. For many different reasons, our languages contain silences, and we are raised to know how to interpret them. Sometimes they make us unsure, often because of discord on an individual level or as a result of circumstances involving conflict. Respect, fear, ignorance: silence tells us things without words, and often it does so very eloquently.

But in some cultures, it is not so easy to classify silence as one of the functions of speech. There are societies in which silence is not just appreciated but also pervasive in communications, and even in relationships. Keith Basso1  was a pioneer in highlighting silence as a cultural form of behaviour among the Western Apache, and alongside them, other Amerindian groups. In her classic study, she tries to explain why, in certain circumstances, the Western Apache avoid talking and do so for reasons that go beyond the socially determined verbal behaviour. Basso outlines a few of these scenarios: meetings with foreigners, which involve social distance; first meetings between lovers (which sometimes may go on for months); reunions between parents and children after long absences, where silence is attributed to the need to rebuild the relationship; in the face of insults or off-key words from people who are not in control of their mental faculties (for example, due to an excess of alcohol), as if reasoning is not possible, it is better not to talk; during mourning rituals, where speaking would be inconsiderate to those who are sad; and, finally, during healing ceremonies, where a patient for whom the specialist is chanting cannot be spoken to out of respect for the powers and spirits and to avoid these having a harmful effect on the patient. Despite this disparate range of situations, Basso believes that the conditions are always the same: silence, in the culture of Western Apache, is associated with social situations in which participants perceive their relations towards another person as ambiguous and/or unpredictable. Rather than being a question of this or that contextual situation, it is one of the social relations perceived, which are not structural but dynamic. But there are other ways to conceive of such silences.

In 1991, I started to conduct fieldwork in the Cóngoma community, one of the seven Tsachila communities in the western lowlands of Ecuador. At the time, there were about 2,000 people—in 2018, there are now about 2,800—and they were the only speakers of Tsafiki, a South Barbacoan language. They are a society that used to hunt, cultivate and fish but that was transitioning to a market economy. They maintained an animist view of the world, in spite of the relatively recent efforts of Catholics and Protestants. At the time, bilingualism was still the preserve of men, and everyday life occurred in Tsafiki2. It was common to see two interlocutors sitting with their backs to one another and one suddenly rising and leaving without a goodbye. Or, at most, the person leaving would say “I’m going”(majinayoe), which would be met with a rhetorical reply of, “Are you going?” (majinayun). And that’s all they would say.

A sociality that avoids conflict

Tsachila men and women are very restrained in their linguistic displays when they meet or say goodbye, unlike other ethnic groups that highly value the rhetoric in some of their stereotyped dialogues—for example, salutations. The Tsachila salutation at a crossroads boils down to a matter of good manners that has a meagre semantic content: “Where are you going?” (nunchi?), to which one responds with the obvious: “I’m going up that way” (fechi) or “I’m going down that way” (pelechi). To be sure, spotting a person coming from afar can trigger a communicative exchange focused on general information about the visitor, the host, and their respective surroundings, but in general, the arrival of a visitor to a home happens very discreetly, almost imperceptibly. Likewise, during group outings to the village, the cemetery, a shaman’s home, parties or meetings, or to go fishing, it is not unusual for it to be the case that, while women above all drift into informal and cheerful conversations, long silences also take hold, as if it were necessary to avoid bothering the others, and these silences are never considered signs of social tension. Avoiding disturbing others could be considered rather a sign of respect for one of this society’s key values, namely the appropriate weighing up of words. This corresponds to what I said earlier about goodbyes. Restraint over words, which certainly reduces the possibility of conflicts, is also expressed in other aspects of communication, such as the avoidance of eye contact during dialogue, where clearly the words are more important than the expression that accompanies them.

Linguistic exchanges are therefore very discreet, and oratory is not valued, unlike other societies in the lowlands of South America. People who attended public events such as major ceremonies during other times were not accustomed to speaking in public; happiness was expressed through singing and music, especially when fermented drinks had begun to have their effect. Today, in collective events such as community meetings, the leaders, the only people who are believed to be proficient in spoken discourse, are also laconic: they avoid big speeches and instead seem embarrassed to have the floor or be the main figures in meetings, especially if there is a high attendance. Embarrassment about being the focus of public attention is patent during interactions that correspond to situations of inferiority. Embarrassment (lu, red; lu tenka kano “be embarrassed,” literally “red heart take”) suggests discretion and shyness to us.

Weighing up the right words

The Tsafiki language system3 shows us the importance of the authority of speech, a fundamental aspect of weighing up the right words. Tsafiki has a marker for indirect speech, ti, which may be repeated to indicate up to three sources between the speaker and the original event. The indirect style seems designed to locate the source of information beyond the speaker, without this carrying connotations linked to suspicion, as would be the case in English or other European languages. Furthermore, this linguistic system allows an extremely high level of personalization of information sources (“he/she says that he/she said that he/she said that …”), and it therefore restricts the possibilities of invention even where rumours and gossip continue to exist, fostered by the great distances that separate homes from one another. Nevertheless, talking too much (tsanke epele palakiman4) is met with condemnation, and a large number of individual actions are intended to avoid revealing private acts to others, precisely so as not to give rise to rumours. These acts include preferring short cuts to main paths, walking late at night or very early in the morning in order to maintain discretion over movements, or even erasing tracks left when walking. Mythology itself punishes excessive use of words, and its repertoire offers constant demonstrations of this precept. The tale of Biali (“otter”) is a good example of this:

Biali helped a Tsachi with fishing, and he agreed to never tell anybody about it. But one day, while inebriated, he broke the agreement. After that, he lost his good relationship with Biali who had helped him to fish and, consequently, the ability to catch fish in abundance that he had been given 5.

But on the other hand, the ability to speak is not anathema. One of the terms that means stupidity, ignorance or irresponsibility is, paradoxically, epela, which in a strict sense is equivalent to “dumb, one who does not speak.” It is worth noting that this correspondence is found in other parts of the Amerindian world, particularly where people have experienced colonial domination most intensely, such as the Andes. According to Howard-Malverde (1990), a lack of mastery of Spanish was to the conquistadores equivalent to ignorance; but the insult of “dumb,” as well as the refusal to speak before the Spaniards, would have been accepted by the victims of this contempt themselves. It does not seem that this was the case regarding the Tsachila, for whom colonial pressure was less far reaching. For them, talking is a sign of wisdom, and although it is not possible to make a direct correlation, it seems pertinent to note that the body organ that allows words to be produced, the tongue, is called mikaka (“The eye or the fruit of knowledge”).

Moreover, silence as an expression of a communication barrier is also a response of rejection in situations of disagreement. Tsachila behaviour that guides most cultural practices leads, as mentioned above, to the resolution of conflicts by means of withdrawal, rather than by open confrontation. Keeping quiet is also a coded response for the Tsachila. They may decline to attend an appointment without apologizing if they consider that they have not received fair treatment. Silence, which can sometimes be difficult to maintain, can be replaced by the concealment of the truth, which can mean the same thing: denying the interlocutor good communication. However, escape and avoidance of conflict cannot be satisfactorily explained only by colonial rule or inferiorization. Everything seems to indicate that these behaviours are part of a cultural ethos taken to extremes during times of crisis. The acceptance that the Tsachila externalize in their everyday relations does not prevent prudence or protection of their privacy.

And in the field of ritual there is also a value associated with silence: secrecy. In any matter involving spirits, it is necessary to avoid causing them to lurk around or bothering them. In any remedy or treatment from shamans, people have to maintain secrecy and discretion. Secrecy and an absence of words are conditions of the ritual, unlike in other Amazonian societies, where remarkable communication takes place, especially in the aural form of recitations, songs, music and different sounds. It is necessary to avoid calling certain spirits by name, and the relationship with them must remain within individual privacy or be shared only with a small number of people. The shaman does recite litanies, but sometimes in a cryptic language, or directly in Spanish, or even Quechua, languages removed from everyday speech. Shamans learn these litanies elsewhere, and their literal sense may not be understood by the shaman: their therapeutic efficacy is not in the meaning of the words but in their aural effect, and this is explained by the power that this society attributes to otherness. Gestures and sound dominate over words in the shamanic ritual act.

