1 Introduction

The purpose of this paper is to explore the changes that have resulted in Texistepec Popoluca (TX), a language of the Zoquean branch of the Mixe-Zoquean family of Veracruz, Mexíco, through Spanish influence. TX is a dying language, currently spoken only by older generations and not being learned by children. Because relatively little work has been done onTexistepec, and virtually none that deals specifically with language contact, this paper is intended to give an overview of some of the effects of long-term contact with Spanish on TX. I do not investigate any single aspect of language change (e.g., lexical change, syntactic change, etc.) in particular, but rather I present a variety of data that gives an overall impression of the nature of Spanish influence on TX.

The data consists of three stories told by R., who is in his seventies and bilingual, in Mexico in the summer of 1994. The first story involves a man whose wife does not keep house, nor feed him properly. His friend advises him that a piece of leather should be used to "correct" the wife. The second story is about a man (the señor) who has a wife and two lovers. He asks hisfriend (the compañero) to visit these women and pretend he is the señor (I use the Spanish terms to refer to the two men in order to distinguish between them). The women discover the deception and proceed to physically attack the compañero with various household objects. The third story deals with R's encounter with the devil and the ensuing trip to hell. This is supposedly a true story. The examples given in this article are taken from the first two stories; however, preliminary examinations of the third text suggest that it conforms to the generalizations made below.

2 Lexical borrowing

As is expected of a language under severe and long-term cultural pressure from another language, TX has borrowed a great deal of "content" words from Spanish. Borrowed nouns include tumeeñ 'money' (> Spanish tomines), kuc‡eeñ 'knife' (> Spanish cuchillo), and c‡apaatus 'shoes' (> Spanish zapatos). TX morphology is freely added to Spanish nouns; e.g., kuc‡eeñ-da÷a 'big knife' (knife + AUG), y-bac‡eete 'his knife' (3P + machete), raato-c‡e 'a little while' (little while + nothing more). TX morphology (e.g., person markers, aspectual markers) is not directly added to Spanish verbs, but maybe added to adjectives to create verbs; e.g., taarde-ham 'it's already late' (late+ CMP).

Spanish transitive verbs are borrowed into a construction consisting of the TX verb wat 'to do, to make' plus the Spanish transitive verb; e.g., wat lastimar 'to hurt', wat dominar 'to dominate', wat educar 'to educate', wat maltratar 'to mistreat'. While the construction itself is not unusual -- i.e., other Mesoamerican languages, such as Pipil (Campbell 1985), have similar constructions, the semantic nature of the verbs which are borrowed into this construction display an interesting pattern. That is, the Spanish verbs borrowed in this way typically involve an unequal power relationship. Ofcourse, to some extent all transitive verbs involve an unequal relationship (i.e.,one participant is acted upon, and the other participant is the actor); however, the particular verbs which participate in the wat construction reflect socio-cultural relationships in which one person somehow dominates another, whether that relationship involves physical domination ('to hurt', 'to mistreat'), or involves one person having authority over another ('to advise', 'to educate', 'to wait on'). The transitive verbs borrowed into the wat construction contrast with native TX transitive verbs, which do not necessarily display the above semantic characteristic. TX transitive verbs include bo÷n 'to cover', pik 'to grab', do÷ba¢ 'to turn on', ÷oota÷ 'to talk to someone', and huy 'to buy'.

The wat construction may exemplify what Hill and Hill (1986), in their investigation of Malinche Mexicano (a Uto-Aztecan language of the Nahua subgroup spoken in Central Mexico), term the "power code." That is, in indigenous Mesoamerican communities, such as TX or Malinche Mexicano, there often exists a choice between two codes -- Spanish, which represents power and authority, and the indigenous language, which represents solidarity. Certain contexts will lend themselves to the use of Spanish; e.g., government and law. In other, more domestic, contexts, the indigenous language may be used, but will be open to Spanish diffusions that symbolize power (cf. also section 5 for discussion of code-switching). Words which refer to relationships involving power and authority may therefore tend to be borrowed from Spanish. It must be noted that the hypothesis that borrowed Spanish verbs in TX display a special semantic characteristic is preliminary in nature; a more thorough investigation of transitive verbs in TX is necessary to confirm it.

3 Spanish syntactic constructions

A discussion of the influence of Spanish on TX syntax is limited by the lack of documentation of earlier stages of TX, as well as by the preliminary nature of comparative studies of Mixe-Zoquean syntax. An understanding of the effect of Spanish on TX syntax will progress concomitantly with comparative Mixe-Zoquean research. In the meantime, syntactic constructions which involve borrowed Spanish morphemes can be identified.

