MA thesis
Spanish in Texas,
Language contact.
About the development of Spanish as a first and second language in Texas.
MA in European Linguistics
Department of Applied Linguistics
Faculty of Arts
University of Groningen
August 31, 2016
Supervisor
dr. N. Haug Hilton
Second reader dr. D.G. Gilbers
Andrés Talaván Basilio
S2771659
a.talavan.basilio@student.rug.nl
Table of Contents
Chapter 1:
Introduction………..………...5
1.1 Historical background: from Mexico to the USA………..………....7
Chapter 2: Literature
review………..……….………..10
2.1 Basic concepts on language contact studies………11
2.2 Sociolinguistics views on
Gender…..……….13
2.3 The age factor………...13
2.4 About the influence of
region………….………14
2.5 A generational
study………...………14
2.6 Linguistic behaviors during the childhood and adulthood……….16
2.7 Self-Confidence and anxiety…….
……….16
2.8 Attitude and opinion towards a
language………17
2.9 Research
Questions...………..………...17
Chapter 3: Methodology………..………..………19
3.1 The corpus: “Spanish in
Texas”..………19
3.2 Nominal agreement in Spanish and adjective placement………20
3.3 Exposition of the data……….
4.4 Results based on the nationality of the participant’s parents….……….30
4.5 Results based on language used with parents, siblings and friends………31
4.6 Results based on
self-confidence………..………..35
4.7 Results based on attitude towards Spanish……….37
Chapter 5:
Discussion...……….….………40
5.1 No gender for language in
Texas………41
5.2 On the non-significant effects of age………..41
5.3 Homogeneity of Spanish within
Texas………...41
5.4 The influence of our participants’ parents birthplace……….42
5.5 Siblings and friends, a closer reality to our participants’ language than their parents………..42
5.7 A division in
Self-Confidence……….43
5.8 Attitude towards Spanish and its significance………44
Chapter 6:
Conclusion………...……….……….………...46
References………..………..………...49
Acknowledgments………...………..………...52
Appendix: Questionnaire data filled by
Language contact
The aim of the following thesis report on an analysis of how language contact affects the languages used by one of the largest immigration groups in the world, the Mexicans and their descendants living in the United States of America, more specifically on the state of Texas. The effects of language contact have been a hot topic for linguists since the beginning of the 20th century when British archaeologist Richard Dawkins studied the Modern Greek in Asia Minor (Dawkins, 1916). Dawkins was followed in this quest of understanding language contact by American scholars like Charles Fries and Robert Lado or Lithuanian Uriel Weinreich (1953) in the 1950s. In recent years there has been a considerable increase on the amount of research about the topic, linguists like Terence Odlin (1989), Donald Winford (2013) or Gillian Sankoff (2001) have been instrumental in the development of the study of language contact, giving special focus to the sociolinguistic factors that can affect language in that context.
One of the most important areas where language contact take place in the present time, is the United States of America, involving two of the most spoken languages in the world, English and Spanish. It is worth to mention, and it will be explained in the coming section, that approximately, by the year 2014, there were more than 10.000.000 immigrants living in the US who arrived from a Spanish speaking country. This collision of languages involving such an amount of speakers creates a need to understand how these communities and individuals communicate with each other. This thesis will study and analyze the influence of some classical sociolinguistic factors such as age or gender, but it will dig deeper, analyzing factors such as origin of parents, languages used during the childhood or adulthood and confidence, just to name a few. As all those factors are studied and evaluated for every single participant and instance, as it will also be shown later, this thesis will provide accurate and trustful information on the linguistic behavior of our participants, being also valuable information for all the Spanish speaking communities living on the US.
When it comes to language contact in the context of this study, it is also interesting trying to comprehend how a language like Spanish, which is the first language of around 500 million people around the world, is used and evolves on a community where it is not the official first language, with the unique exception of the city of El Cenizo, Texas, where Spanish was also named official language in 1999 (Garcia, 1999). The connections that tie Mexico and United States go farther away in time than just recent years, in order to understand where and how Mexicans stand in the society of the US. In section 1.1 the reader will be provided a historical and current background about how Mexican immigration started and developed in the US and more specifically in Texas.
After the historical background, in chapter 2, the focus shifts to the “Literature review” of the selected texts that have been chosen as theoretical background. The texts included are all relevant to understand the outcome of language contact in general, featuring texts from Terence Odlin, Gillian Sankoff or Donald Winford. This chapter also includes the research questions that motivated this work.
7
Chapter 1: Introduction
research, in chapter 3, the reader will find the grammar guidelines of Spanish agreement as described by the Real Academia Española (‘Royal Spanish Academy’). The reader can find as well a brief theoretical approach to the Spanish syntax related to these linguistic phenomena that surely will help understand the aspects of Spanish that has been studied in this research. Chapter
3 will also present the corpus chosen for this research as well as the data gathered from it.
The statistical results obtained from the data will be thoroughly presented in chapter 4. Followed in chapter 5 by an analysis of those results in relation with the arguments presented in the literature review and finally, in chapter 6, a conclusion of this study.
1.1 Historical background: from Mexico to the USA.
The relationship between Mexico and USA in terms of immigration has been characterized for having its ups and downs and dates back to the 19th century when states like California, Texas or New Mexico, among others, still belonged to Spain and migrating within those locations was easier (Congressional Research Service, 2012). At the beginning of the 20th
century the number of Mexican immigrants living in the US was just over 100.000 inhabitants, representing only 1% of the total immigrant population living in the States (Figure 1). The first important influx of Mexican immigrants arrived in 1907 when as the lack of cheap Asian labor increased, the possibility of a better job and life motivated Mexican workers to set their sights on their northern neighbor, doubling the number of immigrants in just 10 years, as by 1910, 220.000 Mexican born immigrants lived in the US (CRS, 2012).
Although recent tensions and problems may lead to think that the Mexican immigrants who crossed the border were the only beneficiaries from the relocation, both the governments from the US and Mexico have signed several treaties and agreements through all these years so Mexican workers could legally have a job in the US. The beginning of this sometimes symbiotic relationship came with the start of the First World War, Mexico agreed to send workers to the US while the North American country was fighting outside their own borders and lacked men. This political situation did not last for long, as once the war was over and the need for working men decreased, the U.S Government created the U.S. Border Patrol with the purpose of controlling and slowing down the Mexican influx of immigrants.