The refusal to share information, confinement of language, silence and secrecy result from the same caution. It might even be suggested that forgetting, which manifests itself in extreme form in passivity and a refusal to transmit myths or certain traditions, may be an expression of this attitude. If this set of practices can be explained by the same logic, that logic would be a predominant sensitivity towards silence, the avoidance of conflict, and discretion: a cultural silence beyond linguistic pragmatics and verbal and nonverbal practices, as well as beyond the specific circumstances of interaction.

  1. Keith Basso, “To give up on words: Silence in western appache culture”, Southwestern Journal of Anthropology, 26 (3) (1970): 213-230.
  2. Montserrat Ventura i Oller, Identité, cosmologie et chamanisme Tsachila. À la croisée des chemins, l’Harmattan, Paris, 2009. Spanish version: En el cruce de caminos. Identidad, cosmología y chamanismo tsachila, FLACSO / Abya-Yala / IFEA, Quito, 2012.
  3. Connie Dickinson, “Mirativity in Tsafiki”, Studies in Language 24(2) (2000): 379-421.
  4. Taken from an unpublished interview by A. Aguavil of J. Aguavil and C. Calazacón, Congoma, 1997. This expression has been translated as, “One who speaks for the sake of it.” I am grateful to the Pikitsa collective, a Tsachila group that collects documents about the Tsachila language and culture, for passing it on to me.
  5. This story was explained by Florinda Aguavil in Cóngoma in 1997.

27. Parler des couleurs sans termes de couleurs

Alexandre Surrallés
Laboratoire d’anthropologie sociale, Collège de France


En ce qui concerne la diversité linguistique et culturelle, un sujet saillant est la variabilité des façons d’exprimer l’expérience sensible suscitée par les textures, les couleurs, les odeurs et tout ce qui s’offre à nos organes des sens. Or, cet objet de recherche se trouve d’emblée confronté au problème de ce qu’on appelle en anthropologie linguistique l’ineffabilité, c’est-à-dire la difficulté ou même l’impossibilité d’expliquer par des mots des expériences pour lesquelles les concepts nécessaires sont soit inexistants, soit moins évidents, un domaine récemment constituée à part entière (Levinson i Majid). L’existence de fossés sémantiques concernant en particulier la dénomination de perceptions se pose de surcroît comme un des problèmes les plus importants pour une anthropologie intéressée par le sensible. Car ce n’est pas la même chose de décrire l’odeur d’une fleur, par exemple, que de décrire la forme d’une figure géométrique. Une dimension de cette interaction sensible échappe à toute conceptualisation ou codage culturellement stabilisé par des conventions sémantiques.

L’ethnographie de la façon dont la langue amérindienne parlé par les Candoshi de la haute Amazonie évoque les couleurs des objets sans termes de couleurs pour les nommer, m’a offert la possibilité de réfléchir à l’ineffable. En effet, ces chasseurs, cueilleurs et horticulteurs forestiers parlant une langue de la famille jivaro évaluent les impressions sensorielles relatives aux couleurs sans l’un des principaux outils descriptifs utilisés à cet effet : les noms de couleurs. Cette donnée a suscité la controverse puisque l’anthropologie a en général considérée qu’on dispose partout d’une notion de couleur et de termes pour nommer celles ci. Cela n’a rien d’étonnant pour les « universalistes » (Berlin, Kay, Rosch) qui affirment l’existence d’une terminologie des couleurs de base commune à toutes les langues. Mais leurs détracteurs, qui prônent une approche relativiste et signalent la diversité culturelle des manières de découper la gamme chromatique, ne nient pas non plus l’existence partout de la notion de couleur et de termes pour les couleurs (Davidoff, Roberson, Sahlins). Cette unanimité semble cependant s´ébranler avec l’apparition d’une série de publications récentes d’anthropologues et de linguistes de différents horizons et sans concertation préalable (De Vos, Everett, Levinson, Saunders, Senft, Wierzbicka) confirmant la conclusion formulée par Harold Conklin dans un texte publié en 1955 selon lequel la couleur, dans le sens qu’on lui donne en Occident, n’est pas un concept universel et n’existe pas comme tel dans nombre de langues. Mon expérience chez les Candoshi, aboutit à la même conclusion. La question est alors de savoir comment les Candoshi parlent des couleurs sans noms de couleurs et sans que cela ne les empêche ni de percevoir des variations chromatiques, ni de pouvoir les exprimer.

En effet, la langue candoshi ne dispose pas de terme général pour nommer la couleur et le mot « couleur » est absent du seul dictionnaire candoshi publié à l’heure actuelle (Tuggy), sans qu’il s’agisse là d’une erreur ou d’une omission. Des notions telles que « multicolore » ou « coloré » ne semblent pas non plus exister, non plus qu’une forme de référence attributive de la perception d’une couleur comme par exemple dans la phrase « ce pot d’argile est rouge». Bien entendu, mes entretiens ne procédaient pas avec des questions du style « de quelle couleur est cet objet ? », laquelle ne peut pas être formulée dans la langue vernaculaire. La question que je posais en montrant par exemple une vignette de couleur était « ini tamaara ? » (déictique, ini, suivi d’un terme se traduisant par « comment est-il ? »). Par ailleurs, s’il n’y a pas de terme pour la couleur ou les couleurs, leur connaissance ne constitue pas un domaine autonome et, par consequent il n’y pas d’intérêt à en parler de façon abstraite. Les Candoshi, pourtant, s’intéressent aux couleurs et je dirais même que les couleurs les passionnent. Quand ils parlent des animaux, des plantes et des minéraux, ou notamment des fruits et des plumes, c’est-à-dire de choses spécifiques dans des situations concrètes, une façon d’évoquer les couleurs sans termes de couleur se fait jour. Si, par exemple, on présente à un locuteur candoshi une pomme de terre qu’il ne connaît pas bien parce qu’elle est cultivée à une plus haute altitude, et qu’on lui demande ini tamaara ?, sa réponse ne consiste pas seulement à essayer de la décrire superficiellement. Il saisit le tubercule, le sent, le pèse et le fait pivoter, de sorte que sa réponse exprime une expérience polysensorielle recourant à des comparatifs pour souligner les propriétés de la pomme de terre qui l’intéressent. Il s’agit d’une description hésitante, volontairement subjective, interactive, évolutive et relative. La couleur des choses, ou plutôt la similitude des couleurs entre les choses décrites et d’autres, apparaît liée à des caractéristiques autres que visuelles : au toucher, au goût et à l’odeur qui sont alors mobilisées dans le cadre d’une action intentionnelle propre à un environnement donné. En résumé, il semble que la couleur comme caractéristique isolée des autres propriétés d’une chose, ne soit pas pertinente pour la décrire, l’identifier, ni même la classifier. Peut-être parce qu‘elle n’est est pas pertinente en soi dans un monde où les objets, les animaux et les plantes sont inextricablement liés à celle qu’ils possèdent. Car ce qui importe à un chasseur ou à un pêcheur qui veut décrire une espèce est un ensemble de propriétés perceptives que l’espèce en question transmet, accompagné d’une description de son comportement et de son habitat écologique spécifique.

