Optimizing Display Other Formats:      PDF - 782k

Volume 10 Issue 1 (2012)        DOI:10.1349/PS1.1537-0852.A.409

Note: Linguistic Discovery uses Unicode characters to represent phonetic symbols. Please see Optimizing Display for requirements to accurately reproduce this page.

Acquiring Complexity: The Portuguese of Some Pirahã Men

Jeanette Sakel

University of the West of England, Bristol

The Pirahã language has been claimed to have no syntactic complexity. What happens when speakers of this language come into contact with another, more complex language? This paper reflects on the Portuguese used by a group of men of the Amazonian Pirahã people.

My study shows that when speaking Portuguese, most Pirahã speakers employ simple syntactic constructions, characterised by juxtaposition of main clauses rather than embedding. Yet, the more proficient speakers utilize constructions that on the surface look more complex. These involve Portuguese subordinating conjunctions and complement clauses, both instances that could be analysed as complex constructions. While the subordinating conjunctions can be explained in terms of transfer and discourse marking functions, one particular speaker uses a Portuguese complement clause that could be analysed as a syntactically intermediate structure between Pirahã juxtaposition and Portuguese embedding.

1. Introduction ^[1]

This paper explores the way in which complex concepts are expressed in the learner varieties of Portuguese spoken by some Pirahã men. While most members of the language community are monolingual in Pirahã, a number of middle-aged men speak basic varieties of Portuguese that they have acquired in order to communicate with outsiders. I refer to these speakers as ‘gatekeepers’ in the sense that they are taking over communication with the outside world and thereby act as gatekeepers within their community. The Portuguese they speak could be analysed as a pidgin, as it is aiding communication with outsiders for trade and related purposes. Yet, the speakers have very different proficiencies, and focusing on the different learner varieties is more appropriate for the purpose of this study. Essentially, some speakers only know a few words, while others are able to use a range of Portuguese constructions.

The question arises how speakers of a language with very little complexity deal with recursive structures in a second language, such as they occur in Portuguese. When speaking about ‘recursion’, I use the term in a general way to refer to syntactic complexity, which can be seen in the subordination of clauses. I do not wish to embark on a theoretical discussion of the concept of recursion within generative grammar. As most gatekeepers have low proficiencies in Portuguese, we would not expect much complexity in their language. Yet, there are a number of indicators of more complex ‘recursive’ Portuguese structures being used by Pirahã speakers.

2. Pirahã

Pirahã is the last surviving member of the Muran language family, spoken by approximately 450 native people on the Maici River in the Brazilian state of Amazonas. My corpus consists of data from two different Pirahã settlements: the majority of the data are from the village of Forquila Grande on the upper Maici River, where I conducted fieldwork. I am including additional data from one speaker from a Pirahã settlement at the mouth of the Maici River, which have been made available to me by Dan Everett, who recorded the data during his fieldwork in 2009. ^[2]

The Pirahã phonology is simple in that there are only three vowel phonemes and maximally seven consonant phonemes (women, it is argued, use only six consonant phonemes, Everett 1979). Yet, there is considerable allophonic variation, even across phonemes in some idiolects, e.g. between p, t and k in koʔopai - koʔotai - koʔokai ‘stomach’ (Everett 1986: 316). This variation is also frequent in Portuguese loanwords, which are all fully integrated into the sound system of Pirahã. ^[3] Pirahã has additionally two phonemic tones (Everett 1986: 312; 1979). Following Everett’s (1979; 1986) conventions, I indicate the two-way phonemic tonal distinction by an acute accent on phonemic high tones, while phonemic low tones are left unmarked.