The 1920s meant another turn of events as both countries enjoyed a friendly decade, United States had a period of economic inflation and required cheap hands to supply the high demands of fruit and other basic products; the solution to their problem was filling those vacancies with immigrants, who arrived legally or illegally in American soil. Back then, Mexicans were not the main group of immigrants in the country, as they only represented the 3.5% of the total immigrants of the nation. With the arrival of the Great Depression, problems arose again and Mexican workers, even those who were legal residents, were deported back to Mexico. When in 1930 the total amount of Mexican born immigrants was close to 650.000 inhabitants, by 1940 the number decreased almost in half to 350.000 (Harvard Magazine, 2007)
This economic situation did not last for long, with the Germans stomping Europe again in World War II and the Japanese attacking Pearl Harbor, the necessity of new cheap labor, favored a new agreement between both governments. By signing this agreement, the government of Mexico sent large amounts of farm laborers to the US so they could work on sustaining the supply of food that the American country needed while they were on war again; this agreement was known as “the Bracero Program”1. Initially, the Bracero Program brought 100.000 more
Mexican-born immigrants, and by the end of the agreement, Mexicans already represented 7% of the immigrants in the whole country. This cooperation would last longer than World War II, more precisely until 1954, when with the end of the Korean War, the US government launched the “Operation Wetback”2. This operation consisted in tightening the security on all the southern
states but especially on the area of Rio Grande, as a result, about a million Mexicans were captured. It would not until the 1960’s, with the rising of the civil-rights movement, when the Bracero Program was concluded as it was considered degrading and dishonest with the community of Mexican immigrants. In 1965, with the approval of the Hart-Celler Immigration Reform Act, well immigration quotas based on race were removed, complicating the matter more for Mexicans, as they also had to compete with other areas and countries for Visas (Keely, 1971). There was still a dire need for labor in the US in the 1970s and with the political and economic situation worsening by the moment in Mexico, immigrants still saw the idea of moving to the US as a good opportunity and still massively tried to cross the border illegally, even with the new extra difficulties.
The political relationship between both countries stalled until the 1980s. In 1986, the Immigration Reform and Control Act (IRCA) was passed, the IRCA amnestied 2.3 million Mexicans that were already residing in the United States while strengthening and tightening the
1 Bracero means in Spanish: “one who works using his arms”.
2 The derogatory term “Wetback”, or “espaldas mojadas” in Spanish, was initially used to refer to illegal Mexican
9
Chapter 1: Introduction
security on the borders. Just between 1981 and 1990, the number of deported Mexican immigrants is estimated to be approximately of one million, per year (Meier and Rivera, 1993). The security measures not only have not stopped but from 1986 to 2007, the Border’s Patrol enforcement budget has increased by a factor of 10 and the number of officers has tripled (Harvard Magazine, 2007). However, the number of Mexican born population has also increased exponentially, almost doubling from those 2.300.000 that were censed in 1980 to 4.300.000 (21.7% of the immigrant population in the US) in 1990 and doubled again in a decade when in 2000, a total of 9.175.000 Mexican born inhabitants were listed. The total number of Mexican inhabitants kept rising during the 2000s until reaching its peak in 2012 of 11.175.000 inhabitants born in Mexico, representing almost 30% of the immigrant population in the US.
The figures explained in the previous paragraphs are of great relevance as they only take into account Mexicans born in Mexico and not their descendants. Nowadays, there is an estimated population of more than 41.000.000 inhabitants who either are natives from Mexico or are descendants from Mexicans in the US, resulting in the not inconsiderable figure of 15% of the total population of the country (United States Census Bureau, 2016). Besides the Mexicans, the Spanish speaking population is huge in the US, and includes, as an example, 5 million Puerto Ricans, 2 million of Cubans, 1.5 million of Salvadorans, 1.5 millions of Dominicans or 1 million of Colombians and Guatemalans each. Focusing the attention on Texas in the present day, Hispanics or Latinos represent 38.8% of the total population of the state, being the second highest ethnic group in the area in front of black or African Americans who represent 12.5% of the total and Asians who represent 4.7%. The proportion of Latinos in Texas, although not homogeneous as the following figures show (Table 1), denotes that there is a huge influence of Latinos in the communities of the state, as they represent the majority of the population in some of the biggest and most important cities like San Antonio and El Paso, and are also almost half on the population in the biggest city of the area, Houston. There are also places where although they are well represented as in Pearland or Austin, the number of Latinos is not comparable to cities like Edinburg, Weslaco or Brownsville where almost all the inhabitants have Latino roots.
Table 1 Percentage of Latinos and total population of the participants' area
Austin El Paso Edinburg San Antonio
Houston Brownsville Weslaco Irving Pearland 35.1% 80.7% 88.2% 63.2% 43.8% 93.2% 85% 41.1% 20.5% 811.456 649.133 74.541 1.327.551 2.109.372 174.982 39.474 233.607 108.821
Chapter 2
Preparing a study of this magnitude requires an intensive study of the relevant literature previously published on language contact and language transfer. In this chapter the reader can find the literature review about the topics that are important to understand the outcomes of language contact. The principal guidelines and ideas for this thesis come from several works: “Language transfer, cross-linguistic influence in language learning” by Terence Odlin (1989), Donald Winford’s text about “Social factors in contact languages” that is part of “Contact Languages: A Comprehensive Guide” (2013) by Peter Bakker and Yaron Matras; and “Linguistic Outcomes of Language Contact” by Gillian Sankoff (2001). In the following paragraphs and sections the reader will get an idea of where some of the most prestigious linguists stand on language contact and language transfer as well as how their ideas are applied for this project.
2.1 Basic concepts on language contact studies.
Before reviewing the main variables that are going to be taken into account in this research, it is also worth to explain some basic concepts on language contact. Odlin (1989) stated that language contact situations appear when speakers with different native languages have the need to communicate between each other. In this situation, users who have not mastered the language of the interlocutor or by extension a lingua franca that both understand, can decide to use signs, gestures or any other communicative means that they may have at their disposal in order to deliver a message that can be understood clearly. Unfortunately, those alternatives are not always sufficient, especially when the addresser wants to transmit certain kind of information (like abstract concepts), creating the need of a second language that is understood by both addresser and addressee (Odlin, 1989).