J’ai compris que les Candoshi ne cherchaient pas une catégorie mais une ressemblance, quand je leur ai montré l’une des vignettes du nuancier chromatique. Un débat vif s’est produit entre deux d’entre eux qui participaient aux conversations autour des vignettes disposées sur la table. Je me suis intéressé aux termes de la discussion et comme je ne parvenais pas à comprendre, l’un de mes interlocuteurs a pris un type de gingembre et en a coupé un morceau pour faire apparaître la couleur et la comparer à la vignette. Le débat était de savoir si la couleur orange ambre de cette vignette ressemblait davantage à un type de gingembre ou bien à une substance de couleur similaire secrétée par un poisson (probablement de la famille des Loricariidae) lors du frai. J’ai compris alors qu’un malentendu culturel survenait. Ils répondaient à mes questions leur demandant de donner des noms de couleurs associées aux vignettes au moyen d’un autre exercice : trouver dans l’environnement des choses de la couleur la plus ressemblante possible.

Quand un Candoshi décrit un objet, il le compare à d’autres en faisant ainsi référence à ses propriétés pertinentes dans le contexte de la description. Et comme la gamme des choses susceptibles d’être évoquées pour la comparaison sont toutes celles existant dans le monde, les possibilités descriptives apparaissent infinies et le résultat peut donc être d’une grande précision – cette forme de communication étant seulement possible si les locuteurs ont une connaissance partagée de l’environnement. Dans ces conditions, des termes pour les couleurs limiteraient les grandes possibilités expressives qu’offre la comparaison entre les choses. En somme, du point de vue candoshi, la couleur des choses est subsumée sous d’autres dimensions perceptuelles et entremêlée avec elles. Les couleurs sont perçues sans être nommées parmi d’autres caractéristiques sensibles des objets dans un acte de perception ou technique que j’appelle « perception contrastive » et qui consiste surtout à les comparer ; une technique alternative à la « perception catégorielle », souvent considérée la seule façon de décoder les sensations.


Pour approfondir sur le sujet voir :

Jones, Nicola, 2017, “Do You See What I See? “,   Sapiens, Anthropology/Everythink human, 9-Feb-2017.

Surrallés, Alexandre, 2016, “On contrastive perception and ineffability: Assessing sensory experience without colour terms in an Amazonian society“, The Journal of the Royal Anthropological Institute (incorporating Man). Vol 22(4). : 810-829.

 Couple candoshi et la passion pour les couleurs

26. Australia’s indigenous wordscape

Nicholas Evans
Australian National University


Australia is a predominantly monolingual country with a deep multilingual past, located at the epicentre of the world’s linguistic diversity. During its first forty to fifty millennia, indigenous cultures developed a mosaic of over three hundred languages in which high levels of multilingualism were the norm and which evinced great interest in language in all its forms, and grammars of formidable grammatical sophistication – Kayardild weighs in with twenty cases (to Latin’s six), the same freedom of word order that classical poets could exploit, and a terseness that allows one to express something like ‘(watch out), lest it get away from the one belonging to your opposite-sex sibling’ in a single word, kularrinkarranmulanharranth.

The phrase ‘epicentre of linguistic diversity’ is not chosen lightly. Of the world’s roughly 7,000 languages, a fifth are spoken in the southwest Pacific — some thousand on the island of New Guinea, 250–400 in Australia depending on the measure (and these are languages, not dialects — counting the latter sends the figure much higher), over 130 in Vanuatu (the world champion in Gross Linguistic Product at close to one language per 2,000 speakers). And among the world’s top dozen countries measured by number of endemic languages, half are in the neighbourhood — #1 Papua New Guinea, #2 Indonesia, #4 India, #5 Australia, #10 Philippines and #12 Vanuatu. The linguistic prodigality on the island of New Guinea alone is comparable to that of Eurasia as a whole, from Ireland to Japan, from Siberia to Sri Lanka — a statement that holds up whether one counts the number of languages, the number of language families, or the amount of ‘disparity’ in language structures. Languages like Iau (in West Papua) with nine tones sit cheek by jowl with others with no tones at all, and the language with the largest sound inventory in the western Pacific (Yélî-Dnye on Rossel Island) is just a couple of hundred kilometres from that with the smallest, Rotokas on Bougainville Island.

This voluptuous linguistic landscape is one reason for the thriving linguistic scene in Australia. For many linguists working here, however, there are other more personal motives — a wish for a more authentic view of who we are in this part of the world, grounded in the intricate and diverse cultural products of fifty millennia of human occupation and the mosaic of world-views these have elaborated. Add to this the fact that so many nonindigenous Australians grow up with an aching sense of unconnectedness to their land, stemming from the invisibility and inaudibility of Aboriginal culture and the peremptory way its insights were briskly swept aside by the British colonisation process. This makes linguistic research — and one day, I hope, the broader cultural and educational awareness that grows from it — an opportunity to create a type of culture that so far we have failed to nourish in this country.

I spent a lot of my childhood in the bush around Canberra, whether after school in the bush behind Campbell or on long camping trips. Nonetheless, I am probably typical of non-indigenous Australians in the shallowness of what I learned about my environment, and in the mismatch between my monoglot English upbringing and the inchoate feelings I held for my surrounds. In northern Australia, on other hand — where I was lucky enough to have a ‘second childhood’ under the welcoming instruction of my many indigenous teachers — almost every plant and bird now bears a vivid charge. Not only have I carefully been taught their names, in Dalabon or Bininj Kun-wok or other local languages, but also their uses, what their flowering says about the availability of food resources, and a whole rich panoply of myth.

The web of life, in languages like this, is mirrored in the web of words, from different verbs for the distinct hopping gait of every different macropod species, male, female and child, to retriplicated nouns for ecozones dominated by a particular plant (e.g. Kunwinjku mi-djoh-djo-djo ‘mixed scrub with wattle, acacia difficilis, dominant’ from an-djohacacia difficilis’). This is mingled with a rich affective lexicon for the sensations and emotions the landscape brings out — words such as, from Dalabon, karddulunghno ‘smell of first rains’, or from Iwaidja, angmarranguldin ‘change in environmental conditions, bringing back memories and inspiring longings for an absent person or place through the recollection of the smell of the sea or of a dying bushfire as the wind turns.

There is also the intriguing phenomenon of ‘sign metonymies’, which signal the fact that one natural phenomenon is a guide in space or time to the presence of the other — e.g. in Gun-djeihmi alyurr denotes the Leichhardt’s grasshopper (Petaside ephipigera), two herb species which it eats (Pityrodia jamesii and Cleome viscosa ) and whose location is thus the best way to locate these grasshoppers, and the lightning spirit, which starts to manifest itself in the first monsoonal storms at the same time as the herbs are ready for these grasshoppers to eat. At the time of the first monsoons, Leichhardt’s grasshopper is said to don its sumptuous orange and blue outfit and go looking for the lightning. Local cave paintings depict lightning spirits with axes on their heads – the grasshopper’s antennae. A central place in Yolngu symbolic thought is held by likan, literally ‘elbow’ but also ‘joint, connection’ — close to what would be called tropes in the Western tradition. ‘Likan names’ are used, in contexts of art and ceremony, to indicate more allusive readings to the culturally knowledgeable.

Alyurr: Leichhardt’s grasshopper, herb and lightning man

These examples illustrate how indigenous languages contain a vast network of knowledge about the natural world, but also how the cultures that nourished them were fascinated by language and developed a range of metalinguistic terms, practices and products.

Few aspects of indigenous culture better illustrate the intellectual sophistication of indigenous Australian traditions than the special auxiliary linguistic systems they created. Many of these were linked to initiation rites, making clear that the passage to adulthood was not just a matter of physical trials and self-discipline, but also of attaining a new understanding of how language articulates with the world.

