Pragmatic Uses of Negation in Chipileño Spanish (Mexico)

: This paper discusses two negation types (standard negation (SN), negative doubling (ND)) in Chipileño Spanish, a variety that has emerged as a result of contact between Spanish and Veneto (an Italo-Romance language) in Mexico. In Veneto, negation can be formed in two ways: preverbally (SN) and as a negative doubling (ND). Based on sporadic observation, bilingual speakers of Spanish and Veneto transfer a ﬁnal no while speaking Spanish, a language that does not allow repetition of the same negator in the postverbal position. Using both a spontaneous and a controlled tasks, the results show two possibilities: preverbal negation only ( no vino ‘[S / he] did not come’) and sentence ﬁnal ( no me gusta no ‘I do not like’) in both tasks. This study compares the ﬁndings from Chipileño Spanish to the other Romance varieties that exhibit similar cases of negation, while discussing its scope and relevance to discourse-pragmatic factors.


Introduction
According to Zeijlstra (2007), "a universal property of natural language is that every language is able to express negation . . . but it differs to quite a large extent as to how each language expresses this negation" (498). The word 'extent' refers to cross-linguistic variation in the form of the negative element, the position of the negative element, and its interpretation. It can also be applied to diachronic change or the three stages of Jespersen (1917) negation cycle, which focuses on the form and position of negators. The Jespersen cycle refers to a cyclical process, by which negative markers or sentence negators weaken, strengthen, and/or are replaced by another negative marker in a specific historical period (Bernini and Ramat 1996;Zanuttini 1997;Schwenter 2006;Zeijlstra 2007;Biberauer and Cyrino 2009). In his analysis of English, French, and Danish, Jespersen (1917) identified three stages or types of sentential negation to demonstrate how this process occurs: Phase 1: NEG I: NEG + Verb Phase 2: NEG II: NEG + Verb + NEG Phase 3: NEG III: Verb + NEG As seen above, the position of the negators corresponds to each phase of the cycle. In, Chipileño Spanish, a variety spoken in Mexico by descendants of Northern Italian immigrants, negation can be formed differently than in standard Spanish. In (1), the negation is formed preverbally, similar to the one observed in the standard variety. However, previous observations and the literature (Barnes 2009) also mention another alternative, as in (2) The cases of negative doubling (ND), demonstrated in (2), in which the final no has a falling intonational contour, a characteristic of a declarative sentence, are ungrammatical in standard Spanish but are possible in Veneto. 1 Given that Chipilo is a bilingual community of Veneto and Spanish speakers, bilingual speakers seem to transfer the final no from their first language, Veneto, where both negative markers are used: one preverbally and the second one at the end of the utterance, as in (3) Based on such observations, Chipileño Spanish seems to exhibit two cases of negation and 'belongs' to two separate stages of the Jespersen cycle. Therefore, this paper aims to answer the following research questions: (a) What is the typology of negation in Chipileño Spanish? (b) What is the pragmatic function of negation in Chipileño Spanish?
The paper is organized as follows: Section 1.1 provides an overview of negation in the Romance languages, focusing on the three stages of the Jespersen cycle; Section 2 describes the methodology of the project, including the two tasks and the participants' criteria. Section 3 presents the results of the two tasks, followed by Section 4, which discusses the findings, their limitations and future work.

Negation in Romance Languages
The Romance languages, such as standard Spanish, Portuguese, Romanian, and Italian, belong to the first stage of the Jespersen cycle, in which negation occurs preverbally. As seen in (4-7), the negator precedes the finite verb. Two other Romance languages, French and Catalan, known for ND, occupy Phase 2 of the Jespersen cycle. As in examples (8) and (9), the first negator precedes the finite verb, whereas the second one follows it. It is important to mention that in spoken Catalan, pas is omitted so that the language forms part of Phase 1 (Juan no parla italià 'Juan does not speak Italian'). The reason for the apparently retrograde change of Catalan from Phase 2 to Phase 1 may be the geographical proximity with Spanish, which only has one preverbal negator no. In other words, it is possible that the negators in some varieties (see also spoken French and some varieties of Italian) can weaken and/or strengthen over a period of time, due to close language contact between the two dialects or languages. As a result, these varieties can correspond to two phases of the Jespersen cycle.
Northern varieties of Italian, specifically western Piedmontese/Lombard, Valdôtain, Milanese, and Pavese, also form part of Phase 2 of the Jespersen Cycle, where both preverbal and postverbal negators are used, i.e., NEG II (Zanuttini 1997 However, the use of some postverbal negators (pa in Piedmontese, pa in Cognen, mia in Pavese and minga in Milanese) brings a different interpretation and function to the negative sentence, specifically a presuppositional reading. Implicit presupposition refers to the contexts in "which [the] non-negative counterpart of the proposition expressed by the sentence is assumed in the discourse", i.e., entailed by the common ground (Zanuttini 1997, p. 61 Other postverbal negators in these dialects have no discourse status and are used as 'regular' negators without any presupposition. In addition, they occupy a different position than presuppositional postverbal markers. For instance, when adverbs are present, presupposition markers always occupy a structurally higher position than those markers, which do not carry any interpretation (Zanuttini 1997, p. 60). In this paper, I treat presupposition as a negative mention.
Finally, the dialects of Cembra and Lisagnago, spoken in the Trentino region, also form part of Phase 2, NEG II. These dialects contain two negators, one preverbally and another sentence finally, as in (11) and (12) have ever no more a pair of socks NEG 'I don't even have a pair of socks anymore.' (Zanuttini 1997, p. 97) According to Zannuttini's observations (Zanuttini 1997), the final no in these dialects is used in restricted presuppositional contexts only, specifically in focused emphatic constructions (96-97).
As for Phase 3, NEG III, in informal speech, speakers of standard French and Italian omit the preverbal negator and only use the postverbal one. In French, ne is omitted so that French is considered part of Phase 3. In a study on Quebec French, Sankoff and Vincent (1980) found that ne was deleted 99.5% of the time, in comparison to a previous study of 1967, where the deletion rate was only 38%. An unspecified spoken variety of Italian, which is not considered a standard variety, also omits the preverbal negative marker and uses a postverbal marker instead (Zanuttini 1997). Furthermore, Bernini and Ramat (1996) show that for the past century there has been an increased use of NEG III in Italian colloquial popular varieties, specifically in the north (13), which omit the preverbal non.