While most of the grammar of Pirahã is simple, the verbal morphology is relatively complex. The verb is not marked for tense or person, yet there are a variety of markers for mood (including evidentiality), aspect, interrogative markers, negation, incorporation and optional temporal distance markers. The latter relate the action or event to the actual time of speaking and can be roughly translated as ‘when’. In (1), the temporal distance marker - so indicates distance from the present, i.e. the action is either taking place in the past or the future (there is no marking for whether this action is realis or irrealis):

This marker was originally analysed by Everett (1986: 263) as a temporal adverbial clause marker. In his more recent approach, Everett (2005: 630) still analyses ‑- so as temporal, but not engaging in embedding, a point which was challenged by Nevins et al. (2009: 379). My fieldwork data on Pirahã confirm that - so is used to express that something is not the case at the time of speaking, either in the past or future. For this reason I refer to - so as a temporal distance marker (Sakel and Stapert 2010: 9). The marker is, however, optional in Pirahã and does not have to be present to express temporal relations between clauses.

Situational, aspectual or evidential suffixes in the Pirahã verb can take over roles that are commonly expressed by embedding in languages such as English. In this way, the Pirahã verbal suffix - há can express a meaning covered by more complex constructions in other languages, e.g. the complex English mental verb construction ‘be sure that’ (2) (Sakel and Stapert 2010: 8):

Where languages such as Portuguese and English would use complex clause combinations, such as to express causal and temporal relationships, Pirahã merely juxtaposes two simple clauses. The relationships between these are generally deducible from the speech context. In this way, the causal and simultaneous temporal relations between the two clauses in examples (1) and (3) are communicated by mere juxtaposition, without any syntactic complexity:

At the discourse level, Pirahã makes use of quotation markers, such as the formula higaisai. This marker and the quotation that follows can again be analysed as two juxtaposed clauses, in that the quotation is expressed in direct speech. This is similar to the English marker of youth speech like in she was like “sure” (Tagliamonte 2005).

The examples above show little - if any - syntactic complexity. The question as to whether there are syntactically complex structures in Pirahã, however, is highly disputed. The language is at the centre of a debate about whether its structures counterprove generally assumed linguistic universals, following Everett’s (2005) claim that Pirahã lacks syntactic complexity, in particular the generative notion of recursion. This claim challenges alleged universals of human language presented by Hauser, Chomsky and Fitch (2002). There have been numerous responses to the claim over the last years, including Sakel (2007a), Slobin (2007), Nevins, Pesetsky and Rodrigues (2009) and Sakel and Stapert (2010). Some authors, including Nevins et al. (2009), have argued that Everett’s older publications on Pirahã, in particular a grammatical sketch from 1986, show that the language is clearly recursive. In his claim, Everett (2005) discusses the absence of recursion and a range of other constructions in Pirahã in relation to the Immediacy of Experience Principle, which entails that the Pirahã only talk about what they have experienced themselves or what has happened to people they know. This principle, Everett (2005) says, can also be used to explain why the Pirahã have remained largely monolingual, even though they have been in contact with Portuguese-speaking outsiders for several centuries.

My findings show that monolingualism has not limited lexical borrowing from Portuguese, which can be seen in the wealth of Portuguese loanwords in Pirahã (e.g. in Everett’s 1986 data). Introducing loanwords to a language does not necessarily require bilingualism, and indeed the majority of Pirahã speakers are monolingual. The bulk of Portuguese loanwords denote items introduced by outsiders, such as houses, boats, coffee and sweets. These items were referred to in Portuguese by traders and other visitors, and the Pirahã readily adopted the new words into their language, amid phonological integration.

3. Portuguese

Portuguese is a world language with approximately 178 million speakers (according to the Ethnologue). This number includes first language speakers only, so if we should include L2 speakers (such as the Pirahã gatekeepers) the overall number of speakers would be somewhat higher. About 80% of Portuguese speakers speak Brazilian Portuguese (Azevedo 2005: 1).

Like other Romance languages, Portuguese uses complex grammatical means to express some relationships between clauses. There are complement clauses, relative clauses and adverbial clauses. For example, to express causal relationships, Portuguese uses subordinating conjunctions such as como, introducing an adverbial clause:

Cf. Azevedo (2005: 131)

There are other ways in which syntactic complexity is expressed in Portuguese. Under some conditions, the verbs of Portuguese subordinate clauses appear in the subjunctive (Azevedo 2005: 139). Yet, in vernacular Brazilian Portuguese the indicative is used almost indiscriminately (Azevedo 2005: 241). This has implications for the complex constructions we can expect the Pirahã to know, as they have probably not heard the subjunctive being used much at all.