Take the problem of antonymy. Giving ‘up’ as the opposite of ‘down’ or ‘tall’ as the opposite of ‘short’ are trivial. But semantic textbooks remain mute on the question of where antonymic oppositions stop — an errant omission in a world seeking to decompose all representation to binary code. What is the opposite of ‘mother’ — ‘father’, or ‘child’? Or of ‘kangaroo’, or ‘countryman’, or ‘(s)he’? The antonym of ‘deaf’ is evident, but what about ‘see’? The special register known as Jiliwirri, learned by Warlpiri initiates, is as far as I know the only case in the world’s intellectual history of a thoroughgoing investigation of antonymy applied to the entire lexicon. To speak it, you must replace all lexical items (though not grammatical affixes other than pronouns) with their opposites. As the following example shows, to convey the proposition ‘I am sitting on the ground’, you use a Jiliwirri utterance which would translate literally into everyday Warlpiri as ‘someone else is standing in the sky’. Jiliwirri has been used to investigate antonymy in Warlpiri lexical semantics, including such nonobvious issues as whether the perception verbs ‘see’, ‘hear’ etc. have antonyms, and how one determines antonyms for natural species names like ‘red kangaroo’.


Even more spectacular is a special initiation register known as Damin, which was taught to Lardil men on Mornington Island as part of their initiation as warama (second degree initiates). Damin is said to have been created by an ancestor known as Kaltharr (Yellow Trevally fish), and has a rich inventory of sounds, supposed to echo what ‘fish talk’ would sound like. In fact, its phoneme inventory is unique among the world’s languages and employs types of sound not found anywhere else, such as the ‘ingressive lateral fricative’. There are also a range of click sounds, like those found in Southern Africa. Because grammatical affixes are simply taken over from everyday Lardil, it is only the lexical roots that display these special sounds, as can be illustrated by the following sentence equivalents from everyday Lardil (2a) and Damin (2b): Damin substitutes ŋ͡!aa for ngada, didi for ji– and m͡!ii (with a clicked m) for werne, but leaves the grammatical suffixes intact.


But it is the semantic structure of Damin which represents a true tour de-force in metalinguistic analysis. Since the time of Leibniz philosophers and semanticists in the Western intellectual tradition have been seeking an ‘alphabet of human thought’ which would allow all meanings to be decomposed into a small stock of elements. Damin comes close to achieving this goal — out of nowhere in terms of prior philosophical traditions, and without drawing on any tools of written logical notation. It maps the thousands of lexical items of everyday Lardil onto around 200 words by a combination of highly abstract semantics, extended chains of meaning extensions, paraphrase, and supplementation by hand signs.

Thus in the above example, ŋ͡!aa does not simply correspond to ngada ‘I’. Rather, it can denote any group including ego. Now everyday Lardil has eight ways of translating English ‘we’ — given by the three-dimensional binary matrix of ‘inclusive’ (i.e. we, including you) vs ‘exclusive’ (we, but not you), ‘dual’ (two) vs ‘plural’ (more than two) and ‘harmonic’ (referents in even-numbered generations with respect to each other, such as siblings, or grandkin) vs ‘disharmonic’ (odd-numbered generations such as parent and child or great-grandkin). This exuberant semantic specificity in the everyday language is mapped onto the sober, highly abstract Damin word ŋ͡!aa ‘I, we, here’, opposed to ŋ͡!uu ‘you, (s)he, they, there’. Integration with gesture is an important part of what makes communication possible in Damin — as well as ‘there’, ŋ͡!uu can also mean ‘north’, ‘south’, ‘east’ and ‘west’ in Damin. The distinction between these is indicated by pointing in the appropriate direction while uttering the word — in the process giving a valuable insight into how a type of language functions in which the communicative load is more evenly distributed between speech and gesture.

As another example of how Damin semantics works, the rich particularity of verbs in the everyday language are mapped onto highly general designators in Damin, reminiscent of attempts at semantic decomposition of verbal predicates which linguistic philosophers began experimenting with in the 1960s. Thus the Damin verb didi takes in, among many other correspondents, jitha ‘eat’, but also all actions producing a physical change on their object, such as barrki ‘chop’, betha ‘bite’, bunbe ‘shoot’, and kele ‘cut’. Another word diidi, which sounds similar but has a long vowel, includes all actions of motion and caused motion, such as waa ‘go’, jatha ‘enter’, murrwa ‘follow’, jidma ‘lift’, and kirrkala ‘put’. Sometimes the motion is to be understood metaphorically, such as a change in possession (wutha ‘give’, wungi ‘steal’), a transfer of information (kangka ‘speak’), or the movement of food from outside to inside the body (jitha ‘eat’). The net effect is to produce a totally indigenous analysis of the semantics of the entire vocabulary into a small number of elements, and Hale justifiably refers to Damin as a ‘monument to the human intellect’. Elsewhere he has drawn attention to the fact that its association with rituals outlawed by the missionaries in power on Mornington Island meant that its transmission was interrupted well before the transmission of everyday Lardil, as well as to the invisibility of this achievement to the outside world:

The destruction of this intellectual treasure was carried out, for the most part, by people who were not aware of its existence, coming as they did from a culture in which wealth is physical and visible. Damin was not visible for them, and as far as they were concerned, the Lardil people had no wealth, apart from their land.

It is a task for our times to make awareness of such intellectual wealth widespread. Never before in human history have we been losing languages so fast, but never before have we had such a dawning realisation of their value, or the means – conceptual, human, technological – to do something about it. In writing this article, I hope that Catalan readers, who have so valiantly maintained their magnificent tongue against historical odds, will be inspired to join the growing number of people who see the conservation of the world’s wordscape as one of the great quests of our time.


This is an abridged and modified version of Evans, Nicholas (2017).
Ngurrahmalkwonawoniyan. Listening here. Humanities Australia 8:34-44. The reader is referred to that publication for bibliographic references

25. Measuring sociolinguistic groups: The most significant transformations in Catalonia and the Balearic Islands

Natxo Sorolla
Universitat Rovira i Virgili, Xarxa CRUSCAT-Institut d’Estudis Catalans i
Centre de Recerca en Sociolingüística i Comunicació (CUSC) de la Universitat de Barcelona


Alongside Xavier Vila, I have been working for some time on sociolinguistic groups in Catalan language area. A single person’s linguistic usages may vary greatly from one area of life to another. An individual might speak in Catalan at home, in Catalan and in Spanish with friends in quite similar proportions, only in Spanish with the doctor, and only in Catalan with the person who serves him or her at the bank. The idea behind sociolinguistic groups is to create a usage index of spoken Catalan in 10 fields (IUP-10), which we believe is much more precise than more general propositions such as usual language or percentage of use in everyday interactions. The idea has certain parallels with studies on happiness, and above all, with what Daniel Kahneman, the only Nobel Prize-winning psychologist, suggests in relation to the polemics accompanying Easterlin’s studies on whether money can buy happiness (see the excellent episode on this from the Catalan TV show Economia en colors). These discussions centre on how to measure happiness, and how things differ if we ask about our state of happiness in general, as opposed to measuring happiness from moment to moment. To a certain degree, when we ask about happiness memories, we are more generous when it comes to our lives than we are when we have to note each moment of happiness, because even though we think that we are (in general) fairly happy, a good part of our day involves situations of stress, sadness or boredom. The idea that we strive for in sociolinguistics is to a certain extent that, beyond the memory that we have of our language usage, it is much more useful to measure and summarize what language we speak in specific areas or with specific interlocutors, such as the language used with friends, spouses, children and neighbours, or when in large commercial establishments, and so forth. The first way to measure happiness is certainly more related to what we would like to do than is the second way, which ought to be more realistic in terms of what we really do.

And so this index has been presented as very useful when summarizing information on language usage in a single variable, observing different sociolinguistic groups based on people’s family linguistic origin, and above all comparing different situations. We have started to work with the current methodology, which has allowed us to compare different waves and contexts, with data for Catalonia (Sorolla & Vila 2015). We first of all used the methodology to detect that a large part of the population predominantly uses its initial language, but there is a very significant non-Catalan-speaking population group that predominantly uses Catalan (7.3% of the population), which indicates the vitality of the language in the region (see the chart below). But at the same time, we have been able to detect that in the last decade there have been significant losses in the use of Catalan in Catalonia, concentrated mostly around the metropolitan area of Barcelona. These losses are not only caused by demographic issues such as the arrival of immigrants or the fact that the younger generation comprises fewer Catalan speakers, but also by sociolinguistic factors. And it is the latter that should raise greater concerns: members of the same generation with a shared social composition make less use of Catalan now than they did 10 years ago. For example, older Spanish-speaking generations seem to be using Catalan less than they were a decade ago, possibly due to changes in their everyday lives.