Spoken Italian:
Politicamente [non] erano niente. Politically NEG be-PAST.3PL nothing 'As for politics, there were not worth anything. ' (cited in (Kim 2000, p. 21)) Overall, this subsection discussed different positions and interpretations of negative elements in the Romance languages, based on the three stages of the Jespersen's negation cycle. The next subsections focus specifically on negation in the two varieties that are in contact in Chipilo: Veneto and Spanish.

Negation in Veneto, Trentino Region
To my knowledge, there has not been a study analyzing the distribution of ND and factors favoring contexts with ND in Veneto. Cecilia Poletto (2016), who specializes in syntactic analysis of negation in Veneto, notes that the variety has not changed a lot with regard to negation patterns. She identifies three negators (miga, gninte, and no), which all occur postverbally, but serve different functions in a sentence. Miga, similar to mica in standard Italian, is strictly a presuppositional marker, whereas no has a semantic implicature favored in contexts with reinforcement and question-answer pairs. In Poletto's work (Poletto 2008), the final negator no in Veneto is a focus negator, which emphasizes or reinforces the negation, as in (37). It is important to keep in mind that the final no is not the main licensing negator; final no is not interpreted as an additional negator, but rather functions as negative concord licensed by a preverbal negator, which is interpreted. 14. Veneto: No ghe vado NO! NEG there go-1SG NEG 'I will not go there!' (Poletto 2008, p. 181) Examples like (14) refer to evidential contexts where both the speaker and the hearer share common ground about the negated event; in other words, both interlocutor and hearer know about the place where the interlocutor does not want to go. This is consistent with Poletto (2016), who claims that the use of the final no is favored with a previously negated context, which should be self-evident to both speaker and interlocutor. With regard to previous adjacent elements, Poletto claims that the final no is adjacent to and favored in contexts with the verb as a previous constituent, i.e., preceding the final no. This is similar to example (14), where vado 'I go' occurs before the final no.

Negation in Spanish
In standard Spanish, negation is preverbal, as in (15) The English translations of (16) and (17) are identical. In Spanish, similar to standard Italian, the position and interpretation are different. First, in (15), nadie 'nobody' occurs after the verbal phrase, whereas in (17), it is in the subject position, preceding the verb. In addition, in order for (16) to be a grammatical sentence with nadie 'nobody' in the postverbal position, the presence of preverbal no, a negative operator, is necessary. In contrast to (16), nadie 'nobody' in (17) does not require another negator to be grammatical, as it appears preverbally. Moreover, it would be ungrammatical to include no in the preverbal position with another negator 'nadie' in the subject position. 2 The position of the negators in the sentence can, however, be different, according to the focus of the sentence. For example, sentences similar to (16) are usually broad focused, whereas sentences which start with negative words, such as 'nadie', for example, can have a narrow focus on the negative constituent, require specific intonation, and are restricted to certain contexts, as in (18) Based on example (18), the interlocutor wants to emphasize ninguna 'none' by putting focus on the negative word, and therefore places ninguna in preverbal position. In contrast to (18), having the object nada 'nothing' in the postverbal position allows for a wider focus reading; therefore, it is more common to say nada 'nothing' in the original postverbal position: In summary, the main negator in Spanish is preverbal. To deepen the knowledge and understanding of negation in the variationist context, the next subsection discusses other cases of ND in the Romance languages.

