The complex clause in Boruca

This paper describes the complex clause in Boruca, a Chibchan language once spoken in the southern territories on the Pacific coast of Costa Rica. The present study concentrates on the behavior of complex clauses, following a major distinction between paratactic and hypotactic constructions. Paratactic constructions are divided into asyndetic and syndetic constructions. The latter are quite frequent and it is common to find the syndetons: mang, ní, ihchí and iné in the combination of sentences in regular speech. Hypotactic constructions also have asyndetic and syndetic constructions usually functioning as object complements. There is also evidence for reported speech, indirect questions and adverbial clauses (classified at least into seven different types).

LA ORACIÓN COMPLEJA EN LA LENGUA BORUCA

Damaris Castro
Universidad Nacional de Costa Rica, Costa Rica
dcastro@una.ac.cr

Artículo de revisión recibido 22-02-2011, artículo aceptado 04-04-2011

---

Abstract

This paper describes the complex clause in Boruca, a Chibchan language once spoken in the southern territories on the Pacific coast of Costa Rica. The present study concentrates on the behavior of complex clauses, following a major distinction between paratactic and hypotactic constructions. Paratactic constructions are divided into asyndetic and syndetic constructions. The latter are quite frequent and it is common to find the syndetons: mang, ní, ihchí and iné in the combination of sentences in regular speech. Hypotactic constructions also have asyndetic and syndetic constructions usually functioning as object complements. There is also evidence for reported speech, indirect questions and adverbial clauses (classified at least into seven different types).

Key words: asyndesis, complex clause, hypotaxis, parataxis, syndesis.

---

Resumen

El artículo describe la oración compleja del boruca, una lengua chibcha que se habló en los territorios del Pacífico Sur en Costa Rica. El estudio se centra en el comportamiento de las proposiciones complejas, a partir de la distinción principal entre construcciones paratácticas e hipotácticas. Las construcciones paratácticas se dividen en sindéticas y asindéticas. Estas últimas son bastante frecuentes, y el uso de los conectores mang, ní, ihchí e iné es también bastante utilizado en la combinación de frases del lenguaje común. Las construcciones hipotácticas también presentan asindesis y sindesis, las cuales suelen funcionar como complementos de objeto. Hay también evidencia de discurso directo e indirecto, preguntas indirectas y construcciones adverbiales (clasificadas en siete tipos diferentes).

Palabras clave: asindesis, hipotaxis, oración compleja, parataxis, sindesis.

---

Introduction

BORUCA, AN INDIGENOUS language of Costa Rica, has only recently begun to be considered a dead language. Although the language is not spoken in the communities, the indigenous reserve still occupies the same region. The people are divided mainly into two reserves: Boruca and Curré. In contrast to what happened with other indigenous groups here, this group still remains in the original territories that they occupied before colonial times. They were not moved to other lands, as were the Bribris, for example. According to Solano (2000), the Boruca population amounts to 2017 inhabitants: 1386 in Boruca and 631 in Curré (see figure 1). The last native speaker of the language died in 2003, thus allowing the language to reach an extreme point in its obvious decay, which was evident since the 1980s, and giving way to its imminent death. Currently there is one remaining semi-speaker of the language; the other one (his wife) passed away in 2007. This last elder semi-speaker grew up having quite a strong contact with the language, but this contact was not strong enough for this language to be his first and most important means of oral communication. Although there are many people in the community who claim to partially understand the language, they are not able to produce sentences or even phrases in Boruca. Spanish is, by far, the language of choice for people in the community.

Figure 1. Map of the indigenous languages and their territories in Costa Rica.
Source: Centro Centroamericano de Población (2000).

This is the first in-depth study on the complex clause in Boruca. There are several other very general descriptions of the language: Constenla & Maroto (1986) and Quesada Pacheco (1995, 1996). These serve as secondary sources supporting this analysis of the Boruca complex clause. Only Quesada Pacheco (1995) discusses complex sentences, albeit to a rather simple extent. The present study addresses clause linkage in Boruca, in relation to complex clauses, following the traditional distinction between parataxis and hypotaxis. These phenomena will be dealt with along the lines of Lehmann (1998, p. 181), who points to desententialization of the second clause as a phenomenon closely linked to parataxis, and argues in favor of the hierarchical downgrading of the dependent clauses, a matter closely related to hypotaxis. Thus, complex sentence formation will be conceived as "varying between the poles of elaboration and compression." Elaboration will be represented here through various types of syndetic constructions while compression will be associated with asyndetic constructions. Following this organization, the information presented below will be divided into two major sections: paratactic and hypotactic constructions. Paratactic constructions refer to the coordination of elements of a sentential nature while hypotaxis has to do with a dependency relation of one clause or sentence on another clause or predicate, thus understood as subordination. Paratactic constructions are in turn divided into asyndetic (juxtaposed) and syndetic constructions, the latter usually linked by the connectives mang, iné, ihchí and ní. The section on hypotactic constructions will be further divided into complement and adverbial clauses. The first are not too common and they represent arguments of the predicate. The latter, although exhibiting more variety, do not usually represent arguments of predicates. Both syndetic and asyndetic hypotactic constructions are described on the basis of their functions, as arguments in the case of complement clauses or as adjuncts in the case of adverbial clauses. Relative clauses are considered nominal modifiers; thus, given that they are not adverbial or complement clauses, they are not included as part of this section.