Another trait of vernacular Brazilian Portuguese is the use of simple, juxtaposed constructions in cases where more formal varieties of Portuguese would use subordinations. In this way, sequential clauses can be strung together with the sequential discourse marker aí. According to de Oliveira e Silva & de Macedo 1992), this marker can be described as an ‘oral paragraph’. Each appearance of aí reflects a new sequence (or ‘paragraph’) in the events (5):

4. The Portuguese Learner Varieties of the Gatekeepers

I interviewed a range of Pirahã men who I would classify as ‘gatekeepers’. I furthermore analysed recordings by a film crew preparing a documentary on Dan Everett’s work. ^[5] Overall, the Portuguese proficiency of the Pirahã gatekeepers differs, ranging from knowing a number of simple formulas to using more advanced lexical and grammatical structures. The two gatekeepers with the highest proficiency are GK1 and GK2:

• I interviewed GK1 in Forquila Grande, a village on the central Maici River. He was my main point of contact, which I attribute to the fact that he acted as the ‘main gatekeeper’ in the village. He always came to help out when I interviewed other gatekeepers or other speakers, unless he was otherwise occupied. His community is in regular contact with speakers of Portuguese, but this contact is usually short-lived.
• GK2 lives at the mouth of the Maici River, an area with more frequent and sustained contact with speakers of Portuguese. I evaluated video recordings of him negotiating with outsiders in front of his entire village. He stands out as the main gatekeeper of his community.

Common to all gatekeepers is the fact that the phonological structures of their learner varieties are influenced by Pirahã. Most speakers only use a very restricted number of vowels and consonants with allophonic variation, similar to Pirahã.

All gatekeepers who know more than a few formulas make use of a range of Portuguese discourse markers, such as the sequential connector aí. This marker is used very frequently to link clauses and introduce new ideas, but it is also employed as a hesitation marker and a marker of other discourse functions. Its popularity with Pirahã speakers may be due not only to its function, but also the fact that it fits well into the phonological system of Pirahã without undergoing any sound changes. ^[6]

Grammatically, the learner varieties display simple structures, some of which can be traced back to the Pirahã substrate. For example, Pirahã lacks numerals (Gordon 2004, Frank et al. 2008) and expresses quantities by using a three-way distinction of ‘few’, ‘some’ and ‘many’. The gatekeepers appear to transfer the conceptual system of Pirahã into the learner varieties (Sakel 2012). As an example, a gatekeeper answered the question of how many children he has using muito ‘much, many’ (6), (Sakel 2012):

The language of the gatekeepers is also characterised by a degree of code-switching in situations where they can adopt a bilingual mode (Grosjean 2008), such as when the interviewer understands both Pirahã and Portuguese, for example (7):

5. Complexity in Learner and Contact Varieties

Before addressing the way complexity is expressed in the learner varieties of the gatekeepers, I want to look briefly at complexity in incipient second language acquisition. According to the processability model in second language acquisition (Pienemann 1999), syntactically complex structures appear at a late stage in the acquisition process. Early learner varieties, as well as pidgins, often display simple syntactic structures with little or no embedding (Holm 1988: 6). Yet, some pidgins have been argued to have subordinate structures, but these have usually developed into expanded pidgins, used as vernaculars among the speakers, such as Tok Pisin (Mufwene 2008: 76). Incipient pidgins, on the other hand, reply heavily on the pragmatic context instead of using grammatical structures (Mufwene 2008: 91).

Another feature common to early learner varieties is a heavy reliance on transfer for the L1 (see Lefebvre, White and Jourdan 2006). This is indeed also the case with the gatekeepers, for example in expressions of quantity (Sakel 2012), see (6) above.