When the use of a second language is not an efficient method of communication between users that do not share the same native tongue, they can be inclined to use language mixing. Language mixing is described by Odlin as being formed by several linguistic phenomena, including the raw merging of two languages and their features or others such as code-switching or borrowings. The last option included in language mixing is code-switching, it does not require to add new terms to a language as users just interchange words, phrases or sentences of two or more languages with each other. This option is more feasible when the speakers understand two languages that are closely related, as it could be Portuguese and Spanish or Swedish and Danish. Another point of view was offered by Winford (2013), who used some of Uriel Weinreich’s most famous quotes in order to introduce language contact to the reader. The first of these quotes concerns the goals of language contact studies which are “to predict typical forms of interference from the sociolinguistic description of a bilingual community and a structural description of its languages” (Weinreich, 1953). Another one of Weinreich’s quotes that Winford (2013) mentioned explains that in order to understand language contact, a broad psychological and sociocultural setting is needed. This is closely related with the nature of our study, as the reader will see on the methodology chapter, this research is based on analyzing as many variables related with the participants as possible, from their education to their confidence or opinions. Weinreich also provided a classification of the two kind of non-structural factors that appear on both the macro (societal) level, and those that work on the micro (individual level).
13
Chapter 2: Literature Review
or the other language as their mother tongue, demographic facts, social and political relations between these groups; the prevalence of bilingual individuals with given characteristics of speech behavior in the several sub groups and stereotyped attitudes toward each language (“prestige”); indigenous or immigrant status of the languages concerned (Weinreich, 1953)
On the other hand, the following factors are identified with the micro-level: the speaker’s facility of verbal expression in general and his ability to keep two languages apart, relative proficiency in each language, specialization in the use of each language by topics and interlocutors, manner of learning each language; attitudes toward each language, and whether idiosyncratic or stereotyped (Weinreich, 1953). Winford notes that unfortunately this last group of factors are not taken into account as we are far from achieving Weinreich’s vision of unifying linguistic, social and psychological aspects of language contact. In the study of this research, one of the main objectives is to analyze both the macro and micro level factors with the information given about the participants by the University of Texas.
Odlin (1989) also reviews the nonstructural factors that affect transfer, Odlin comes to the conclusion that researchers need to avoid generalizations over the language of an individual as there are not two people who speak exactly the same. For Odlin researchers need to take into account that differences in voice quality, intonation and vocabulary choice are among the most common distinguishers of individual speech. Odlin also criticizes contrastive analysis between languages as “they consist of idealized descriptions of both languages”, leading frequently to inaccurate predictions of individual performances.
Additionally, when studying language contact and bilingualism, Winford states that some factors that are usually associated with it are geographical isolation (Gaelic speakers in the Scottish Highlands), urban segregation (Hispanics in the US), the persistence of ethnic minority enclaves (Basques in Southern France), and a tradition of limited cultural contact (Pennsylvanian Dutch in the US). He also claims that in these kind of situation, the first native language of the minority group can be heavily influenced by the language that is more predominant; for example, at the beginning of the process of language contact, the minority language is more prone to receive borrowings from the dominant language. These minority groups tend eventually to become bilingual and speak both languages and in some cases they switch to the predominant language; however, this does not happen always, other groups like some Greek communities in Asia Minor were able to resist the adoption of borrowings for long periods of time (Dawkins, 1916).
These findings along with other researches mentioned by Winford, introduce one of his final claims: “the greater the intensity of contact during the phase of bilingualism and shift, as well as the asymmetry in power and prestige of the languages involved, promote increasing structural influence from the dominant language on the subordinate ancestral language” (Winford, 2013). In the case of this study we will see if the Spanish of our participants is affected by the surrounding and more predominant English language.
Another important aspect of languages and language contact in the 21st century is pointed out by Sankoff (2001), as she states that linguistic minorities have arisen in the 21st century both
outcomes of language contact in an area, supporting the need of contextualize any discussion with a socio-historical perspective. When it comes to the socio-historical context that affects language contact, Sankoff states that historically a large part of this contact has been produced under conditions of social inequality resulting from wars, conquests, colonialism, slavery and migrations; it is also worth mentioning that benign contacts as the one created by the desire of trading is also documented as a possible influence on language contact. As the reader probably has noticed by now, explaining in depth the recent, and not that recent, history between México and USA was a key factor in order to evaluate and explain the analysis carried out in the following sections.
Regarding the methodology to be chosen on researches of this nature, Sankoff follows the three main guidelines that Weinreich used in his study on language contact (Weinreich, 1953), first, her research is based on studying the language from a community instead of the individual language; secondly, she studies the linguistic results of contact and thirdly she seeks to elucidate the social structuring of diversity internal to the speech community. This thesis share two of those objectives, as it will focus on the Spanish speaking community in Texas and we will obtain the results of language contact in that region.
2.2 Sociolinguistic views on Gender
The first issue that needs to be addressed when talking about gender, at least on sociolinguistic studies, is about the use of the terms “gender” or “sex” over the other. In order to avoid any kind of misinterpretation, in this thesis the term that will be used is “gender”, gender is understood as “the state of being male or female” and it is also the term used on the “Spanish in Texas” corpus that will be described on chapter 3.
As a variable, during the years, gender has been one of the hottest and most controversial topics in sociolinguistics. Exemplifying how controversial and contradictory studies based on gender can be, one of the earliest views on the variable appeared in the 1970s with Robin Lakoff’s work Language and Woman’s Place, where she adduced to the classification of a “women’s language”. Some of the characteristics of “women’s language” are the use of “hypercorrect” grammar, to use standard pronunciations, or on the lexical level, the use of “empty” adjectives, among other features (Lakoff, 1975). Supporting part of these claims, Trudgill (1983) stated that “women, allowing for other variables such as age, education and social class, produce on average linguistic forms which more closely approach those of the standard language or have higher prestige than those produced by man”. On the other hand and focusing on prestige forms, Abd-el-Jawad (1987) proved that on Arabic-speaking countries of the Middle East, men use prestige forms more frequently than women. It is important to take into account that those three studies had different sociolinguistic backgrounds, as will also be the case in this study so different results could be expected.
These contradictory results between studies that analyze a linguistic feature such as the use of prestige forms, show that other variables such as region, ethnicity or culture need to be taken into account. In this research a specific group of participants is aimed as targets, allowing us to obtain specific results for this speech community.
15
Chapter 2: Literature Review
Age is one of the main focuses of attention in linguistics. Hypothesis and ideas regarding the matter have been under study over decades, an example of this is the hypothesis of the “critical period” which states that young learners have an advantage over older leaners on mastering the phonology and syntax of one language. From a sociolinguistic point of view, in this thesis, we will study the effects of age using as a guideline a research carried out by Bahrick et al. (1994). In this research, the Spanish and English language skills of 801 Cuban and Mexican immigrants who had been living in the US between 4 months and 50 years was tested. The participants of that study were examined on the following linguistic tasks: text comprehension, grammar, vocabulary recognition, oral comprehension, category generation, Cuban Anglicisms and Mexican Anglicisms. In their results they found out that the performance of the subjects remain almost equal along the years, even when compared with the control subjects, who had been only in the US for 3 months. As an exception, they found out that the performance of the oldest group was worse on vocabulary recognition and category generation, where they scored lower results than the youngest groups.