1. Paratactic Constructions

1.1. Asyndesis

Asyndetically coordinated sentences in Boruca may or may not have elements in common. An example of the latter is illustrated in (1); note that the clauses do not have arguments or verbs in common. In the remaining examples below, there are shared elements; these can be arguments as in (2) í', or (3) yubú. Arguments can be overtly expressed as in (2), or they can be expressed anaphorically, via ø anaphora as in (3) and (4). The shared element can also be a verb or verb phrase, where the verb can be overtly expressed as in (5).

(1) Káhwik   Kwasráng   ki        ká       o-krá         yo     brú'    ki.      Kak       ki         ro'-krá.

     Far        Kwasráng   SPEC   from    finish-PNT   rain    big    DEF.    Sun      DEF      shine-PNT.

     'Far from Kwasráng, the big rain finished. The sun shined.' (Constenla & Maroto, 1986, p. 94)

(2) Drik        kráng      ki         i'       wíkra           di'       ka,       i         ki        de-krá          ihnéng.

      Money    AUG       DEF    3SG      take-PNT     river    to,       3SG    SUBJ    leave-PNT     nothing.

      'He took a lot of money to the river. He left with nothing.' (Quesada Pacheco, 1996, p. 64)

(3) I'         ramróhk      é'dé       yubú'           yáng-íra,       yubú'       ki       oh-íra.

     POSS   woman        with        chicha¹    drink-DUR,     chicha     DEF    drip-DUR

     'With his wife (he) drank chicha, (he) dripped (prepared) chicha.' (Quesada Pacheco, 1996, p. 71)  

(4) E'tsi     ramróhk       ki          kawí'ra          we'é.    Chubú          kam       kí       bing-íra        breát       breát.

     One     woman         SPEC      live-ACTL      here.    Cotton         thread     DEF    warp-DUR    a.lot         a.lot.

     'One woman lived here. (She) warped a lot of cotton thread.' (Quesada Pacheco, 1996, p. 70)

(5) We'é    i'         ramát    ki          koht-krá,   wé'é    í          ki       kohd-ra.

     Here    POSS   wife       SUBJ      die-PNT,    here    3SG    SUBJ   die-ACTL.

     'His wife died here, he (will) die here.' (Constenla & Maroto, 1986, p. 58)

1.2. Syndesis

Syndetic constructions behave similarly to asyndetic constructions, the only difference being the presence of a syndeton. The most common syndetons in Boruca are: mang, ní, ihchí and iné. Boruca exhibits a conspicuous tendency to use syndetic constructions. Clauses which have no elements in common can be linked as in (6) and (7). There are also cases where there can be elements in common. If the shared element is an argument, it can be expressed as an anaphora: either as a personal affix (i, 3SG) as in (8), a pronoun as in (9), or a zero anaphora as in (10). It is also possible to share a verb or verb phrase as in (11), where the verb may be left unexpressed (12) without rendering the sentence ungrammatical.

(6) Bak       róhk     ki        de-krá    yaba'   í',   mang    i-ng,           i'       aht-krá       i'       reshí.

     Owner   PL        DEF    go-PNT    hill      to,  and      3SG-FOC,    3SG   stay-PNT    3SG   alone.

     'The owners went to the hill, and he, he stayed alone.' (Quesada Pacheco, 1996, p. 60)

(7) Dabag-ír-í-ng             róhk,   iné     káng     k         si'án-írá.

    Arrive-DUR-3PL-AGR    PL,     and    stone    DEF     disappear-DUR.

    'They arrived and the stone had disappeared.' (Constenla & Maroto, 1986, p. 122)

(8) Mang   abí        bóhk            ki      i        kún     tsít    ki        kruh-krá,  mang    i          káng     Tsít     ki       tsáng-kr-i         ká,    mang    i         shúng-kr-i           ká.

    Then    person   young.man   DEF   POSS  sling    DIM   SPEC    take-PNT, and      POSS    stone    DIM    SPEC   put-PNT-3SG     in,     and     3SG     throw-PNT-3SG    to.

    'Then the young man took his sling, and he put his little stone in and he threw it at her.' (Quesada Pacheco, 1996, p. 48)

(9) Ramróhk   ki       trú'-ir-í-ng                i       e'dé,    mang    ing            cohd-íra      i'        é'dé.

     Woman    DEF     join-DUR-3SG-AGR    3SG   with,     and      3SG-AGR     die-DUR     3SG     with.

     'The woman, she joined him and she died with him.' (Quesada Pacheco, 1996, p. 71)

(10) I        kí       ang     daba-krá       di'        kák     tá     ihchí    é'tse     wá'     ki           ba-krá.

       3SG   SUBJ   FOC    arrive-PNT    POSS     place   in     and      one      child   SPEC     have-PNT.

       'He arrived in our town and (he) had one child.' (Constenla & Maroto, 1986, p. 96)

(11) A'd     beyáng-r-i-ng,               ní      át      ki        i'       beyáng-ra      i        do'sh.

      1SG    want-ACTL-3SG-AGR,    and    1SG    SUBJ   3SG    want-ACTL    3SG    too.