A different picture emerges when looking at findings on the typology of grammatical borrowing in longer-term bilingual situations (Matras 2007; Matras and Sakel 2007a; Sakel 2007b). We see subordinating conjunctions - alongside other function words - being frequently borrowed in language contact situations, are often embedded in their source language constructions. This indicates that some complex structures are easily borrowable when speakers have a command of both languages. For example, the Amazonian language Mosetén (Mosetenan, Bolivia) has borrowed a number of Spanish subordinating conjunctions together with the Spanish strategy of embedding (8a). The Spanish strategy differs substantially from the Mosetén construction shown in (8b) in that the Spanish subordinate clause is introduced by a subordinating conjunction, while Mosetén uses a postclitic that appears on the main verb of the subordinate clause:

Yet, Spanish elements and their constructions borrowed into Mosetén often display different functions from their source-language originals. For example, the Spanish adverbial clause marker hasta ‘until’ is used in Mosetén to express both ‘until’ and ‘when’. This can be explained by contact-induced grammaticalisation, where a construction can go through some changes after being replicated in another language (cf. Heine and Kuteva 2005; Matras and Sakel 2007b).

6. Complexity in the Language of the Gatekeepers

What do the gatekeepers make of the complex structures they encounter in Portuguese? We could assume that, speaking a language that does not have recursion, the gatekeepers would not use any of the Portuguese complex structures.

As expected, I found that complex relations are frequently expressed by the juxtaposition of two independent clauses, relying on the pragmatic context for clarification. However, some of the gatekeepers also make use of Portuguese subordination markers.

Starting with the first, juxtaposition of independent clauses is very frequent in my data and is used to express a wide range of meanings. For example, a causal relationship between two clauses can be expressed by juxtaposing the two independent clauses. The causal meaning is understood in the context of what is said, rather than marked grammatically (9):

The speaker says that ‘the dog can not hunt because it is ill’, which is the speaker’s reply to my question if the dog will follow them on their hunt the next day. He expresses this by juxtaposing the clauses ‘now he does not hunt’ and ‘he is ill’. The optional Portuguese sequential discourse marker aí appears at the end of each clause in this example and indicates that a new sequence is to follow. Likewise, a hypothetical conditional meaning can be expressed by juxtaposing two clauses, even without the use of aí (10):

The overall meaning in (9) and (10) is deducible from the speech context, but in other cases, the context is not enough to reveal the exact intended meaning of the juxtaposition of two clauses. For example, the last clause in (11) could be understood as ‘it is good to eat’ or ‘it is good, I eat it’. Only the use of the sequential discourse marker aí suggests that the second reading is more appropriate, as aí appears before a new sequence:

The strategy of juxtaposing two clauses can also be used when dealing with verbs such as ‘want’, which would be expressed by a complement clause in target-like Portuguese (12):

Example (12), expressed by the proficient GK2, consists of three simple clauses, one after the other: the first one gives the setting ‘CIMI (an organisation) is leaving now’, the next gives the wish ‘CIMI to enter’, followed by ‘I want (it)’. Intonation presents us with a vital clue in this case, as the latter part of the example appears almost like an afterthought, and instead of expressing this as a complement clause as in Portuguese, the main clause ‘I want (it)’ is juxtaposed to the subordinate clause ‘CIMI enters’. However, a point to consider is the use of the infinitive form entrar in the example. As we are dealing with one of the most proficient speakers, I would have expected him to use the finite verb form entra in a simple clause, in particular since this speaker uses the appropriate inflections in other contexts, e.g. deixa ‘it leaves’ and quero ‘I want’. ^[10] Admittedly, the two forms entrar ‘enter.INF’ and entra ‘enter.3SG’ are very close in pronunciation, and the speaker may have used the infinitive assuming it was a finite verb form. This possibility aside, we may have an indication that we are dealing with more than just juxtaposition, arising complexity in the Portuguese of a gatekeeper, using a non-finite clause as a complement to the verb ‘to want’. In this case, we would observe that the resulting construction is very different from the Portuguese source. Yet, it would be expected, in line with contact-induced grammaticalisation in other contact situations (Heine and Kuteva 2005; see also the discussion on Mosetén above).