Another important finding, although it came with no surprise, on Bahrick et al. (1994) showed that the participants who were older and had been under more Spanish education performed better than those who were younger and had less Spanish education, they also proved that older immigrants were more prone to remain Spanish dominant while younger immigrants tend to use the host language.
In this study, although all our participants were born on American soil, we will see if there are differences in their Spanish or if following the results obtained by Bahrick et al, age will be considered as a non-significant factor that has little to no effect in this language contact situation.
2.4 About the influence of region
The next variable to be explained and reviewed is the influence of the region where our participants lived at the time of the interviews. Region was described by William Kretzschmar Jr. (2011) as a “a location in time and space in which people behave in some particular way, and in which we may find physical evidence related to that behavior”, in linguistics this is known as a speech community.
Kretzschmar mentions, among others, a research carried away by Ronald and Barbara Horvath across nine localities in Australia and New Zealand and the rate of /l/ vocalization in those areas (Horvath and Horvath, 2011). They found out that overall, in both countries, there was a 33% rate of /l/ vocalization, however there was a significant difference between the nations, New Zealand had a rate of 58% of instances while Australia had a way lower rate of 15%; this kind of differences were also found from a local perspective, a region like Brisbane (Australia) showed a rate of 3%, meanwhile the region of Wellington (New Zealand) a 58%. As it can be seen by the figures, the closer we look into the map, the more precise and accurate the results will be. Therefore, in this study, we will see individual results for several areas within Texas instead of studying it as a whole, allowing us to make a more accurate analysis.
is different within regions in Texas, if proven to be a significant variable, this would be really interesting, as we could asses with statistics, as those presented in Table 1, if factors such as demographics or distance with the Mexican border play a role in our participants’ Spanish.
2.5 A generational study
The explanation of the variable age bring us to one of the most interesting studies on language contact which was carried out in East Harlem (New York City, USA). This study was focused on the loan words used in the Spanish of that speech community (Poplack and Sankoff, 1984). They found out that the overall language used by the younger speakers was closer to English than to Spanish when in comparison with older speakers. This comes with no surprise as it should be expected that newer generations have been living during a longer part of their lives in the country and are also more immersed into the American culture. The case of East Harlem is related to the one that we find in the present study, as we have bilingual speakers of English and Spanish, living in the US, on a community whose influx of immigrants is steady and who are willing to maintain a situation of stable bilingualism.
Sankoff also stated in a newer study (2001) that immigration historically has resulted in rapid linguistic assimilation; Sankoff considers that a major variable would be the duration of the contact, whether linguistic assimilation is relatively rapid (it often happens in one generation) or relatively slow, possibly over many generations. Elaborating on this idea, a study carried out by Thompson and Lamboy (2012) found a significant difference in the use of either Spanish or English across generations of Hispanics living in the US. Those of the first generation claimed to think mostly in Spanish, with a rate of 65%, while those who are part of the second generation showed a similar rate, 63%, but thinking on English; the last group, the third generation, see only an increased value on the use of English as the language of thought with a rate of 80%.
Focusing on phonology and referenced by Sankoff, it is worth to mention a study carried out by Fought (1999). Fought found out that only some of the second generation from the Mexican American community in Los Angeles participate in the u-fronting typical of the Anglo community, producing a [u] that is readily identified with the Spanish high back vowel.
Based on these studies, it will be interesting to see if the participants’ Spanish performance is different, in a significant way based on the origin of their parents (which will make them first generation, second generation…) or if by the contrary, our participants produce normative and non-normative instances at similar rates, regardless of the generation that they belong.
2.6 Linguistic behaviors during the childhood and adulthood
17
Chapter 2: Literature Review
Following this line of thought, on Bahrick et al. (1994) it was found that the participants who used Spanish predominantly over English were also the ones with the lowest scores on Spanish reading comprehension and Spanish oral comprehension, fact that they attributed to them being younger immigrants instead of linguistic talent. In this study we will also test if the use of Spanish is a variable that hinders the Spanish of our participants or if by the contrary, they will obtain better results.
On another study, this time by Marjohan (1988), it is claimed that “the language used by speakers is influenced by the interlocutor they are speaking with”; in relation with our study, we have information about our participants’ linguistic behavior in their interpersonal relationships, from the language use in their childhood with their parents and siblings, to the language used in the present day with their families and also friends. Another study along this topic was carried out by Flege (1995) who discovered that the pronunciation of English consonants by Italian immigrants who arrived in Canada in their early childhood vary depending on how often they speak Italian. The data presented in the appendix shows that every participant had a different kind of behavior and that the use of English or Spanish varied depending on the interlocutor.
These studies provide valuable information about the influence of the languages spoken during the childhood and adulthood. In the following chapters we will examine the language of our participants and we will identify if their Spanish is more influenced by the language spoken with their parents or as some studies suggest, the language spoken with their friends and siblings.
2.7 Self-confidence and anxiety
One of the important points in this thesis is to study and measure the influence of self-confidence in our participants. As it will be explained in the next chapter, all our participants answered questions regarding the confidence on their Spanish skills. Several recent studies have proven that learners base their linguistic attitudes and anxieties on previous experiences (Frantzen and Magnan, 2005).
Analyzing the variable of confidence and anxiety will takes us back to the Mexican/Cuban study carried by Bahrick et al. (1994). In their study they found that participants who rated their understanding of English higher than their understanding of Spanish, tripled their comprehension scores in English than in Spanish. In our study, although we do not have information about the self-confidence of our participants in English, we will be able to see the effects of their confidence of Spanish on their own Spanish.
One of the most recent studies on self-confidence and bilingual abilities was carried out by Starks (2015). In her research she tested 120 participants from four of the largest communities in the island of Pasifika (New Zealand), each one of them having an own language in contrast with English. Starks found out that language shift was affecting the four communities and all community members regardless of their confidence in their native tongue.
2.8 Attitude and opinion towards a language
On another study, this time carried out by Holmes et al. (1993) it was found that in communities and individuals with positive views towards the community language and strict rules about language use at home, it can take up to four generations for language shift to happen. In the case of this research, we will see how all of our participants regard Spanish really high on their level of importance and we will have the chance to see if in this case the language shift is slower or faster.
2.9 Research Questions
As it can be inferred in the previous sections, the present research aims to answer the following research questions:
- Can we consider gender as an element with significant influence on the production of non-normative instances in our participants?
- What is the role played by age in the Spanish of our participants?
- Does the geographical location where our participants are located have any influence on their Spanish?