      'He loves me and I love him too.'(Quesada Pacheco, 1996, p. 77)

(12) Entonces,  i'          wá'    bush     ki         shit-krá     ní      wá'   bóhk    ki        i'       dó'sh.

       Then,       POSS   child    girl       SUBJ   laugh-PNT   and   child   boy    DEF    3SG    too.

       'Then her girl laughed and the boy did too.' (Quesada Pacheco, 1996, p. 73)

As illustrated in (11) and (12) above, ní coordinates clauses; however, this morpheme also serves to subordinate a clause, as illustrated in (13). Whether it is a coordinator or subordinator is expressed by the meaning of the conjuncts or the subordinated sentences in their corresponding functions in the given context.

(13) Ramróhk    búsh     ki     kraé'kra          ni       i'          wá'       ki         baw'íra.

       Woman     young   DEF   observe-PNT   that     POSS   child     SPEC     be.born-DUR.

       'The woman observed that her own child was being born.' (Quesada Pacheco 1995, p. 111)

Moreover, iné aside from expressing coordination as in (7) above, may also express opposition; in such cases, the context determines the semantics and functions of this morpheme. See (14) below for an example of contrastive information.

(14) I         shang-chá-ír-í-ng,                iné    i            bot           ki       yikrá    do'i'sh-ira.

       3SG    shoot-want-DUR-3SG-AGR,    but    POSS    shotgun    SUBJ     fire     give-NEG-DUR.

       'He wanted to shoot him, but his shotgun didn't fire.' (Literally: 'didn't give fire') (Constenla & Maroto, 1986, p. 88)

2. Hypotactic Constructions

2.1. Asyndetic Complement Clauses

Subject complement clauses are not identifiable in Boruca. Asyndetic complement clauses are usually objects of the complement, with verbs of saying, perception and others. Observe how in (15) below there is no connector to introduce the complement clause (i' wá' ki baw'íra); in the second case (16), the complement clause preceding the verb (beyangí'shíring) does not have the connector either.

(15) Ramróhk   búsh     ki      kraé'kra          i'         wá'    ki        baw'-íra.

       Woman    young    DEF   observe-PNT   POSS   child   SPEC   be.born-DUR.

       'The woman observed (that) her child was being born.' (Quesada Pacheco, 1995, p. 112)

(16) Íní'     í-ng        yá'shi      kawí'kra   róhk, kóngróhk   róhk  krabé'  a  vér   si    í-ng        i     kruhg-írá     róhk   yét,   ramróhk   róhk   ang    yabáhg-írá   yá     Róhk

      Since 3PL-AGR   like.this   'live-PNT   PL,   man          PL     watch to see   if   3PL-AGR   3PL  catch-DUR   PL     PURP, woman     PL      CONN  pass-DUR    DEM   PL 

      ki         beyang-í'sh-ír-i-ng               róhk.

      SPEC    want-NEG-DUR-3PL-AGR       PL.

      'Since they lived like this, watching men to see if they could catch them, they didn't want the women (that) passed by there.' (Constenla & Maroto, 1986, p. 62)

2.2. Syndetic Complement Clauses

Syndetic complement clauses are also objects of complement-taking verbs and they are introduced by the connective ní (ni) (17). Towards the end of the existence of the language there was a rather widespread use of the Spanish connective ke ('that') in substitution of the Boruca form ní; see (18) below.

(17) Ramróhk    búsh      ki        kraé'kra          ni     i'          wá'     ki          baw'íra.

       Woman     young    DEF    observe-PNT    that   POSS   child   SPEC    be.born-DUR.

       'The woman observed that her own child was being born.' (Quesada Pacheco, 1995, p. 111)

(18) At      ki     te-krá        ke      do'    do'     kráng    róhk   ki       og-íra.

       1SG   subj   say-PNT   that   soon   soon   tree      PL     DEF    disappear-DUR.

       'I said that soon the trees (will) disappear.' (Quesada Pacheco, 1995, p. 111)

2.2.1. Reported Speech

There are two rather common ways of expressing reported speech in Boruca. Reported speech clauses can be divided into non-literal reported speech and quoted reported speech. In the first, complement clauses indicating reported speech are, in terms of syntax, objects of the verbs used to express indirect speech. The most common verb used to introduce reported speech is tek, te ('say') as in (19) and (20). In these constructions the clause expressing reported speech follows the main clause.

(19) Te-kr-i-ng               róhk: 'tsu'-krá         yá        ki      tebek    i'          wá.

        Say-PNT-3PL-AGR    PL:    'scape-PNT   DEM    SPEC   snake    POSS    son.

       'They said: The son of that snake escaped.' (Constenla & Maroto, 1986, p. 58)

(20) Antónces   bó's        róhk   ki      ang   te-rá:       'deg-rá        di-ng          i       kwin-í       kawik.

       Then       wizards     pl      def   FOC   say-actl:   'go-actl       1PL-agr       pl     put-ING    far.