3.1 Subordinate clauses

TX subordinate clauses are often based on those of Spanish. Several Spanish subordinating conjunctions have diffused into TX, including temporal markers such as antes que 'before' and hasta 'until', the purposive marker para que 'so that', and the contrastive marker sino que 'but'. The Spanish gloss is that given by R.

(1) ma÷ y-put-a÷ piiñ ante ke y-wat-a÷-p c‡e÷ ÷e÷eñ

PST 3E-take.out-APP man before 3E-do-APP-INC thing some

the man took it out before doing anything

lo sacó el hombre antes que hacer alguna cosa

(2) ma÷ buh-k-togeñ n-kopik ata ma÷ ÷uk di÷

PST wet-enter 1P-head until PST drink water

it wet my head until I drank a little bit of water

se mojó su (mi?) cabeza hasta tomé poquito de agua

(3) ma÷ ny-bo÷n ny-kehke÷ paka wa÷ ky-÷evs‡pik

PST 2E-cover 2P-chin so.that NEG 2A-know

you covered your chin so that she wouldn't know you

lo tapaste su quijada para que no te conoce

(4) y ÷iñdye ma÷ dik y-ti˜ kuy

and NEG PST go 3E-cut.down tree

and he didn't go to cut down trees

y no fue a derribar

sino ke ÷iñdye ma÷ y-pik hepe ÷aac‡a

but NEG PST 3E-grab that ax

but instead he grabbed that ax

sino que agarró ese hacha

The TX clauses conjoined by the Spanish conjunctions appear to retain the syntactic configuration of TX.

3.2 Relative clauses

Relative clause formation still displays a native pattern; that is, the native relativizer -pu÷ is still used (5, 6). The Spanish relativizer que, however, is sometimes used in addition to -pu÷ (5). I have no examples of relative clauses which are only marked by que. Of course, this does not mean they do not exist. The relativizer -pu÷ is not the only native means by which headless relative clauses may be formed. The TX word c‡i?, meaning 'what, something' can introduce headless relative clauses (7).

(5) pero ÷a÷y y-wat-p ke n-dima÷-paa-pu÷

but if 3E-do-INC that 1E-say-NOM-REL

but if you do what I'm saying

pero si lo va a hacer que voy a decir

(6) porke ma÷ pu÷ kny-c‡e÷ buurru-baa

because PST REL 2-->l-give burro-DIM

because the burro that you gave me

porque burro que me diste

(7) ÷a÷m c‡i÷ ma÷ k-wat-a÷ tu amigo

look what PST 1A-do-APP your friend

look what your friend did to me

mira que cosa (me hizo) tu amigo

It is unclear whether (7) represents a native usage of the word c‡i÷ (translated as 'what?, some' in Wichmann, forthcoming), or whether it results from influence of Spanish que. The expansion of the usages of native elements based on Spanish que is known elsewhere in Mesoamerica. In particular, in two Nahua languages native complementizers have become true relative pronouns, based on parallels with Spanish que. In Malinche Mexicano, tlen is now a true relative pronoun, whereas in precontact Mexicano, tlen could only be used to introduce headless relative clauses (Hill and Hill 1986). In Pipil, a Nahua language of El Salvador, ka(h) (> Proto-Nahua *i:-ka) has shifted from a complementizer to which headless relative clauses had been juxtaposed to a true relative pronoun (Campbell 1987). However, in absence of evidence to the contrary, it is premature to assume that the use of c‡i÷ as a complementizer that introduces relative clauses is based on Spanish que.

4 Discourse markers

Hill and Hill (1986: 177) note the prevalence of Spanish discourse markers in Malinche Mexicano.

A striking feature of Mexicano is the proliferation of Spanish function words and particles, including discourse and conversational elements, logical connectives, conjunctions, prepositions, exclamations, and hesitation forms. Borrowing of Spanish particles has been reported for all modern varieties of the language.

The ubiquity of Spanish discourse markers is characteristic of other indigenous Mesoamerican languages as well (e.g., Tojolabal, Brody 1989; various dialects of Mexicano, Suarez 1977).

Like Mexicano, TX is rife with Spanish discourse markers. Here I discuss the function of several of these markers.

4. 1 The connectors y and pero

Spanish y occurs frequently within the TX texts, and most frequently expresses an additive relationship between idea units in TX.