The same proficient speaker uses querer ‘want’ in another way, which again suggests we are dealing with more than juxtaposition. In this case, the construction looks suspiciously like a complement clause (13):

In (13) the verb ‘want’ is in the conditional, followed by a second verb in the infinitive. Such a construction is commonly used by a range of gatekeepers. In these examples, the verbs always have the same subject as a catenative verb construction (14): ^[11]

Yet, the catenative verb construction of the type ‘you would want to come here’ is not a likely reading for (13). ^[12] Rather, throughout the conversation, the gatekeepers tries to convince his Brazilian interlocutor that he wants the organisation (CIMI) to come to the area. The most likely reading for (13) is thus ‘I want CIMI to come here’, which is a complement clause in that the subject of ‘want’ is different from the subject of ‘come’. In favour of this analysis is that the finite verb form quería is in the 1^st person singular, ^[13] and that this speaker generally uses the appropriate inflections in finite verbs.

When we compare this construction to complementation in Portuguese, various differences become apparent. The latter appears with an overt complementizer que and an inflected subordinate verb form. Furthermore, an overt personal pronoun is present where the referential context is not entirely clear. The construction in (13) is thus not target-like, but a manifestation of GK2’s interlanguage, undergoing a form of contact-induced grammaticalisation. I suspect that this example has its roots in catenative verb constructions as in (14), which were then extended to be used for different subjects. So far, however, this is the only example I have found in my data.

While juxtaposition of clauses alone can express complex constructions, many gatekeepers also use three Portuguese subordination markers: porque ‘because’, quando ‘when’ and por isso ‘therefore’.

I have only two examples of porque ‘because’, both uttered by the same speaker in close succession as parts of question-answer pairs as in (15), and hence not examples of embedding:

In the same way as porque, the reason marker por isso ‘therefore’ is used by a gatekeeper (in this case one of the most proficient speakers). In (16) the gatekeeper uses it to explain why he would never go to Humaitá, the nearest city, by boat:

The marker por isso takes the preceding statements as its antecedent, and could be analysed as a marker of syntactic embedding. However, por isso is often used as a conclusive discourse marker in Portuguese (similar to the English marker so), appearing at the beginning of an utterance. This type of marker could appear with or without an antecedent, but rather than being a case of syntactic complexity, it could be analysed as subordinate to the pragmatic context (cf. Schiffrin 1988: 191ff). I think that this is the most likely reading in this case, which is also supported by the intonation in this example. Yet, this use of por isso in (16) is only a small step away from a syntactically complex construction.

Out of the three possible subordination markers in my corpus, quando ‘when’ occurs most frequently, in most cases uttered by the proficient speaker GK1. Example (17) shows the temporal marker quando ‘when’ being used between two clauses:

In (17) quando refers to an event or action that is not the case at the time of speaking. In that way, its use overlaps in meaning with the Pirahã temporal distance marker ‑ so. It is highly likely that rather than using quando in its Portuguese sense, the gatekeeper transfers the functions of ‑ so to its Portuguese equivalent. This would mean that even though an adverbial clause marker is used in (17), the example does not show a subordinate structure. It merely follows the system of Pirahã temporal distance marking. Another example of possible transfer due to an overlap in meaning between quando and - so is (18) (from Sakel and Stapert 2010: 9):

Yet, the use of quando by GK1 does not always merely follow the Pirahã pattern. In (19), quando is used to contrast two hypothetical situations without directly relating them to the speech context: ^[15]

This example (19) could be seen as a step away from the transferred Pirahã structure based on the temporal distance marker ‑ so. Yet, there is no convincing case to be made that this is a syntactically complex structure.