Other Cases of ND in Romance Varieties
This subsection focuses on five other varieties-Brazilian Portuguese, Palenquero, Chocó Spanish, the Spanish of Argentinian Corrientes, and Minorcan Spanish-that also exhibit ND. Two varieties, the Spanish of Argentinian Corrientes and Minorcan Spanish are in close contact with other languages, namely Guaraní and Catalan, respectively. Brazilian Portuguese, unlike the above two varieties, historically has been in contact with a multitude of languages including African varieties (e.g., Kikongo), and today also exhibits ND (Johnson and Schwenter 2019;Schwegler 2016;Schwenter 2005). Palenquero creole and Chocó Spanish, the two varieties spoken in Colombia, also present different negation pattens (Schwegler 1991(Schwegler , 2016(Schwegler , 2018. Therefore, it is crucial to report data on these varieties, relevant as they are to the present work, so as to investigate the frequency and the contexts of ND usage in these varieties and compare them with the case of Chipileño Spanish.

Negation in Vernacular Brazilian Portuguese
Vernacular Brazilian Portuguese exhibits three ways to express sentential negation (Bernini and Ramat 1996;Biberauer and Cyrino 2009;Martínez 2006;Schwenter 2005Schwenter , 2006Teixeira de Sousa 2011, 2012Johnson and Schwenter 2019). According to Johnson and Schwenter (2019), all three variants convey the same semantic content (120).  From Sousa's analyzed corpus (Table 1), Sousa (2007) concludes that preverbal negation, as in (20), is the most frequent type (84%), whereas the other two cases of negation are not that common. Only preverbal negation as in (20) can occur in all contexts, and it is considered the canonical negation that can be used in 'out of the blue' contexts, when new information is presented (Johnson and Schwenter 2019, p. 120); the other two forms, however, are related to a specific information structure or context. According to Teixeira de Sousa (2012), cases with two negators or ND, as in (21), emphasize the sentential negation, where final no serves as a marker of polarity focus; by contrast, cases like (22) are used to deny or correct given information (Teixeira de Sousa 2013, p. 2). Schwenter (2005) also studied negation in Brazilian Portuguese and concluded that sentences with ND were restricted to occurrences with 'common ground' and propositional denials, to ensure the interlocutor interpreted the sentence (as in 23). Furthermore, the use of the sentence-final negator não is "restricted to denials of activated, salient discourse-old propositions" (Schwenter 2006, p. 13).

A.
Você viu esse programa? You-NOM watch-PAST that program? 'Did you watch that program?' B.
Não vi não NEG see-PAST,1SG NEG 'I did not watch it.' In example (23) the first interlocutor 'activates' or introduces the topic by posing a question; this becomes interpreted as a hearer-old context, therefore licensing ND in the response, B. (Schwenter 2006). Based on Schwenter (2005) claim, the second, final, não has similar pragmatic usage as mica in Standard Italian.
In a more recent study, while comparing NEG2 and NEG3, Johnson and Schwenter (2019) argue that both variants require dialogic contexts, typically denials of an interlocutor's previous assertion, but more generally require access to opposing viewpoints in discourse (120). Furthermore, ND is felicitous in inferable contexts, whereas NEG3 must be explicitly 'activated' to be felicitous. In (24), both ND and NEG3 are felicitous in Answer 1, because the proposition has been explicitly 'activated', and so both variants are licensed. As for Answer 2, only the option with ND is possible, since the negated proposition of "not going" was not overtly activated in the preceding question. Since ND is possible with inferable propositions, its use is licensed and felicitous.
In this paper, I adopt a slightly different approach, which investigates whether the final no is used as a reinforcement marker, and not as a denial of proposition. My previous pilot study (Tararova 2016) did, however, show that ND was used in contexts which had to be 'activated' or introduced by the previously immediate negated clause (standard negation (SN)).

Palenquero Creole
Palenquero creole, a variety that has been in contact with Spanish for several centuries exhibits three cases of negation, preverbal and two postverbal types of negation: ND and clause-final only negation, as seen in (25), in declarative, interrogative and imperative sentences 3 (Dieck 2000;Lipski 2018;Schwegler 1991Schwegler , 2016Schwegler , 2018.