The second study that we have undertaken has been in the Balearic Islands (Vila & Sorolla 2017). It has just been published. The work that we undertook adopted the same technique, and it allowed us to observe that, once again, the most significant sociolinguistic groups are those in which linguistic use combines with family origin. But, unlike in Catalonia, the group of Catalan speakers from non-Catalan-speaking backgrounds is proportionally smaller (2.8%), and in contrast, its size is similar to the group which predominantly speaks Spanish but is of Catalan-speaking origin (2.5%), which is undergoing the opposite process. But in addition to not observing Catalonia’s vitality, we detected in Palma, Ibiza and Formentera that a lack of integration on the part of newcomers has unleashed a widespread process of Spanish use, even among some sectors of young Catalan speakers, through which they are abandoning use of their own language. See the chart below: a fifth of young people habitually use Spanish, despite being of Catalan-speaking origin.

Finally, and as a complement, we have reviewed the history of this process to create sociolinguistic groups. Xavier Vila began developing sociolinguistic groups through exploiting the Survey of Language Use for Catalonia 2003 (Vila 2005), and we continued the development with analysis for the Andorra 2004 (Sorolla, Bretxa &Vila 2006; Sorolla 2011), Franja 2004 (Sorolla 2014), and Catalonia 2008 (Sorolla & Vila 2011) surveys. But it was not until the 2013-15 wave of surveys that we consolidated the methodology by proposing that sociolinguistic groups should be defined theoretically, and not on the basis of empirical results, in order to be able to compare surveys conducted in different years or in different territories. Moreover, this approach has also been adapted to sociolinguistic surveys of Galician by Xaquín Loredo (Loredo 2015; Monteagudo, Loredo & Vázquez 2016).

Sorolla, N., & Vila i Moreno, F. X. (2015). Els grups d’ús lingüístic i els grans canvis en els usos entre el 2003 i el 2013. In Generalitat de Catalunya (Ed.), Anàlisi de l’Enquesta d’usos lingüístics de la població 2013. Resum dels factors clau. (Generalitat de Catalunya, Departament de Cultura, p. 30-33). Barcelona.

Vila, F. X., & Sorolla, N. (2017). La realitat i l’evolució a partir dels grups d’orígens i usos lingüístics. In J. Melià & M. del M. Vanrell (Ed.), Enquesta d’usos lingüístics a les Illes Balears. 2014 Anàlisi (Conselleria de Cultura, Participació i Esports (Govern de les Illes Balears) ­- Departament de Cultura (Generalitat de Catalunya) – Universitat de les Illes Balears, p. 106-124).

24. What is language revitalization about? Some insights from Provence*

James Costa
Sorbonne Nouvelle / UMR LACITO (CNRS), Paris


Should you find yourself in Provence this summer, you might wonder why some villages have bilingual signs at the entrance. Your surprise would be forgiven, since you are unlikely to have heard anything but French in most places, and likely a lot of English as you approach the Mediterranean. But if you listen more closely, observe more closely, you might come across a world that is fast vanishing, but that is still present. You might stumble upon a concert in a language that you cannot identify, or wonder why some street names don’t sound French. You might even hear people speak Occitan — for this is what it is, a language also known as Provençal, one which many locals will refer to as “Patois” (a derogatory term in France to refer to anything other than French traditionally spoken in the country).

This sort of experience might happen to you in Provence, but not only. Across the European Union, several million people speak a language that is not the official language of the state they live in. Across Europe, there are language advocates who defend and promote the right to speak one’s language. This struggle for language rights also extends to Latin America, North America, Australia, and many other places. This, many scholars assert, is a consequence of globalization—a backlash against uniformity if you like. A way of being oneself, of finding meaning locally in a world that seems to be getting smaller. In my recent book, Revitalising Language in Provence: A Critical Approach, I argue otherwise. Those movements are not a reaction to globalization—they are, on the contrary, a way of taking part in this process, on the very terms defined by those who define what globalization is (and not on their own terms, as Leena Huss [2008, 133] asserts).

But let’s start from the beginning. This book focuses on Provence, home to what is perhaps the earliest language reclamation movement, or at least one of the earliest. Poets had already started writing texts in defense of Gascon, Provençal or Languedocien (all dialects of what most scholars of Romance linguistics view as Occitan) back in the 16th and 17th centuries. This is perhaps a consequence of an increasingly aggressive move to promote French in all administrative domains at the expense of Latin and Occitan, which had been in use for official usage for centuries in what is now Southern France. But it was after the French Revolution Terror government (after 1793) sought to eradicate the “patois” that a genuine interest was born in various parts of France, resulting in the south in a rediscovery of the poetry of Medieval Troubadours and in a scholarly interest in the history of Provence and Languedoc before their annexation to France. It wasn’t, however, before the 1850s that an organized language-based movement was formed, under the aegis of poets such as Frederic Mistral or Joseph Roumanille.

The Felibrige was the name they gave to their movement, a name whose origin remains mysterious. The Felibres sought to revive the Provençal or Occitan language (which was still almost universally spoken in all of Southern France) through poetry and literature. And indeed, Mistral published a series of long, epic poems that were hailed across Europe as monuments of literature. Mirèio is probably his most well known poem, a love story set in the Crau region of Provence and an allegory of the language revival movement. Mirèio was acclaimed in Paris as a chef d’æuvre, and was prefaced by Lamartine.

I recount parts of the history of the movement in the book but for this article, suffice it to say that while successful on a literary level, it never succeeded in political terms. Provençal was long banned in education, and despite a strong Occitan movement throughout the 20th century, the use of Provençal continued (and continues) to decline. But the story I tell in this book isn’t the story of the language movement. Instead, following a 2 year ethnographic study in Provence, I ask why the movement was based on language at all, like so many others afterwards—but, crucially, none before, or at least none before the 1840s.

What struck me as I started the work on the PhD thesis that led to this book, back in 2006, was that anthropologists had long written about what they called “revitalization movements”. Anthony Wallace was perhaps the scholar who coined the term back in 1956, but others before him such as Ralph Linton had talked about nativist movements. Those movements, anthropologists argued, could use religion as their basis, or cultural elements, or political institutions. In 1943 Ralph Linton wrote: “The avowed purpose of a nativistic movement may be either to revive the past culture or to perpetuate the current one, but it never really attempts to do either” (Linton 1943, 231). In 1956, Wallace stated that revitalization movements are “a deliberate, organized, conscious effort by members of a society to construct a more satisfying culture. Revitalization is thus, from a cultural standpoint, a special kind of culture change phenomenon: the persons involved in the process of revitalization must perceive their culture […] as a system […]; they must feel that this cultural system is unsatisfactory; and they must innovate not merely discrete items, but a new cultural system, specifying new relationships as well as, in some cases, new traits” (Wallace 1956, 43). For Wallace, Revitalization thus starts with a steady state period which gets shattered by an outside event (such as colonization), leading to individual stress and later cultural (collective) distortion which then prompts the revitalization process to happen, often triggered by a Prophet. In turn, revitalization is meant to lead to a new steady state.