7. Discussion and Conclusion

My findings show that some degree of syntactic complexity can indeed enter the language of speakers with a syntactically simple L1 that is in contact with a more complex L2. Yet, the majority of gatekeepers only use simple constructions in their Portuguese learner varieties, in line with findings from other early learner varieties and pidgins. They generally juxtapose simple clauses to express complex concepts, relying on the pragmatic context to clarify what is meant. Even a variety of cases where gatekeepers use Portuguese subordinating conjunctions cannot be argued to be complex structures. For example, the conjunctions por isso ‘therefore’ and porque ‘because’ are used as discourse markers in the learner varieties. The properties of the marker quando ‘when’ can, at least in part, be explained in terms of conceptual transfer from Pirahã.

It is not surprising to find little syntactic complexity in the language of the gatekeepers, as even the vernacular Brazilian Portuguese input uses simple constructions, where more formal language varieties would employ subordinations.

Yet, my data show that one speaker, GK2, uses a complement clause which appears suspiciously complex. It can be analysed as an extension of a catenative verb construction and could allow for an analysis as a syntactically complex structure. The resulting construction is not similar to the target-like Portuguese way of expressing complementation. Rather, the construction is intermediate between Pirahã and Portuguese, arguably undergoing contact-induced grammaticalisation. The gatekeeper who uses this construction is one of the most proficient Portuguese speakers among the Pirahã and he is in regular contact with outsiders. The arising syntactic complexity in his variety of Portuguese is probably something that we will eventually also see in other speakers as well. This is because the language situation in the Pirahã region is likely to change considerably over the coming years, due to prolonged contact with the outside world, including a new hospital and education projects. The Portuguese proficiency of the gatekeepers is likely to increase, and with it - more than likely - also the complexity of their learner varieties.

Abbreviations

1: 1^ST PERSON; 2: 2^ND PERSON; 3: 3^RD PERSON; AD: ADESSIVE RELATION ‘AT’; AMC: ASSOCIATED MOTION ‘DO AN ACTION ON THE WAY TO THE DEICTIC CENTRE’; AMS: ASSOCIATED MOTION ‘DO AN ACTION AFTER ARRIVAL TO THE DEICTIC CENTRE’; COMP_CERT: COMPLETE CERTAINTY; CON: CONTRASTIVE CONNECTOR; COND: CONDITIONAL; DEF: DEFINITE ARTICLE; DM: DISCOURSE MARKER; F: FEMININE; GK: GATEKEEPER, FOLLOWED BY INDIVIDUAL IDENTIFICATION NUMBER; IMP: IMPERATIVE; INF: INFINITIVE; INTENS: INTENSIFIER; INTENT: INTENTION; IR: IRREALIS; LOC: LOCATIVE; M: MASCULINE; NEG: NEGATION; NEG.IMP: NEGATIVE IMPERATIVE; O: OBJECT; PAST: PAST TENSE; PERF: PERFECTIVE; PL: PLURAL; REL_CERT: RELATIVE CERTAINTY; REM: REMOTE; RD: REDUPLICATION; S: SUBJECT; SG: SINGULAR; SPAN: SPANISH FORM; SUP: SUPERESSIVE RELATION ‘ON’; TAG: TAG QUESTION MARKER; TEMP: TEMPORAL RELATIONAL MARKER; UNCERT: UNCERTAINTY; VERB: VERBALIZER.

References

Bakker, Peter. 2009. ‘Creoles versus languages with little morphology’. Conference presentation at Morphologies in Contact, 3.10.2009, Bremen, Germany.

Chomsky, Noam. 1995. The minimalist program. Cambridge: MIT Press.

de Oliveira e Silva, Giselle M. and Alzira Tavares de Macedo (1992) ‘Discourse markers in the spoken Portuguese of Rio de Janeiro’ in Language Variation and Change 4: 235-249. doi:10.1017/s0954394500000776

Everett, Daniel. 1979. Aspectos da fonologia do Pirahã. MA thesis at the Universidade de Campinas, Brazil.

-----. 1986. ‘Pirahã’ p. 200-325 in Desmond Derbyshire and Geoffrey Pullum (eds.) Handbook of Amazonian languages, vol. 1. Berlin: Mouton de Gruyter.