NEG VP:
Ele nu ten monasito. NEG VP NEG: Ele nu ten monasito nu. VP NEG: Ele ten monasito nu. s/he-NOM NEG have-1SG child NEG 'He/she doesn't have a child.' (Schwegler 2016, p. 232) According to Schwegler (1991Schwegler ( , 2016Schwegler ( , 2018, these negators differ with regard to their frequency of use, as well as discourse-pragmatic and information-structural factors. Specifically, the preverbal negation appears in unmarked factual, matter-of-fact assertive statements, whereas both ND and NEG3 have to be 'activated' from a previous dialogic context, topic, or proposition, and used to contradict statements or previous assumptions. The two types differ, however, when it comes to the relative frequency of use and functional differences. As such, NEG3 is a more common and frequent negator in oral speech, than ND (Schwegler 2016, pp. 236-37). In his more recent article, Schwegler (2018) explains that Palequero's contemporary high use of NEG3 (especially for younger generations) is a result of their ethnic pride. Since ND and NEG3 are ungrammatical in standard Spanish, Palenqueros use these two variants, instead of SN, as it is associated with the Spanish language (275). 4 However, based on Lipski (2018) results, he found that young Palenquero speakers are different from adults when it comes to processing Palenquero clause-final negation. Specifically, his results show that the processing rate of NEG3 among the younger group was significantly lower than that of the older group. 3 Most often, ND in Palenquero occurs in imperative sentences and rarely in non-imperative utterances (Lipski 2018, p. 46). 4 Historically, all three NEG patterns were part and parcel of the creole grammar. This was true well into the 1990s; in the last two decades the efforts from locally-born teachers and middle-aged community leaders were put forward to revitalize the creole language (for more information, see (Schwegler 2018)). Furthermore, he concludes that young heritage Palenquero speakers who receive the creole input at home resemble more adult speakers, whereas the young speakers who receive the input from school only exhibit more Spanish-like morphosyntax when speaking in Palenquero (Lipski 2018, p. 54). As far as it concerns the functional differences, ND in Palenquero is felicitous in contexts with inferable or implicit propositions, as in (26) (Schwegler 2016, p. 247) In the example (26) above, there was no indication that A missed the party, NEG3 in B2 is infelicitous, thus resulting in only one possibility, ND. Furthermore, Schwegler concludes that the two types of negation are triggered by a differential focus; specifically, ND "tends to invite the possibility of further discussion or (re) consideration of TRUTH . . . meanwhile, NEG3 seeks to RESET the VERUM of a proposition for which there is an opposing viewpoint" (Schwegler 2016, p. 253).
Before moving to the next section, it is important to mention that Palenqueros are bilingual speakers of creole and Spanish. Schwegler (2018) briefly comments that these heritage bilinguals while speaking Spanish use SN for the most part, and ND sporadically with low frequency. Therefore, it seems like Palequeros are able to separate the domains of use of two languages, and not transfer the negation patterns from creole to Spanish. This observation is significant for the understanding of the present findings, particularly the frequency of use of the variants.

Chocó Spanish
Chocó Spanish, an Afro-Hispanic dialect, spoken in the Pacific lowlands of Colombia also exhibits cases of SN and ND negation (27), yet there is discrepancy in the number of variants present in Chocó speech (Schwegler 1991(Schwegler , 2018Sessarego 2017). Schwegler (1991Schwegler ( , 2018 discusses that all three types of negation exist in Chocó Spanish (28), whereas Sessarego's corpus (Sessarego 2017) (Sessarego 2017, p. 225) According to Sessarego (2017), ND is used as a rejection to affirmation and felicitous in contexts where a presupposition is available on the basis of the common ground. The author, furthermore, mentions that cases of NEG2 are still stigmatized, because they are characterized as Bozal speech or the speech of the slaves.
In contrast to Sessarego (2017), Schwegler (2018) maintains that all three types of negation are present in Chocó speech, but that their frequency of use varies considerably. Specifically, ND is used far 5 The editors of the journal note that a fluent adult Palenquero typically does not include the perfective particle 'a' in negative utterances (for more discussion, see (Dieck 2000). more than NEG3. The author notes, however, the fact these two variants (ND and NEG3) are heavily stigmatized, so these variants are rarely used in a recorded speech. In (29), the final no is clause-internal (with L%); in other words, it is not preceded by a pause, and therefore differs from Spanish tag questions (with H%), where the final no is separated by a pause from a previous constituent. In an experimental study of laboratory-controlled speech in Spanish, De Prada Pérez (2008) measured the F0 height of the final no to investigate whether the final no is clause internal (ND) with L% or clause external (tag questions) with H%. Given that this construction is attested only in Minorcan Catalan, the language that exhibits ND, and not present in standard Spanish, she proposed that L1, Catalan, had an effect on speakers' production, specifically in their production of final no. Her results show that speakers with L1 Catalan, in fact, produced utterances with a final contour, a characteristic of ND, whereas L1 Spanish speakers only produced affirmative sentences, such as utterances with a tag. In other words, she concludes that there is a transfer from L1 Catalan to Spanish of these bilinguals.