(8) y-su÷k-a÷-p ny-pak

3E-kiss-APP-INC 2P-foot

she is going to kiss your foot

lo va a besar su pie

y y-su÷k-a÷-p ny-ki÷

and 3E-kiss-APP-INC 2P-hand

and she is going to kiss your hand

y lo va a besar su mano

The marker y can also signal a contrastive relationship between idea units.

(9) ma÷ pooy-put ma÷ dik

PST run-leave PST go

he went out running; he left

salió afuera y se fue

y y-dohkuy ma÷ y-¢ak hoho

and 3P-pants PST 3E-left inside

but he left his pants inside

pero sus pantalones dejó adentro

Spanish pero is another means of marking a contrastive relationship between clauses.

(10) ma÷ y-dima÷ "yaka-hi÷ day upa n-tikeñ hoho"

PST 3E-say turn.out-IMP light PRG ZE-enterinside

he said "turn out the light, I'm going to come inside"

entonces lo dijo "apaga la luz, ahora voy a entrar adentro"

pero yoom no es pendejo

pero mujer no es pendejo

but the woman isn't stupid

pero la mujer no es pendejo

Spanish discourse markers are the only means, other than markerless juxtaposition, by which additive and contrastive relationships between clauses are indicated in the TX texts.

4.2 The sequencer entonces

Spanish entonces 'then' (pronounced [´ntonse] in TX) occurs frequently in the TX texts, joining temporally sequential idea units.

(11) ma÷ y-bo÷n y-kehke÷ entonse hepe yoom u y-dim

PST 3E-cover 3P-chin then that woman PRG 3E-say

he covered his chin then that woman is saying

tapo su quijada entonces ese mujer dice

(12) huc‡ ky-dik entonse en segido ma÷ pooy-put

where 2A-go then at.once PST run-leave

"where did you go?" then immediately he went out running

donde fuiste entonces en seguido salió corriendo

The borrowed element entonces is the only overt marker whose primary function is to indicate temporal sequentiality.

4.3 Spanish bueno: a marker of change

The marker bueno has a variety of functions (cf. Hill and Hill 1986), but they all appear to share the feature of signaling that a shift in the discourse is imminent. For example, bueno can mark the end of a topic unit. In (13) (from story 2), bueno marks the end of a topic unit in which the compañero visits the señor's lover. She realizes that she has been deceived, and he runs out of the house, knocking over a pot of water. The señor arrives and a new topic begins. In(14) (from story 2), the compañero pays a nocturnal visit to the señor's wife, she realizes it is not her husband, and the compañero goes running out, leaving his pants inside. Then the señor returns home, beginning a new topic unit.

(13) ma÷ buh-k-tokeñ y-dohkuy y y-kopik

PST wet-enter 3P-pants and 3P-head

his pants and his head got wet

se mojó su pantalon y su cabeza

y como y y-÷eec‡a di÷ ÷a˜ki÷i ÷iñdye ma÷ poy y bweeno

and as and 3E-havewater outside NEG PST run and good

and as she has water outside he didn't run and good

y como tiene agua afuera no corrió

ora hes y-du÷k hepe tu˜kak piiñ

now when3E-anive that other man

now when the other man arrived

ahora cuando Ilegó otro hombre

(14) y-dohkuy ma÷ y-¢ak hoho hembe÷ yoom y-tik-i bweeno

3P-pants PST 3E-leave inside there woman 3E-house-to good

he left his pants inside good

sus pantalones dejó adentro

y hes ma÷ du÷k-am hes du÷k y-da÷a˜

and when PST arnve-CMP when arnve 3P-husband

and when he arrived when her husband arrived

y cuando llegó cuando llegó su marido

The marker bueno can also indicate a return to the main story line. For example in (15) (from story 1), the man has given background information, explaining that he has come to see his friend because his wife does not cook for him and he is hungry. The story then returns to the present time, with the two men discussing what is to done about the recalcitrant wife.

(15) ya taarde-ham y bweeno y heks‡ hepe-ka÷y ma÷ k-beñ

already late-CMP and good and now therefore PST 1A-come

already it's late and good and now that's why I came

ya está tarde y bueno ahora por eso vine

Although I did not have the opportunity to tape conversation between TX speakers, many examples of embedded conversation occur within the texts. From these examples, it is apparent that bueno can mark the resolution of an argument. In (16) (from story 2), the señor is trying to convince his friend to once again risk life and limb by pretending to be the señor. The compañero hesitates, but when offered a pig as compensation, agrees to the señor's wishes.