-----. 2005. ‘Cultural Constraints on Grammar and Cognition in Pirahã: Another Look at the Design Features of Human Language’, in Current Anthropology 76,4: 621-646. doi:10.1086/431525

Frank, Michael; Daniel Everett; Evelina Fedorenko and Edward Gibson. 2008. ‘Number as a cognitive technology: Evidence from Pirahã language and cognition’ in Cognition 108(3): 819-824. doi:10.1016/j.cognition.2008.04.007

Gordon, Peter. 2004. ‘Numerical cognition without words: evidence from Amazonia’ Science 306: 496-499. doi:10.1126/science.1094492

Grosjean, François. 2008. Studying bilinguals. Oxford: Oxford University Press.

Hauser, Marc D.; Noam Chomsky and W. Tecumseh Fitch. 2002. ‘The Faculty of Language: What Is It, Who Has It, and How Did It Evolve?’ in Science 298: 1569-1579. doi:10.1126/science.298.5598.1569

Heine, Bernd and Tania Kuteva. 2005. Language contact and grammatical change . Cambridge: Cambridge University Press.

Holm, John. 1988. Pidgins and creoles, vol. 1 theory and structure. Cambridge: Cambridge University Press.

Jarvis, Scott and Aneta Pavlenko. 2008. Crosslinguistics influence in language and cognition. New York: Routledge.

Karlsson, Fred. 2007. ‘Constraints on multiple initial embedding of clauses’ International Journal of Corpus Linguistics 12(1): 107–118. doi:10.1075/ijcl.12.1.07kar

Laury, Ritva and Ono Tsuyoshi. 2010. ‘Recursion in conversation: what speakers of Finnish and Japanese know how to do’ p. 69-92 in Harry Van der Hulst (ed.) Recursion and human language. Berlin: Mouton de Gruyter. doi:10.1515/9783110219258.69

Lefebvre, Claire, Lydia White and Christine Jourdan. 2006. ‘Introduction’, p. 1-14 in Lefebvre, Claire, Lydia White and Christine Jourdan (eds.) L2 acquisition and creole genesis. Amsterdam: Benjamins.

Matras, Yaron. 1998. ‘Utterance modifiers and universals of grammatical borrowing’. In: Linguistics 36(2): 281-331. doi:10.1515/ling.1998.36.2.281

-----. 2007. ‘The borrowability of structural categories’ p. 31-73 in Yaron Matras and Jeanette Sakel (eds.) Grammatical borrowing in cross-linguistic perspective. Berlin: Mouton de Gruyter.

-----. 2009. Language contact. Cambridge: Cambridge University Press.

Matras, Yaron and Jeanette Sakel. 2007a. (eds.) Grammatical borrowing in cross-linguistic perspective. Berlin: Mouton de Gruyter.

-----. 2007b. ‘Investigating the mechanisms of pattern replication in language convergence’ in Studies in Language 31,4: 829-865. doi:10.1075/sl.31.4.05mat

Mufwene, Salikoko. 2008. Language evolution - contact, competition and change . London: Continuum.

Nevins, Andrew, David Pestsky and Cilene Rodrigues. 2009. ‘Pirahã exceptionality: a reassessment’ in Language 85: 355-404. 10.1353/lan.0.0107

Pienemann, Manfred. 1999. Language processing and second language development: processability theory . Amsterdam: John Benjamins.

Pullum, Geoffrey and Barbara Scholz. 2010. ‘Recursion and the infinitude claim’ p. 113-138 in Harry Van der Hulst (ed.) Recursion and human language . Berlin: Mouton de Gruyter.

Slobin, Dan. 2007. ‘Context and comments on Dan Everett’s claims, In: Letters to the Editor, Human Development 13. [http://www.online.karger.com]

Sakel, Jeanette. 2007a. ‘On Pirahã and Everett’s claims about recursion’. In: Letters to the Editor, Human Development 13. [http://www.online.karger.com]

-----. 2007b. ‘Types of contact phenomena: matter and pattern distribution’ in Yaron Matras and Jeanette Sakel (eds.) Grammatical borrowing in cross-linguistic perspective. Berlin: Mouton de Gruyter.