Negation in Argentinian Spanish-the Case of Corrientes
The Argentinian variety of Corrientes Spanish (CS) exhibits cases of ND, possibly due to contact with Guaraní, an indigenous language where ND exists. In their study, Cuervo and Mazzaro (2013) analyzed the alternative ways of using negation with the presence of negative words. In most dialects of Spanish, the example in (30) is the only possible variant to express the above negative sentence Nadie abrió la puerta 'Nobody opened the door', where negative word nadie 'nobody' occurs in the preverbal position, preceding the verb. 6 Example (31) is ungrammatical in most varieties, based on the interaction between the presence of a preverbal negative word and the presence of negator no in the Spanish negative concord system. Sentence (31) can only be grammatical in most varieties as a marked case with a double negation reading. However, in the Spanish of Corrientes, both examples are grammatical. Cuervo and Mazzaro (2013) found that three linguistic variables 6 In the given example, it is also possible to place nadie 'nobody' in the postverbal position: No abrió la puerta nadie 'Nobody opened the door'. Given that doubling in Corrientes Spanish consists of two negative elements in preverbal position, the variant with postverbal nadie is not discussed.
(specificity value of a negative word, distance in number of words between the negative word and preverbal no, and type of the negative word) favored the use of ND. Social factors (age and sex) however, had no effect on elicitation of ND, suggesting that the variation in Corrientes is stable and occurs with all social groups.
Taking these studies into consideration, this study also discusses a case of language contact, similar to other varieties such as the Spanish of Corrientes and Minorcan Spanish. Regarding the structure and position of the negators, declarative sentences with ND in Brazilian Portuguese and Minorcan Spanish have a falling contour on the final no, similar to Chipileño Spanish. Finally, Schwenter (2005Schwenter ( , 2006 and Schwegler (1991Schwegler ( , 2016Schwegler ( , 2018 proposals on pragmatic use of negators are taken as base for discussing negation in Chipilo.

Materials and Methods
One of the main challenges in examining morphosyntactic variation is obtaining a sufficient number of tokens necessary to analyze the data. The interview method might not be the ideal method, as a variable might not be used in a spontaneous 'free-topic' interview. Therefore, this study combines elicited conversational speech and experimental work to gain better insight on negation patterns in Chipileño Spanish. 7

Participants
A total of 79 bilingual participants of Veneto and Spanish were recruited for the study. The criteria to participate in the study were that they be born in Chipilo, over eighteen years old and currently reside in Chipilo. All subjects gave their informed written consent for inclusion before they participated in the study. The protocol (30514) was approved by the Social Sciences, Humanities, and Education Research Ethics Board (REB) at the University of Toronto, Canada.

Elicited Speech Conversation
The first recorded task was a short interview or elicited speech conversation, which sought to elicit spontaneous use of negation. This task included a total of twenty questions-both open and closed-ended-as aids for getting to know the participants better, and potentially eliciting cases of negation. Some of the questions are shown in (Q3-Q6): Participants spent roughly 20 min on this task. This task was used first in order to 'relax' the participant's speech and familiarize them with the procedure. All productions were recorded using a Marantz PMD561 recorder and a unidirectional condenser microphone.

Sentence Completion Task
The third task was the Sentence Completion Task, which was approximately 10 min long. Participants heard nine incomplete scenarios interrupted by dog barking and had to complete each 7 This project includes four tasks but due to the nature of this paper, I limit the discussion to two tasks only.
sentence. This task tested for first negative mention, where negative clause was presupposed but not explicitly used, as in (32), or second negative mention negation, where the negator no was explicitly used after SN, as in (33) In (32), since no appears only at the end of the utterance right before the dog barking, it is treated as a first negative mention. In this case, since there is no second negative mention in the given sentence, ND is not expected. In (33) on the other hand, since single negation with first negative mention is introduced, the last clause, which participants have to complete, is an example of second negative mention. Therefore, ND will be expected to occur.

Data Analysis
Interviews were partially transcribed using the ELAN software program (ELAN 2019). Transcriptions included all utterances in which participants used negation (one negator or two negators) in order to observe and compare the scope of variation of ND. They also contained utterances of the preceding context and the actual sentence containing the negation, in order to understand better the environment for ND use and better code the linguistic variables. Excerpts from the interviews were then analyzed in Praat (Boersma and Weenink 2017) in order to determine a final boundary tone. If the final boundary tone was HL%, it was treated as a characteristic of ND; if the final boundary tone was LH%, it was treated as a tag question, which is part of SN. In order to determine the type of contour, I compared the pitch movement associated with the preceding lexical item and the final no: if the final pitch was lower than the pitch associated with the preceding lexical word, a typical falling contour in a declarative sentence, it was a characteristic of ND. In contrast, if the final pitch on the final no had a rise associated with 'no', it was analyzed as a tag question. Given that speakers in interviews produced longer and faster sentences, it was sometimes hard to measure pauses.
For the Sentence Completion Task, responses were transcribed in ELAN. I classified the final contour auditorily and confirmed it with Praat analysis afterwards. The utterances were coded for the first and second mention.