(16) n-c‡e÷-p tum yooy y

1E-give-INC one pig and

"I'll give a pig" and

voy a dar un cochino y

bweeno ÷a÷y kny-c‡e÷-p si entonse ÷eepu÷ k-dik-kak

well if 2-->l-give-INC yes then another.time 1A-go-ITER

"well if you give me [a pig] yes then I'll go again"

bueno si me va a dar si entonce voy a ir de vuelta

4.4 Markers of peak

A particularly interesting discourse marker is the grosería ('cruderemark, expression') ahh hijola chingada. The patterning of this phrase is not random; rather, R. used it to mark the peak of the narrative. Each text contained several peaks, many of which were marked by ahh hijo la chingada. The quoted speech in (17) illustrates one such usage. In(17) (from story 2), the compañero takes the señor's burro as payment for deceiving the señor's lovers, and the burro dies.

(17) ma÷ pa÷˜-ksa y-bumbe keñkuy

PST put-without.effect 3P-all snout

he put his whole snout

se metió todo su trompa

y heks‡ ma÷ ka÷-ham ijola chingada y-dim

and now PST die-CMP hiiola chingada 3E-say

now he's already dead "hijola chingada" he said

ahora ya se murió "¡hijo la chingada! " dijo

In (18) (from story 2), one of the señor's lovers discovers the deception when she illuminates the compañero's face and recognizes him.

(18) y-÷eec‡a÷ tum y-÷eec‡a÷ fooko lampara de mano y ma÷ y-do÷ke÷

3E-have one 3E-have flashlight and PST 3E-light

she has one she has a flashlight and she lit it

tiene uno tiene foco de lampara de mano y lo alumbró

hem y-pooy-ka÷ hembe÷ y-÷es‡pik ahh ijola chingada

there 3E-run-INT there 3E-know ahh hijola chingada

he was running and she knew him; ahh hijo la chingada!

estaba corriendo y conoció ¡ahh hijo la chingada!

Other markers of peak include ora si and baya la jodido. In (19) (from story 1), the man's friend tells him to go home and when his wife asks him three times if he is hungry to take out the piece of leather and beat her. Like entonces, ora si marks temporal sequentiality, but unlike entonces, it is a marker of peak as well. In (20) (from story 2), another lover discovers the compañero's identity, and she hits him with the tizón.

(19) y tre veces hes ÷u ky-÷aa˜wa÷k-kak ora si

and three times when PRG 2A-ask-ITER now yes

"and three times when she asks now yes"

y tres veces cuando pregunta ora si

hiikput-a÷ hepe pyal y daks-i÷ daks-i÷daks-i÷

take.out-APP that skin and hit-IMP hit-IMP hit-IMP

"you will take out that skin and hit her hit her hit her"

va a sacar ese piel y pegar pegar pegar

(20) baata ma÷ k-daks n-kok-yuku he÷ba÷a hepx-i˜-kuy

alot PST la-hitlFhead-above with burning.wood

"she hit me good in the head with the burning wood

basta le pegó en la cabeza con el tizon

baya la jodido ma÷ k-put ÷a˜ki÷i

vaya al jodido PST 1A-leave outside

¡vaya al jodido! I went outside"

vaya la jodido salió afuera

In TX, Spanish discourse markers tend to cluster in the "'important" sections of the discourse, i.e., the peak (ahh hiio la chingada, ora si) and the evaluation (bueno). Hill and Hill (1986: 103) argue that the distribution of Spanish loan words in Mexicano discourse is indicative of the nature of Spanish as a symbol of "power." That is, in narrative "Spanish loan material and code-switching to Spanish is often concentrated in what Labov (1972) has called 'evaluation': the sections of a narrative which show that it is important and worth listening to" (Hill and Hill 1986: 103).

Native TX discourse markers are still used, and in fact, Spanish and TX discourse markers often cooccur in discourse marker phrases (see 15); however, a full discussion of the patterning of indigenous particles is beyond the scope of this article.

5- Code-switching

One of the most salient characteristics of the TX texts is the extensive code-switching between TX and Spanish. Hill and Hill (1986:357) argue that borrowing and code-switching serve similar purposes in Malinche Mexicano:

For Malinche Mexicano speakers, the Hispanicization of speech serves similar functions in communication, whether this Hispanicization is achieved by code-switching or by the use of high densities of materials which should be classified as borrowings.