-----. 2007c. ‘Mosetén borrowing from Spanish’ p. 567-580 in Yaron Matras and Jeanette Sakel (eds) Grammatical borrowing in cross-linguistic perspective . Berlin: Mouton de Gruyter.

-----. 2012. ‘Transfer and language contact: the case of Pirahã’ in Jeanette Sakel and Jeanine Treffers-Daller (eds.) ‘Transfer and language contact’, special issue of International Journal of Bilingualism.

Sakel, Jeanette and Eugenie Stapert. 2010. ‘Pirahã - in need of recursive syntax?’ p. 3-16 in Harry Van der Hulst (ed.) Recursion and human language. Berlin: Mouton de Gruyter.

Schiffrin, Deborah. 1988. Discourse markers. Cambridge: Cambridge University Press.

Tagliamonte, Sali. 2005. So who? Like how? Just what?: Discourse markers in the conversations of young Canadians. Journal of Pragmatics 37 (11): 1896-1915.

Thomason, Sarah and Daniel Everett. 2001. ‘pronoun borrowing’ in Proceedings of the Berkley Linguistic Society 27: 301-315.

Tomalin, Marcus. 2007. ‘Reconsidering recursion in syntactic theory’ Lingua 117(10): 1784-1800. doi:10.1016/j.lingua.2006.11.001

Wray, Alison. 2002. Formulaic Language and the lexicon. Cambridge: Cambridge University Press.

Van der Hulst, Harry. 2010. (ed.) Recursion and human language . Berlin: Mouton de Gruyter.

Author’s Contact Information:
Jeanette Sakel
Linguistics
University of the West of England
Frenchay Campus
Coldharbour Lane
Bristol
BS16 1QY
UK
jeanette.sakel@uwe.ac.uk

^[1] I am grateful to Dan Everett, Eugenie Stapert, Mike Frank, Ted Gibson, Nigel Vincent, Jeanine Treffers-Daller and Raquel Guirardello-Damian, with whom I have discussed aspects of this paper.

^[2] I am grateful to Dan Everett for making these data available to me.

^[3] In my transcription I give the pronunciation used by the speaker.

^[4] Portuguese elements in the speech of the gatekeepers are presented in capital letters. In this example, the discourse marker AÍ is borrowed from Portuguese (cf. also discussion further below).

^[5] These recording were kindly made available to me by Dan Everett.

^[6] Indeed, due to its phonology and frequent use by the Pirahã speakers I worked with, I assumed it was a native Pirahã discourse marker at first (Sakel & Stapert 2010; Sakel 2012).

^[7] Portuguese elements are given in capital letters in the examples.

^[8] These data are from my own fieldwork corpus of Mosetén.

^[9] Examples 10, 12, 13 and 14 were recorded by Dan Everett at the Pirahã settlement at mouth of the River Maici in 2009.

^[10] Cf., however, example (10), where the same gatekeeper uses the 3^rd person singular form pode with the first person pronoun eu.

^[11] Other examples are modal verbs + infinitive constructions. In my data these are pode vir ‘can come’, pode quee ‘can stay’, pode aader ‘can walk’ and tene preeder Pirahã ‘must learn Pirahã’. These were probably first acquired as formulas (in terms of Wray 2002).

^[12] In some dialects of English the phrase you’ll want to come could be understood as ‘you should come’ or ‘I want you to come’, i.e. little to do with the wishes of the listener, but rather those of the speaker. This reading is not possible in Brazilian Portuguese and it is thus unlikely to be the interpretation in this example.

^[13] For the sake of completeness the reader should notice that due to paradigm syncretism this is also the form of the 3^rd person singular conditional, though in this context there is no likely antecedent for a 3^rd person singular.

^[14] The translation of (16) is ambiguous, as the verb chegar ‘go away’, can both mean ‘go there’ and ‘come here’ in the pidgin, while traer ‘bring’ can also mean ‘give’.

^[15] The code-switches in this example appear because the gatekeeper explains the meanings of two Pirahã words to me.