Two elements (at least) interested me in this body of work. First, revitalization could happen to something other than language. Second, while the very fact that steady state periods ever existed is itself probably an illusion, what I did observe during my ethnographic fieldwork in Provence was that the language movement did tell a story that adhered closely to the pattern described by Wallace. Those two observations led me to ask two sets of interrelated questions:

  • If revitalization can be about many things, why language? When (and why) does language become something that one can revitalize? Is it, as many sociolinguists claim, because of the realization that many languages are now endangered, and on the brink of disappearance? Or is there something else?
  • Whatever actually happens in the revitalization process, aren’t the movements that generate and support revitalization primarily involved in creating a narrative of revitalization, in order to provide (and ultimately impose) a particular reading of the world? In that sense, aren’t revitalization movements not like most other social movements, seeking to provide collective interpretations for individual events, in this case using language as a focus to take part in a struggle over categorization (to use a term employed by Bourdieu [1991])? And if this is the case, which categorizations are those movements seeking to challenge?

In the book, I thus propose that if language revitalization movements should not be dismissed offhand as romantic pipedreams or as ethnic inward-looking enterprises, they shouldn’t be idealized either as leading a resistance against globalizing and uniformizing forces. Instead, they should be analyzed as human enterprises that seek to provide collective interpretations of the world, using language as a prism. Neither romanticized nor dismissed, language revitalization movements should nevertheless be studied more extensively as social movements. This needs to be done not with the aim of vindicating or extolling their actions, but primarily because their multiplication in the 21st century tells us something about what it means to be human, to struggle as humans, in the world.

I thus propose the following approach in order to problematize the types of social actions that actors involved in language revitalization engage with:

  • First, language revitalization movements are about “groupness,” to use a term coined by Rogers Brubaker to describe how groups are always collective projects. As he writes, “a group is a variable, not a constant; it cannot be presupposed” (Brubaker 2004, 4). Groups, in other words, are what we want to explain, “not what we want to explain things with” (Brubaker 2002, 165). In that sense, language revitalization isn’t about regenerating pre-existing groups, but about inventing new ones, on new terms, while drawing on a construction of those groups as timeless or ancient. What, then, are categories of language and groupness mobilized for, and by whom?
  • Second, language revitalization is fundamentally a struggle over classifications—“struggles over the monopoly of the power to make people see and believe, to get them to know and recognize, to impose the legitimate definition of the divisions of the social world and, thereby, to make and unmake groups” (Bourdieu 1991, 221). The struggles implied here concern what the group that is the object of revitalization is, what are its defining features, what counts as language, as legitimate language, and what rights and duties members have towards each other.
  • Third, language revitalization is a consequence of social contact. Such an approach allows us to shift our gaze from the group that the language movement apparently addresses to the interstices revitalization generates between the minority group and a majority group that gets constructed at the same time. The consequences here are twofold: revitalization is neither an inward-looking movement, nor a way to deal with oppression on the minority group’s own terms. Rather, it is a way to renegotiate the very terms of contact between groups that are created through contact, more often than not on terms imposed by the dominant group (as already suggested by Sanford 1974). In order to exist, in other words, minority groups have to appeal to categories that already have currency among those who they view as the majority. In order to address the cultural elite in Paris, Frederic Mistral had to appeal to categories that they viewed as legitimate: high literature and a codified language with a prestigious, legitimate past.
  • Finally, language revitalization is ultimately not primarily about saving languages, but it is perhaps essentially about finding a terrain to frame other types of claims and projects—social, political, related to land rights etc. In that sense, it is important to consider conflicts within language-based movements not as something that must be overcome, but as an indicator of tensions locally and globally over what the world should look like, who should be allowed to take part in its daily affairs, under what moral and political conditions.

In the book, using those preliminary propositions and applying them to the Provençal case, I seek to show how we, as scholars, can reframe questions relating to language revitalization. While the case is approached ethnographically, I intend to raise questions beyond revitalization in Provence, and to generate a wider discussion as to why so many language-based social movements have emerged worldwide since the 1990s. Because it has been so well documented over the past 150 years, the Provençal revitalization movement is a good point to start this discussion. What then, if anything, has this explosion got to do with the fall of the Soviet Union? If we accept that revitalization is about renegotiating terms of contact on terms imposed by dominant groups, what does this interest in language tell us about global ideologies of language? How does language, rather than, say, race, religion or politics, allow minority groups worldwide to frame their plight in a way that affords them a voice—through Unesco for instance?

* An earlier version of this note appeared in The Philological Society Blog (


Bourdieu, Pierre. 1991. Language and Symbolic Power. Cambridge: Polity Press.

Brubaker, Rogers. 2002. “Ethnicity Without Groups.” Archives Européennes de Sociologie 43 (2): 163–89. doi:10.1017/S0003975602001066.

———. 2004. Ethnicity Without Groups. Cambridge [MA]: Harvard University Press.

Huss, Leena. 2008. “Researching Language Loss and Revitalization.” In Encyclopedia of Language and Education, edited by Kendall A King and Nancy H Hornberger, 10: Resear:69–81. Springer.

Linton, Ralph. 1943. “Nativistic Movements.” American Anthropologist 45 (2): 230–40.

Sanford, Margaret. 1974. “Revitalization Movements as Indicators of Completed Acculturation.” Comparative Studies in Society and History 16 (4). Cambridge University Press: 504–18.

Wallace, Anthony F C. 1956. “Revitalization Movements.” American Anthropologist 58 (2). Blackwell Publishing on behalf of the American Anthropological Association: 264–81.

23. Alghero’s Catalan in Perspective

Francesc Ballone
Philology Section, Institute for Catalan Studies


Last May, the main results of the Survey of Language Use in Alghero 2015 were presented (report available at:

In this note, I discuss some of the 2015 survey data that provide information pertaining to the (possible) future of Algherese.

Discrepancies between declared and real language skills

During the analysis of the 2004 survey the difficulty of defining in a clear way, a posteriori, the level of linguistic competence in the four main skills (listening, speaking, reading, writing) based on dichotomous questions such as “Do you know how to speak Algherese?” emerged. (See: <, especially pp. 35-37). For this reason, gradated questions were added to the definition of the research protocol for the 2015 survey to be able to qualify different competency levels according to certain typical communicative situations encountered in Alghero. In addition, a prior definition of the minimum threshold in order to establish whether the informant in question has or does not have active competence in the skill under consideration was created. For example, in the case of “knowing how to speak,” a minimum threshold was applied of the ability to carry out an everyday conversation in Algherese without too much difficulty. This was an adaptation of a criterion already used by the European Commission to investigate the linguistic knowledge of its citizens. If we make comparisons between the values from the dichotomous and the gradated questions, we find obvious discrepancies in the results.

At first glance, it seems that the informants tend to overestimate their own Algherese language skills when they are asked a closed question such as “Do you understand Algherese? Yes/No.” In this regard, a more thorough comparative analysis between gradated and dichotomous questions regarding knowing how to speak reveals how the question “Do you know how to speak Algherese?” also received an affirmative response from the 19.6% of interviewees who speak it with difficulty and are normally not able to start and finish a conversation in that language.

The people of Alghero’s opinions on their own language

One possible key to interpreting the unexpectedly high positive percentages of responses to dichotomous questions is that users attach a positive value to knowledge of Algherese. This possibility was further confirmed by questions about linguistic opinions, which indicated that the resident population is strongly in favour (92.3%) of the introduction of Algherese and Sardinia’s other languages in the Sardinian school curriculum. In addition, a large majority (77.2%) would like Algherese to be the main or secondary language used in everyday activities. Moreover, the prejudice that holds that knowledge of local languages can cause problems when it comes to learning Italian and foreign languages, which was very strong even a few years ago, is now rejected by 86.8% of Algherese people.

The future of the language

The chart provides a stylized representation of the vitality of Algherese from a diachronic perspective. The green and red lines indicate the percentage of residents of Alghero who have Algherese, respectively, as a first language and as an acquired one.