      'Then the wizards said: We will go to drop them far (away).'(Constenla & Maroto, 1986, p. 66)

The second type of reported speech mentioned above is quoted or literal reported speech. The first possibility is to have verbs such as yuak ('yell') or tek ('say') immediately followed by a direct quote; see (21) below. The second type, and by far the most common way of expressing quoted speech is through the use of the postposition ka ('to') in a clause where the main verb has been deleted. Most commonly the subject and the object are expressed in the first clause; the verb, however, is omitted in this clause. In these main clauses, full nouns may function as the subject of these clauses and pronouns as objects (21), pronominal subjects and objects are found (22), or the sentence may have only the object of the clause, leaving the subject unexpressed (23). Finally, the order or these clauses may alternate, presenting the reported speech clause following the main clause as in the examples of the non-literal reported speech above; or they may have the reported speech clause in sentence initial position as in (24).

(21) I        ki         kasá'-írá: '¿í          ugé'      b-ing          at       churuká'-ra.'

       3SG   SUBJ     yell-DUR: '¿what     for      2SG-AGR     1SG     tickle-ACTL'

       'She yelled: "Why are you tickling me?".' (Constenla & Maroto, 1986, p. 72)

(22) Kuréwá     ki         i       ká:   'at     ki       deg-í'sh-a.'

       Kuréwá    SUBJ    3PL    to:   1SG   SUBJ   go-NEG-ACTL

       'Kuréwá to them: "I am not going".' (Literally 'I do not go') (Quesada Pacheco, 1996, p. 54)

(23) I       róhk   ki         i       ká: 'chí'       suhróhk    tsít     abí      déng     ka    di'róhk     ki        do'á-ra.

       3PL   PL     SUBJ    3SG   to:   'DEM    elder        DIM    EMPH   behind   to    1PLPL      SUBJ    CONT-ACTL.

       'They to him: "We come behind (following) that little old man".' (Quesada Pacheco, 1996, p. 81)

(24) Mang    i'         be        ki        ka,  'mí'rang      ka,    mí'rang       ka.'

       Then    POSS   mom    SPEC   to,    'go            to,     go             to

       'Then, to his mom: "Go, go".' (Quesada Pacheco, 1996, p. 73)

(25) 'Moréng-r-i-ng',                 i            ká      Loréto       ki.

        'Good-ACTL-3SG-AGR',     3SG        to       Loréto      SPEC.

         '"That's good," he (said) to Loreto.' (Quesada Pacheco, 1996, p. 114)

2.2.2. Indirect Questions

Indirect questions are also present in Boruca. These are usually marked by the Spanish complementizer ke ('that') which usually integrates the clause with the main verb as in (tekrá), in example (26) below. In the second example (27), the Spanish form sábe ('who knows?') is used to express the indirect question; in this case the connector that is used is ang.

(26) Di       abí         te-krá           ke          kákba     nengwáhk     ke        chah-rá?

       Who    EMPH     say-PNT       CONN      day        first             DEF      today?

       'Who said that today was the first day?' (Quesada Pacheco, 1996, p. 152)

(27) Sábe      í         ang        i'             óng-kra?

       Know    3SG     CONN     3SG        do-PNT?

       'Who knows what he did?' (Constenla & Maroto, 1986, p. 116)

2.3. Adverbial Clauses

Simultaneous and causative clauses are the only types of clauses that allow the apodosis to appear before the main clause; the rest of the clauses exhibit a protasis-apodosis order. The following represent the main types of adverbial clauses in Boruca.

2.3.1. Conditional Sentences

The original conditional sentences are formed using the word angká ('then','however', 'on the contrary') at the beginning of the apodosis. As in (28) and (29), the adverbial clause always follows the main clause in these constructions.

(28) Yó       ki           tru'rá,         angká     ógé     róhk     ki          ya'      sé-ra.

        Rain   SUBJ       fall-ACTL,     then       all       PL       SUBJ     REFL    get.wet-ACTL

        'If it rains, then everybody gets wet.' (Literally: 'It rains then everybody gets wet.') (Quesada Pacheco, 1995, p. 117)

(29) Juan    ki          dabag-rá,    angká    di-ng           i'          wí'-ra          dí'        bang    ká    i.

       Juan   SUBJ    arrive-ACTL,  then     1PL-SUBJ      3SG      take-ACTL    river     bank    to     in.

       'If Juan comes, then we take him to the river bank.' (Quesada Pacheco, (1995, p. 117)

Nevertheless, these conditional sentences are rarely found in the available Boruca literature. They have been progressively replaced by forms that look increasingly like the Spanish conditional sentences, even adopting the Spanish conditional marker si ('if'), at the beginning of the first clause. Observe how in (30) si ('if'), is used in the first clause although the second clause still keeps the Boruca marker angká at the beginning of the second clause. However, in example (31) the Boruca marker angká is eliminated and the conditional is expressed through the use of the Spanish marker 'si' only. In these examples, we can see the progression of the Boruca language loss and the transition to Spanish.

(30) Si     ya'sróhk     bush    ki         che'át-irá,  angká     i      bahdí            ki          kwíng          kri'éh-irá.

       If     young       girl      SUBJ     sick-CONT,  then     3SG   punishment    SPEC      AUG           grow-du

       'If the young girl was sick, then her punishment was greater.' (Constenla & Maroto, 1986, p. 124)

(31) Si    yá      ki        tebek     tsít      róhk     ki         ba'-ra              we'é,   wá       ki          ahd-rá              yabá.