Interviews/Elicited Conversational Speech
The first task was elicited conversational speech. Based on preliminary research (Tararova 2016), I anticipated occurrences of ND in interviews. Table 2 shows the distribution of SN and ND among 79 speakers. This accounts for a total of 310 tokens.  Table 2 shows that standard negation (a single sentential negation or a tag) was the preferred variant in the community, whereas the ND variant occurred only 1% (n = 4) of the time. Most of the instances with SN occurred with first negative mention, and only 26 of them occurred after the second negative mention. As for the ND cases, all four tokens of ND occurred with second negative mention.
I now turn to discussing excerpts from the spontaneous speech in more detail. Below, I illustrate the use of ND with a second negative mention, immediately preceded by a first negative mention. Three of the four examples show the use of ND in contexts with different verbs when we compare the verbs in SN and ND. The identity of the verb is determined in relation to the immediately previous used verb, either in the same sentence or in the previous one.

34.
Interviewer In (34), the participant responds that monolingual speakers do not notice any difference between them and Chipileños. In other words, the first sentence 'activates' the participant's response, whereas the second sentence reinstates the same idea with the use of second negative mention. This supports the idea that ND would be favored in the context of second negative mention. With regard to the verb, in both sentences, the main verb is different, creer 'to believe' and notar 'to notice'.
Example (35) demonstrates that ND is also used in context with a second negative mention, after the speaker negates the idea of character traits in specific individuals.

35.
Interviewer: ¿Y qué es lo que no te gusta de Chipilo? 'And what is there that you do not like about Chipilo?' Participant: Pues a veces el carácter de algunas personas no me gusta pero nada inconveniente, que -que no se puede arreglar no. 'Well, sometimes the character of some people does not please me, but there is nothing inconvenient that one cannot resolve NEG.' Again,in (35), the participant first introduces his idea, using SN in the first clause, followed by a continuation of the same idea in the following clause, i.e., second negative mention. Similar to (28), ND was predicted to occur within this context. With regard to type of verb, the second verb arreglar 'fix' is different from the first verb gustar 'like'. In the present example, the speaker uses the final no after the first negative mention with SN. My preliminary research found a few instances where ND occurred with a different verb from a previous sentence. In example (35), this is again the case. Example (36) is similar to example (35) in the sense that ND is used with a different verb from the preceding negated phrase, as well as a second negative mention.

36.
Participant: Como soy moreno, nada más se ríen, no me van a entender, o sea no me creen no. 'Given that I am dark skinned, they [non-Chipileños] just laugh. They won't understand me, in other words they don't trust me NEG.' In (36), the participant uses the two different verbs in two separate clauses. Similar to the other two examples, ND occurs in the context of second negative mention, which is preceded by first negative mention. In both cases, the participant talks about the same idea. Therefore, ND was predicted to occur.
Example (37) also shows that ND occurs in the clause with second negative mention. In (36) What is interesting about example (37) in comparison to the previous examples, is that the first clause does not have an explicit SN (NEG V), but rather the verb 'criticize'. The word critican '(they) criticize', however, seems to bring a negative connotation to the sentence, so it might be the case that ND is used as a marker of the implicit negation. In this case, critican could be treated as SN, followed by the ND. Future research can determine whether these types of verbs affect the occurrence of ND.
In summary, ND in examples (34-37) occurs with a verb as a previous constituent. ND, in three examples (29-31), contains a different verb from SN. All four of these examples show the occurrence of ND with second negative mention. Moreover, as mentioned at the beginning of this section, there were also 26 SN cases out of a total of 306 instances that appeared with the second negative mention. The next two examples (38-39) show that ND does not always occur in a context with the same negative idea, although both first negative mention and second negative mention are present. It is important to mention though, that the 300 cases happened in different linguistic environments, so we can assume that SN does not occur with second negative mention that often.

38.
Interviewer: ¿Y cuáles cosas no te gustan de Chipilo? 'And what things do you not like in Chipilo?' Participant: Unas cosas que no me gustan . . . uhm . . . la entrada de narco . . . es una de las cosas que no me gusta nada. 'Some things that I do not like . . . hmm . . . the 'entrance' of drug dealers . . . it is one of the things I do not like at all.' Participant: . . . A veces usan unas palabras que nosotros, los jóvenes, ya no usamos o al revés, pues nosotros usamos palabras que los viejos ya no usan. 'Sometimes [the older speakers] use some words that us, young speakers do not already use or vice versa, well we use some words that the elders do not use already.' In (38) and (39), ND could potentially have been used in contexts with a second negative mention, and with the verb as a previous constituent. If we compare (38) and (39), however, in (38) there is no second final no; instead in (38) the participant uses nada 'nothing' as a marker of reinforcement, in place of no. In other words, what we see in (38) and (39) is the following: although each speaker expresses the same idea in two sentences, in both of which negative mention is shown, participants produce SN in both sentences. Therefore, as seen from the distributional analysis, ND is an infrequent variant in this task. Thus, overall, the conversational speech produced only weak support for the hypotheses about negative mention and the linguistic factors.
Most directly relevant to this study however, four of the participants commented on the use of the final no, as seen from (40-41).

39.
Participant: A veces hay repetición de palabras que las personas que no tienen el véneto, no lo hacen. Normalmente habla de . . . cuando se niega algo usamos a veces el no. 'No vine el otro día contigo no'. 8 Por ejemplo, con mi novia: quiere pronunciar algunas cosas en véneto pero no le sale . . . 'Sometimes, there is a repetition of words, which people who do not speak Veneto, do not do. Normally, speaking of . . . when something is negated, sometimes we use no. 'I did not come with you the other day NEG'. For example, with my girlfriend: she wants to pronounce some things in Veneto, but it does not come right . . . ' Interviewer: Y este 'no' que mencionaste, ¿qué significado tiene?
'And this 'no' that you have mentioned, what meaning does it have?' Participant: No sé. Yo creo que es para reiterar la idea o refirmar la idea pero no sabré decirte cuál es. 'I do not know. I believe it is used to assert the idea or to reconfirm the idea but I wouldn't know which one it is.' 41.
Interviewer: ¿Y cuáles son unas cosas específicas del español de los Chipileños? 'And what are some specific things of Chipileño Spanish?' Participant: Ah, por ejemplo, cuando decimos 'no te lo presto no'. . . . es como afirmar la negación, como que estar seguro de que no es. 'Ah, for example, when we say "I do not lend it to you NEG." It is like to confirm the negation, like to be sure that it is not.' Based on the participant's responses, the purpose of the final no was to reconfirm or reinforce the negation. These observations seem to be consistent with the four actual ND uses under discussion.
Summarizing these findings, negation in Chipileño Spanish predominantly occurs preverbally, the only possible pattern in standard Spanish.

Sentence Completion Task
The other task was the controlled sentence completion task. Table 3 shows the distribution of forms among the four speakers that showed variability, with a total of 36 tokens. The other tokens (n = 877) were used by bilingual speakers with an invariant SN pattern. Table 3. Overall distribution (in absolute numbers and percentages) of ND and SN in the Sentence Completion Task, for participants exhibiting variation.

Use of Negators SN ND Total
First negative mention 56% (n = 20) 11% (n = 4) Second negative mention 25% (n = 9) 8% (n = 3) Total number of tokens 29 7 36 As seen in Table 3, the results show that ND occurred only in seven cases out of 36 analyzed tokens in contexts with both first and second negative mentions. In other words, we find no support for the occurrence of ND with second negative mention only.
Comparing contexts with first and second negative mentions, the results did not show a significant effect of second mention on ND use (p = 0.429, Fisher's Exact Test). It is important, however, to investigate the cases which did elicit ND. Three of the seven ND utterances (41-43) show that ND was used with second negative mention. 42. Luis pinta todos los días y María se supone que practique el piano. Ella no toca el piano suficiente porque prefiere salir con sus amigos. En cuanto al arte y la música lo importante es practicar. Luis pinta a diario pero María no . . . 'Luis paints every day and Maria is supposed to practice the piano. She does not play the piano enough because she prefers to go out with her friends. As for the art and music, the important part is to practice. The participant in (44) does not use final no after quiero 'want', but he uses it sentence-finally with another verb gustar 'like', reinforcing the idea of negation and favoring ND. In this case, the first clause is the first negative mention with explicit SN, which is immediately followed by the second negative mention.
The next examples (45)(46)(47)(48) show the use of ND in contexts which are distinct from the previous examples. The participant in (45) uses ND, even though there is no explicit previous use of SN. 45. Hay tres trapos en la mesa. Luisa estaba limpiando todo el día pero se enfadó porque solo un trapo limpia bien sin dejar ninguna mancha. En cuanto a limpiar bien, un trapo limpia perfectamente bien pero los otros dos no . . . 'There are three cleaning rags on the table. Luisa had been cleaning the whole day but she got frustrated because only one rag cleans well without leaving any stains. In terms of cleaning well, one rag cleans perfectly well but the other two NEG . . . ' Participant: pero los otros dos no limpian no. 'but the other two do NEG clean.' Additionally, it is important to mention that the participant in (45) uses the same verb in ND as he/she did in the previous clause. Moreover, the use of ND following the affirmative clause (limpia . . . bien) suggests that ND might be a marker of contrastive focus. The example in (46), similar to (45) is not a case with explicit first negative mention, so the use of ND was not expected. The production of ND in (46) could then be explained by the inclusion of sin embargo 'however' or 'nevertheless', which could be treated as some sort of explicit negation or possibly as a first negative mention. Since this project does not focus on connectors or adverbs, which carry contrastive or negative connotations, this merits future research.
The last example (47)  It is important, however, to compare the two responses in (46). The first participant produces her response by adding an explicit SN followed by ND. In other words, she "activates" a first negative mention, which triggers the second negative mention.
As for participant 2, the use of ND is similar to that in example (47), where the participant uses ND as a marker of contrastive focus, since the previous clause is affirmative. Similar to (45), the second participant in (47) uses the same verb as they did in the first clause.
In summary, this task provided very few tokens of ND. Again, SN was the preferred variant among all the participants. In the participants' speech that exhibited variant behavior, four cases with ND occurred with second negative mention, or incorporation of an additional negative sentence to reinforce the negation. The other cases with ND seem to have occurred due to contrastive focus. However, due to the limited data pool used, it is important to study this proposal further.

Discussion and Conclusions
According to the present findings, negation in Chipileño Spanish seems to exhibit two types of negation, but with different frequencies. The preverbal negation (SN) occurred with the highest frequency. Given that the preverbal negation is the standard form of negation in both Veneto and Spanish, this result is not surprising. The other variant also occurred but was limited to restrictive contexts only. Thus, it is possible to conclude that Chipileño Spanish occupies Stage I in the Jespersen Cycle, similar to other Romance varieties.
It is important to acknowledge that during the elicited conversational speech, a few participants mentioned the existence of the ND phenomenon and its frequent use among many bilingual speakers in both Veneto and Spanish. Given my role as a researcher, and not a local Veneto speaker, it is possible that some participants may not have felt comfortable enough to speak freely about their life, taste, and other topics, and thus their spontaneous speech might not have been fully natural. This is consistent with observations made by King (2000), who claims that it is very challenging to gather and reveal the full repertoire of individual speakers, much less the entire speech community, in sociolinguistic interviews or other traditional data elicitations, if the interviewer is not local (p. 54). Secondly, Schwegler (2018) claims metalinguistic awareness about specific features plays an important role in participant use of a variant. In the case of Chipileño Spanish, even though the speakers in the community are proud of their heritage and language, some potentially realize ungrammaticality of ND in Spanish. Therefore, future analysis should focus on 'free-topic' spontaneous speech with a local speaker leading the discussion.
Although the elicited conversational speech and sentence completion task did not have a large total number of tokens with ND, the results in both of these tasks also showed a tendency to use ND with a previous negative mention, compared to contexts with no previous negative mention. These findings are consistent with my previous research (Tararova 2016(Tararova , 2017(Tararova , 2018, as well as the research on ND in Italy (Poletto 2008(Poletto , 2016. Poletto (2008Poletto ( , 2016, in her research on ND in Veneto in Italy, also claimed that ND was favored in contexts with a previous negative mention. Regarding Schwenter (2005 and Schwegler (1991Schwegler ( , 2016Schwegler ( , 2018 proposals of ND use in contexts with denial of proposition, this project showed that ND was mostly used as a marker of reinforcement of a previously negated clause (i.e., used as second negative mention), rather than as a marker of disagreement. Moreover, ND in Brazilian Portuguese and Palenquero is restricted to specific pragmatic contexts, specifically to denials of activated, salient discourse-old propositions, which are similar to Zanuttini (1997) observations on the use of miga in Northern Italian varieties or mica in Standard Italian. It is interesting to mention, however, that there were three similar cases in the controlled task, which exhibited the use of ND as a marker of a contrastive focus. Additionally, similar to Schwenter (2005Schwenter ( , 2006 findings, my project did find more cases of ND after an immediately previous negated clause (SN), i.e., second negative mention than first negative mention.
There are still certain aspects that are not clear about negation in Chipileño Spanish. First, even though ND occurred with second negative mention, it is unclear whether ND and SN are in free variation. The participants that used ND in the elicited speech did not seem to produce SN with the second negative mention or with similar contexts, as they used ND. However, since there were not many cases of negation altogether, future work should analyze natural speech further, to study the frequency of the variants.
To conclude, ND in Chipileño Spanish was favored in contexts with a second negative mention. This supports my previous observations, as well as Poletto (2008) work on ND in Veneto, which seems to suggest that ND in both varieties is used in similarly negative contexts. Additionally, it is possible that ND has been transferred from Veneto to the Spanish of some bilingual speakers, along with its pragmatic function. This prediction still merits further research, because the effect of negative mention was not consistent among the bilingual speakers across the different tasks.
Regarding future research, more data is needed to analyze the variants in greater detail. In the current project, I have not found any cases of NEG3 (V-NEG). However, in her Italo-Mexican dictionary, MacKay (2002) shows that this third form of negation with the falling contour is a possible variant in Veneto in Mexico, but there is no mention about frequency or pragmatic function of the variant. Since NEG3 is ungrammatical in Spanish, Chipileños might be more conscious about their speech and separate the two grammars. It is still not clear, however, whether NEG3 is prominent in their Veneto grammar. If the variant is infrequent, then the transfer into Spanish will not occur. Therefore, further work is required to establish the distribution of the V-NEG structure with the falling intonational contour in Chipileño Spanish. A comparison with the same structure in Veneto from Chipilo, specifically its frequency and distribution, will determine whether the phenomenon of negation is a change in progress (Jespersen 1917), or whether some variants (SN, ND, V-NEG) have their own stable contexts of use. Based on the findings from the present study, negation in Chipileño Spanish falls under the first stage of the Jespersen Cycle (NEG V), since SN occurred at the highest rate among all the participants.
Funding: This research received no external funding.