- Does the origin of our participant’s parents play a role in their language skills?
- Can we consider confidence in their language abilities as an important factor? Is the confidence in all the aspects of language equally significant or are there factors that play a bigger role than others?
- Is the opinion of the speaker about the inherited language relevant in their production of
19
Chapter 3: Methodology
In this section the reader can find a complete guide of the methodology followed by the author of this thesis. The presented items in this chapter are the corpus chosen for this research, the grammar guidelines followed to test the normativity of the utterances of our participants and lastly a summary of all the data gathered for this project.
3.1 The corpus: “Spanish in Texas”
As a researcher wondering about how Spanish is spoken as a first and second language in the United States of America and more specifically in a state with the demographics of Texas, finding the right corpora to work with is the first key element to take into account. A study of this significance requires information regarding the participants’ background and their past experiences with the language; the more accessible information, the deeper study and clearer results will be shown.
In this regard the corpus of the University of Texas called “Spanish in Texas” stood out above the rest of possibilities as not only it provided a free and accessible source of data to work with, but it also resulted to be the perfect tool to use for this study for it assists the researcher with a wide range of information from the participants - information needed to approach and analyze every instance gathered in their files from a macro and micro level perspective. Along the fundamental information that this corpus procures are basic details from the participants as age, gender, birthplace, birthplace of both parents as well as the region within Texas where they lived at the moment of the interview. The amount of data that can be found about the participants does not end there, researchers can also find appropriate information about the past and present use of languages of the participants in contexts like education or personal life, including language habits with parents, siblings and friends. The last bits of information provide the participants’ self-evaluation of their oral, written, reading and listening dexterity with Spanish and also their opinion about the importance of the language.
The “Spanish in Texas” corpus has a total of 96 interviews recorded and transcribed between 2011 and 2013 with subjects from different background and origins (Bullock, 2016). For this study, only participants who were born in the USA, regardless of their gender or location of the interview, were chosen to be analyzed. All the data from the participants was stored on excel tables so it could be later studied individually and also put in comparison with the rest of the interviewees.
21
Chapter 3: Methodology
of the participant that produced the utterance, giving the opportunity to examine the tokens in detail.
On the other hand, some instances were labeled as non-normative, these instances are formed by those with a non-normative nominal agreement between one or more items on the same linguistic structure. An example of this is “los personas” (‘the* people’) that is formed by a plural masculine definite article and a plural feminine noun; other non-normative instances are instances where code-switching appeared (those with an English noun or an English adjective found with a Spanish article or noun), and also instances where the speaker makes use of an incorrect noun or an incorrect adjective from both the lexical and semantic point of view.
As consistency is primordial in this kind of study, the normative or non-normative character of the instances is based on the rules of concordance established by the RAE (Spanish Language Royal Academy) that will be explained in section 3.2 Regarding vocabulary, two dictionaries were used as a measuring stick, the “Diccionario de la lengua española” (RAE, 2016c) from the RAE and “Diccionario del español de México” (Diccionario Español México, 2016) created by ¨El Colegio de México¨.
After ordering and analyzing every instance on another Excel sheet, the data was transferred to SSPS in order to get the statistical value from the instances and participants. The formulas used to dictate if the variables of this study are significant or not were the Pearson’s Chi-Square test, in the cases where Chi-Square was not a possibility due to the own limitations of the token numbers, other trustworthy methods like Fisher’s Exact Rate and Likelihood Ratio were used.
3.2 Nominal agreement in Spanish and adjectives placement.
The linguistic feature that has been chosen in this research to evaluate the Spanish language of the participants is nominal agreement. It is important to note that in Spanish there is also verbal agreement, however, only the nominal one will be taken into account. The main reason for choosing this feature is the lack of nominal agreement in English (the other native tongue of our participants) and that although Spanish native speakers tend to produce non normative instances when it comes to verbal agreement (Barnés, 2013), they do not produce non-normative instances of nominal agreement frequently.
The Real Academia Española, generally abbreviated as RAE, describes agreement as “the obligatory coincidence of inflection (gender, number and person3) between the different
variable elements on a sentence” (RAE, 2016b). As it will be seen in this section, there are
certain rules and guidelines regarding agreement in Spanish that should be followed in order to achieve a normative communication. One of the first things that a non-native Spanish speaker should know about the matter is how articles work in Spanish, as well as their inflections.
Articles are divided in two groups, definite articles (Table 2) which are put in front of a noun indicating that the noun is already known by the speaker: El equipo de Philadelphia vence siempre (‘Philadelphia’s team always wins’) and indefinite articles (Table 3) which are also located in front of a noun, but in this case, the noun is not specific: Un equipo de Philadelphia vence siempre (‘A team from Philadelphia always wins’).
Table 2 Spanish definite articles.
Singular Plural
Masculine el (el perro, ‘the dog’) los (los perros, ‘the dogs’) Feminine la (la silla, ‘the chair’)
el4 (el águila, ‘the eagle’)
las (las sillas, ‘the chairs’) Neuter lo (lo mejor, ‘the best’)
Table 3 Spanish indefinite articles.
Singular Plural
Masculine un (un perro, ‘a dog’) unos (unos perros, ‘some dogs’)
Feminine una (una silla, ‘a chair’) un (un águila, ‘an eagle’)
unas (unas sillas, ‘some chairs’)
When it comes to agreement it is also important to note that, in Spanish, demonstrative pronouns and demonstrative determiners require also to be inflected and follow the same rules as the articles. The following table (Table 4) shows all the demonstratives that have inflection in Spanish:
Table 4 Spanish demonstratives with inflection
Singular Masculine
Singular Feminine
Neuter Plural Masculine Plural Feminine
Este Esta Esto Estos Estas
Ese Esa Eso Esos Esas
Aquel Aquella Aquello Aquellos Aquellas Tanto Tanta Tanto Tantos Tantas
Tal Tal Tal Tales Tales
The last aspect of agreement that needs to be clarified is the one regarding gender in Spanish. As it could be understood from Table 1, where el águila is used as an example, gender is not always marked by the noun’s inflection; to think that because a noun ends in –o is masculine, or –a is equal to a feminine noun, is a common mistake made by new learners of the language or for those who do not know a noun’s gender. In reality, there are masculine nouns that end in –a like día (‘day’), feminine nouns that end in –o like mano (‘hand’) and other endings that are used for both masculine and feminine as –e in words like héroe (‘heroe’) and serie (‘series’), –i as in alhelí (‘wallflower’) or hurí (houri), and also words ending in –u like ímpetu (‘vigour’) or tribu (‘tribe’) (RAE, 2016a).