       If    DEM   SPEC    snake    odim     PL       SUBJ     be.born-ACTL    here,   DEM     SPEC      remain-ACTL      lake.

       'If these snakes are born here, this will become a lake.' (Quesada Pacheco, 1995, p. 117)

2.3.2. Purposive Clauses

There are two types of purposive clauses in Boruca: implicit and explicit clauses. Implicit clauses lack a purposive marker, and according to Quesada Pacheco (1995, p. 120), the second verb lacks TAM markers as well, but it could optionally carry the infinitive marker (-í); see (32) and (33).

(32) Rahd-rá            di-ng          ábu'    dí'        kahk     í.

       Leave-ACTL    1PL-SUBJ     swim    river     bank     in.

       'We left to swim in the river.'(Quesada Pacheco, 1995, p. 120)

(33) Daba-kr-í-ng               kab-í            di'      kába.

      Come-PNT-3SG-AGR     sleep-ING     1PL     house.

     'He came to sleep in our house.' (Quesada Pacheco, 1995, p. 120)

Explicit purposive clauses are different from the above clauses in that they have purposive markers. Among these, by far the most common purposive clauses, chá ('for', 'in order to') is placed at the beginning of the clause and most commonly (but not always) yét ('for') comes at the end, as in (34). The subject of these clauses is always marked by the focus marker ang.

(34) Iné        i-ng              i'         wí-kra         i'        ú            í,      ch-i-ng               i         dobóhg-írá    yét.

       Then     3SG-SUBJ      3SG     take-PNT     3SG    house     to,    PURP-3SG-SUBJ    3SG    raise-DUR     PURP.

      'Then, he took him to his house, for him (the father) to raise him (the child).' (Constenla & Maroto, 1986, p. 96)

The regular order of purposive clauses is altered in certain cases. In one type of construction, the purposive clause can appear after the subject of the main clause (35), and the main verb appears after the purpose clause.

(35) Brúngkahk    róhk   ki,     chá     si'kwa       róhk    ang  ya'       cóng-írá             róhk    ki       yét,     ing             kúshtan-írá  róhk   siní'

       Boruca        PL     DEF,  PURP    foreigner    PL      FOC  REFL    get.scare-DUR     PL      SPEC   PURP,   3PL-SUBJ     roar-DUR     PL      wild.boar  

       róhk     dóhkre    kurá'    róhk    dóhkré     kwíng     tsa'ongká.

       PL        like         tiger    PL       like           AUG      loud.

       'The Borucas, for the foreigners to get scared, roared really loud, like wild boars, like tigers.' (Constenla & Maroto, 1986, p. 76)

Although the following do not represent the most common constructions, it is also possible to find purposive clauses where the arguments of the clause are moved to the right of the purposive clause, probably with the intention of emphasizing the purpose; see (36).

(36) Dí'      kahk    tá, a-ng            ba       dená-ra,       ch-a-ng                ba      yét,   bú'k    só't         sá'-ra              más.

       River  bank    on, 1SG-SUBJ     2SG     wait-ACTL,    PURP-1SG-SUBJ     2SG    PURP, two    Bluegill    catch-ACTL      more.

       'On the river bank, I (will) wait for you, so that I, for you, two more bluegills will catch.' (Constenla & Maroto, 1986, p. 62)

Finally, it is also rather common to have purposive constructions where the second, final purposive marker yét is not present (37). One could hypothesize that these constructions resemble the Spanish purposive construction where the purposive marker is placed only at the beginning of the clause and does not require a second marker.

(37) Mang   curá'    róhk    ki         ká'wing-rá      di-ng,          ch-í-ng                 i tú'-ra              róhk.

       Three  tiger     PL      SPEC    call-ACTL        1PL-SUBJ     PURP-3PL-SUBJ      3SG eat-ACTL     PL

       'We will call three tigers, for them to eat her.' (Constenla & Maroto, 1986, p. 70)

Furthermore, the following sentence shows an even stronger influence of Spanish through the use of the purposive expression a ver si ('to see if'). In (38), the speaker marked this sentence with a ver si at the beginning and with yét at the end of the clause, probably not aware of the double purposive marking.

(38) Íní'       í-ng        yá'shi         kawí'kra      róhk,  kóngróhk   róhk    krabé'    a    vér    si    í-ng         i       kruhg-írá      róhk   yét,

      Since    3PL-AGR   like.this     live-PNT      PL,    man           PL      watch   to    see    if    3PL-AGR   3PL   catch-DUR    PL     PURP,

      ramróhk    róhk   ang    yabáhg-írá     yá     róhk   Ki beyang-í'sh-ír-i-ng                 róhk.

      woman      PL      FOC   pass-DUR       DEM   PL     SPEC want-NEG-DUR-3PL-AGR     PL.

      'Since they lived like this, watching men to see if they could catch them, they didn't want the women (that) passed by there.' (Constenla & Maroto, 1986, p. 62)

2.3.3. Causal Clauses

The original causal clauses expressing circumstances or reason contain the postposition úge', at the end of the causal sentence. The causal clause may appear preceding (39) or following (40) the protactic clause.

(39) I          turi     róhk   ki          i         dená-'shi-krá     úge',       í        ki         i     dí-'shi-kra               áyi'.