Blom and Gumperz (1972) distinguish between situational and conversational code-switching. In situational code-switching, aspects of the speech event determine code-choice. For example, in Malinche Mexicano certain topics, such as religion, government, money, and time, tend to be talked about in Spanish (Hill and Hill 1986:357-362). Spanish is also used for these topics in TX.

In conversational code-switching, pragmatic factors govern code-choice. In Malinche Mexicano, Spanish is frequently the language of inauthenticity and nonsense, as well as the language of authority, and often occurs at the narrative peak or for reiteration (Hill and Hill 1986). Spanish serves similar functions in TX discourse. The functions of code-switching are indicative of Spanish as a power code, which carries authority, but also reflect the conflicting feelings of indigenous people toward Spanish (Hill and Hill1986: 117-19). That is, while Spanish is a language of power, indigenous people do not have only positive emotions toward the use of Spanish. It therefore is used for "negative" contexts; e.g., nonsense, anger, and grosería(see 4.4 above).

Spanish is often put in the mouth of untrustworthy characters, while TX is put in the mouth of "true" friends. For example, the compañero in the second story, who helps to deceive the señor's lovers -- and the señor as well-- is often quoted in Spanish.

(21) entonse un amigo ma÷ y-÷aa˜-k-¢o˜ y como ba acer eso amigo

then a friend PST 3E-answer and how will you do that friend

then the friend answered "and how will you do that friend?"

entonce el amigo contestó "y como va a hacer eso amigo"

(22) y ma÷ y-dima÷ ke tal amigo

and PST 3E-say how are you friend

and he said "how are you friend?"

y dijo "¿que tal amigo?"

Contrast (22) with(23). They "mean" the same thing, but (22) is spoken by the "true" friend of story 1 who gives his compañero excellent advice on how to keep his wife in line.

(23) c‡i÷ y-wãt haytiiw

what 2E-do brother

"what are you doing brother"

"¿que hace?"

Words of anger are often in Spanish. In (24) the señor (story 2) realizes his friend has deceived him by paying his wife, rather than his lover, a nocturnal visit. The señor's wife discovers the same fact, confronts her husband (25) and tells him how stupid he is (26).

(24) entonse y-dima÷ hep piiiñ

then 3E-decir that man

then the man says

entonces dice ese senyor

ahh ijola chingada este me vino chingar este cabron me enganyó

"ahh ijola chingada he came to screw me over this guy tricked me"

(25) ÷u y-dima÷ mira y-dim ÷a÷m c‡i÷ ma÷ k-wat-a÷ tu amigo

PRG 3E-say look 3E-say look what PST 1A-do-APP your friend

she says "look!" she says "look what your friend [did to me]"

dice "¡mira!" dice "mira que cosa [me hizo] tu amigo"

(26) entonse ma÷ y-dima÷ y-duuñe÷ como eres pendejo

then PST 3E-say 3P-wife how you.are stupid

then his wife said "how stupid you are"

entonces dijo su mujer "como eres pendejo"

Hill and Hill (1986) argue that the use of Spanish adds "weight" to an utterance; i.e., Spanish lends the utterance a note of authority. This may explain why words of anger are often in Spanish. Other sections of TX narrative which carry the weight of Spanish authority include the narrative peak and evaluative discourse. The examples above exemplify this as well; e.g., (25) occurs at the narrative peak and (26) is an example of evaluative discourse, in which the señor's wife sums up her reaction to the events of the main story line.

In fact, it is difficult to separate the functions of particular instances of code-switching, since one utterance may serve many roles. However, it appears that the use of Spanish is especially frequent when one utterance has several of the above characteristics; i.e., conveying anger, signaling peak or evaluation, etc. This is true of borrowings as well. For example, the quoted speech in (17) (ahh hijola chingada) marks the peak, but is also evaluative, as it expresses the compañero's attitude toward the event, and expresses anger as well.

6 Relatively unaffected domains

6.1 Morphology

While TX morphology may be added to Spanish loan words, Spanish morphology has not been borrowed into TX. That is, TX lexical items remain relatively free of Spanish morphology. I have identifed only one Spanish element which has infiltrated the TX tense-aspect system -- duru (from Spanish duro 'hard'), which has become grammaticalized to function as a marker of progressive aspect.

(27) hes du÷k duru y-÷oota÷

when arrivePRG 3E-speak

when [the woman] is speaking

cuando (la mujer) está hablando

(28) entonse heks‡ hepe tu˜kak y-¢ay duru y-s‡ek-a÷

then now this other 3P-lover PRG 3E-laugh-APP

then now the other lover is laughing

entonce ahora la otra querida está riendo

It is significant that this element fits in well typologically with the TX tense-aspect system. That is, TX has several native tense-aspect preverbal particles, including a progressive marker ÷u. Spanish duru has therefore been accommodated into a native tense-aspect system.