If it were possible to take a snapshot of Algherese’s state of health on any one day of its history, we would note that this language was not only the L1 (mother tongue) of the majority of the population, but also that it had a strong ability to attract new speakers (vertical arrows) and transformed allophone newcomers into Catalan speakers in a relatively short time period. Although the state languages (Spanish and, later, Italian) exercised a strong attraction on Alghero’s elite—who nevertheless continued using Catalan in local-level relations—our language’s capacity for assimilation was still very strong until the early fifties of the previous century, a time when, as elderly people recall, “practically everyone spoke Algherese.” Since that period, however, a twofold factor has caused a significant fall in speakers. On the one hand, the people of Alghero have stopped passing on the language to their children, creating a reduction in native speakers of Catalan. And on the other hand, newcomers have not felt the need to integrate linguistically, given that Italian fulfils all communicative functions well. In this regard, the chart shows how Algherese has been transformed into a minority language in the space of only two/three generations (1950≃ — 2015).

Another aspect that does not bode well for the future of our language is the increasingly obvious progression of the process of intergenerational linguistic interruption, given that only 3.6% of the youngest parents use Algherese with their children as a primary language.


The data discussed in this note make it clear that today Algherese is a language that has a strong drawing power, given that the percentage of residents who would like to know and use this language in a significant way (77.2%) is more than double the percentage of people who know how to speak it fluently (30.7%) In addition, Algherese society exhibits a desire to incorporate the language in educational institutions in a structured way, overcoming the tendency of our educational system to conduct activities in Catalan (and Sardinian) sporadically and without long-term planning.

On the other hand, the “strong drawing power” mentioned in the preceding paragraph does not seems to be in itself sufficient to transform the people of Alghero’s positive attitude towards their own language into a proactive attitude, given that the tendency to use Italian with the new generations is becoming increasingly apparent.

22. Over the rainbow

Don Kulick
Uppsala University


Note: this essay is a chapter from a forthcoming book titled The End: how a language dies. The book is a series of short stories that document the author’s thirty years of linguistic anthropological research in Papua New Guinea, on an isolate Papuan language called Tayap. Tayap is spoken in a village called Gapun (pop. about 200 people). The language is dying: villagers under the age of 35 all speak Tok Pisin, a creole language that is Papua New Guinea’s most widely-spoken national language. It currently has fewer than 50 active speakers.  


One morning after a night of heavy rain, a wide vibrant rainbow appeared in the sky. On my way to wash my clothes in one of the village waterholes, I looked up and saw it, and I realized that I didn’t know what a rainbow was called in Tayap. So I asked the first person I saw – Michael, the village prayer leader, who I knew was fluent in the language – what the word for rainbow was in Tayap.

Renbo”, he responded, without missing a beat.

Um, no, I told him, that must be the Tok Pisin word – the Tayap word had to be something else.

Oh, he said. In that case he didn’t know. I should go ask his father, sixty-five-year-old Mone.

So carrying my plastic bucket with dirty clothes and bar of soap to wash them with in it, I went off to find Mone. Little did I know that my innocent query about the word for rainbow would spark a month of acrimonious debate from one end of Gapun to the other.

Mone was sitting in his usual morning spot, on his veranda chewing betel nut. I said good morning and I told him I had a question. What is the Tayap word for rainbow?

Instead of simply saying the word, as I expected he would, Mone paused and put a finger to his chin. He pondered. After a minute, he told me he couldn’t remember the word offhand; he needed to think about it. I thought that was odd. On the other hand, though, it isn’t as though rainbows are exactly common occurrences in the rainforest. I only saw that one the entire nine months I was in the village that year. So I thought that maybe Mone was just having a senior moment or had been caught off guard with my unexpected question about a word that villagers didn’t have occasion to use very often.

It turned out that Mone’s thinking about rainbows took several days. Finally, when I passed by his house late one afternoon on my way to take my end-of-day shower at the water hole, he called me over to his house and told me that “rainbow” had no single word in Tayap. Instead, “rainbow” was expressed through a verb phrase which meant “cloud is marked with color”.

This sounded reasonable to me, and I duly recorded it. But when I repeated it to other people in the village to check their reactions, I was universally met with disdain. Em giaman – “He’s lying”, everyone sneered, using their favorite expression to dismiss another speaker’s expertise in Tayap. Even though no one could think of the correct term themselves, they all told me they knew that the phrase Mone had volunteered was wrong.

I had encountered this kind of collective disagreement several times before. People disagreed testily on the word for “caterpillar”, for example. And then there was the wind problem. There are four named winds in Tayap: awar, ngamai, mbunim and mbankap. On this all the older villagers are agreed. They are also agreed that the winds are differentiated primarily by the directions of their origin. What they absolutely could not agree upon, however, was what those directions of origin are. One old man was adamant that the ngamai wind came from the mountain to the south of Gapun. An old woman was equally adamant that the wind came from the sea, which lies to the north of the village. Likewise, the awar wind was held by some people to come from the mountain (i.e. the south) and others to come from the mangrove lake (i.e. the north). Old people argued vigorously with one another whenever this topic came up, but they never resolved it.

By sheer luck, the four winds are listed and defined in a list of 125 Tayap words that was published in 1938, by a German missionary and anthropologist named Georg Höltker. In 1937, Höltker traveled to Gapun in the company of another missionary, thereby becoming one of the few white people to ever actually visit the village. Höltker and his companion spent only three hours in the Gapun. He took two photographs and collected a word list. A year later, he published the list, together with the weary remark that “it will be awhile before any other researcher ‘stumbles across’ Gapun, if only because of the small chances of worthwhile academic yields in this tiny village community, and also because of the inconvenient and arduous route leading to this linguistic island”.

Aside from Australian linguist Don Laycock’s unpublished word list that he gathered from two Gapun villagers whom he met in another village in about 1973, and my own work on the language, Höltker’s word list is the only documentation that exists on Tayap. For having been gathered in three hours by someone who had never before heard Tayap (and who would never hear it again), Höltker’s list of 125 words is impressively accurate. To resolve the controversy regarding the four winds, I decided, therefore, to go with the definitions listed by Höltker. He had, after all, spoken to language informants who still lived in a completely Tayap-speaking village. Also, one of the oldest speakers still alive in Gapun defined the winds as Höltker does in his wordlist. So the problem of the winds was solved.

Unfortunately though, “rainbow” wasn’t one of the words on Höltker’s list.

Days went by and no one could come up with the Tayap word “rainbow”. Old villagers explained to me that their parents and relatives had warned them about rainbows, saying they should never walk underneath one, because if they did, their minds would become clouded and their sense of direction confused. But even though they remembered these cautions, nobody could recall the word for rainbow that their parents and relatives had used while articulating them. The word for rainbow, villagers told me, “i hait”– it was hiding.

Eventually Mone’s old wife, Sopak, had a dream in which she said the true word for rainbow was whispered in her ear by a dead ancestor. The word, the ancestor had revealed, was mɨnuomb – a word that otherwise means “large round lake”. Sopak said that the way to say “rainbow” in Tayap was to say akɨnnɨ mɨnuomb utok, “a round lake appeared in the clouds”.

I told other old people in the village about Sopak’s revelation. They were unmoved. “Em giaman”, they all intoned impassively.

A few days after Sopak recounted her dream, one of the oldest men in the village told me that he had remembered the word – it was wagurmos.

The other speakers’ judgement fell predictably: “Em giaman”, they all pronounced. They explained that wagurmos meant the white veil of stars that appears in the sky at night – in other words, it is the Tayap word for the Milky Way. It doesn’t mean rainbow at all. Many of the people I asked about that word also took the opportunity to disparage the linguistic knowledge of the old man who had offered wagurmos. That man may be old, they said belittlingly, but he’s “lapun nating” – he’s grown old without having learned anything. All he has, people said, is “bebi sens” – the sense of a baby.

Weeks passed and frustration grew. Finally, having heard about the old peoples’ disagreements and disputes over the rainbow, a man in his thirties came to my house one day told me that he remembered once having heard his grandfather, old Kruni, say the word for rainbow. Kruni had been one of the old people who had taught me Tayap in the 1980s; he died in the early 1990s. For the last few decades of his life, Kruni had been universally respected and vaguely feared as an elder who knew everything about Gapun’s history and who spoke flawless and eloquent Tayap.