       POSS    cow    PL     SPEC     3SG    wait-NEG-PNT    because, 3SG   SUBJ    3PL   look.for-NEG-PNT    again.

       'Because the cows were not expecting him, he didn't look for them again.' (Constenla & Maroto, 1986, p. 86)

(40) Móro  róhk   di-ng             i'      róhk     ki          ka'yéng-ra,   ká'        béyang-í'shi-kr-i-ng             Róhk      úge'.

      Moro   PL      1PL-SUBJ      3PL   PL        SPEC     call-ACTL,     name     want-NEG-PNT-3PL-SUBJ     PL         because

      'We called them móros because they didn't want a name.' (Constenla & Maroto, 1986, p. 84)

There is a rather common strategy to present reasons whereby the causal clause is introduced by a causal marker yabúge'; in these constructions the causal marker úge' is indexed to a demonstrative and focus marker to present a reason. These causal clauses can also precede (41) or follow the protactic clause (42).

(41) Y-ab-úge'             sí'kwa      róhk    ki       nani'-ch-í'shi              brúngkahk   róhk   deng      ka.

       DEM.FOC.cause    foreigner   PL      DEF     go.down-want-NEG     boruca       PL      behind   of.

       'For that same reason the foreigners didn't want to go behind the Borucas.' (Constenla & Maroto, 1986, p. 76)

(42) Wá     ki        yuakrá     í'      abí      kóngróhk  ang,  yabúge'              di'   róhk  ki        ya-í'sh-a        nunca    yá     ke'.

       DEM   SPEC   say-PNT  3SG   EMPH   man        FOC,  DEM.FOC.cause   1PL  PL    SUBJ   go-NEG-ACTL  never    DEM    in.

         'That very man said that, for that reason we never go in there.' (Constenla & Maroto, 1986, p. 94)

Unfortunately, regarding causal constructions there are many examples with the Spanish causal marker porque ('because') substituting the original Boruca form (ugé') altogether, as in (43) below:

(43) I'     rúd-ra         di-ng        pórque    i     bag-r-i-ng                  ki       tebek  tsít   róhk-ra.

      3SG  burn-ACTL   1PL-AGR   because  3PL  have-ACLT-3SG-AGR   SUBJ   snake  DIM   PL-ACTL

        'We will burn her because she will have them, little snakes.' (Constenla & Maroto, 1986, p. 56)

2.3.4. Temporal Clauses

Temporal clauses are expressed in Boruca through the use of the forms nengwahk ('first') to express anteriority as in (44) and (45), or ki dengí ('later') to express posteriority as in example (46).

(44) Wé'   duríh   róhk  ki     te-krá     que  néngwahk  tebek   ki       ái'-ishkong-írá,     iné     i'      ramróhk  ki       néngwáhk   k-i-ng              wá'   bagírá       tsú-ishkong-ír-í-ng.

       DEM wizards PL   SUBJ  say-PNT  that  first         snake   SPEC  die-have.to-DUR  and   POSS  woman   SPEC   first          that.3SG-SUBJ  child  have-DUR  burn-DUR-3SG-AGR.

       'These wizards said that first the snake had to die and that the first child, had to be burned.' (Constenla & Maroto, 1986, p. 94)

(45) Yá     ki       ishdúa  bag-ír-í-ng               róhk   désde   néngwáhk.

       DEM  SPEC   seen    have-DUR-3SG-AGR   PL      since   before.

       'There, they had seen that since before.' (Constenla & Maroto, 1986, p. 76)

(46) Yá     ki       déngí   abí       róhk    ki      múla   róhk   ki     dohk-máng-kra        fing-í           yá     ta   bígát.

       DEM  SPEC   later   people   PL      SUBJ   mule   PL     DEF   hear-be.used-PNT   scream-ING   dem    in   at.night.

       'Later, the people were used to hearing the mules scream there at night.' (Constenla & Maroto, 1986, p. 76)

In addition to the above ways of expressing temporal clauses in Boruca, the most common way of expressing temporal sequences is through coordination, especially through the use of the morpheme mang ('then') as in (47).

(47) Bá    ki      kup   ki      bruh-rá,      mang  yá     ki       kup    ki       é'ra           ba-ng       é'tsi    Kuré   ta, mang   ba-ng      dí'        é'ra          i        ká.

      2SG  SUBJ  corn  DEF   grind-ACTL, then    DEM  SPEC   corn  SPEC   put-ACTL   2SG-AGR   one     Pot    in, then    2SG-AGR   water   put-ACLT  3SG    in.

      'You have to grind the corn, then you put that corn in a pot, then you put in water.' (Quesada Pacheco, 1996, p. 86)

2.3.5. Locative Clauses

In Boruca locative clauses are formed using the adverb yé' at the beginning of the apodosis and the postposition ta at the end of this clause as in (48). Although Quesada Pacheco (1995:127) claims that the subject of this clause should be marked with ang, this is not always the case, as can be observed in the second locative clause in sentence (49).

(48) Yá'shi       i-ng          dabag-írá   róhk, yé'        i-ng         ya-chá-írá         róhk   ta.

       Like.this    3PL.SUBJ  arrive-DUR   PL,   where   3PL.SUBJ   go-want-DUR     PL      to.