4 El is only used as a feminine article with feminine names that begin with the sound /a/. This form is always
23
Chapter 3: Methodology
After this small recap about gender in Spanish that hopefully will be useful for all the readers that are not familiar with the language, the focus turns back to agreement. As only nominal agreement will be taken into account in this research, in the next lines, the reader will find a complete summary of how agreement works in Spanish, also as defined by the RAE:
3.2.1 Nominal agreement (agreement of gender and number): It is the agreement established between a noun and an article or the adjectives that accompany it: El libro rojo (‘The red book’); Esas mesas blancas (‘Those white chairs’); also the agreement of a pronoun with its antecedent or subsequent: A sus gatas las vi esta tarde (‘I saw his/her *female* cats this afternoon); Les dí el premio a tus amigos (‘I gave your friends the award’); the agreement of the subject with the attribute, with the predicative complement or with the participle of the verb on a periphrastic passive sentence: Su tía es una belleza (‘His aunt is a beauty’); El chico se encontraba contento (‘The boy was happy’); Esa ciudad fue destruida completamente (‘That city was completely destroyed’).
Besides these specific rules for nominal agreement, there are two more rules that not only are applied on nominal agreement but also on verbal agreement:
3.2.1.1 The agreement between two or more singular nouns or singular pronouns, if each of the pronouns is referred to a different entity, forms a group that agrees in plural with the adjective or the pronoun, or with the verb of the sentence in which they are subjects; Ese pan y ese vino son míos (‘That bread and that wine are mine’).
3.2.1.2 The coordination of two or more nouns or pronouns of different grammatical gender forms a group that agrees in masculine with the adjective or the pronoun; La tía y el tío estan cansados (‘The aunt and the uncle are tired’)
As it could not be any other way, where there are rules, there are also special cases and exceptions, the following points explain the ones that are relevant for this project:
3.2.2.1 One determiner for several nouns. When there are two or more specific nouns whose references are different entities, it is recommended that each of them is preceded by their own pronoun. In the case of having just one determiner for two or more coordinated nouns, it must agree on gender and number with the closest placed noun. Escoja una herramienta o instrumento (‘Choose a tool or utensil’).
3.2.2.2 An adjective postposed to several nouns. When an adjective describes two or more coordinated nouns and it appears after them, it is recommended that the adjective is inflected in masculine and plural. As in the case of agreeing only with the noun placed in last position, it would create ambiguity as it could be understood that it is only referring to the closest noun. Nevertheless, when the coordinated nouns are understood as a whole unit, the adjective can only agree in gender and number with the closest one: Las personas de origen y habla inglesa predominan en Nueva Zelanda y Australia. (‘People who are from English origins and that speak English predominate in New Zealand and Australia’).
recommended to use a singular noun and, if it has a determiner, repeat it before each adjective: El Antiguo y el Nuevo Testamento (‘The Old and The New Testament’).
The last syntactic point that may need some clarification is the placement of adjectives in Spanish. In this case the rules that are summarized are from the University of Washington (Petersen, 2016). There are three different kind of adjectives considering the position that they take in a sentence: the ones that follow a noun, the ones that precede a noun and those which can either precede or follow.
The first group of adjectives are the ones that follow the noun, they are in charge of differentiating the noun from the group to which they belong. For example: un águila verde (‘a green eagle’). By adding the adjective after the noun we are saying that the eagle of which we are referring, constitutes a different entity as it has a specific trait; within this group of adjectives, there are also included those adjectives related with science and those with a technical role: los fundamentos psicológicos (‘the psychological fundamentals’). Also part of this group are those adjectives who express nationality: el pastor alemán (‘the German shepherd’) and color5: un tazón dorado (‘a golden bowl’). There are certain associations between a noun and an adjective that are invariable, for example: Nos vamos en Semana Santa a Philadelphia. (‘we are going on Easter Week to Philadelphia’). The last of the cases are adjectives which are preceded by an adverb: Es un equipo realmente potente (‘It is a really strong team’).
The second group of adjectives are those which precede the noun, they are known as restrictive adjectives and in their meaning they involve the idea of number or quantity. Within this group are numerals, demonstratives or possessive adjectives: aquella mesa (‘that table’); esos son nuestros instrumentos (‘those are our instruments’), las cuatro primeras semanas (‘the first four weeks’). Also part of these groups are those adjectives that appear on exclamatory sentences that are introduced by Qué (‘what’) and those exclamatory sentences in which the speaker makes a value judgment: ¡Qué espectacular estampa! (‘what a spectacular image’) and ¡Infeliz invento! (‘poor invention!). Adjectives that are part of set phrases that are renowned because of its use: Mi hijo estudia bellas artes (‘my son studies Fine Arts’).
The last group of adjectives is the ones which can either precede or follow a noun. There are several cases when an adjective can be located before or after the noun, for example when the adjectives express a quality that is implicitly associated to a noun like in la blanca nieve (‘the white snow’) or in some cases when the quality expressed is commonly associated to the noun as in El frío agua del Polo Norte (‘The cold water from the North Pole’).
As anticipated before, for stylistic or poetic reasons some adjectives can be placed in front of the noun: La suave brisa de esta templada primavera (‘the soft breeze of this mild spring’). Adding to this group of adjectives are those who are used on protocol or in a ceremonial way: El honorable profesor Velasco Martín (‘the honorable professor Velasco Martín’). Also part of this group are those used to gradate subjectively within a scale a noun, they are numerical, evaluative or to express quantity. In this case the adjective only alludes a grade within a scale that is attributed to a human or thing as if the adjective was an intrinsic part of the noun: Se puso a gritar en pleno teatro (‘He started shouting in the middle of the theatre’). It is also significant to note that inside this last group there are some adjectives which depending on the
25
Chapter 3: Methodology
position will change their meaning as it happens with the adjective grande (‘large’), when preceding a noun this article can mean deep, imposing, real or great along other possibilities.
3.3 Exposition of the data.
In this section the reader can find a summary of the data that was gathered to accomplish this research. The complete raw data with details of every single token and instance, as well as the information about the participants and the conversation in which the tokens appear, can be found in the Appendix. All the data was collected from the Corpus del Español en Texas (‘Spanish in Texas Corpus’) provided by the Department of Spanish & Portuguese from the University of Texas at Austin. From the 96 total conversations that are available, 54 conversations were chosen. All the participants were native of Texas but from different regions within the state (Figure 2). 14 represented the area of Austin making it the area with the most participants, 9 of them are from El Paso and other 9 are from Edinburg, 8 are from San Antonio, 5 are from Houston, Brownsville is represented with 4 participants, Weslaco with 2 and finally Irving and Pearland that have 1 participant respectively.