6.2 Word order

TX remains a verb-initial language (29), with postpositions (30) and prenominal adjectives(31). Relative clauses may be preposed (see 6 above), as well as postposed, to the noun. Spanish exhibits different word order patterns: the basic word order is SVO, nouns precede adjectives, prepositions express locative-directional relations, and relative clauses must follow the noun.

(29) ma÷ du÷k tum piiña÷

PST arrive a man

a man arrived

llegó un señor

(30) ¢e÷es-ki÷i

bed-under

under the bed

abajo la cama

(31) ti¢ ÷aan

hard tortilla

hard tortilla

tortilla dura

Obviously, this is not an exhaustive account of TX word order. A more in-depth analysis of TX syntax would potentially uncover Spanish influence in the realm of word order.

7 Conclusion

TX provides evidence that the use of Spanish as a "power code" is not restricted to Malinche Mexicano (Hill and Hill 1986). TX does not necessarily display any unusual characteristics in the distribution of Spanish diffusions; however, it contributes to our knowledge of the possible manifestations of the Spanish power code. That is, the nature of Spanish as a language of authority is evident not only in the clustering of Spanish loanwords and code-switching in certain sections of the discourse (i.e., the evaluation and peak), but also in the semantic properties of borrowed transitive verbs.

The TX data also points to the multifunctionality of discourse markers. For example, ahh hiiola chingada typically marks the narrative peak, but may also encode evaluation. Multifunctionality may be especially characteristic of borrowed markers, whose patterning is determined not only by linguistic factors, but also by sociocultural factors. That is, Spanish discourse markers serve not only to demarcate topic units, encode temporal sequentiality, etc., but may also lend a note of authority to the discourse, or indicate negative evaluation of participants and/or actions. Furthermore, the use of groserías as discourse markers in TX suggests that the investigation of similar expressions in informal discourse in other languages would be fruitful.

TX also gives additional evidence that a language can be nearing extinction, yet not exhibit some classic signs of language death, such as morphological reduction and morphological borrowing (Dorian 1978). In spite of intense contact with Spanish, TX itself has undergone only "slight structural interference" (on Thomason and Kaufman's 1988 hierarchy of borrowing). This does not mean that Spanish has not influenced TX greatly; extensive code-switching and lexical borrowing lends a heavily -Hispanicized flavor to the Texistepec texts.

References:

Blom, J.P. and J.J. Gumperz. 1972. Social meaning in linguistic structures, in Directions in Sociolinauistics, ed. by J.J. Gumperz and D.H. Hymes, pp. 407-34. New York: Holt, Rinehart and Winston.

Brody, Jill. 1989. Discourse markers in Tojolobal Mayan, in Chicago Linguistics Society 25, ed. by Music, Bradley, et al, pp. 15-29.

Campbell,Lyle. 1985. Pipil. Berlin:Mouton. 1987. Syntactic Change in Pipil. International Journal of American Linguistics 53:253-80.

Dorian, Nancy. 1978. The fate of morphological complexity in language in language death: Evidence from East Sutherland Gaelic. Language 54:590-609.

GriImes, Joseph. 1975. The Thread of Discourse. Mouton: The Hague.

Hill, Jane & Kenneth Hill. 1986. Speaking Mexicano. Tucson: University of Arizona Press.

Kaufman, Terrence. 1964. Mixe-Zoque diachronic studies. Unpublished manuscript. University of Pinsburgh, Pittsburgh.

Labov, William. 1972. Language in the Inner Citv. Philadelphia: University of Pennsylvania press.

Longacre, Robert. 1983. The Grammar of Discourse. Plenum Press: New York.

Schiffrin, Deborah. 1987. Discourse Markers. Cambridge University Press.

Thomason, Sarah & Terrence Kaufman. 1988. Language Contact, Creolization, and Genetic Linguistics. Berkeley: University of California Press.

Wichmann, SØren. 1993. Mixe-Zoquean linguistics: a status report. Paper presented at the 13th International Congress of Anthropological and Social Sciences, Mexico City, July 29-August 5, 1993.

________________In press. The relationship among the Mixe-Zoquean languages of Mexico.

________________Forthcoming. Diccionario analítico del Popoluca de Texistepec. Mexico, D.F.: UNAM.