       'Like this they arrived where they wanted to go to.' (Constenla & Maroto, 1986, p. 100)

(49) Ya-rá        di-ng          bi'      kwing-í,  yé'-ra           kwíng   abí       róhk    ang  tá, yé'ra             dí'       krí'   ta.

       Go-ACTL   1PL-SUBJ    2SG    put-ING, where-ACTL   many   people   PL      FOC  in, where-ACTL   water   big   in.

       'We are going to place you, where there are many people, where there is a big river.' (Constenla & Maroto, 1986, p. 68)

Moreover, according to Quesada Pacheco (1995, p. 127), the postposition ta can at times be substituted by ki as in (50).

(50) yé'       di'       tsasúh       róhk  ang   kawí'írá ki

       Where  POSS  ancestors    PL    FOC   live-DUR?

       'where our ancestors lived' (Quesada Pacheco, 1995, p. 127)

There is also evidence of Spanish influence in this type of clauses. Note that in (51) the Spanish preposition desde ('from') is used to express location in sentence (51), tá still appears at the end of this clause. Yé', however, does not appear in this construction.

(51) Mang  i-ng         rahd-rá,   mang  ing           uhd-rá        i       ká     désde    i          kuráskwa    tá.

      Then   3SG-AGR  come.out, then   3SG-AGR   curl-ACTL   3SG   on     from      POSS    feet           on

      'Then, he comes out, then he curls on her, from her feet.' (Constenla & Maroto, 1986, p. 54)

2.3.6. Simultaneous Clauses

Simultaneous clauses are introduced by the form í'k ('when'), at the beginning of the apodosis and é' ('then') at the end of this clause; see (52). However, a common strategy is found to express simultaneous clauses, perhaps following the Spanish syntactic pattern, thus containing í'k at the beginning of the clause and lacking the second Boruca marker é' as in (53) below. There is also evidence of a widespread use of Spanish cuando ('when') in many constructions expressing simultaneous actions (54).

(52) Í'k      í       ki      bohgát-írá   é',     i        tsasúh          róhk   ki       i'      wí'-írá          drik     suát     déng    ka.

      When 3SG   SUBJ  young-DUR  then, POSS  grandparent   PL     SUBJ   3SG   take-DUR     metal   White   behind  of.

      'When he was young, his grandparents took him to look for (behind) the white metal.' (Constenla & Maroto, 1986, p. 106)

(53) Í'k        i-ng           daba-krá     yá     ta,    kú'         ki       ká'wing-kr-í-ng         róhk.

       When   3PL-SUBJ    arrive-PNT   DEM   in,   alligator   SUBJ   call-PNT-3PL-AGR     PL.

       'When they arrived there, they called it alligator.' (Constenla & Maroto, 1986, p. 72)

(54) Cuándo   i-ng          i      ká      wa-rá          róhk,  iné    ramróhk   ki      tamáño-rá.

       When     3PL-AGR   3SG  about  know-ACTL   PL,    then   woman    DEF   big-ACTL.

       'When they knew about it, the women were older.' (Constenla & Maroto, 1986, p. 64)

2.3.7. Causative Clauses

These clauses have the causative adverb iní ('since') at the beginning of the apodosis, which precedes the main clause (55). These clauses may, at times, include the form do'sh (like) at the end of the secondary clause. The last example (56) shows how more than one type of adverbial clause can appear in one sentence; here the causative clause serves as protactic clause for a purposive construction.

(55) Iní'     yá     ki       sutáng-ra            do'sh,   yá     ta   ír-i-ng            róhk.

      Since  DEM   SPEC  savannah-ACTL    like,      DEM   in   DUR-3PL-AGR   PL.

      'Since that was like a savannah, they were there.' (Constenla & Maroto, 1986, p. 86)

(56) Iní'    i'     róhk   ki      y-ab-é'           káng    ki     sodíh-írá     róhk  dó'sh, y-ab-úge'           i-ng         i       kachí'-kra   yé'        i-ng          i      réhg-írá    róhk    ki      tá

      Since 3PL  PL     SUBJ  DEM-EMPH-in   stone  DEF   make-DUR   PL     like,    DEM.FOC.cause  3PL-AGR   3SG   cover-PNT  where   3PL-AGR   3SG   cut-DUR   PL      SUBJ   in

      y-abí           surá'   tá; yá      ki      tsihré'kra       káng    dó'sh.

      DEM-EMPH   mud     in; DEM   SPEC  harden-PNT   stone   like.

      'Since they made the stone like mud, for that reason they covered it where they cut it; that one hardened like a stone.' (Constenla & Maroto, 1986, p. 106)

This closes the account of adverbial clauses and complex clauses in general in the Boruca language. Some of the structures presented above may be observed in the sample text provided in Appendix A.

Conclusion

Several studies have been conducted on the Chibchan languages of Costa Rica but only recently have they gone into further depth. At this moment a research group of the Universidad Nacional of Costa Rica has taken it upon itself to describe the grammar of each of the indigenous languages in Costa Rica. Quesada (2000) had already written a grammar for Teribe. The remaining grammars are in progress at this time, and particular emphasis is being given to the analysis of the complex sentence in each one of the languages. This paper provides a complete, detailed account of the complex sentence in Boruca. It offers a new perspective using the categories of paratactic and hypotactic constructions; this approach makes possible a type of analysis that had not been applied in previous studies, with a broader view of how the language functions. This research complements a series of other studies that the author has carried out in an aim to complete a description of the Boruca grammar.