Figure 2 Distribution of participants by region
Besides the demographical information, the Spanish in Texas Corpus provides information about the language habits of the participants from their childhood until the day the data was recorded, with the sole exception of one participant that did not provide any information about his habits and another one that is missing the information related with his education6 (Table 5). Thanks to this information we know that only 8 participants went to an elementary school where Spanish was the main language, 36 attended a school where English was the main language and 8 went to a bilingual (Spanish and English) school. The next step after elementary school is middle school, it can be seen that the participants are adapting more and more to the native tongue of the country in which they live as only 4 of them had middle school in Spanish and 1 went to a bilingual middle school. At the same time 47 students studied on an English only middle school. The last educational step that our participants were asked to answer concerned their high schools; 3 of them went to a high school where Spanish was the main tongue, 49 of them studied at a high school where English was the main language and none of them studied at a bilingual high school.
Table 5 Language of education of the participants
Elementary Middle School High School
Spanish 8 4 3
English 36 47 49
Bilingual (Sp & En) 8 1 0
Speaking still about the factors that can affect the development of a language, there is also information about the language habits that the participants had in their everyday life (Table 6). For example, 37 of them only used Spanish to speak with their parents during their childhood, just 5 only used English with them and 11 of the participants used both. Just by the numbers, it can be seen that the communication with parents is not the same as with the siblings, as 21 participants admitted communicating only in Spanish with their siblings, 13 communicated only in English and 18 of them used both languages. As it has been advertised in previous sections, there is also data about their present use of Spanish with their parents, siblings and even friends. 29 participants still only use Spanish with their parents, while 6 more switched to English raising the number of participants that only speak English with their parents to 11 and 13 that use both languages. One of the figures that may draw the most attention is the fact that most of the participants have stopped speaking exclusively in Spanish with their siblings as just 5 used only Spanish with them, on the other hand the number raises again for the use of just English with 28 participants using it with their siblings and 19 who use both languages. Finally, the last piece of information regarding the participants’ habits of communication is the language used with friends - the most selected option is both languages with 29 participants, 19 of them only use English with their friends while just 4 used only Spanish.
Table 6 Childhood and today's language use of the participants
C/Parents C/Siblings T/Parents T/Siblings T/Friends
Spanish 37 21 29 5 4
6 In another few cases in this figures and in the figures related with use of language in the environment, one or two
27
Chapter 3: Methodology
English 5 13 11 28 19
Both 11 18 13 19 29
The last bits of information about the participants that were provided are their opinions about how important an aspect of the Spanish language is in their lives and their self-evaluation of their skills in those aspects (Table 7). From a higher grade to a lower grade and always with 5 as the maximum grade, our participants gave 4.9 and 4.85 points at importance of listening and oral skills respectively. So it can be said that they believe that being able to hold a conversation in Spanish is essential in their lives, however, while they value their listening skills higher than any other skill with a 4.76 grade, they are not as confident with their oral skills, where they grade themselves with 3.98. Reading seems to be quite important too as they graded it with a 4.62 making it also a priority which can also be said about writing that although it represents the lower grade in importance with a 4.4 it is still a high grade of indispensability. Unfortunately for them, our participants do not trust their writing skills as they barely pass their abilities with a 3.37 grade overall, on the other hand they have a healthy 4.27 when it comes to reading.
Table 7 Self-evaluation and importance of Spanish language.
S-E Oral Imp Oral S-E Listening Imp Listening Grade 3.98 / 5 4.85 / 5 4.76 / 5 4.9 / 5
29
Chapter 4: Results
The present chapter features the complete results obtained in this study, including the to-tal amount of normative and non-normative instances produced by our participants and linked to each and every variable. As an introduction and clarification to all the tables that will appear in this chapter, there were a total of 1620 instances recorded, 1080 produced by women and 540 that were produced by men; these instances were divided depending on their normativeness re-lated with the grammar explained in section 3.2, resulting in 2 different kinds of tokens, norma-tive and non-normanorma-tive.
4.1 Results on gender.
The first variable that is going to be analyzed with our data is gender. A total of 54 participants, consisting of 36 women and 18 men, were chosen to take part of this study. Although there is no parity in terms of number, as the women double the amount of men, thanks to the extensive quantity of tokens gathered for each participant, it was still possible to present a proper statistical evaluation of the data.
In the following table (Table 8) the reader is provided with a total distribution of the instances and gender including percentages of instances which were normative and non-normative.
Table 8 Distribution of instances and gender
Gender
Female Male Total Normative count
% within Normative Type % within Gender % of Total 994 66.2% 92.0% 61.4% 507 33.8% 93.9% 31.3% 1501 100% 92.7% 92.7% Non-normative count
% within Normative Type % within Gender % of Total 86 72.3% 8% 5.3% 33 27.7% 6.1% 2.0% 119 100% 7.3% 7.3% Total Count
% within Normative Type % within Gender % of Total 1080 66.7 100% 66.7 540 33.3% 100% 33.% 1620 100% 100% 100% Fisher’s Exact Test: 0.190 (Not significant) / Likelihood Ratio: 0.172 (Not significant)
Although the amount of tokens produced by the females doubles the ones produced by the men, it can be understood just by checking the mere percentages that the production of normative and non-normative tokens is fairly equal between both genders. This fact is also corroborated by the Fisher Exact Test and the Likelihood Ratio that found the relevance of gender for this hypothesis as non-significant.
4.2 Results on Age
participant is 86 years old. As the range of ages between the participants is too ample for the statistical tests to be run properly, in this study they have been grouped by decades. They have been classified on table 9 as: below 20, in their 20s, 30s, and so on.
Table 9. Age and instances
Age 0-20 20-29 30-39 40-49 50-59 60-69 80-89 Total Normative Count % within N. Type % within Age % of Total 138 9.2% 92.0% 8.5% 660 44.0% 91.7% 40.7% 115 7.7% 95.8% 7.1% 142 9.5% 94.7% 8.8% 332 22.1% 92.2% 20.5% 86 5.7% 95.6% 5.3% 28 1.9% 93.3% 1.7% 1501 100% 92.7% 92.7% Non-N. Count % within N. Type % within Age % of Total 12 10.1% 8.0% 0.7% 60 50.4% 8.3% 3.7% 5 4.2% 4.2% 0.3% 8 23.5% 7.8% 1.7% 28 23.5% 7.8% 1.7% 4 3.4% 4.4% 0.2% 2 1.7% 6.7% 0.1% 119 100% 7.3% 7.3% Total Count % within N. Type % within Age % of Total 150 9.3% 100% 9.3% 720 44.4% 100% 44.4% 120 7.4% 100% 7.4% 150 9.3% 100% 9.3% 360 22.2% 100% 22.2% 90 5.6% 100% 5.6% 30 1.9% 100% 1.9% 1620 100% 100% 100% Pearson Chi Square: 0.54 (Not significant)
The figures on the table indicate that the most proficient group of speakers are those who are in their 50s and in their 30s as they produce normative instances more than 95% of the time. However, the results obtained by the Chi Square formula proved that in the case of our
participants, age cannot be considered a significant factor in the production of non-normative instances within the studied context.