The young man reported that as a child, he had once been in a canoe together with Kruni when they paddled through the mangrove lagoon. In the middle of the lagoon they met a canoe full of women from the neighboring village of Wongan who were talking about rainbows. In the Kopar language spoken in Wongan, rainbows are called mamor. The young man remembered that the women had called out to Kruni and asked him what the word in Tayap was. Kruni told them that it was mamar.

Rather than being the happy breakthrough that I thought this was, mamar, too, was rejected. “It means ‘banana’”, all the old people responded dryly when, without telling them why and hoping to jar their memory, I asked them to define mamar.

And indeed, the word does mean a kind of banana. But lots of words in any language are homonyms, like the word “mole” in English, which has at least three different meanings: a small burrowing animal, a raised blemish on the skin, and a unit of measurement in chemistry. Couldn’t mamar, in a similar way, mean more than one thing? Might it not maybe also mean rainbow?

Nope. Kruni giaman. Or the young man who reported what Kruni said giaman. Somebody, in any case, was lying, the old speakers were agreed.

In the end, after a month of squabbling, unable to come up with a word or expression that satisfied them all, and undoubtedly growing annoyed at my persistence in questioning them, the older villagers begrudgingly allowed that mamar must be the word for rainbow, since Kruni apparently (and here several of them rolled their eyes furtively) had claimed it was.

My own conclusion is that mamar probably is the correct Tayap word for “rainbow”. Tayap and Kopar are completely unrelated languages, even though the villages where they are spoken are only two hours apart. But because speakers of the two languages have been in contact with one another for a very long time, they share quite a few lexical items. The kind of slight phonetic variation between mamor and mamar are common in the words shared by Tayap and Kopar. For example:

Because I had already recorded similarities like these, I told the villagers that I would enter mamar in my dictionary as the word for rainbow. This announcement was met with muttering.

The villagers’ inability to agree on proper Tayap is a feature of village life that is contributing to the language’s demise. I was continually struck by how vigorously (and, to my mind, how gratuitously) the old speakers of Tayap discounted and ridiculed one another’s linguistic competence. Early on during my stay in the village, I stopped trying to discuss Tayap in groups of old people because any discussion of any aspect of the language would inevitably result in bickering. Speakers might eventually grouchily agree on whatever it was I was asking them about, but later on, they would always arrange a private moment with me to heartily dismiss the knowledge and opinions of their fellow speakers.

It escapes no one’s attention in Gapun that Tayap is a tiny language spoken nowhere else but there. But a difference between Gapun and many other communities around the world is that language in Gapun is not regarded as a communal, shared possession. Like everything else in the village, knowledge of language regarded as private property. Gapun villagers would shake their heads in absolute bewilderment at the persistent Western stereotypes about how a rainforest-dwelling people like themselves supposedly eschew ownership and magnanimously share their natural resources in a kind of prelapsarian socialist ecological bliss.

On the contrary. In Gapun, nothing is communal, nothing is equally owned and shared by everyone. Everything – every area of land, every sago palm, every coconut palm, every mango tree, every pot, plate, axe, machete, discarded spear shaft, broken kerosene lamp, and every anything else one can think of – is owned by someone. This includes people’s names and the right to bestow them, as well as knowledge of myths, songs, and curing chants. Villagers always know who owns what. They have to know who owns what in order to take things freely, or steal them. They guard their rights of ownership energetically and they defend them fiercely. I have heard bitter arguments and shouts that “It’s not yours, it’s mine!” over objects as trivial as a discarded piece of string that a woman who had thrown it away saw her sister salvage from the rainforest.

Understandings like those of possession and proprietary ownership have consequences for language: they mean that the Western truism of a common “shared” language has little purchase in Gapun. In their own view, villagers don’t “share” a language. Instead, each speaker owns his or her own version of the language. And the older those speakers become, the more they regard their version as the proper one and everyone else’s as “a lie”. This absence of an understanding that regards a common language as something “shared” means that speakers are predisposed to not regard the loss of Tayap as particularly traumatic.

Fluent elder speakers still have ‘their’ Tayap; if younger speakers don’t possess a version of it as, well, wari bilong ol, that’s their problem.

21. The languages of the poets in northern Egypt

Dwight F. Reynolds
University of California, Santa Barbara  


When I was still a graduate student studying in Cairo in the early 1980’s, my mentor, a renowned folklorist, sent me out into the countryside of the Nile Delta to experience first-hand some of the oral artistry to be found in villages there.  He suggested several sites and I eventually found myself in a village that was known as “the village of the poets” because it was home to so many singers of the epic song of the Bani HIlal Bedouin tribe, known in Arabic as Sirat Bani Hilal.  To my astonishment there were fourteen households of professional, hereditary epic singers, men whose only occupation was to perform this oral epic sing in vernacular Egyptian dialect at weddings, in cafes, at festivals, at private gatherings and other occasions.

On that first visit I recorded some samples no longer than a half-hour each, and, I must confess, I scarcely understood a word of what they sang.  Though I had read written versions of the poetry, I found it difficult to understand the verses when they were sung, and, I soon learned, the versions these poets sang were marked not only by the local dialect, but also by a level of speech that might be termed ‘artistic colloquial.’  It is neither the language of everyday conversation, nor the ‘standard’ or ‘classical’ form of written Arabic, but rather something in between, a linguistic register used primarily in various forms of verbal art.

For nearly 15 years I made repeated trips to visit this village, eventually living there for a year and on other visits sometimes staying for several weeks at a time.  I recorded approximately 75 performances and took notes during many conversations and interviews (the poets, as with everyone else in the village, would under no circumstances allow me to record our casual conversations).  I also took lessons from a singer, Shaykh Taha Abu Zayd, whom many in the village deemed to be the best and most knowledgeable when it came to repertory and history.  I learned to play the rabab (a two-stringed spike fiddle) and to sing a small fragment of the epic.  The poets of this village had repertories that ranged from as little as 20 hours to over 120 hours of material.  And all of them were completely illiterate, unable even to spell their names (they commonly used a thumbprint when a signature was required, as did many of the older villagers).

Over time my understanding of the various unusual turns of phrase in the epic song grew and I was able to follow the story without difficulty.  But I soon stumbled upon a completely different language issue that involved the poets and their families.

The epic singers are from a special, distinct social group, similar to (but not directly related to) the Roma of Europe.  They call themselves the “wilaad halab” or “wilaad halaba,” a term that seems to mean “the children of Aleppo” since Aleppo is ‘Halab’ in Arabic.  But this makes no historical sense and the poets have not preserved any narratives that link them to Aleppo in northern Syria.  Scholars have hypothesized that the word might be a colloquial form of “halaqa” (a ring or circle), since the poets performed in the center of open circles to outdoor audiences, or that it refers to “milk” (‘haleeb’ in Arabic), and indicates a link to grazing animals.  But there is no real evidence that supports any of these ideas.

In any case, members of this social group do not marry outside the group except in very rare cases, and one of the features that sets them apart is their use (only among themselves) of language or argot that they refer to only as ‘ratana’ (gibberish).  Some of the young men in the poets’ community thought that it was fun to teach me to say things in ‘ratana,’ but elders in the community found out and put a stop to this.  This American researcher seemed perfectly nice and respectable, and it was fine that he wanted to record and learn to sing the epic, but as for matters that should be kept within their community (such as speaking ‘ratana’), that was off limits.  Everyone continued to help me with the epic project, but I never heard another word of ‘ratana’ during the remaining years of my research.  By 2000, all of the epic poets I had known had passed away, and none of their children had taken up the craft.  They had instead been sent to schools, learned to read and write, and now held more respectable jobs.  Though the epic lives on in other regions, in this village, the epic has disappeared and now lives on only in the recordings we made between 1982 and 1995.

Audio recordings, photographs, and selected translated texts can be found at: