Trans Internet-Zeitschrift für Kulturwissenschaften 15. Nr. Juni 2004

6.1. Standardvariationen und Sprachauffassungen in verschiedenen Sprachkulturen | Standard Variations and Conceptions of Language in Various Language Cultures
HerausgeberIn | Editor | Éditeur: Rudolf Muhr (Universität Graz)

Buch: Das Verbindende der Kulturen | Book: The Unifying Aspects of Cultures | Livre: Les points communs des cultures

Indians, languages, and linguistic accommodation in modern Chiapas, Mexico(1)

John B. Haviland (Ciesas, Mexico / Reed College, US)


1. Archive of the Indigenous Languages of Chiapas

The research on which I report in this short paper is part of a linguistic documentation and database project funded by CONACyT (the National Council on Science and Technology, Mexico) which aims to produce electronically accessible databases of the three most widely spoken indigenous languages of the state of Chiapas: Tzotzil, Tseltal, and Chol.(2) All three languages are members of the same branch of the Mayan family, and all three are, in different degrees, "endangered" in the sense that some communities of speakers have begun not to impart the Indian language to children, with many families preferring instead that their children learn Spanish first and the native language only second, if at all.

The Archive project involves the compilation of texts, lexical, discursive, and dialect databases, and audiovisual recordings of interaction in a wide variety of social and ethnographic settings. Equally important is the education and involvement of native speaking linguists, who not only enrich the database by contributing to the selection of "representative" material from different speech contexts in their communities, but also become active agents in the continuation and extension of the database in the future. We selected Internet access (supplemented by traditional and electronic publishing in various formats) as the appropriate means for distributing archival results both because of the inherent flexibility and power of electronic representations, and to insure the widest possible access in the future. (Current versions of the Archive can be consulted at and


2. A traditional view

My topic is the changing linguistic repertoires in modern Chiapas - especially in the relations between these three major languages, their dialects, and the dominant national language Spanish - in the face of other sorts of rapid, indeed cataclysmic, social change. Let me begin with a traditional representation of the demographic and geographic situations of the languages involved: a language map.

Figure 1: Traditional language map of the most widely spoken Indian languages of Chiapas

Here the areas in which Tzotzil, Tseltal, and Chol are spoken are shown as partially overlapping and partially contiguous colored areas on a map of the southeastern part of Mexico, principally the state of Chiapas. The number shown for each language corresponds to approximate numbers of speakers as reported by the Mexican national census of 2000 (INEGI 2000). It can be seen that both Tzotzil and Tseltal have more than a quarter million speakers each, and Chol slightly more than half that number. This simple graphic representation of the languages in question invokes several traditional assumptions which require critical re-examination.

When modern linguistic and ethnographic studies began in Chiapas, in the late 1950s and early 1960s, a vision shared by almost all researchers held that language, territory, culture, and identity in Indian Chiapas were roughly coextensive: a Zinacantec, for example, was a person who spoke the local Zinacantec dialect of Tzotzil, who lived in the geographic confines of the township of Zinacantán, who participated in the central institutions of Zinacantec society (the system of religious offices, for example, or the practice of typical Zinacantec occupations such as milpa farming, transport, flower cultivation and selling), and who dressed in distinctive hand woven Zinacantec clothes. This same vision persists today in institutionalized and bureaucratic ideas about etnias or "ethnic groups," usually construed on linguistic grounds. On such a view it is largely the unifying power of a "language" which gives its corresponding "ethnic group" its supposed identity and cohesion.

To be more explicit, the following assumptions are involved in this traditional view. (1) First is an assumed correspondence between language, territory, and culture-embodied in the Spanish term etnia or "ethnic group" common in both official and non-official discourse. That is, a language map suggests that in a single contiguous expanse of territory can be found the speakers of a particular language; this in turn is consistent with the notion that the group of speakers thus defined is culturally homogenous, and therefore that the coincidence of language, culture, and space constitutes some sort of relevant ethnic identity (whose numerical strength can thus also be estimated on the basis of the total number of speakers of the corresponding language). Other elements of the traditional view of Indian languages in Chiapas include (2) a distinction between Indian municipios or townships, administrative units whose inhabitants are largely if not exclusively "Indians" - that is, speakers of Indian languages - and ladino or non-Indian townships (such as the regional market center, the Colonial town of San Cristóbal de las Casas). With respect to these Indian townships, there is the further assumption (3) that just as the community has a distinctive dialect of its language (for example, a distinctive Chamula or Zinacantec variety of Tzotzil corresponding with the municipios of San Juan Chamula or Zinacantán), so will it have at least traditionally other distinctive ethnic trappings, such as a characteristic costume (whether still widely worn or not), and probably relevant features of a social system, such as a hierarchy of civil or religious cargos or traditional community positions of authority with associated ritual obligations. Going along with a marked local "dialect" are also metalinguistically available stereotypes (often lexical) of dialect/geography: in Zinacantec Tzotzil one might, for example, hear the word volje 'yesterday,' whereas in the Chamula dialect speakers would use the metathesized vojle. Such stereotypes are easily and frequently conjured by speakers themselves as marks of dialect difference.


3. Community linguistic profiles

We may begin to deconstruct this traditional model with a series of observations. First, a more detailed map of the modern day locales where speakers of different Indian languages are reported to live shows that such speakers are not limited to townships "traditionally" considered to be "Indian." (See Figure 2 on next page). In fact, in the case of Tzotzil, now the most widely spoken Indian language of the state of Chiapas with officially almost 300.000 speakers of five years of age or older, the four Chiapas townships with the largest absolute numbers of Tzotzil speakers include three traditionally "Indian" communities, Chamula, Zinacantán, and Chenalhó. However, in second place is the central ladino or non-Indian town of San Cristóbal de las Casas, its outlying barrios or neighborhoods transformed into poor Indian shantytowns.

Figure 2. Speakers of Indian languages by locality (1990)

Figure 3 shows the absolute number of Tzotzil speakers in the townships traditionally thought of as "Tzotzil communities," and all those other townships with comparable or greater numbers of Tzotzil speakers.

(Each bar in the graph shows first the number of monolingual Tzotzil speakers, then the number of Tzotzil speakers also bilingual in Spanish. To complete the graph, each bar then also shows the number of reported Tseltal and Chol speakers in each municipio.) Non "Indian" communities on the resulting list include not only San Cristóbal but also the state capital, Tuxtla Gutiérrez, as well as two other communities normally thought of as non-Indian towns: Ocotzocuautla and Teopisca.

Fig. 3. Municipios with the largest numbers of Tzotzil speakers, 2000.
(Source: INEGI.)

Figures 4 and 5 show comparable population distributions for Tseltal and Chol, with total official populations of speakers 5 years of age and older at around 280 thousand for Tseltal, and just over 140 thousand for Chol.

Fig. 4. Municipios with large numbers of Tseltal speakers, 2000. (Source, INEGI.)

The centers of large populations of Chol speakers continue to be the "traditional" townships of Tila, Salto de Agua, Palenque, and Tumbalá, although there are significant pockets of Chol speakers in Ocosingo, Huitiupán, and Yajalón.

Fig. 5. Municipios with large numbers of Chol speakers, 2000. (Source, INEGI.)

In the case of Tseltal, in addition to such "traditionally" Tseltal communities as Cancuc, Tenejapa, Oxchuc, and Chilón, the greatest number of speakers are spread across the immense municipio of Ocosingo, which includes the once sparsely populated Lacandón rainforest. Again, a large number of speakers (mostly bilingual, according to the census) live in San Cristóbal. There are also large proportions of speakers of Tseltal in communities ordinarily thought of as Tzotzil townships - Pantelhó, Huistán, and Venustiano Carranza, for example - and in others normally thought of as Chol townships, such as Palenque and Tila (where the Tseltal community of Petalcingo is administratively located). This distribution reflects a large scale outmigration of Tseltal speakers from their traditional communities in recent decades.

Two obvious conclusions can be drawn from the census statistics: (1) an important proportion of the Tzotzil, Tseltal, and Chol "ethnic groups" now lives in townships traditionally considered "not Indian"; and (2) there are many communities (or at least political entities at the level of municipio or township) where more than one Indian language is spoken by significant numbers of people. Specifically, there are communities with a large populations of both Tzotzil and Tseltal speakers, and others with significant populations of speakers of both Tseltal and Chol. There are significant numbers of speakers of the three most important Indian languages of the state in the regional center of San Cristóbal de las Casas, once considered to be a thoroughly ladino town, at least after dark when Indian buyers and sellers at the market had gone home to their own villages.


4. Changing relations between Spanish and Indian languages

Let me now turn to my principal theme, the changing social relationships among languages of the region and their dialects. It is perhaps easiest to start with Spanish and its intermingling with Indian languages. I shall concentrate on relationships between Spanish and Tzotzil, the case I know best. Over the five centuries of contact between Spanish and the Indian language, one can distinguish several strata in the linguistic sediments. For example, there are (1) contact era borrowings into Tzotzil that survive only in archaic and ritual forms of speech. From the middle of the last century comes (2) the characteristic "Kastiyero" (< Spanish castillero 'speaker of Castilian') Spanish that grew out of different kinds of commercial relations between Indians and ladinos. More recently there appear (3) new kinds of code mixing in the speech of a new generation of bilinguals.

With regard to (1) consider the archaic Spanish in Tzotzil ritual couplets. Like other Mesoamerican languages, Tzotzil has a rich generic tradition of ritual speech which exhibits characteristic parallelism (Haviland, 1996). In prayer, in formal declamation, in song, even in certain kinds of argument, Tzotzil speakers arrange their words into closely matching parallel lines, differing often only in doublets and sometimes triplets of words or phrases that incorporate what has been called a "stereoscopic" cultural imagery. The genre appears to be the verbal expression of power, efficacy and affect: Tzotzil words arranged into ritual parallel structures can bring about changes in the state of the Tzotzil universe. Within the fixed though vast repertoire of ritual doublets are many lexico-cognitive survivals: archaic Tzotzil words which occur now only in ritual speech, many Spanish "mates" for native Tzotzil words, including archaic Spanish no longer in common use.

Here, for example, is a fragment from a shaman's prayer at a curing ceremony. The prayer was designed to cure a patient from recurrent fainting spells. Spanish loanwords are shown in boldface:

(1) Curing prayer

1. chkom o ti yo latoje // chkom o ti yo lakantela
Your lowly pine will remain // your lowly candle will remain.
chkom o ti yo lajelole // ti yo lalok'ole
There will remain your lowly substitute // your lowly replacement.
mu teyuk nox ti ipe // mu teyuk nox ti k'ux une
May the sickness not merely be there // may the pain not merely be there
ja` jch'unoj o tal k'op uk // ti mantal uke
I have come obeying the word // the order
mi lekil ch'ambil // mi lekil nupbil
Will it be well received? // will it be well met??
ti yo stoj une // ti yo lakantela une
Your lowly pine // your lowly candle?
ti yo lajelol une // ti yo lalok'ol une
Your lowly substitute // your lowly replacement
ja` xa me chaj-ja` xa me chajk'opon o // chajti`in o
For that reason I talk to you // for that I speak to you
mi xakejan aba // mi xapatan aba
Will you kneel // will you prostrate yourself
mi xa`ok' // mi xa`ava:n
Will you weep // will you cry?
ja` me jta o ti ta-ti yolon yoke // ti yolon sk'obe
Thus will I (we?) come beneath his feet // beneath his hands
ti isak' ch'ul maretik // isak' ch'ul ajvetik une:
of Isak' holy oceans // of Isak' holy lords
k'opojkotik un // ti`ijkotik u:n
Let us talk // let us speak
jk'antik i pertonal
let us ask for pardon

A closer look at a couple of lines reveals the structure here. The fragment begins with the curer's reference to the offering of candles she is leaving in front of the crosses where she prays. Instead of simply saying "candle" - the Tzotzil word is kantela, a clear loan from the Spanish candela (no longer in use in Chiapas Spanish, where instead the modern word is vela) - she produces a standard doublet, combining the Tzotzil toj 'pine' (as in 'pine tree') with the loanword kantela.

ch-kom o ti yo l-a-toj-e // ch-kom o ti yo l-a-kantela
ASP-remain REL ART humble ART-2E-pine-CL // ASP-remain REL ART humble ART-2E-candle
Your lowly pine will remain // your lowly candle will remain.

The two parts of the line are exactly parallel, with only the substitution of the second word of the doublet for the first. Both of the words for candle appear with a 2nd person possessive prefix a-, a reference to the fact that the candles are offerings to the curer's official addressee - ancestral guardian spirits. Line 14 is a reference to the location where the curing ceremony is taking place, a sacred cave called Isak'tik "Potatoes." The doublet the curer uses to refer to the cave combines the name Isak' with two images of power: mar (a loan from Spanish meaning 'sea') and the pan Mayan root instantiated by the Tzotzil ajv 'lord, owner' (cf. Tzotzil kajvaltik, lit. 'our lord,' i.e., "God").

A much later form of interaction between Spanish and Tzotzil can be seen in the speech of a well known "kastiyero" or speaker of Castilian. The man in question had learned his Spanish in the context of commercial interactions with the non-Indian world as a youth, and later in his life was a much sought after "Tzotzil lawyer" - that is, a Zinacantec intermediary in legal dealings in which fellow Zinacatecs needed help in negotiating the often bewildering Mexican legal bureaucracy. I will refer to this man by his nickname, jtubtavanej "the spitter" - a joking reference to the common opinion that his speech was a rapid and almost incomprehensible mixture of Tzotzil, Spanish, and spit.

In fragment (2), recorded in 1988, "the Spitter" (shown as "b") is telling another elder Zinacantec about land squabbles. Again, I have put the Spanish loan expressions (and the corresponding English translations) in the Spitter's performance (and that of his interlocutor, shown as "p," who otherwise is left little floor space) in boldface. The Spitter's nephews have been complaining that they have not received a large enough share of the inheritance due them from their grandfather, the Spitter's father. In describing how he and his brothers decided to try to resolve the issue by dividing remaining land into five equal portions (line 3), he demonstrates his characteristic form of code switching: whole phrases in Spanish, some apparently almost pre-formed although only semi-grammatical, interspersed through a wider Tzotzil matrix. Both he and his interlocutor standardly use loan interjections and connectives (cf., eso in line 4, or b's porque in line 7), even when the rest of the utterance is Tzotzil, but the Spitter goes considerably farther.

(2) The "spitter"

1 b: ora li buch'u xa muk' bu xch'ake che`e
2 ijch'akkotik ta komun
3 en cinco parte la-
The ones who weren't given any land, we have now divided it in five parts
4 p: eso
5 b: ikom lavie
6 sé que un día li`one
I know that one day, I am here
7 porque lavie, oy kosiltak
Because now, I have my lands
8 para qué voy a estar pegando con mi hermano
Why am I going to be fighting (lit., hitting) with my brother?)

It seems clear that the Spitter is showing off his Spanish competence, as a member of a generation of people in which fluency in the non-Indian language was an asset few Zinacantecs commanded. In this conversation with another distinguished but non-Spanish speaking dispute settler (the two men were, in fact, waiting for one of the disputants in a land quarrel to appear so they could begin a settlement procedure), there was little or no referential need for such code switching, something that makes the Spitter's semi-grammatical displays all the more marked. (In line 3 in the following segment, he misconjugates the verb, and in line 5 the adjective casadas 'married (plural, feminine)' fails to agree with its implied subject in either gender or number.) There is, indeed, a certain irony in his message here: he argues that his nephews are not really entitled to the land they have inherited, since they have declared an implicit allegiance to the non-Indian world (living outside the Indian community in Tuxtla, and married to non-Indian women); but he chooses Spanish to make his point.

(3) The "spitter" part II

1 b: ora li skremotik une
now, his sons...
2 k'u cha`al chkal e
as I say
3 ya vivan en Tuxtla
they live in Tuxtla
4 y ora
and now
5 es casadas con la Juchiteca
they're married to non-Indian women
6 una es de Comitán
one is from Comitán
7 p: cara:jo
8 b: ora k'ajom xa jun skreme
9 batz'i te x`i`o:n tzsa` k'op
now just one son is making trouble

Let me jump ahead to 2003. Although I cannot catalogue here the full range of socio-political changes that characterize highland Chiapas in the first years of the new millennium - and especially since the Zapatista rebellion of 1994 - it is obvious to even the most casual observer that processes of profound social upheaval abound. Among other things, the Colonial town of San Cristóbal has been transformed from a non-Indian market center, with a transient Indian population frequenting the daytime markets but returning to their villages after dark, to a socially and ethnically mixed metropolis where multilingual street vendors who live in poor shantytowns on the edges of the city hawk their wares to European tourists in their own languages and speak fluent Spanish to Mexican tourists, but confer with one another in Tzotzil.

(4) Spanish Tzotzil mixing in radio talk, 2003.

One symptom of the newly re-Indianized society of modern highland Chiapas is "Indian radio."

1 mi o to buch'u la spas llamar a ver mi stak' xjelav al aire,
if someone wants to call, let's see if the call will go on the air
2 porque a veces tzpas fallo jutuk,
because sometimes there is a small failure
3 sk'an vaklajuneb minuto sk'an xk'ot ta vo'ob hora,
4 únicamente . dieciseis minutos para las cinco de la mañana,
it is just 16 minutes before 5
5 lek oy amigos amigas
Ok, (male & female) friends
6 chabanukoxuk, junuk avo'ontonik amigos
Greetings, be well, friends

One of San Cristóbal's two radio stations broadcasts in Indian languages in the early hours of the morning, and the main locutores or announcers are well-known personalities throughout the Indian highlands, spoken of as familiar friends and celebrated for their linguistic abilities in even remote villages. The most popular announcer, Mateo, is a native speaker of the Chamula dialect of Tzotzil. His Spanish is also quite fluent; indeed, his Tzotzil is mixed with so much Spanish that it is almost a special dialect in itself, as the following illustrative fragments suggest.

The integration between the two languages in this performance is much more complete than in the case of the code switching in the Spitter's speech, and at the same time the radio announcer's reliance on Spanish lexical resources is greater. In addition to set Spanish phrases (such as al aire 'on the air' in line 1) imported directly into the overall Tzotzil frame, and in addition to the ubiquitous Spanish loan connectives, the speaker imports Spanish verbs by combining their infinitive or nominal forms with the appropriately inflected Tzotzil pro-verb pas 'do' (line 1 pas llamar 'do calling, i.e., call', and line 2 pas fallo 'do a failure, i.e., fail'). He similarly observes contemporary political correctness by importing not only the Spanish word "amigos" for his virtual addressees, but both its explicit gender specific halves.

(5) Morphological mixing in radio talk, 2003

This announcer also produces a new sort of morphological mixing between the two languages.

1 jechal amigos amigas, amigos indígenas,
so, friends (m&f), Indian friends
2 amigos ta kampo, pues kolavalik
friends from the fields, well thanks
3 buch'utik vo'oxuk chatak tal akartas,
to whoever has sent your letters
4 chatak asaludo,
or sent your greetings
5 ja' ta jk'antik amigos,
that's what we want, friends
6 lek oy, junuk avo'ontonik
that's good, be happy
7 takik tal akartas, takik tal asaludo
send your letters, send your greeting(s)

He deploys fully calqued Spanish constructions, like amigos ta kampo (line 2) whose only nod to Tzotzil is the all-purpose Tzotzil preposition ta 'in, on, at, from, to&ldots;' Moreover, he seems willing to combine fully inflected Spanish plurals like kartas (=carta-s 'letters') with the possessive prefixes of Tzotzil, here the second person a- (though he eliminates the Spanish plural formative on saludo 'greeting'). Little wonder that many non-Indian monolingual Spanish speakers congratulate themselves that they can "understand Tzotzil" on the basis of listening to a few minutes of early morning "Tzotzil" radio.


5. Changing relations between Indian languages themselves

Let me now turn to the more subtle accommodations between different varieties of Indian languages themselves. In pre Conquest times, the highlands of what is now Chiapas and its immediate environs were characterized by varying degrees of contact between at least three distinct major language families (Mayan, Mixe-Zoquean, and Oto-Manguean). Despite the monoglot view of Indian communities, mentioned at the outset and fostered by post-Conquest resettlement policies, even in the last century Indians in many townships traveled widely across routes that brought them into regular contact with speakers of other languages, some distant and some close to their home tongues. However, in recent decades highland Chiapas Indians have largely been monodialectal, clinging to the distinctive ways of speaking in their hometowns much as they clung to the traditional handwoven costumes of the same place. Now, as these same costumes give way to the anonymous polyester and bluejeans of mass produced clothing, Indian dialects, too, begin to transform themselves. Once again, let me survey different eras of multilingualism and multidialectalism among Indian languages.

First, it is possible to identify what might be termed "traditional" monolingual accommodations - those small adjustments that speakers of different languages and dialects with an essentially monolingual repertoire would make to each other in the neutral, courteous interactions circumstances might force upon them. A different situation is that of people whose work trajectories (or perhaps contacts through marriage) engendered partial polyglotism in Indian languages. A new kind of multilingualism from the last three decades has also emerged in recent settlements, particularly in the Lacandón rainforest, which were multilingual from their foundation. Such communities grew up where families from different Indian townships with different languages joined together to settle newly opened rainforest areas. Finally, the cultural merging and homogenization in modern urban life in places like San Cristóbal de las Casas have characteristic linguistic expressions. I will take each phenomenon in turn with brief exemplifications.

Consider first what I have called "traditional" monoglot accommodation. In circumstances in which speakers of different Indian languages or dialects came together in the past - in casual encounters on the path, or in the market, for example - when interactive demands exceeded mere formulaic exchanges, speakers might accommodate to each others' lexicons but maintain a strict phonological or morphological adherence to their own dialects. For example, in conversational fragment (7), a Tzotzil speaker from Chalchihuitán (shown as x) is telling another Tzotzil speaker from Zinacantán (p) about an encounter with a supernatural demon who had adopted the shape of a goat. The two dialects are rather divergent, and one notable morphological difference between them is the 1st person plural absolutive marker. In Zinacantec Tzotzil this is suffixed to the verb stem, whereas in Chalchihuitán it is prefixed. (See 6) As a result the forms have superficially quite different shapes and sounds.

(6) 1st person plural inclusive form of cham 'die'


In the exchange (see lines 104-105), each speaker simply uses his own form, and the morphological "translation" is made swiftly and effortlessly.

(7) Supernatural goat (Chalchihuitán and Zinacantec Tzotzil)

97 p: e pero:.. le`e:
But those things (demon-goats)
98 ja` nox chk'ojvane
They just gore you
99 mu nan xmilvan a`a
They don't kill you, indeed.
100 chk'ojvan nox
They just gore you.
101 x: a mu xmilvan
Ah, they don't kill you.
102 p: mu xmilvan
They don't kill you.
103 x: ji`i
104 p: pero bueno chichamotik yu`un
But, well, we can die from it.
105 x: chijcham mi lijxi`e:
We can die if we get frightened.

Further evidence of accommodation is provided by the backchannel (Yngve 1970) that Tzotzil interlocutors offer one another in polite interaction, characteristically by "repeating" parts of previous utterances as part of proper interactive listening (Haviland 1988). The repetitions show that interactants are parsing (and thus following) what has been said. They also indicate adjustments to the other's dialect (as well as failures to adjust). Once again the topic is a witch in animal form. C, a speaker of Chamula Tzotzil, tells P, a Zinacantec, about advice he was given by his father about what to do if he met such a creature on the path. The advice was: "shoot it straight away," and this advice is recounted in line 2, repeated almost verbatim in line 2 by the interlocutor, and then repeated again by the original narrator in line 4 (who is in turn acknowledged by the interlocutor in line 5) - a highly typical pattern.

(8) "Buried" (Chamula and Zinacantec Tzotzil)

1 c: k'usi chanup ta bee
"Whatever (animal) you meet on the path
2 ak'bo me tuk' xi
"Shoot it," he said.
3 p: ak'bo tuk' xi
4 c: ak'bo tuk' xi
5 p: bweno
6 c: va`i un
So listen,
7 te nox ch'ay . ta . yut lume
"It will just disappear underground.
8 ak'bo tuk' ta banamil
"But shoot it anyway, underground.
9 te . mukul ta banamil chata xi
"You'll find it buried underground," he said.
10 p: te mukul ta banamil xi

Of somewhat more interest is the Zinacantec's lexical accommodation in line 10, which again repeats verbatim the Chamula narrator's turn in line 9. The crucial fact here is that whereas Chamulas say banamil 'earth', Zinacantecs insist on balamil, and in fact this is one of the highly stereotyped lexical features that the two dialects use to characterize the other. (In imitating Chamula speech, for example, a Zinacantec will jokingly use the Chamula pronunciation banamil.) Here, however, in a polite exchange the Zinacantec adopts Chamula usage as part of what the Zinacantec involved, P himself, explained to me was common courtesy. Both lexical accommodation and morphological resistance can be observed in exchange (10) between the Chalchihuitán man and his Zinacantec interlocutor. The subject is the former's father, whose adventures on return from work in the lowland coffee plantations, or fincas (in Tzotzil pinka), are being recounted. The relevant differences between the two dialects are a lexical difference between the inchoative denominal verb meaning 'get old' and completive verb morphology, as illustrated in (9).

(9) Lexical and morphological variants

Chal. malub 'get old' = Zin. molib
Chal. la-s-ta COMP-3ERG-find = Zin. (i)-s-ta

In the conversation - see lines 3 & 4 - the Zinacantec interlocutor courteously adopts the Chalchihuitán lexical item malub (which does not even exist in his dialect with the expected meaning), but he resists, whether consciously or unconsciously, the morphological difference in his backchannel repetition of line 8 at line 9.

(10) Getting old on the fincas: Chalchihuitán and Zinacantec Tzotzil

1 x: yu`un te sbe o te k'al chk'ot ta pinka
That was his route for going to the fincas.
2 p: te chk'ot to ta pinka
That was his route?
3 x: te malub tal ta pinka
He got old at the fincas.
4 p: te malub tal
He got old there?
5 x: ji`
6 yich'oj tal xa . 40 50 anyo sut tal ta pinka
He was 40 or 50 when he returned from the fincas
7 p: jijola mol xa
Very old
8 x: te lasta tal yajnil
He brought his wife from there
9 p: aa te sta tal yajnil
Ah, he brought his wife from there?

Such examples illustrate graphically an apparent difference in the relative levels of ethnolinguistic awareness (Silverstein 1981) of lexicon and morphology on the part of native speakers, a difference which in turn contributes to the nature of interdialect accommodation.

For reasons of space I will not present transcripts of the speech of the sorts of new multilinguals whose linguistic skills result from individual work experiences or the new polyglot settlements mentioned above. Nonetheless, I will cite a couple of anecdotal exemplars. Among my colleagues at CIESAS are Petul, from Tenejapa, and Manvel, from Larraínzar, both of whom are virtually bilingual in Tseltal and Tzotzil (though the native languages of each are different) because of work experiences that have brought them into prolonged contact with the other speech community. More striking is the case of Pascual, from Col. Zapata in the Lacandón rainforest, a quintalingual Chol speaker. Born into a Chol family, he grew up in a "new" community settled in the 1960s by migrants from both Chol and Tzotzil traditional communities, acquiring both languages "naturally." He then attended primary school in a neighboring new settlement whose founders spoke Tseltal, and afterwards did secondary school in a Tojolabal community near Comitán, in the process acquiring fluent Spanish.

I shall return for my final exhibits to the Indian language radio program. In the previous examples we observed the Spanish and Tzotzil mixing of a semi-bilingual talk-show host, Mateo, a native speaker of Chamula Tzotzil who receives phone calls in various Indian languages, including other dialects of Tzotzil, as well as Tseltal, Chol, and sometimes Tojolabal, and also Spanish. Consider the following fragments of a phone call Mateo receives from a speaker of Tenejapa Tseltal. In the early stages of the conversation, there is a complete lack of accommodation on the announcer's part: he simply continues to speak in Tzotzil while the caller speaks in Tseltal. (Because the languages are close sisters, the apparent differences are small - focused in these small samples of simple communication on the lexicon, and some major phonological and morphological differences.) In the fragment, V - the Tenejapan - speaks Tseltal mixed with Spanish, and M - the Chamula - his mixed variety of Tzotzil. The insistence on speaking ones own language persists even in the repetitive feedback given by one to the other. For example, in lines 4 and 5, V asks in Tseltal if they are "on the air" using the Tseltal existential ay 'exist,' and M replies in Tzotzil using the Tzotzil equivalent oy.

(11) Opening of a Tzotzil/Tseltal talk-show phone call

1 V: ja'at wan Mateo.
Is that you, Mateo
2 M: ¿k'usi chaval?
What do you want to say?
3 vo'on amigo.
It's me, friend.
4 V: bueno ay wan ta al aigre Mateo.
Are we on the air, Mateo?
5 M: a claro que sí oy al aire.
Of course we are on the air.
6 V: yu'un ay bi ya kalbeel . Mateo.
There's something I want to tell (her), Mateo
7 M: aja, k'usi chaval o
Aha, what do you want to say?

However, by the time the short conversation comes to an end, Mateo assays a final (formulaic?) switch to Tseltal, or at least pidgin Tseltal. In fragment (12) Spanish loans are shown in boldface and markedly Tseltal expressions are shown in small caps. (An unmarked font corresponds to Tzotzil or to words which could be either Tzotzil or Tseltal.)

(12) Closing of a Tzotzil/Tseltal phone call

1 M: bueno
2 bin abiil te jiche amigo?
What's your name, then, friend?
3 V: Jo'on A. G. L.
I am A. G. L.
4 teun ta banti komunidad Tsakibil jok'e.
I am in the community of T.J.
5 M: aay, muy bien. bueno, lek ay jichuk.
Ahh, very good, ok, good.
6 V: jichuk ich'a asaludo Mateo.
Ok, receive a greeting, Mateo
7 M: bueno.

As can be seen, Mateo asks his interlocutor for his name, in line 2, "politely" switching to Tseltal to do so. The caller never departs from his standard Spanish-mixed Tseltal. However, Mateo goes on to use a couple of characteristic Tseltal words in his pre-closure at line 5. (The Tzotzil equivalent, in Mateo's native dialect, would have been lek oy jechuk 'it's good like that.')

My impression is that Mateo, probably the most widely known and listened to highland Chiapas Indian, has a limited command of Tseltal, but what he knows he deploys in this characteristic fashion, welcoming his callers, as it were, by welcoming their ways of speaking and appropriating them to himself. By the same token, by his example there is set a new standard for "excellence" or "mastery" in Tzotzil, one that includes polyglot accommodation to other Indians linguistic codes.


6. Trilingual literacy and changing linguistic repertoires in modern Chiapas

Let me end with two final vignettes, to illustrate the directions that language change and dialect accommodation are taking. When I began learning Tzotzil, nearly forty years ago, I imagined neither that one day, in my own lifetime, I would be envisioning the "death" of the language, nor that it might be displaced not only by Spanish but even by, say, English, nor that the distinctive Zinacantec way of talking that I was taught would begin to compete with other, equally distinctive Indian linguistic codes. But the pace of language change, in a Chiapas beset by revolution, neo-liberalism, and mass migration, is dizzying.

I have begun to study the new kinds of Tzotzil that are emerging among modern, urban speakers who live exclusively in town rather than in their communities of origin (or the communities of their parents), who intermarry across township and language boundaries, and who mix dialects willy-nilly in everyday speech on the job, in the market, and in school. One family of four, who run a market stall in the new "popular" market on the south edge of San Cristóbal, consists of a husband from Zinacantán to the west, a wife from Huistán to the east - homes to two widely divergent Tzotzil dialects (Delgaty & Ruíz Sánchez 1978) - and two children born in San Cristóbal, a girl about 8 and a little boy about 4. The market stall is sandwiched between other Tzotzil speaking vendors from Chamula and Zinacantán. The little girl is hardly willing to utter a single word in Tzotzil, preferring to speak Spanish at least in public. The little boy speaks Tzotzil, and spends most of his days playing with a contemporary from the neighboring Chamula vendors. As a result he is multidialectal in Tzotzil, and it is hard to know, as his aunt remarked to me, "what kind of Tzotzil he speaks." Nonetheless, his attitudes are not left to chance. His Zinacantec father, demonstrating to me that the boy knew both "our" word for "seated" (since my own Tzotzil is markedly Zinacantec), namely chotol, and also his mother's dialect equivalent nakal explicitly asked the kid to choose "which one is better." Predictably, the boy chose his father's word, though it seems inevitable to me that he will end up speaking the composite "Standard Average Chamula" which seems to be the de facto lingua franca of the market.

Finally, to show how linguistic fieldwork can come full circle, consider the following recent e-mail from a native Tzotzil speaker from the northern municipio of Chenalhó now resident in Canada, addressed to Robert M. Laughlin, distinguished Tzotzil lexicographer. This remarkable short text gives clear evidence of a combination of Tzotzil, (Spanish,) and English competences, both spoken and written, and it also suggests how modern technologies of language are extending to highland Chiapas.

"But this is the reason that I am sending you this message, my brother has asked me to develop/create a computer program which will be a spell check for Tzotzil language. He has been involved for many years with other linguists working on the Tzotzil Language. I know that you have had written dictionaries or help to, the reason that I am asking you this is that if I have to input 30, 40 thousand words into the computer I much preferred to copy and paste if there is already something on electronic form. Any help that you can provide me with I will really appreciate. Kolabal. In peace."

Aside from a certain orthographic doubt (the final voiced fricative in what I would write as kolaval 'thank you' is labio-dental, unlike the bilabial fricative represented by b in standard Spanish orthography) and some minor infelicities in the English, the message is a stunning testimony to the changing linguistic horizons in Indian Chiapas.


7. Questions and conclusions: the swirl of tongues

In the present context of diaspora, expulsions, armed conflict, refugees, and an economic crisis without relief in Indian Chiapas which accelerates Indian urbanization, the situations of contact and admixture among Chiapas Indian languages continue to grow. The processes of linguistic accommodation I have begun to note here are just one symptom of a more global social-historical process - and as I have noted, the symptom is not new. The emergence in San Cristóbal (and surely in other Chiapas cities as well - there was at least one linguist studying urban Tzotzil in Tuxtla Gutiérrez, the state capital, in 1998) of Indian lingue franche is a phenomenon of great social as well as linguistic interest. The notes I have presented here about contact and accommodation between dialects and languages in situations less extreme than those found in modern Chiapas cities, suggest a division between those linguistic features relatively easy to accommodate - most obviously, stereotyped lexical diagnostics - and those features that resist such accommodation, such as verbal or pronominal morphology in the present cases. (As I have already mentioned, the continuum is reminiscent of a similar range of more or less "conscious" metalinguistic features described long ago in Silverstein 1981.) Nonetheless, the details of the development of new linguistic and communicative resources requires a much more detailed study that goes well beyond my own limited observations. A central question, not at all resolved in my own mind, is whether the new "interlanguages" of Chiapas previsage the emergence of a new consciousness, identity, and corresponding set of linguistic standards for a pan-Indian (or pan-Mayan) consciousness, or whether, on the other hand, they simply signal a brief intermediate phase in the inevitable extinction of the native languages of Chiapas.

© John B. Haviland (Ciesas, Mexico / Reed College, US)


(1) Oral versions of this paper were presented at the Seminario Interno, CIESAS-Sureste, Chiapas, Mexico, and at the Annual meetings of the American Anthropological Association in November 2004. I am indebted to colleagues at those events for comments and criticism.

(2) I follow orthographic conventions for language names, and in general for practiocal orthographic representations of the languages cited, established by indigenous Chiapas writers.


Delgaty, Alfa H. and Agustín Ruíz Sánchez (1978): Diccionario Tzotzil de San Andrés (Larraínzar) con variaciones dialectales. México, DF: Instituto Lingüístico de Verano.

Haviland, John B. (1981): Sk'op Sotz'leb: El Tzotzil de San Lorenzo Zinacantán. Mexico City: Universidad Nacional Autónoma de México.

Haviland, John B. (1986): "Con buenos chiles: talk, targets, and teasing in Zinacantan." Text, 6(3):249-282.

Haviland, John B. (1988): "A father-mother talks back: the micro-creation of context in Tzotzil." Paper presented to the workshop "Contextualization of Language," Konstanz, October 1988.

Haviland, John B. (1990): "The grammaticalization of space (and time) in Tzotzil." Working Paper Number 2, Cognitive Anthropology Research Group at the Max Planck Institute for Psycholinguistics, Nijmegen, the Netherlands.

Haviland, John B. (1996): "We want to borrow your mouth": Tzotzil Marital Squabbles." In: Briggs, Charles L. (ed.): Disorderly discourse; narrative, conflict, and inequality. 158-203. New York, Oxford: Oxford University Press.

Haviland, John B. (1997): "Owners vs. bubu gujin: Land rights and getting the language right in Guugu Yimithirr country." Journal of Linguistic Anthropology, 6(2):145-160.

Hopkins, Nicholas A. (1970): "Estudio preliminar de los dialectos del tseltal y del tzotzil." In: McQuown, Norman A. / Pitt-Rivers, Julian (eds.): Ensayos antropológicos en la zona central de Chiapas. 185-214. Mexico: INI.

INEGI (2000): Instituto Nacional de Estadística Geografía e Informática, XII Censo General de Población y Vivienda, 2000, Mexico, D.F.: INEGI.

Kaufman, Terrence S. (1962): "Tseltal-Tzotzil comparative grammar, Part I: Phonology." In: Microfilm Collection of Manuscripts on American Indian Cultural Anthropology. Series IC, No. 55:125-310. Ms., University of Chicago.

Laughlin, Robert M. (1975): The Great Tzotzil Dictionary of San Lorenzo Zinacantan. Washington, D.C.: Smithsonian Institution.

Silverstein, Michael, (1981): "The limits of awareness." Working Papers in Linguistics, No. 84: Austin: Southwestern Educational Laboratory.

Viqueira, Juan Pedro (1998): "El lento, aunque inexorable, desmembramiento del señorío de Zinacantán." Ms., CIESAS-Sureste (SCLC, Chiapas, Mexico).

Yngve, V. H. (1970): "On getting a word in edgewise." In: M.A. Campbell (ed.), Papers from the Sixth Regional Meeting, Chicago Linguistics Society. 567-78. Chicago: Dept. of Linguistics, Univ. of Chicago.

6.1. Standardvariationen und Sprachauffassungen in verschiedenen Sprachkulturen | Standard Variations and Conceptions of Language in Various Language Cultures

Sektionsgruppen | Section Groups | Groupes de sections

TRANS       Inhalt | Table of Contents | Contenu  15 Nr.

For quotation purposes:
John B. Haviland (Ciesas, Mexico / Reed College, US): Indians, languages, and linguistic accommodation in modern Chiapas, Mexico. In: TRANS. Internet-Zeitschrift für Kulturwissenschaften. No. 15/2003. WWW:

Webmeister: Peter R. Horn     last change: 22.6.2004     INST