4.3 Results according to region.
As mentioned in the literature of section 2.4, geography can have a big influence in cross-linguistic studies and in this case it will not be an exception. If the reader takes a look again at Table 1 (Percentage of Latinos and total population of the participants' area) it will probably be understood that even if in all the areas that are studied in this research Spanish has a big role, the amount of contact that our participants is exposed to is different depending on their location.
Table 1 Percentage of Latinos and total population of the participants' area
Austin El Paso Edinburg San Antonio
Houston Brownsville Weslaco Irving Pearland 35.1% 80.7% 88.2% 63.2% 43.8% 93.2% 85% 41.1% 20.5% 811.456 649.133 74.541 1.327.551 2.109.372 174.982 39.474 233.607 108.821
31
Chapter 4: Results
total of 88.7% normative instances -the state where speakers produced the most non-normative tokens. On the top of the region, regarding production of normative instances, we can find the area of Brownsville, which is almost flawless in their production of normative tokens, as 99.2% of the instances are part of that type for only 0.08% that are non-normative. Regarding the results of the Chi-Square and the possible significance of region in the Spanish of the participants, the results show that with our data, there is no significance with this variable; this comes with no surprise after comparing the results of Table 1 and Table 10, for example, a region like Pearland, the one with the lowest Latino population of all them (20.5%) and that is located 498 kilometers away of the boarder, has a production of normative instances of 96.7%; on the other hand an area like El Paso with 80.7% of the population being Latinos or descendent of Latinos, and being located just on the border between United States and Mexico, has a lower level of accuracy with 90.4% of the instances being normative.
Tables 10 & 11 Regions of Texas and instances.
Region
Austin Brownsville Edinburg El Paso Houston Normative Count % within N. Type % within Region % of Total 389 25.9% 92.6% 24.0% 119 7.9% 99.2% 7.3% 283 18.9% 94.3% 17.5% 244 16.3% 90.4% 8.5% 138 9.2% 92.0% 8.5% Non-Normative Count % within N. Type % within Region % of Total 31 26.1% 7.4% 1.9% 1 0.8% 0.8% 0.1% 17 14.3% 5.7% 1.0% 26 21.8% 9.6% 1.6% 12 10.1% 8.0% 0.7% Total Count % within N. Type % within Region % of Total 420 25.9% 100% 25.9% 120 7.4% 100% 7.4% 300 18.5% 100% 18.5% 270 16.7% 100% 16.7% 150 9.3% 100% 9.3% Region
Irving Pearland San Antonio Weslaco Total Normative Count % within N. Type % within Region % of Total 29 1.9% 96.7% 1.8% 29 1.9% 96.7% 1.8% 213 14.2% 88.8% 13.1% 57 3.8% 95.0% 3.5% 1501 100% 92.7% 92.7% Non-Normative Count % within N. Type % within Region % of Total 1 0.8% 3.3% 0.1% 1 0.8% 3.3% 0.1% 27 22.7% 11.3% 1.7% 3 2.5% 5.0% 0.2% 119 100% 7.3% 7.3% Total Count % within N. Type % within Region % of Total 30 1.9% 100% 1.9% 30 1.9% 100% 1.9% 240 14.8% 100% 14.8% 60 3.7% 100% 3.7% 1620 100% 100% 100% Pearson Chi Square: 0.20 (Not significant)
From the 54 participants and as it should be expected on a study of this matter, the com-bination of parents that is repeated with the highest frequency is the one where both parents are Mexicans, involving 28 of our participants. The next combination that is repeated the most, for 12 of our participants, is one parent being Mexican and the other one being American, closely followed by 11 participants whose parents are both Americans. The remaining three participants do not have a Mexican parent involved, there is a Colombian pair of parents, another pair of par-ents from El Salvador and the last interracial couple of parpar-ents that involve an American and a Spaniard. On the literature review we saw how in some of the previous researches being first generation on a new country or being second or third generation was a variable of significance value but we also saw in point 3.1 that age was not a significant role in this matter. Thanks to the information provided about the parents in the “Spanish in Texas” corpus, the following data was collected (Table 12).
Table 12. Origin of parents and instances produced.
Birthplace of parents
Col-Col ElSal-ElSal Mex-Mex USA-SPA USA-USA Total Normative Count % within N. Type % within Countries % of Total 27 1.8% 90.0% 1.7% 28 1.9% 93.3% 1.7% 789 52.6% 93.9% 48.7% 30 2.0% 100% 1.9% 278 18.5% 84.2% 17.2% 1501 100% 92.7% 92.7% Non-normative Count % within N. Type % within Countries % of Total 3 2.5% 10.0% 0.2% 2 1.7% 6.7% 0.1% 51 42.9% 6.1% 3.1% 0 2.2% 0.0% 0.0% 52 43.7% 15.8% 3.2% 119 100% 7.3% 7.3% Total Count % within N. Type % within Countries % of Total 30 1.9% 100% 1.9% 30 1.9% 100% 7.4% 840 51.9% 100% 51.9% 30 1.9% 100% 1.9% 330 20.4% 100% 2.4% 1620 100% 100% 100%
Pearson Chi Square: 0.00 (Significant)
Starting with the analysis related to the parents of the participants, it can be seen that the ones that produce the most cases of non-normative results in our research are those participants whose parents were born in the USA, making them at least, second generation Americans. In fact, the test Pearson Chi Square points out that the results are significant and that the nationality of the parents plays a big role in the language of our participants, making this factor important when analyzing language contact situations. In the case of those participants whose parents (or at least one of them) were born in a Spanish speaking country, from the lowest figure of 84.2% of normative instances obtained by the participants whose parents were Americans, the rest of our group achieve at least, a 90% of normative utterances. The most skilled group, if we do not take into account that participant who has one parent from Spain, are those participants whose parents are from Mexico, achieving an accuracy of almost 94% on more than 840 instances possible and proving that the closer to the roots within your family language, the more skilled.
4.5 Results based on language used with parents, siblings and friends