Abbreviations

Appendix A
Sample Text
Wá' yabá

E'tsi ramróhk ki kawí'íra we'é, Brúngkahk ta. Abí bush. Yá ki abí bush ki che'átkra, é'tsi kóngróhk ki beyángkra, mang kongróhk ki i' ahtkra kúa é'dé. I' be ki i yebéht ki breát i báhdira,kráng úngíra i ka róhk, káng úngíra i ka róhk, kúa ta, ching i' ai'ra yét. Mang ing dekrá i co'kráng e'dé Rurúgíh, cóm déng ka, káu'turinát déng ka. Dekring é'tsi kak yá ki i kúa é'dé Rurugíh, kwando ing warí'íra ong yá'óng shihshí, yabíh ki yá' sentíkra king che'át wáhdirá, warí' San Joakín ta. Dabakríng yá ki dí kahk tá, kwandíng yá' túh túh, túhkrá wishkú', wishkú' i' shikríng, sino i' wá' abíng bau'krá. Ya'sróhk ki, ramróhk bush ki kraékra ní i' wá' ki srek srek srek, yá' bawíra. Yabíshé' ramróhk só' ki wá' ki kruhkrá i yé'tsa ta, mang ing shúngkra yá ki dí kahk ta, é'tsi yabá tá. Yá ta yá ki wá' ki ahtkrá yiri' ká. Ora abí róhk ki i dóhgra boí. Yá ki yabá só' ki i' úra. I' yebéht ki i' be ki tekrá ki mang, seguramente que yá ki yabá súhkra abíng i' wá tsít ki shékri ká. Abí róhk ki ní i' íshtkra i yuré'tsít ki rahtkrá. Diz que kabáng abí beyángira abí tsít ki pero i' be róhk ki tekrá: No, DURíh ki tekrá ning yá ki dí' suhkra tsít ra.Yá ki aldredo i shékra i' be kí ka. Quesada Pacheco (1996, p. 31).

Wá'   yabá

Child   pond

Free Translation

One woman lived here in Boruca. She was a young woman. She was sick, she wanted a man and he left her pregnant. Her mother and father beat her a lot, with a stick with stones on her stomach, they wanted to kill her. Then, she went with her basket to Rurúk to get zapote and cacao. One day she left to Rurúk, with her big stomach. She went down the hill carrying the heavy basket, she felt sick when she was getting to San Joaquín. When she got to the river bank, she fell on her knees and started to pee, but she was not peeing, she was giving birth to her own child. The young woman saw that her child was crying and crying, then he was born. Then the woman grabbed the child in her hands and threw him in the pond. That is where the child remained alive. Now, people hear him cry. That pond is his house, his father and his mother know the story and for sure they want to remove their child's spirit from the pond. People have seen his little hand come out of the water. People say that what the child wants is a priest, but the elder women say that the wizards say that it is the little spirit of the water. That is why he was taken away from his mother.

---

^* This analysis was carried out at the Universidad Nacional de Costa Rica, within the framework of research on the indigenous languages of Costa Rica and Central America. It contributes information to the linguistic community on Boruca, a dead language, an original member of the Chibchan Family. The article depicts a descriptive characterization of the complex clause in this language. The English version of the glosses and English translations throughout the article are provided by the author of this article. Given the state of the language, the author has relied primarily on these secondary sources, complemented by examples of sentences produced by Nemesio González, one of the last semi-speakers of the language. Only when the examples are taken from secondary sources is their respective citation given; examples with no reference citation correspond to Nemesio's production and were collected personally by the author of this article.

¹ Chicha is an alcoholic drink made of fermented corn.

---

References

Castro, D. & Murillo, J. M., & Quesada, J. D. (2010). Evolución morfosintáctica en las lenguas chibchas. Letras, 45,25-68.

Castro, D. (2008). Brúnkahk Tek: An extinct language. Letras, 43,51-74.

Castro, D. (2010). The noun phrase in Boruca: Studying a recently extinct language. Sprachtypologie und Universalienforschung, 63,196-220.

Castro, D. (in press). The Boruca verbal dimension. Amerindia.

Constenla, A. & Maroto, E. S. (1986). Leyendas y tradiciones borucas. San José, Costa Rica: Editorial Universidad de Costa Rica.

Lehmann, C. (1988). Towards a typology of clause linkage. In J. Haiman & S. Thompson (Eds.). Clause Combining in Grammar and Discourse, pp. 181-225. Amsterdam: Benjamins.

Quesada Pacheco, M. A. (1995). Hablemos Boruca (Chá ding di' tégat tegrá). San José, Costa Rica: Ministerio de Educación Pública.

Quesada Pacheco, M. A. (1996). Narraciones borucas. San José, Costa Rica: Ministerio de Educación Pública.

Quesada, J. D. (2000). A Grammar of Teribe. Munich: Lincom-Europa.

Solano Salazar, E. (2000). La población indígena en Costa Rica según el censo 2000. San José: Centro Centroamericano de Población. Retrieved May 14, 2007, from http://www.ccp.ucr.ac.cr/noticias/simposio/pdf/solano.pdf